Full Proceedings GTSD2022-IsBN 978-604!73!9622-1-VNUHCM Press

PROCEEDINGS OF
TH
20226 INTERNATIONALCONFERENCE
ONGREENTECHNOLOGY
ANDSUSTAINABLEDEVELOPMENT
(GTSD)
July29-30,2022-NhaTrangCity,Vietnam
VNUHCM PRESS
Proceedings of
2022 6th International Conference on Green Technology
and Sustainable Development (GTSD)
July 29-30, 2022 - Nha Trang City, Vietnam
VNUHCM PRESS
2022 6th International Conference on Green Technology and Sustainable Development (GTSD)
Table of Contents
Editors Preface ..................................................................................................................................................... viii
GTSD2022 Organizing Committee......................................................................................................................... ix
A Novel Mechanism Design following Augmented Objective with Flexible Energy Control ......................... 1
Xuan Phu Do and Van Chi Le
Evaluating Green Marketing Trending Determinants by a Text Mining Approach ...................................... 5

Phan-Anh-Huy Nguyen
A Simple Synthesis of Antibacterial and Antifungal ZnO Nanorods ............................................................... 8

Khanh Son Trinh and Vinh Tien Nguyen
An Assessment of Complementary Energy of HPFRCs under Tension ......................................................... 15

Duy-Liem Nguyen, Tien-Tho Do, Thi-Ngoc-Han Vuong and H. T. Tai Nguyen
A Study of the Scavenging Process in a Two-stroke Free Piston Linear Engine at Low Velocity Using
CFD and DPM ..................................................................................................................................................... 21
Nguyen Huynh Thi, Nguyen Van Trang, Huynh Thanh Cong, Dao Huu Huy, Huynh Van Loc, Truong Hoa
Hiep, Ngo Duc Huy and Vo Bao Toan
Investigating the Relationship between Workers’ Needs and Commitments to Garment Enterprise ........ 27
Tu Tran
Optimization of Aggregates in Concrete Brick with Recycled Materials from Stone ................................... 32
Nguyen Thang Xiem, Ho Minh Chau, Tran Doan Hung and Truong Thanh Chung
A Comparison on the Flexural Capacity and Dynamic Performance of a Reinforced Concrete Beam
and a Steel Beam in a High-rise Building .......................................................................................................... 36
Tham Hong Duong
Optimization Model for Biomass Supply Chain Planning: A Case Study in Mekong River Delta –
Vietnam ................................................................................................................................................................ 44
Thi-Be-Oanh Cao, Duc Duy Nguyen, Thanh-Tuan Dang and Chia-Nan Wang
Developing a Digital Competence Performance Assessment Platform for University Students Based on
the DigComp Framework ................................................................................................................................... 50
Anh Tho Mai, Thi Kim Oanh Duong and Anh Tuan Ngo
The Performance of Geonet Reinforced Straw Rolls as a Flexible Waterbreaker for Riverbank
Protection ............................................................................................................................................................. 57
Minh-Duc Nguyen, Le-Nhat-Huy Nguyen and Tran-Phuong-Thao Hua
SSD21, Educational Toolbox for Static, Stability, and Dynamic Analysis of Frame .................................... 63
Truong Thanh Chung, Le Nguyen Anh Vu, Le Cong Lap and Nguyen Thang Xiem
Digital Competence of University Students: A Comparative Study at Three Universities in Vietnam....... 67

Anh Tho Mai, Quynh Trang Mai and Anh Tuan Ngo
i
Load Capacity Evaluation of Simple Reinforced Concrete Girder Bridges with Considering the
Corrosion of Reinforcement and Concrete ....................................................................................................... 75
Tran The Truyen, Tran Thu Minh, Nguyen Dac Duc, Tran Duc Manh and Nguyen Quoc Cuong
Effect of Biodegradable and Metallic Mordants on Dyeing Cotton Fabric with Spent Coffee Grounds .... 80
Tuan Anh Nguyen
The Importance of Green Technology for Sustainable Development Education: A Case Study at Lac
Hong University ................................................................................................................................................... 85
Nga Hong Thi Doan and Truong Van Nguyen
Unique Competitive Advantages of Vietnam’s Garment Industry in the Sustainable Development

Trend ................................................................................................................................................................... 89
Quang-Tri Tran, Thanh-Nhan Nguyen, Tho Alang, Tuyet-Anh Truong, Kim-Chi Le and Nguyen Thi Le
Driving Factors of Green Economy for Smart Cities in the Context of Developing Countries ................... 95
Tiep Nguyen, Nghia Hoai Nguyen, Leonie Hallo and Bao Van Pham
Impacts of Green Training on Green Competencies of Employees: Empirical Case of Industrial

Manufacturers in Dong Nai Parks ................................................................................................................... 101
Thanh-Lam Nguyen, Doan Thi Chuyen, Nguyen Thi Phuong Thao and Doan Van Ly
Mathematical Modelling of Combined Infrared and Heat Pump Drying of Squid .................................... 107
Pham Van Toan, Phan Nhu Quan, Nguyen Hay and Le Anh Duc
Improving the Tensile and Compressive Strength of Cement-Based Materials by Hybrid Electrospun
Nanofibers .......................................................................................................................................................... 114
Tri N. M. Nguyen, Xuan Tung Nguyen, Thanh Toan Dao, Huy Q. Nguyen and Jung J. Kim
Controlling Crystal Morphology via Crystallization Processes, Cases Studied of KDP and Zinc Lactate
............................................................................................................................................................................. 118
Tam Le-Minh, Cuong Nguyen Van and Venkata Subbarayudu Sistla
A Green Solution for Kitchen Waste Treatment Using Earthworm, Experimental and Mathematical
Approaches......................................................................................................................................................... 122
Tam Le-Minh, Phuong Pham Thi Hong and Nhu Vo Thi Thu
Effect of Reinforcement Corrosion on Crack Development in Concrete Under Load ............................... 127
Vo Van Nam and Tran The Truyen
Study on Synthesizing and Size Controlling of Silver Nanoparticles by Using a System of Two
Protectants Trisodium Citrate and Polyvinylpyrrolidone ............................................................................. 132
Hien Chuc Mai, Quynh Nguyen Thi Nhu, Thuan Hoang Duc, Du Cao Van, Cuong Ngo Van and Dung Duong
Thi Ngoc
Study on Chemical Composition and In-vitro Biological Activities of Salvia officinalis L. in Lamdong,
Vietnam .............................................................................................................................................................. 137
Thao Tran Thach, Cuong Ngo Van and Xuan Nguyen Bang
Dynamic Analysis of Plates under Moving Discontinuous Impulsive Load on Viscoelastic Foundation .. 141
The Tuan Nguyen and Trong Phuoc Nguyen
ii
Isolation and Quantitative Determination of Geniposide from Gardenia jasminoides Ellis. Using High-
Performance Liquid Chromatography ............................................................................................................ 147
Vo Thi Nga, Truong Thi Khanh Van and Bui Trung Huu
Oxygen-LPG Torch for Thermal Spraying..................................................................................................... 152

Ngo Thanh Binh, Le Van Canh and Pham Huy Dong
Predicting Land Use Change in Buon Ma Thuot City, Dak Lak Province by Integrating GIS and
Markov Chain .................................................................................................................................................... 158
Nguyen Thi Ngoc Quyen, Nguyen Thi Tinh Au, Nguyen Cong Tai Anh and Tran Thi Xuan Phan
Dissimilar Friction Stir Welded Lap-joint of Aluminum Alloy 6061 and 316 Stainless Steel .................... 165
Huy Huu Ho, Hao Dinh Duong, Nam Hoai Quach, Thuyen Van Phi, and Tra Hung Tran
A Stochastic Half-Car Model for Vibration Analysis with Uncertain Parameters ..................................... 169
Nguyen Van Thuan
Sources of Payment Risks to Contractors in the Vietnam Construction Industry ...................................... 173
Duong Vuong, Thao Huynh and Phu Tran
Pharmacognostic Assessment of Polyscias Fruticosa Leaves in Vietnam ..................................................... 180

Dao Phan Thi Anh, Hue Ha Thi, Trang Le Vu Khanh, Thanh Le Duc, Huong Nguyen Thi Thu and Trieu Ly
Hai
Alkali-Activated Slag/Sugarcane Bagasse Ash Pastes Cured in Room-Air Ambient and in Saturated
Lime Water: A Study on the Compressive Strength and Shrinkage ............................................................ 185
Duc-Hien Le and My Ngoc-Tra Lam
Production Efficiency Improvement Using Value Stream Mapping with Simulation: A Case Study in
Vietnam .............................................................................................................................................................. 191
Xuan-Quang Bach, Thanh-Tuan Dang and Chia-Nan Wang
Study on the Effect of Mixing Ratio of Biodiesel Fuel Made from Animal Fat on Exhaust Emissions of
the Fishing Vessel’s Diesel Engines .................................................................................................................. 199
Pham Dinh Trung, Mai Duc Nghia and Ho Duc Tuan
A Study on Mechanical Properties of Ca-Alginate Hydrogels ...................................................................... 204

Thanh Tan Nguyen, Van Tron Tran, Long Nhut-Phi Nguyen and Nguyen Thi My Le
Mechanical Characteristics of PBT Based Blend ........................................................................................... 208

Hoang-Khang Lu, Ngoc Tran-Nhu Nguyen, Huy Huynh-Nhat Do, Van-Huong Hoang, Van-Thuc Nguyen,
Nga Thi-Hong Pham, Van-Tron Tran, Long Nhut-Phi Nguyen and Thanh-Tan Nguyen
An Evaluation of Corporation Social Responsibility Performance for Vietnamese Contractors .............. 213
Nguyen Van Minh, Ha Duy Khanh, Soo Yong Kim and Chu Viet Cuong
Local Waste Seeds as Organic-Based Coagulant Aids in Water and Wastewater Treatment ................... 218
Nhung Thi-Tuyet Hoang, Anh Thi-Kim Tran and Luu Hong Quang
The Implementation of 5C’s in Online-Foreign Language Teaching for Vietnamese Students in the 4.0
Era....................................................................................................................................................................... 222
Chau Le Thi Bao, Nhu Vo Hoang Nhu and Nhi Ho Yen
iii
Antibacterial Activity of Aqueous Extracts from Marine Sponges Found in Vietnam’s Sea..................... 229
Huynh Nguyen Duy Bao and Nguyen Khac Bat
Digital Competence of University Students: Developing Information and Data Literacy for IT
Students at Ho Chi Minh City University of Technology and Education .................................................... 233
Anh Tho Mai and Quynh Trang Mai
A Simple Design Method for Piled Raft Foundations .................................................................................... 239

Tong Nguyen, Nhat Nguyen Le Anh and Dat Nguyen Thanh
Evaluating Performance of Petroleum Industry Using Data Envelopment Analysis: A Case Study in
Vietnam .............................................................................................................................................................. 247
Kristofer Neal C. Imperial, Chia-Nan Wang, Thanh-Tuan Dang and Nguyen Ngoc Hiep
Fire Resistance Properties and Geopolymer Coating .................................................................................... 254

Van Su Le, Van Vu Nguyen, Artem Sharko, Doan Hung Tran, Petr Louda, Piotr Los, Thang Xiem Nguyen,
Stanislaw Mitura and Katarzyna Buczkowska
Effects of Intake Air Temperature on Power and Emission Characteristics of the HCCI Engine Fueled
with the Blends of 15% Ethanol and 85% Petrol Fuels ................................................................................. 259
Minh Xuan Le and Thanh Tuan Nguyen
Linear Viscoelastic Characterization of an SMA Mixture Using Dynamic Indirect Tensile Test ............. 263
H. T. Tai Nguyen, Hong Ha Mai and Van Hien Nguyen
Effect of the Limestone Powder Content on the Properties of Alkali–Activated Slag Mortar .................. 268
Tai Tran Thanh, Chung Pham Duy, Tu Nguyen Thanh and Hyug-Moon Kwon
The Educational Philosophy of Existentialism with the Development of Personalized Learning for
Learners in Massive Open Online Courses (MOOCs) – The Case of Ho Chi Minh City University of
Technology and Education ............................................................................................................................... 273
Thi Thao Tran, Tran Phuong Thao Hua and Thi Chu Tran
Transportation Infrastructure Strategy for Sustainable Development: A Case Study of Vietnam

Mekong Delta ..................................................................................................................................................... 281
Le Thu Huyen
Study on Behaviour of Short Pile Groups in Soft Ground with Sand Leveling on a Small Scale Model
Using Schneebeli Analogue Soil........................................................................................................................ 286
Sy Hung Nguyen and Thi Phuong Huyen Tran
Research and Preparation of Thinking Strategies for Quality of Textile and Garment Products
Manufactured in Vietnam................................................................................................................................. 292
Nguyen Phuoc Son, Nguyen Ngoc Chau and Nguyen Thi Tuyet Trinh
Building Students’ Self-Reliant on Problem-Based Learning by Embedding Mind of Engineering

Design Concept at HCMC University of Technology and Education ........................................................... 296
Xuan Tien Vo
Modeling and Comprehensive Assessment of Construction Risks: A Perspective of PPP

Transportation Projects .................................................................................................................................... 300
Ha Duy Khanh, Soo-Yong Kim and Nguyen Van Khoa
iv
Fabrication of Graphene Oxide from the Graphite Rod of a Disposed Battery .......................................... 305
Huy-Binh Do, Hoang-Trung Huynh, Thien-Trang Nguyen, Van-Cuong Pham, Tien-Luat Nguyen, Anh-Vu
Phan-Gia and Maria Merlyne De Souza
Design and Optimization of a Compliant Mechanism for Vibration-Assisted Drilling .............................. 309
Hai-Thanh Nguyen, Van-Khien Nguyen, Phan-Khanh-Tam Nguyen, Huy-Tuan Pham, Quang-Khoa Dang
and Pham Son Minh
Eco-friendly Adsorbent Formulated from Rubber Shell Biochar to Remove Chromium (Cr(VI)) from
Aqueous Solutions ............................................................................................................................................. 313
Linh My Nguyen
A Study on the Effect of the Shape of the Center Rib on Thermal Resistance on the Dual-Layer Micro-
channel Heat Sink .............................................................................................................................................. 317
Hung-Son Dang and Thi-Anh-Tuyet Nguyen
Experimental Study of the Effect of Heat Input on Tensile Strength and Microstructure of the Weld
using the Orbital TIG Welding Process .......................................................................................................... 321
Thien Tran Ngoc, Ngoc-Huy Dinh, An-Duong Tra, Kha-Duy Doan, Binh-Minh Ngo and Anh-Duc Pham Duc
A Study of Customer Satisfaction in Online Food Delivery Service Quality During the Covid-19
Pandemic: Baemin’s Case Study ...................................................................................................................... 326
Hong-Xuyen Thi Ho, Ngoc-Tra Tran Thi and Ngoc-Anh Ha Thi
Controllable Green Synthesis and Morphological Properties of Gold Nanostar ........................................ 332
P. Quoc-Duy Huynh, Van-Dung Le, Chi-Hien Dang, Radek Fajgar, The-Ha Stuchlikova, Jiri Stuchlik and
Thanh-Danh Nguyen
Isolation, Selection and Identification of the Probiotic Properties of Lactobacillus spp. from Shrimp
Ponds in Nhon Trach, Dong Nai....................................................................................................................... 336
Doan Thi Tuyet Le, Le Thi Thu Huong, Phan Pham, Pham Minh Thinh, Vo Thi Lan Chi, Nguyen Phuoc
Trung, Huynh Minh Hieu and Do Minh Anh
Production of Cheese from Reconstituted Milk and Soy Milk with the Addition of Microbial
Transglutaminase .............................................................................................................................................. 340
Pham Thi Hoan and Trinh Khanh Son
Studying the Effect of Biodiesel Blend Ratios Derived from Rubber Seed Oil on the Technical
Characteristics of Diesel Engines without Modifying the Structure by Simulation and Experiment ....... 349
Nguyen Manh Cuong and Huynh Phuoc Son
Work from Home during the Covid-19 Pandemic: A Qualitative Research with Women Taking Care
of Young Children ............................................................................................................................................. 355
Hien Phan Thi Thanh, Thuy Nguyen Thi Thanh and Tram Nguyen Thi Mai
Study on Determining the Freezing Mode of Frozen Fillet Bigeye Tuna (Thunnus obesus) ...................... 361
Dzung Tan Nguyen, Linh Khanh Thuy Do, Chuyen Van Hoang and Tuan Thanh Chau
Numerical Investigation of the Optimum Operating Condition in Magnetically Confined Plasma with
Sheared Slab Ion-Temperature-Gradient Model ........................................................................................... 367
Thanh Tinh Tran
v
Using Brake and Engine Torque to Control Traction on Either Side of the Drive Wheel ......................... 371
Tan Tai Phan and Van Nhu Tran
Fatigue Life of Accelerated Corroded Steel Plate .......................................................................................... 377

Dao Duy Kien, Nguyen Thanh Hung, Nguyen Thi Thu Hao, Nguyen Van Hung and Haidang Phan
A Study on Ultrasonic Shear Horizontal Waves in Composite Structures .................................................. 381

Duy Kien Dao, Hoang Ngoc Quy, Truong Giang Nguyen, Ductho Le, Hoai Nguyen and Haidang Phan
Designing and Modeling Pipe Welding Machine ............................................................................................ 385

Tuong Phuoc Tho, Phan Phuc Khang, Tran Thanh Nhon and Pham Phi Long
Investigation of the Mechanical Properties of Lightweight Geopolymer Concrete Using Keramzite as

Coarse Aggregate............................................................................................................................................... 390
Kiet Tran Tuan, Duc Nguyen Phan, Tuan Le Anh, Khoa Nguyen Tan and An Huynh Thao
Researching and Improving the Registration and Treatment Process in Health Care Facility with Lean
Principles and Ergonomics Standards ............................................................................................................. 394
Minh-Tai Le
Application of Lean and Six Sigma Tools to Improve Productivity and Product Quality at Dien Quang
Company, Vietnam ............................................................................................................................................ 401
Minh Tai Le, Hoang Khang Lu and Nhat Huy Do Huynh
Hydrogen Plasma Annealed Gallium and Aluminum co-doped Zinc Oxide Films Applied in Low-
emissivity Glass .................................................................................................................................................. 408
Shang-Chou Chang, Yun-Che Tsai, Huang-Tian Chan, Jian-Liang Lai, Jian-Syun Wu and Wei-Min Wang
Study on Pectinase Production by Bacillus subtilis in Molasses and Its Application for Coffee
Fermentation ...................................................................................................................................................... 412
Ngan K. Le, Duy Q. Nguyen, Nhi Y. Dinh and Phu H. Le
Environmental Sustainability: Exploring Managers’ Attitudes and Behaviours at High-End

Accommodation Businesses in Vietnam .......................................................................................................... 417
Thanh-Nhan Nguyen, Quang-Tri Tran and Tho Alang
Research on Optimal Algorithms Using Experimental Planning to Improve Shoe Production Line
Productivity ........................................................................................................................................................ 425
Minh Tai Le, Thi Cam Duyen Doan and Huynh Thao Vy Nguyen
Application of Multivariable Linear Regression Algorithm to Support Inventory Management ............. 431
Trung Tin Ngo, Minh Tai Le, Nguyen Kim Thoa Vo and Thanh Nam Luong
Ethanol in Gasoline Fuel Blends: Experimental Investigation Effects on Exhaust Emission of the
Homogeneous Charge Compression Ignition Engine ..................................................................................... 437
Minh Xuan Le and Thanh Tuan Nguyen
The Impact of the Industrial Revolution 4.0 on Logistics Businesses: A Case in Mekong Delta ............... 441
Ngo Hong Ngoc and Huang Ying Yin
vi
Enhancement in Dielectric Constant of Poly Vinyl Alcohol by Loading of Strontium Titanate for
Supercapacitor Applications ............................................................................................................................ 447
Anju Yadav, Dinesh Kumar Yadav, Khushbu Meena, Kiran Devat, Narendra Jakhar, Rajesh Sahu, S. K. Jain
and Balram Tripathi
Free Vibration Analysis of Smart FG Porous Plates Reinforced by Graphene Platelets Using
Isogeometric Approach ..................................................................................................................................... 451
Lieu B. Nguyen and Duc-Thien Pham
Free Vibration of MSGT Porous Metal Foam Microplates Using a Moving Kriging Meshfree
Approach ............................................................................................................................................................ 459
P. T. Hung
Impacts of Adverse Weather on Mode Choice Behavior: A case study in Hanoi City, Vietnam ............... 469
Binh Nguyen Mai, Thi Anh Hong Nguyen, Akimasa Fujiwara and Canh Do
Travel Behavior on Ground Access Mode Choices by Introducing a New City Air Terminal: A Case
Study of Vientiane Airport, Laos ..................................................................................................................... 474
Komack Keochampa, Canh Do, Akimasa Fujiwara and Thi Anh Hong Nguyen
A Study of The Tensile Strength for The Mixing Ratio of Eva in Pa6/Eva Blends ..................................... 482
Minh-Nhat Nguyen, Quy-Long Nguyen, Ngoc-Thien Tran, Vinh-Tien Nguyen and Minh The-Uyen Tran
Cytotoxicity of Isoxazole/Pyrazole Curcuminoids against Human Oral Epidermal Carcinoma-KB Cell

Line ..................................................................................................................................................................... 486
Hoang Minh Hao, Ho Dung Manh and Vo Thi Nga
Modeling of Flow Mixed with Polymers in Open Channel Flow: Application on the Blumenau River in
Brazil................................................................................................................................................................... 491
Walid Bouchenafa, Airton Hoenicke, Bruna Luiza Cunico, Huyen Xuan Dang-Vu and Trong Dang-Vu
vii
Preface
This proceedings contains the scientific contributions included in the program of the 6th International
Conference on Green Technology and Sustainable Development (GTSD2022), which was organized on July
29-30, 2022 in Nha Trang University, Khanh Hoa Province, Vietnam. The GTSD International Conference
series is a prestigious bi-annual event created to provide an international scientific research forum in
technologies and applications in the field of Green technology and Sustainable development in the Industrial
revolution 4.0. The areas of GTSD include but are not limited to energy engineering, environmental
engineering, education, digital transformation, new materials and solutions for sustainable development,
advances in computational intelligence and their applications to the real world and so on.
The conference is structurally organized in order to promote the active participation of all attendees and
presenters, via plenary presentation sessions, keynote addresses, interactive workshops and panel
discussions, to find out how to further contribute to and solve various problems in life and manufacture. The
aim was to further increase the body of knowledge in this specific area by providing a forum to exchange
ideas and discuss results.
The program committee members of GTSD2022 come from various countries, and the 269 selected papers
(out of more than 450 submitted papers) are from 27 countries and from 5 continents. This certainly attests to
the widespread and international importance of the theme of the conference. Each paper was carefully
reviewed on the basis of originality, novelty and rigorousness.
We would like to take this opportunity to express our deep appreciation to all authors, participants, keynote
speakers, program committee members, session chairs, organizing committee members, steering committee
members, as well as the organizers for their great efforts and contributions to making the GTSD2022
successful, surging the global care about green technology research for sustainable development.
On behalf of GTSD2022 Organizing Committee
Assoc.Prof. Hoang An Quoc

General Co-Chair of GTSD2022
and
Dr. Nguyen Vu Lan
Publication Chair of GTSD2022
viii
GTSD2022 Organizing Committee

General Chairs Honorary Chairs
A/Prof. Le Hieu Giang, HCMUTE, Vietnam A/Prof. Thai Ba Can, HCMUTE, Vietnam
A/Prof. Trang Si Trung, NTU, Vietnam A/Prof. Do Van Dung, HCMUTE, Vietnam
Prof. Nguyen Ngoc Thanh, WUST, Poland A/Prof. Ngo Van Thuyen, HCMUTE, Vietnam
Prof. Wen-June Wang, NCU, Taiwan Dr. Khong Trung Thang, NTU, Vietnam
Prof. Yo-Ping Huang, NPU, Taiwan
General Co-Chairs Program Chairs

Dr. Quach Hoai Nam, NTU, Vietnam A/Prof. Nguyen Truong Thinh, HCMUTE, Vietnam
A/Prof. Hoang An Quoc, HCMUTE, Vietnam A/Prof. Do Thanh Trung, HCMUTE, Vietnam
Prof. Huann-Ming Chou, KSU, Taiwan A/Prof. Nguyen Minh Tam, HCMUTE, Vietnam
Prof. Chang-Ren Chen, KSU, Taiwan Dr. Huynh Phuoc Son, HCMUTE, Vietnam
A/Prof. Nguyen Vu Quynh, LHU, Vietnam Dr. Tran Thi My Hanh, NTU, Vietnam
A/Prof. Nguyen Thanh Lam, LHU, Vietnam Dr. Pham Thi Thu Thuy, NTU, Vietnam
Program and Steering Committee

Prof. Nguyen Trung Kien, HUTECH, Vietnam Dr. Irma Kunnari, HAMK, Finland
Prof. Surya Narayan Panda, Chitkara Uni., India Dr. Lisa Donaldson, DCU, Ireland
Prof. Arun Patil, SGU, India Dr. Lai Jiang, KULEUVEN, Belgium
Prof. Sid Nair, VIT, Australia Dr. Manoj Kumar Dash, IIITM Gwalior, India
Prof. Pracha Yeunyongkul, RMUTL, Thailand Dr. Le Minh Nhut, HCMUTE, Vietnam
A/Prof. Atul Sharma, RGIPT, India Dr. Nguyen Xuan Vien, HCMUTE, Vietnam
A/Prof. Truong N. Luan Vu, HCMUTE, Vietnam Dr. Diep Phuong Chi, HCMUTE, Vietnam
A/Prof. Le Chi Kien, HCMUTE, Vietnam Dr. Huynh Van Vu, NTU, Vietnam
A/Prof. Bui Van Hong, HCMUTE, Vietnam Dr. Mai Thi Tuyet Nga, NTU, Vietnam
A/Prof. Nguyen Tan Dung, HCMUTE, Vietnam Dr. Tran Quang Huy, NTU, Vietnam
A/Prof. Chau Dinh Thanh, HCMUTE, Vietnam Dr. Jari Jussila, HAMK, Finland
A/Prof. Vo Viet Cuong, HCMUTE, Vietnam Dr. Phan Gia Anh Vu, HCMUTE, Vietnam
A/Prof. Truong Dinh Nhon, HCMUTE, Vietnam Dr. Tran Vu Tu, HCMUTE, Vietnam
A/Prof. Duong Thi Kim Oanh, HCMUTE, Vietnam Dr. Dang Quang Vang, HCMUTE, Vietnam
A/Prof. Trinh Khanh Son, HCMUTE, Vietnam Dr. Nguyen Khac Hieu, HCMUTE, Vietnam
A/Prof. Ngo Dang Nghia, NTU, Vietnam Dr. Vu Van Phong, HCMUTE, Vietnam
Publication Chairs Local Arrangement Chairs

Dr. Nguyen Vu Lan, HCMUTE, Vietnam Tran Nhat Tan, NTU, Vietnam
Vu Thi Thanh Thao, HCMUTE, Vietnam
Do Thi Ngoc Dung, HCMUTE, Vietnam
Media and Publicity Chairs Phung Phuong Thu Thuy, HCMUTE, Vietnam
Chau Ngoc Thin, HCMUTE, Vietnam Nguyen Thi Minh Ngoc, HCMUTE, Vietnam
Nguyen Van Nguyen, HCMUTE, Vietnam Nguyen Thi Hoa, HCMUTE, Vietnam
ix
A Novel Mechanism Design following Augmented

Objective with Flexible Energy Control
Xuan Phu Do Van Chi Le
Mechatronics and Sensor Systems Technology Mechatronics and Sensor Systems Technology
Vietnamese German University Vietnamese German University
Binh Duong, Vietnam Binh Duong, Vietnam
phu.dx@vgu.edu.vn levanchi1991@gmail.com
Abstract: This study presents a newly proposed theory for mechanism. The metabolic cost was studied in [6]. This issue
design of augmented structure, especially in exoskeleton is one of the problems when controlling the mechanism and
design. A general model is proposed with basic elements of a directly affects the structure’s functions [6]. New modification
featured actuator related to mechanism. The calculation of the structure [1] was presented in [7]. This modification [7]
theory is based on dynamical analysis with new modifications concentrated on improving the mechanical function of the
to obtain general equations for design. The new theory is spring at the knee exoskeleton. In addition, the movement of
different from the conventional theory, which always the centroid point in the robotic mechanism was studied in [8].
assumed that the analyzed structure was zero in a static state. When designing the structure, the centroid point is also a part
The augmented theory has broken out the barrier of the of soft compliance calculation. The stiffness of the above
mechanisms is different from the application of smart material
classical model with a new added element related to the
as shown in [9-13] with the same objective in control the
accumulated energy in the structure, especially in the movement.
mechanical spring. After formulating, antagonistic variable
stiffness mechanism is chosen to evaluate the proposed From the above summarization, stiffness control is still a
analysis. The simulation results show that the proposed state-of-art issue when controlling and designing the structure.
theory can meet the requirement in the design of new However, the conventional theory does not fully view the
augmented mechanism. affection of the initial energy in the designed model. Hence it
needs support from the controllers to fulfil the desired
Keywords: augmented mechanism, exoskeleton design, requirements. This is also the main objective of this research
general augmented mechanism, upper exoskeleton, lower which is shown in the next sections.
exoskeleton II. CONVENTIONAL AND PROPOSED THEORY
I. INTRODUCTION
Design of exoskeleton with low power consumption and
high efficacy is always the ultimate objective. The main
design is to develop special calculation and modified
mechanisms based on the conventional structure such as beam
deflection. Normally, the trials in design are always the safe
choice to guarantee the performance of the proposed structure
when manufacturing. However, this also leads to unexpected
results that the efficiency of the structure is lower than the
desired values. This study presents a new view of design
mechanism based on the dynamical parameters to overcome
the disadvantage. Generally, a view of the existing
publications related to the content of this study is carried out. Figure 1. Suddenly energy at the first time of vibration control
A design of a low-power stiffness mechanism was presented
in [1]. The ideal variable stiffness mechanism with its There exists a phenomenon as shown in Fig. 1 in any
potential energy was analyzed based on many assumptions system at the first state. The energy at this state is always
related to the consumption value. To use the internal energy larger than other positions and also affects the control system.
to control the exoskeleton mechanism’s movement was To analyze the phenomenon, two specific models are chosen
presented in [2]. The internal energy [2] was from the muscle in this study. The conventional model of the design
states of a human, and this value belonged to the physical mechanism is shown in Fig. 2. In this figure, three main
health of every person. The disadvantage of the internal mechanism elements are listed: motor, transmission, and
energy is its unstable value and can be attacked by the working dynamic functions. The output of the mechanism includes
environment. As mentioned above, the theory of beam force, stiffness. In some particular cases, the output force can
deflection is main analysis in any research of design be replaced by the output torque belonging to the objective
mechanism. The review of this analysis was presented in [3]. control. This model is also similar to the featured model of an
An extension of the research [2] was studied in [4]. The exoskeleton. The potential energy of the structure is defined
infinity energy in the calculation of the exoskeleton was as follows [1]:
presented in [4]. The analyzed theories [1,2,4] are different
from the conventional analysis [5]. The result [5] was the
background to interpolate the efficient area of the stiffness
Copyright © authors. This work is licensed under a Creative Commons 1

Attribution-NonCommercial-NoDerivatives 4.0 International License.
N li  p , x  always exists in any system, especially in suspension. The

V ( p, x )   
i 1
fi  l  dl (1) variation of the energy also leads to the changing stiffness of
the system as follows:
l0
kOUT  K , x 
Where, fi is the ith force of the structure with its  0, f IN  K , x   0 (5)
equilibrium l0 and the length li at ith position. It is noted that x
Eq.(1) is the general model for any mechanism. The value p Eq.(5) also satisfies the assumption as shown in (3) and
is angular position, and the value x is the translational (4). The stiffness and its energy are not always obtained zero
position of the system. value or equilibrium as seen in the conventional theories.
Hence the criterion for the design of variable stiffness
mechanism with low power can be determined as follows:
d  f IN , k   f IN
n
(6)
Where, n is the order of the mechanism belonging to the

potential energy of the actuator and the desired objective. The
Figure 2. The first model of Antagonistic Variable Stiffness value n can be defined as a fraction or an integer. Normally,
Mechanism (AVSM) the integer value of n is a favorite choice, which will bring a
flexible calculation for design progress.
Energy (1) includes the internal and external energies.
III. APPLICATION TO ANTAGONISTIC
Normally, the internal energies are always assumed zero
value for all calculations. This assumption is used in the VARIABLE STIFFNESS MECHANISMS AND
conventional model to neglect the complicated analysis, and DISCUSSIONS
the obtained results are acceptable. From definition (1), the A. First modification of AVSM
official dynamic parameters related to the movement are
The potential energy of the first model of AVSM as
determined as follows [1]:
shown in Fig. 2 is given by:
V V  2V
f IN  p, x   , fOUT  p, x    , kOUT  2 (2) V ( p, x ) 
1 
k  x  pR    x  pR  
3 3
(7)
x x p 3  
Where, fIN is the input force, fOUT is the output force, and The potential energy (7) is rewritten as dimensionless
kOUT is the output stiffness. In Fig. (1), the transmission function as follows:
element is normally the mechanical spring-like model. In
some exceptional cases, the element is the beam and the
activated actuators. The condition for an ideal elastic element
V ( p, x ) 
1
3

x 3 p2  x 2  (8)
is proposed as follows: x
Where, x  , p  p . Using Eq.(2), the output stiffness
N li  p , K 
R
VIDEAL
VIDEAL ( p, K)  V ( p, K )    fi  l  dl ,
K
0 and the input force of AVSM are found as follows:
i 1 l0 kOUT  x   2 x , f IN  p, x   p 2  x 2 (9)
(3)
The input force (9) can be written as:
Where, K is a symbolized translational position. In the
kOUT  x 
proposed equation (3), the condition energy for an ideal 2
mechanism is different from the classical definition. This is f IN  p, x   p 2   funIN
2
(10)
from the characteristic of the manufactured elements always 4
remains an initial power inside its model. The energy can be Using Eq.(10), the criterion for design AVSM is
seen as lower power, but it is complete affection to combine determined as:
the external energy. This combination brings an unexpected
phenomenon in dynamic response at the first time to apply a d  f IN , kOUT   funIN
2
(11)
controller. The power motor can be defined as follows:
The result (11) shows that the energy mechanism for
p  x, x, x, f IN   UI control is a square exponential function with the input control
 w3P0   w1x  w2 x  P0   w1x  w2 x  f IN  w3 f IN
2
(4) and the stiffness output. The exponential value also reflects
the relation in the connection that the energy for every spring
=  w1x  w2 x   P0  f IN   w3 P0  f IN
2
  cannot dependently. In addition, the source energy must be
large and obtain double value for control the system.
Where, the values of w1 , w2 , w3 are found based on the
experiment data corresponding to acceleration, velocity, and
initial power of the system; P0 is the initial power of the
motor (W). Eq.(4) is designed based on the assumption in
Eq.(3). Eq. (4) also satisfies the phenomenon in the first time
of vibration control as shown in Fig. 1. This phenomenon
This research is funded by Vietnam National Foundation for Science and
Technology Development (NAFOSTED) under grant number 107.02-
2020.13. Corresponding author: Dr. Xuan Phu Do
2
B. Second modification of AVSM potential energy is not zero value if there exists internal
energy of the structure as shown in Eq.(4).
The result of second AVSM is shown in Fig. 5(a,b,c). The
relation of the input force and the output stiffness is linear-
like as shown in Fig. 5(a). This response is from the
connection of the first and the second mechanical springs as
shown in Fig. 3. The affection of the connection is also shown
in Fig. 5(b). It is remarked that the values of the output
stiffness are the actual value of the mechanical spring, which
is different from Fig. 4 as the difference value when
Figure 3. The second model of AVSM controlling. The result in Fig. 5(c) also points out that the
potential energy will increase if there is internal energy in the
The second mode of AVSM is shown in Fig. 3. The structure.
potential energy of the first model of AVSM as shown in Fig.
3 is given by:
1
k l  p, x1, x 2 
2
V ( p, x1, x 2 )  (12)
2
Where, k is the input stiffness of the structure, p is the
angular position, l is the deflection of the spring length, x1
and x2 are the position of the first and the second springs. The
input force and the output stiffness are determined as follows:
 l  p, x1, x2  
2
kOUT  x1, x2   k   
 p
  (13)
 2l  p, x1, x2 
 k l  p , x1, x2  (a)
p 2
l  p, x1, x2 
f IN  p, x1, x2   k l  p, x1, x2  (14)
p
It is remarked that the input stiffness is defined as

k  2 R 2  x1  x2  , where  is a constant related to the
proportionality of the force-deflection springs. The input
force (14) can be written as follows:
f IN  p, x1, x2   funIN (15)
Using Eq.(10), the criterion for design AVSM is

determined as:
d  f IN , kOUT   funIN (16) (b)
The result (16) shows that the mechanism’s energy for
control is first order function with the input control and the
stiffness output. This also points out that the movement is
independent when controlling.
IV. SIMULATION AND DISCUSSIONS
Simulation results of the first and the second AVSMs are
shown in Fig. (4) and Fig. (5). In the first model as shown in
Fig. (4), the criterion for design follows the second-order
function of the input force and the output force. This relation
is reflected clearly in Fig. 4(a). It is shown that the increased
value of the output stiffness from 0 to 2 N/m corresponds to
the input force of approximately 100 N. In this point, the
potential energy is nearly zero value. However, the output
(c)
stiffness changes from 2 to 4 N/m with the potential energy 5 Figure 4. Result of the first AVSM: (a) input force versus output
W as shown in Fig. 4(b). It is remarked that the relation stiffness, (b) potential energy versus output stiffness, (c) potential
between the input force and the potential energy is not linear energy versus input force
as shown in Fig. 4(c). In Fig. 4(c), it is shown that the
3
and lower limb. The theory also gives a new view of design
mechanisms when the energy and the stiffness of the system
cannot be predicted.
REFERENCES
[1] Vincent Chalvet, David J.Braun, “Criterion for the design of
low-power variable stiffness mechanisms”, IEEE Transactions
on Robotics, vol. 33, no. 4, pp. 1002-1010, 2017.
[2] Amanda Sutrisno, David J.Braun, “How to run 50% faster
without external energy”, Science Advances, vol. 6, no. 13,
eaay1950, 2020.
[3] Ke Wu, Gang Zheng, “A comprehensive static modeling
methodology via beam theory for compliant mechanisms”,
Mechanism and Machine Theory, vol. 169, pp. 104598, 2022.
[4] Amanda Sutrisno, David J.Braun, “Enhancing mobility with
quasi-passive variable stiffness exoskeletons”, IEEE
Transactions on Neural Systems and Rehabilitation
(a) Engineering,vol. 27, no. 3, pp. 487-496, 2019.
[5] C.E.English, D.Russell, “Mechanics and stiffness limitations
of a variable stiffness actuator for use in prosthetic limbs”,
Mechanism and Machine Theory, vol. 34, pp. 7-25, 1999.
[6] Daniel F.N.Gordon, Christopher McGreavy, Andreas Christou,
Sethu Vijayakumar, “Human-in-the-loop optimization of
exoskeleton assistance via online simulation of metabolic
cost”, IEEE Transactions on Robotics, pp. 1-20, DOI:
10.1109/TRO.2021.3133137
[7] Sung Y.Kim, David J.Braun, “ Novel variable stiffness spring
mechanism modulating stiffness independent of the energy
stored by the spring”, 2021 IEEE/RSJ International Conference
on Intelligent Robots and Systems (IROS), pp. 1-6, DOI:
10.1109/IROS51168.2021.9636339, 2021
[8] Qiang Huang, Chencheng Dong, Zhangguo Yu, Xuechao
Chen, Qingqing Li, Huanzhong Chen, Huaxin Liu, “Resistant
compliance control for biped robot inspired by humanlike
behavior”, IEEE/ASME Transactions on Mechatronics, pp. 1-
(b) 11, DOI: 10.1109/TMECH.2021.3139332, 2022
[9] Do Xuan Phu, Kruti Shah, Seung Bok Choi, “A new
magnetorheological mount featured by changeable damping
gaps using a moved-plate valve structure”, Smart Materials and
Structures, vol. 23, pp. 125022, 2014
[10] D. X. Phu, S. B. Choi, Y. S. Lee, M. S. Han, “Design of a new
engine mount for vertical and horizontal vibration control using
magnetorheological fluid”, vol. 23, pp. 117001, 2014
[11] Do Xuan Phu, Seung Bok Choi, “Vibration control of a ship
engine system using high-loaded magnetorheological mounts
associated with a new direct fuzzy sliding mode controller”,
vol. 24, pp. 025009, 2014
[12] Xuan Phu Do, Huy Ta Duc, Le Tran Huy Thang, Seung Bok
Choi, Dalseong Yoon, “Design of a new inside multi-coil
clutch for knee-exoskeleton structure based on Helmholtz
phenomenon and magneto-rheological fluid”, Proceedings
Volume 10598, Sensors and Smart Structures Technologies for
(c) Civil, Mechanical, and Aerospace Systems 2018, 105983M,
2018
Figure 5. Result of the second AVSM: (a) input force versus [13] Xuan Phu Do, Long Mai Bui Quoc, Seung Bok Choi, Nguyen
output stiffness, (b) potential energy versus output stiffness, (c) Quoc Hung, Ae Ri Cha, “Design of a new exoskeleton based
potential energy versus input force on the combination of two magnetorheological damper”,
Proceedings Volume 10598, Sensors and Smart Structures
V. CONCLUSION Technologies for Civil, Mechanical, and Aerospace Systems
2018, 105983M, 2018
In this study, a new view of the stiffness mechanism is
presented. This proposed theory is designed from the mutation
energy at the first time to control the system. This theory also
modifies the traditional mechanism theory that the energy is
always approximate zero value. The proposed theory also
presents equations related to the stiffness, input energy, and
output energy. These equations point out that the energy
cannot assume the zero value related to the design of the
mechanism. Two AVSM models are used to apply the
proposed theory. The simulation and its results show that the
proposed theory can be developed to design the augmented
mechanism for supporting the movement of the upper limb
4
Evaluating Green Marketing Trending

Determinants by a Text Mining Approach
Phan-Anh-Huy Nguyen
Faculty of Economics
HCMC University of Technology and Education
Ho Chi Minh City, Vietnam
huynpa@hcmute.edu.vn
Abstract: Green Marketing can also refer to the used MCDM to select a suitable green marketing strategy
production and marketing of goods based on environmentally under fuzzy environment. Suki et al. investigated the impacts
friendly products made from renewable materials. This of corporate social responsibility on the links between green
approach attracted many companies in recent years. marketing concerns and consumer purchase intentions [4].
Identifying the trending factors influencing green marketing Chung, Kuo Cheng reported the green marketing orientation
is a complicated task. This paper proposed a novel to achieve sustainable development in hospitality
approach based on text mining techniques to figure out the management [5]. Mujahid, M., et al. analyzed the link
new determinants in green marketing activities. The results between green marketing and performance in small
businesses [6]. In [7], Mohd Noor et al. tried to understand
of this paper can be applied to enterprises in their selecting
the consumers and green product purchase decision in
business strategy development to boost their business Malaysia using a structural equation modeling-partial least
performance. square (sem-pls) approach. In [8], Amaliana, Luthfatul, et al.
analyzed the consistency of bootstrap resampling in structural
Keywords: text mining, green marketing, marketing model with PLS-PM approach with the technology
strategy, data mining acceptance model to evaluate in green marketing
management strategy. In [9], Said, Syahnur, et al. applied
I. INTRODUCTION green marketing practice in purchasing decision home care
Green marketing is an important topic in recent years product. The results of the study showed that the attribute of
since more entrepreneurs focus on the environmental impacts environmentally friendly product design with the green label
on society. Furthermore, it is getting more concerns from is a variable that dominates the selections of consumers when
many organizations in the era of industrial revolution 4.0. In purchasing the products. Furthermore, preference attribute
current years, there are several studies investigating green weight is having the smallest influence on making the
marketing issues in business [1 - 8]. However, most of the purchasing decisions. To the best of our knowledge, there is
previous papers have not found the main themes which are a lack of paper applying text mining to figure out the trends
affected by green marketing. Figuring out the trending in green marking in recent years. Hence, this paper proposed
determinants is critically important to solve those problems. a new approach to handle this issue by evaluating the green
Therefore, this paper implemented a text mining approach in marketing determinants.
order to help the organizations to have better conversations in
the new situation. III. METHODOLOGY
The contributions of this study are as follows: A. Dataset description

In this study, we use a dataset which is collected from
1) A new text mining approach is adopted to analyze the new
4000 related articles from newspapers recently. The
dataset.
following keywords are used for the query to get access to the
2) Four main themes with some related factors have been data: business, experience, data, analysis, technology, skills,
discovered. work, management.
3) The implications are proposed to support the business to B. Text mining method
solve the green marketing issues. We use text mining software to visualize word cloud with
The remaining sections of the study are organized as follows: co-occurrences of term extracted from the main text dataset.
in section 2, the reviews of related works are presented. In this paper, we use topic modeling by co-word analysis. The
Section 3 proposed the methodology. Section 4 shows the frequency of keywords is utilized to create the co-occurrences
experimental results and implications. matrix and the clusters of related topics. Then, the network
visualization is also conducted to show the relationships
II. LITERATURE REVIEW between topics. The algorithm is:
Green marketing has been defined and analyzed in many Step 1: Similarity Matrix construction
studies in recent years. In [1], Peattie and Ken reported the
concepts of green marketing and its main features. Dalton et From the co-occurrence matrix by normalizing the latter
al. drilled down about management techniques when matrix, the similarity matrix is constructed on differences in
applying green marketing [2]. In [3], Jain, Akansha, et al. the total number of occurrences.

Step 2: Mapping technique: The factor “climate change” appears 134 times, although
it is not the factor with the highest frequency, it is an
The aim of mapping technique is to minimize a weighted important factor that shows the relationship with other factors
sum of distances between all pairs of items by the squared in the same cluster 1 such as person, government, planet,
Euclidean. effort. This shows that climate change is an issue that the
Step 3: Translation, Rotation, and Reflection. whole world needs to pay attention to. Climate change will
lead to many serious consequences for our planet, so people
The objective of three transformation techniques is to and governments are both working to limit and prevent
transform the solution obtained for the optimization problem. climate change by saving resources, based on research and
Hence, it can create consistent results. development technology to create new alternative energy.
IV. EXPERIMENTAL RESULTS
The text mining results with discovered topics and
implications are described in this section.
Figure 1. The word-cloud

From the resulting data in the word-cloud, it can be seen
that the keyword "Green" has been updated with a lot of tools, Figure 3. Results for cluster 2 (green)
maybe 5023 from green issues is seen as an issue that many The “market” element appears with the highest frequency
people care about and talk about. The whole issue is updated 620 times among the elements in cluster2. It is not surprising
as “Marketing” with 2194 occurrences for marker work in that the element “market” appears a lot in the analysis results
today's society is very important, especially the Green because the market is the place where marketing strategies
Marketing trend. Around the keywords when it comes to this are implemented. In addition, in this cluster 2, we see the
issue are: product, environment, consumer, company, appearance of these factors such as strategy, research, firm
enterprise, energy, market and new. (company), country, ... The appearance of these factors shows
The keywords related to Green Marketing show that that if companies want to understand the market or dominate
consumers are interested in the company's environmentally the market, it is essential to do research to analyze and come
friendly products, affect the market, how is the energy up with appropriate strategies for each market in different
problem. The keywords related to the issue show that countries and different families. In a market where consumers
consumers are now interested in environmentally friendly are increasingly concerned about the environment,
green products and new products or new energy sources to companies need to make changes in product strategies to
help protect the environment, and companies should grasp meet consumer needs.
this green trend to influence the company's marketing to help
the company understand the market and develop better.
The results obtained after performing data set analysis on
Green marketing appeared 4 main color clusters. In each
cluster, there will be a particular factor that is larger than the
others because it appears more and has a strong association
with other clusters.
Figure 4. Results for cluster 3 (blue)
Product is the element that appears the most with 1837

times and has many close links with factors inside and outside
the cluster to create a strong linked system. The reason that
Figure 2. Results for cluster 1 (red)
6
products are mentioned so many times in the topic of green discovered. From that, we have also analyzed the meanings
marketing is because plastic waste is currently a factor that of each topic with some related determinants. Then some
has a serious impact on the environment. However, most of business implications have been reported from the results.
the plastic waste discharged into the environment today
comes from product packaging, so in order to implement The limitation of this research is the number of text
green marketing campaigns, the first thing that needs to be documents in the dataset is not quite huge. For future work,
mentioned must be the change. Product packaging becomes we could get more data to solve some bias problems. We
more environmentally friendly. could also apply sentiment techniques for implementing
those tasks.
Through the links between product elements (product)
and factors inside and outside the cluster such as consumers, ACKNOWLEDGEMENTS
technology, brands, strategies, etc., we see that brands should This research is supported by Ho Chi Minh City
start making changes to their products by applying new University of Technology and Education, Vietnam.
technologies to research alternative materials in order to
change the product packaging to be more environmentally REFERENCES
friendly by capturing the green living trend that consumers [1] Peattie, Ken. "Green marketing". The marketing book.
are aiming. Certainly, this change will also consume time and Routledge, pp. 595-619, 2016.
money of businesses, but if businesses have specific and [2] Dalton, Marie, Dawn G. Hoyle, and Marie W. Watts. Green
properly executed strategies, they will bring good results such marketing management. South-Western College, 2010.
as gaining advantages in the market as well as building value
[3] Jain, Akansha, et al. "Selection of a green marketing strategy
for the brand in the hearts of consumers.
using MCDM under fuzzy environment". Soft Computing for
Problem Solving. Springer, Singapore, pp. 499-512, 2020.
[4] Suki, Norazah Mohd, Norbayah Mohd Suki, and Nur
Shahirah Azman. "Impacts of corporate social responsibility
on the links between green marketing awareness and
consumer purchase intentions". Procedia Economics and
Finance, vol. 37, pp. 262-268, 2016.
[5] Chung, Kuo Cheng. "Green marketing orientation: Achieving
sustainable development in green hotel
management". Journal of Hospitality Marketing &
Management, vol. 29.6, pp. 722-738, 2020.
[6] Mujahid, M., et al. "Linking green marketing with
performance: Environmental marketing model for small
business". IOP Conference Series: Earth and Environmental
Science. vol. 737. No. 1. IOP Publishing, 2021.
[7] Mohd Noor, Mohd Nazri, et al. "Understanding consumers
and green product purchase decision in Malaysia: a structural
equation modeling-partial least square (sem-pls)
approach". Asian Social Science , vol. 12.9, pp. 51-64, 2016.
Figure 5. Results for cluster 4 (Yellow)
[8] Amaliana, Luthfatul, et al. "The consistency of bootstrap
resampling in structural model with PLS-PM approach:
In this cluster, we focus on analyzing the green product technology acceptance model in green marketing
factor. Like the analysis mentioned above, green product is a management strategy". IOP Conference Series: Earth and
trend in the current market. However, creating green and Environmental Science. Vol. 239. No. 1. IOP Publishing,
environmentally friendly products will cost many times 2019.
higher than conventional products. Therefore, at present, [9] Said, Syahnur, et al. "Green Marketing Practice In Purchasing
green and environmentally friendly products are still very Decision Home Care Product". International Journal of
expensive compared to other common products. That also Scientific & Technology Research, vol. 9.06, pp. 893-896,
creates a huge impact on sales of green products. Hence, 2020.
businesses need to have more strategies, learn to apply [10] A. Ottman, Green Marketing: Opportunity for Innovation,
technologies, and cooperate with suppliers to minimize the Chicago, IL:NTC Business Books, 1997.
cost of producing green products. From there, it is possible to
[11] M. Charter, Greener Marketing: A Responsible Approach to
optimize the price to compete in revenue with other
Business, Sheffield, UK:Greenleaf Publishing, 1992.
conventional products.
[12] M. Polonsky, "An Introduction to Green
V. CONCLUSION Marketing". Electronic Green Journal, vol. 1, pp. 1-10, 1994.
In this study, a novel text mining approach has been
implemented to find the green marketing trends. There are
four main clusters which are the themes that have been
7
A Simple Synthesis of Antibacterial and Antifungal

ZnO Nanorods
Khanh Son Trinh Vinh Tien Nguyen
Faculty of Chemical and Food Technology Faculty of Chemical and Food Technology
Ho Chi Minh City University of Technology and Ho Chi Minh City University of Technology and
Education Education
Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam
sontk@hcmute.edu.vn tiennv@hcmute.edu.vn
Abstract: ZnO nanorods were prepared by precipitation preparation and hence increasing the production costs. In this
method with Zn(CH3COO)2 as the precursor, NaOH as the study, we used a simple chemical method to prepare the ZnO
precipitating agent and soluble starch as the capping agent. nanoparticles (NPs) using inexpensive and commercially
X-ray diffraction analysis showed that the ZnO nanoparticles available reagents including zinc acetate as the precursor,
(NPs) were structured in the wurtzite lattice with an average sodium hydroxide as the precipitating agent, and soluble
crystallite size of 11.38 nm. Field-emission scanning electron starch as the capping agent to control the size of NPs. The
microscopy coupled with energy-dispersive spectroscopy produced ZnO NPs were then tested for antibacterial and
showed that the material was mostly nanorods with an antifungal activity against different species.
average size of 28 nm. UV-vis absorption spectra showed that II. EXPERIMENTAL METHODS
ZnO-NPs strongly absorb UV radiations. The ZnO NRs
demonstrate typical dose-dependent antimicrobial activities A. Synthesis and characterization of ZnO NPs
and inhibited Escherichia coli, Staphylococcus aureus, and ZnO NPs were synthesized based on a reported method
Salmonella sp. 99.7, 99.96, and 100% respectively at 184 [4]. Soluble starch (1.5 g) was completely dissolved in 250 mL
ppm ZnO-NPs. At 553 ppm concentration, ZnO-NPs of distilled water by stirring at 90oC for 10 min. After cooling
demonstrated 100% inhibition for the three bacterial strains. the solution to room temperature, 5.4875 g of
Gram-positivity and Gram-negativity seemed to be Zn(CH3COO)2.2H2O (0.025 mol) was added and stirred until
unimportant for the resistance of bacteria toward ZnO NRs. complete dissolution. After that, 250 mL of 0.2 M NaOH
ZnO NRs at 14740 ppm showed a weak 98.2% growth solution was added dropwise under continuous stirring. The
inhibition against the fungi Colletotrichum gloeosporioides. solution was then left stirred for 2 h to complete the reaction.
FESEM and EDS analyses showed that the cells of bacteria The reaction mixture was then filtered and the solid was
were distorted and damaged when being in contact with ZnO washed with 3x 500 mL of distilled water and dried at 80oC.
The dried solid was then ground into a powder.
NRs in dispersion. The fungal cells in these conditions turned
into spores and were also damaged. The synthesized ZnO UV-vis spectra were recorded using a UH5300 (Hitachi,
NRs can find a wide range of applications as an antibacterial Japan) spectrophotometer.
and antifungal component in inks, paints, coatings,
X-ray diffraction spectra were recorded using a D8
cosmetics, and pharmaceutics. Advance platform (Bruker, Germany) with  = 0.15406 nm at
25oC.
Keywords: ZnO nanoparticles, antibacterial, antifungal,
UV absorption, Staphylococcus aureus, Salmonella The size of crystallites was calculated using the Scherrer
equation [5]
I. INTRODUCTION
𝐾
In recent years, the synthesis and antimicrobial activities 𝐷=
of ZnO nanomaterials are gaining much attention due to their 𝑐𝑜𝑠(𝜃)
low cost, chemical stability, biocompatibility and Where  was the Full Width at Half-Maximum
biodegradability. Recent research on ZnO nanomaterials (FWHM,rad), 2 was the diffraction angle (rad),  = 0.15406
focused on green methods of their syntheses using natural nm was the wavelength of the X-ray, K = 0.9 was a constant,
sources of reagents. For example, plant extracts were used to and D (nm) was the crystallites size.
synthesize ZnO nanoparticles and their antibacterial activity
was tested on E.Coli, Streptococcus sp, Enterococcus sp, and Field Emission Scanning Electron Microscopy (FESEM)
Klebsiella sp [1]. Another study used the powder of leaves of was conducted using an S–4800 (Hitachi, Japan) electron
S. multiflorus to prepare ZnO nanomaterials and tested their microscope. Energy Dispersive Spectra (EDS) were recorded
antifungal activity on Aspergillus niger and Aspergillus flavus using a 7593-H spectrometer (Horiba, Japan). Before FESEM
[2]. ZnO nanomaterials were also synthesized using and EDS measurements, the ZnO sample was sputtered with
diethylene glycol and triethylene glycol and incorporated into a thin layer of Pt to enhance the electrical conductivity of ZnO.
an antimicrobial film against Staphylococcus aureus and B. Antimicrobial tests of ZnO NPs
Proteus vulgaris [3].
Antibacterial activity of ZnO NPs was evaluated based on
Most of the methods using natural sources such as plant a reported method with some modifications [6]. The E.coli
extracts are relatively complicated due to involving the extract bacteria were first activated in a Nutrient Broth medium at

37oC for 24 h. The ZnO NPs were dispersed in a nutrient agar III. RESULTS AND DISCUSSION
medium with different concentrations from 0 to 737 ppm. The In this reaction, zinc ions were first precipitated into
nutrient agar medium was poured into Petri dishes and when Zn(OH)2 and then dissolved in the presence of excessive
the agar solidified, 100 l of the bacteria suspension was hydroxide ions:
spread on the surface of the agar. The Petri dishes were then
incubated at 37oC for 24 h. After that, the density of survived Zn2+ + 2OH- → Zn(OH)2
bacteria was determined by counting the number of bacterial Zn(OH)2 + 2OH- → Zn(OH)42- + 2H+
colonies. The antibacterial activity was calculated using the
formula The complex was unstable and gradually turned into ZnO
𝐴1 −𝐴2 NPs according to the reaction [10]:
%inhibition= × 100,
𝐴1 Zn(OH)42- → ZnO +H2O + 2OH-
where A1 was the bacterial density in the Petri dish without The presence of excessive OH- ions turned soluble starch
ZnO NPs (CFU/mL), and A2 was the bacterial density in the into a negatively charged form and hence stabilized the ZnO
Petri dish with ZnO NPs (CFU/mL). NPs by interacting with the positively charged Zn ions in the
One-way Analysis of Variance (ANOVA) was conducted ZnO crystallites [11].
to test the statistically significant difference between the A. Characterization of ZnO NPs
means with a confidence level of 95%.
XRD spectrum of the synthesized material (Figure 1)
Antifungal activity of the ZnO NPs was evaluated using a shows characteristic peaks with corresponding planes of ZnO:
reported method [7]. A PDA medium was autoclaved at 121oC 31.8424o (100), 34.3866o (002), 36.2351o (101), 47.4113o
for 15 min and then ZnO NPs were added with different (102), 56.605o (110), 62.9608o (103), and 68.0339o (112),
concentrations from 0 to 14740 ppm. The medium was poured which were also reported in another study [12]. These peaks
into Petri dishes and left for solidification. The spores of the indicated the wurtzite structure of ZnO based on the standard
fungi were then put in the centre of each Petri dish. The size JCPDS number 36–1451.
of the fungal colony was measured every day. The percent
inactivation was calculated using the following formula [8]: 1000
(002)
So  Ss
Intensity (a.u)
800
%inhibition  100
(100)
(101)
So 600
(110)
(102)
(103)
(112)
400
Where So was the area of the fungal colony (cm2) in the
control Petri dish without ZnO NPs (cm2) and Ss was the area 200
of the fungal colony (cm2) in the Petri dish with ZnO NPs
0
(cm2). 20 30 40 50 60 70
To evaluate the effect of the presence of ZnO NPs on the 2 theta (o)
bacterial cells, the microorganisms were first activated (in
Nutrient Broth medium for bacteria and in Potato Dextrose Figure 1. XRD spectum of ZnO NPs
Broth medium for fungi) at 37oC for 24 h. After that, 30 mL
The peaks at 2θ of 33.5334o and 59.5205o belong to
of the bacteria suspension was added to 30 mL of a sterilised
Zn(OH)2, according to JCPDS number 38–0356 [13]. This
Nutrient Broth medium containing ZnO NPs of 184 ppm for
result indicated that the temperature of 80oC was not high
bacteria and 3685 ppm for fungi. After 24 h of stirring, the
enough to fully convert Zn(OH)2 to ZnO.
bacterial cells were centrifuged at 2000 rcf for 30 min, put on
a silicon wafer and left for drying under ambient conditions. The crystallites sizes of ZnO NPs were calculated using
The morphology of the cells was observed using the FESEM the Debye – Scherrer equation for different XRD peaks. The
and EDS equipment which was used for ZnO NPs described average of these sizes was 11.38 nm, which is close to the size
above [9]. of the ZnO NPs determined from the SEM images.
A B
Figure 2. FESEM micrograph of ZnO NPs at 50 000 magnification (A) and frequency distribution of particle sizes from
the FESEM micrographs (B)
9
Figure 2A shows that there were several shapes of ZnO The Energy dispersive spectrum (Figure 3) shows that zinc
NPs, including nanorod, nanowire, and nanotube, with the and oxygen were the main elements in the synthesized
majority being nanorods. The nanorod shape of ZnO NPs in material. The presence of carbon and the higher atomic
this study was due to the usage of precursor zinc acetate. It percentage of oxygen compared to zinc were possibly due to
was shown that the acetate precursor of zinc produced mainly the presence of soluble starch absorbed on the surface of ZnO
nanorods, while the chloride and sulfate precursors produced NPs. Because the size of ZnO NPs was small, the surface
nanoprims [14]. Besides, in this study we used dilute 0.1 M interaction between the capping soluble starch and the ZnP
solution of zinc ions, thus facilitating the formation of was strong, therefore it was difficult to completely remove the
nanorods instead of nanoplates [15]. soluble starch by washing with water. The presence of Al was
due to the aluminum holder of the ZnO sample and the
Figure 2B shows that the sizes of nanorods ranged from 10 presence of Pt was due to the platinum sputtering process
to 57 nm with an average of 27.7 nm. This result demonstrated before SEM and EDS measurements.
that the soluble starch effectively protected the NPs from
aggregation and growing in diameter [16].
Element % mass % atom
C 21.38 39.72
O 31.39 43.79
Al 0.43 0.36
Zn 46.57 16.10
Pt 0.23 0.03
Total 100 100
Figure 3. EDS spectrum (left) and element composition (right) of the synthesized ZnO NPs
UV-vis spectrum of the ZnO NPs dispersion in water Table 1 shows that the antibacterial effect of ZnO NPs
(Figure 4) shows a broad absorption peak near 366 nm. This against Gram-negative E.coli was lower than against Gram-
result is similar to that of another study that reported an positive S.aureus, similar to other reports [20-22]. In Gram-
absorption band of ZnO NPs from 355 to 380 nm [17]. The positive bacteria, the cell is protected by a thick peptidoglycan
strong absorption of UV radiations by ZnO NPs can have membrane composed of teichoic acid, lipoteichoic acid, and
several applications such as in sunscreens or anti-aging surface proteins, while in Gram-negative bacteria, the cell is
coatings. protected by a membrane composed of lipopolysaccharide,
porins and a thin peptidoglycan layer [23]. Although E.coli
B. Antibacterial activity and Salmonella are both Gram-negative, E.coli is more
Table 1 shows that a concentration of 185 ppm of ZnO resistant while Salmonella is less resistant than S.aureus to
NPs inhibited at least 99.0% of the growth of the three ZnO NPs. This result indicated that the Gram type of bacteria
bacteria. ZnO NPs exhibited a dose-dependent antibacterial might not be a significant factor in their resistance toward ZnO
activity, which was also reported in other studies [18, 19]. The NP.
high antibacterial effect was due to ZnO NPs because an
equivalent amount of Zn(CH3COO)2 exhibited significantly
lower antibacterial activity [19].
Table 1. Percent inhibition against bacteria by ZnO NPs at
0,6 different concentrations
0,55
Bacteria %inhibition
0,5
Absorbance
0,45
0,4 184 369 553 737
0,35 366
ppm ppm ppm ppm
0,3
0,25 E.coli 99.07a 99.97b 100c 100c
0,2
200 250 300 350 400 450 500 S.aureus 99.96b 100c 100c 100c
Wavelength (nm)
Salmonella 100c 100c 100c 100c
Figure 4. UV-vis spectrum of the dispersion of synthesized ZnO Numbers with different superscript letters are significantly different
NPs in water
(p<0.05).
10
35,00
30,00
25,00
Area of growth (cm2)
20,00
15,00
10,00
5,00
0,00
1 2 3 4 5 6 7
Day
Control 0 ppm 3685 ppm 7370 ppm 11055 ppm 14740 ppm
Figure 5. Area growth of C. gloeosporioides treated with ZnO NPs with different concentrations for 7 days
C. Antifungal activity D. Morphology and elemental analysis of microbial cells

after ZnO NPs treatments
Figure 5 shows that the fungi C. gloeosporioides
continuously grew even in the presence of ZnO NPs. Comparing sub-figures (A) and (B) in Figures 6, 7, and 8
However, the presence of ZnO NPs significantly inhibited the shows that the bacterial cells were shrinkled and distorted after
fungal growth in all 7 days of the experiment. Increasing the the treatment with ZnO NPs. On the cells in sub-figures (B)
ZnO NPs concentration resulted in higher growth inhibition. there were white dots that contained 3-10% zinc atoms based
After 7 days, with ZnO concentrations from 3685 to 14740 on the EDS analysis (data table not shown), which are similar
ppm, the percent inhibition ranged from 82.8% to 98.2%, to other studies [28-31].
which is similar to the results of 93% inhibition against C. It was found that the dispersion of ZnO NPs in the nutrient
gloeosporioides [24] and 84.5% inhibition against medium produced reactive oxygen species (ROS) such as
Erythricium salmonicolor [8]. hydroxide, superoxide, and hydrogen peroxide [29]. These
The high antifungal activity of ZnO NPs was due to the ROS with high reactivity would then react with components
small size of the synthesized material [25]. The mechanism of of bacterial cell walls including proteins, lipids, and DNA,
antifungal activity was found to be the formation of reactive hence leading to cell death. Hydrogen peroxide penetrates the
oxygen species with high oxidative activity, hence lead to the cell wall into the cell, while superoxide ions attack the cell [32,
alteration of membrane permeability and subsequent cell 33]. Another antibacterial mechanism of ZnO NPs was the
death [26, 27]. direct attachment of the NPs on the cell wall and subsequent
interaction with the membrane components such as proteins
We used FE-SEM and EDS methods to observe the and enzymes, which also lead to the dysfunction of the cells
microbial cells without and with the presence of ZnO NPs in and their death [34].
the nutrient agar in the previous antimicrobial tests.
A B
A B
Figure 6. E.coli cells before (A) and after (B) treating with 184 ppm of ZnO
11
A B
E B
Figure 7. S.aureus cells before (A) and after (B) treating with 184 ppm of ZnO NPs
A B
I B
Figure 8. Salmonella cells before (A) and after (B) treating with 184 ppm of ZnO NPs
Before treating C.gloeosporioides with ZnO NPs, one can we found only their spores (Figure 9B), with the presence of
easily find the fungal cell in the FESEM micrograph (Figure approximately 6.5% zinc atoms on the surface.
9A). However, after treating the fungi with ZnO NPs for 24 h,
A B
N N
Figure 9. A cell of C. gloeosporioides before ZnO NPs treatment (A) and a fungal spore after treating the cells with ZnO NPs (B)
The antifungal mechanism of ZnO NPs is different from synthesis of chitin, and β-1,3-D-glucan synthase takes part in
their antibacterial mechanism and is attributed to the attack of the synthesis of β-1,3-D-glucan [35]. Both chitin and β-1,3-D-
Zn2+ ions on N-acetylglucosamine or β-1,3-D-glucan synthase glucan are important components in the structure of fungal cell
(FKs1p) [8]. N-acetylglucosamine plays a vital role in the walls [8].
12
IV. CONCLUSION [11] N. A. Ibrahim, A. A. Nada, A. G. Hassabo, B. M. Eid, A. M.

Noor El-Deen, and N. Y. Abou-Zeid, "Effect of different
In this study, we synthesized ZnO NPs in the shape of capping agents on physicochemical and antimicrobial
nanorods with sizes mostly lower than 60 nm due to the properties of ZnO nanoparticles", Chemical Papers, vol. 71,
capping role of soluble starch. The ZnO NPs demonstrate a no. 7, pp. 1365-1375, 2017.
dose-dependent antibacterial activity toward different bacteria [12] S. Yedurkar, C. Maurya, and P. Mahanwar, "Biosynthesis of
and fungi. Gram-positivity and Gram-negativity seemed to be zinc oxide nanoparticles using ixora coccinea leaf extract—a
unimportant for the resistance of bacteria toward ZnO NPs. green approach", Open Journal of Synthesis Theory and
FESEM and EDS analyses showed that the cells of bacteria Applications, vol. 5, no. 1, pp. 1-14, 2016.
were distorted and damaged when being in contact with ZnO [13] F. V. Molefe, L. F. Koao, B. F. Dejene, and H. C. Swart, "Phase
NPs in dispersion. The fungal cells in these conditions turned formation of hexagonal wurtzite ZnO through decomposition
into spores and were significantly inhibited. With the low cost of Zn (OH) 2 at various growth temperatures using CBD
of zinc acetate precursor and the ease in the synthesis of ZnO method", Optical Materials, vol. 46, pp. 292-298, 2015.
NPs, the material can find a wide range of applications as an [14] J. Hasnidawani, H. Azlina, H. Norita, N. Bonnia, S. Ratim, and
antibacterial and antifungal component in inks, paints, E. Ali, "Synthesis of ZnO nanostructures using sol-gel
coatings, cosmetics, and pharmaceutics. method", Procedia Chemistry, vol. 19, pp. 211-216, 2016.
ACKNOWLEDGEMENT [15] L. Wang and M. Muhammed, "Synthesis of zinc oxide
nanoparticles with controlled morphology", Journal of
The authors acknowledge Ho Chi Minh City University of Materials Chemistry, vol. 9, no. 11, pp. 2871-2878, 1999.
Technology and Education for the facility and equipment [16] G. Krithika et al., "Zinc oxide nanoparticles—Synthesis,
required for this study. We thank Mr Nguyen Duc Manh characterization and antibacterial activity", Journal of
(student ID 15116030) and Mr Nguyen Quoc Tuan Anh Nanoscience and Nanotechnology, vol. 17, no. 8, pp. 5209-
(student ID 15116002) for their helpful technical support. 5216, 2017.
[17] N. M. Shamhari, B. S. Wee, S. F. Chin, and K. Y. Kok,
REFERENCES
"Synthesis and characterization of zinc oxide nanoparticles
[1] M. G. Demissie, F. K. Sabir, G. D. Edossa, and B. A. Gonfa, with small particle size distribution", Acta Chimica Slovenica,
"Synthesis of zinc oxide nanoparticles using leaf extract of vol. 65, no. 3, pp. 578-585, 2018.
lippia adoensis (koseret) and evaluation of its antibacterial [18] A. Akbar et al., "Synthesis and antimicrobial activity of zinc
activity", Journal of chemistry, vol. 2020, 2020. oxide nanoparticles against foodborne pathogens Salmonella
[2] N. A. Al-Dhabi and M. Valan Arasu, "Environmentally- typhimurium and Staphylococcus aureus", Biocatalysis and
friendly green approach for the production of zinc oxide agricultural biotechnology, vol. 17, pp. 36-42, 2019.
nanoparticles and their anti-fungal, ovicidal, and larvicidal [19] T. Shanmugasundaram and R. Balagurunathan, "Bio-
properties", Nanomaterials, vol. 8, no. 7, p. 500, 2018. medically active zinc oxide nanoparticles synthesized by using
[3] P. P. Mahamuni et al., "Synthesis and characterization of zinc extremophilic actinobacterium, Streptomyces sp.(MA30) and
oxide nanoparticles by using polyol chemistry for their its characterization", Artificial Cells, Nanomedicine, and
antimicrobial and antibiofilm activity", Biochemistry and Biotechnology, vol. 45, no. 8, pp. 1521-1529, 2017.
biophysics reports, vol. 17, pp. 71-80, 2019. [20] B. L. da Silva, B. L. Caetano, B. G. Chiari-Andréo, R. C. L. R.
[4] X. Ma, P. R. Chang, J. Yang, and J. Yu, "Preparation and Pietro, and L. A. Chiavacci, "Increased antibacterial activity of
properties of glycerol plasticized-pea starch/zinc oxide-starch ZnO nanoparticles: Influence of size and surface
bionanocomposites", Carbohydrate polymers, vol. 75, no. 3, modification", Colloids and Surfaces B: Biointerfaces, vol.
pp. 472-478, 2009. 177, pp. 440-447, 2019.
[5] K. Barmak and K. Coffey, Metallic films for electronic, optical [21] M. Arakha, M. Saleem, B. C. Mallick, and S. Jha, "The effects
and magnetic applications: Structure, processing and of interfacial potential on antimicrobial propensity of ZnO
properties. Woodhead Publishing, 2014. nanoparticle", Scientific reports, vol. 5, no. 1, pp. 1-10, 2015.
[6] S. S. Shinde, "Antimicrobial activity of ZnO nanoparticles [22] V. V. Shinde, D. S. Dalavi, S. S. Mali, C. K. Hong, J. H. Kim,
against pathogenic bacteria and fungi", Sci Med Central, vol. and P. S. Patil, "Surfactant free microwave assisted synthesis
3, p. 1033, 2015. of ZnO microspheres: Study of their antibacterial activity",
[7] M. Al‐Hetar, M. Zainal Abidin, M. Sariah, and M. Wong, Applied surface science, vol. 307, pp. 495-502, 2014.
"Antifungal activity of chitosan against Fusarium oxysporum [23] M. T. Cabeen and C. Jacobs-Wagner, "Bacterial cell shape",
f. sp. cubense", Journal of applied polymer science, vol. 120, Nature Reviews Microbiology, vol. 3, no. 8, pp. 601-610, 2005.
no. 4, pp. 2434-2439, 2011. [24] P. Arciniegas-Grijalba et al., "ZnO-based nanofungicides:
[8] P. Arciniegas-Grijalba, M. Patiño-Portela, L. Mosquera- Synthesis, characterization and their effect on the coffee fungi
Sánchez, J. Guerrero-Vargas, and J. Rodríguez-Páez, "ZnO Mycena citricolor and Colletotrichum sp", Materials Science
nanoparticles (ZnO-NPs) and their antifungal activity against and Engineering: C, vol. 98, pp. 808-825, 2019.
coffee fungus Erythricium salmonicolor", Applied [25] P. Rajiv, S. Rajeshwari, and R. Venckatesh, "Bio-Fabrication
Nanoscience, vol. 7, no. 5, pp. 225-241, 2017. of zinc oxide nanoparticles using leaf extract of Parthenium
[9] R. K Dutta, P. K Sharma, and A. C Pandey, "Assessing the hysterophorus L. and its size-dependent antifungal activity
conformational and cellular changes of ZnO nanoparticles against plant fungal pathogens", Spectrochimica Acta Part A:
impregnated Escherichia coli cells through molecular Molecular and Biomolecular Spectroscopy, vol. 112, pp. 384-
fingerprinting", Advanced Materials Letters, vol. 2, no. 4, pp. 387, 2013.
268-275, 2011. [26] R. Brayner, R. Ferrari-Iliou, N. Brivois, S. Djediat, M. F.
[10] N. Vigneshwaran, S. Kumar, A. Kathe, P. Varadarajan, and V. Benedetti, and F. Fiévet, "Toxicological impact studies based
Prasad, "Functional finishing of cotton fabrics using zinc on Escherichia coli bacteria in ultrafine ZnO nanoparticles
oxide–soluble starch nanocomposites", Nanotechnology, vol.
17, no. 20, p. 5087, 2006.
13
colloidal medium", Nano Letters, vol. 6, no. 4, pp. 866-870, [31] A. A. Tayel et al., "Antibacterial action of zinc oxide
2006. nanoparticles against foodborne pathogens", Journal of Food
[27] A. S. H. Hameed, C. Karthikeyan, V. S. Kumar, S. Kumaresan, Safety, vol. 31, no. 2, pp. 211-218, 2011.
and S. Sasikumar, "Effect of Mg2+, Ca2+, Sr2+ and Ba2+ [32] M. J. Akhtar, M. Ahamed, S. Kumar, M. M. Khan, J. Ahmad,
metal ions on the antifungal activity of ZnO nanoparticles and S. A. Alrokayan, "Zinc oxide nanoparticles selectively
tested against Candida albicans," Materials Science and induce apoptosis in human cancer cells through reactive
Engineering: C, vol. 52, pp. 171-177, 2015. oxygen species", International journal of nanomedicine, vol. 7,
[28] Y.-j. Liu, L.-l. He, A. Mustapha, H. Li, Z. Hu, and M.-s. Lin, p. 845, 2012.
"Antibacterial activities of zinc oxide nanoparticles against [33] Y. H. Leung et al., "Toxicity of ZnO and TiO2 to Escherichia
Escherichia coli O157: H7", Journal of Applied Microbiology, coli cells", Scientific reports, vol. 6, no. 1, pp. 1-13, 2016.
vol. 107, no. 4, pp. 1193-1201, 2009. [34] R. Wahab, A. Mishra, S.-I. Yun, Y.-S. Kim, and H.-S. Shin,
[29] K. S. Siddiqi, A. ur Rahman, and A. Husen, "Properties of zinc "Antibacterial activity of ZnO nanoparticles prepared via non-
oxide nanoparticles and their activity against microbes", hydrolytic solution route", Applied Microbiology and
Nanoscale Research Letters, vol. 13, no. 1, pp. 1-13, 2018. Biotechnology, vol. 87, no. 5, pp. 1917-1925, 2010.
[30] A. Sirelkhatim et al., "Review on zinc oxide nanoparticles: [35] J. T. Park and J. L. Strominger, "Mode of action of penicillin:
antibacterial activity and toxicity mechanism", Nano-micro biochemical basis for the mechanism of action of penicillin and
Letters, vol. 7, no. 3, pp. 219-242, 2015. for its selective toxicity", Science, vol. 125, no. 3238, pp. 99-
101, 1957.
14
An Assessment of Complementary Energy of

HPFRCs under Tension
Duy-Liem Nguyen Tien-Tho Do Thi-Ngoc-Han Vuong H. T. Tai Nguyen
Faculty of Civil Faculty of Civil Faculty of Civil Faculty of Civil
Engineering Engineering Engineering Engineering
Ho Chi Minh City Ho Chi Minh City Ho Chi Minh City Ho Chi Minh City
University of Technology University of Technology University of Technology University of Technology
and Education and Education and Education and Education
Ho Chi Minh City, Ho Chi Minh City, Ho Chi Minh City, Ho Chi Minh City,
Vietnam Vietnam Vietnam Vietnam
liemnd@hcmute.edu.vn thodt@hcmute.edu.vn hanvtn@hcmute.edu.vn tainht@hcmute.edu.vn
understanding mechanical properties of HPFRCs, including

Abstract: This paper deals with the complementary their complementary energy, is highly expected to widen the
energy of HPFRCs under tension through an experimental application of HPFRCs.
study. The HPFRCs were prepared from an identity matrix
but with different fiber types or volume contents. The fiber
Tensile stress
used in this research is long twisted with a length of 30 mm
and a diameter of 0.3 mm. There were four levels of fiber Hardening energy
volume content as follows: 0.5 vol.%, 1.0 vol.%, 1.5 vol.% Multiple (Ghd)
and 2.0 vol.% to investigate the effect of fiber volume content microcracks
spc
on the energy parameters of HPFRCs. The complementary
Complementary
energies of the HPFRCs were evaluated and correlated to a energy (CE)
number of microcracks as the dosage of the fiber was scc Crack
changed. Localization
Crack tip
toughness
Keywords: HPFRC, Complementary energy, First- (CT) ecc epc
cracking, Post-cracking, Energy absorption capacity Tensile strain
Elastic energy (Gel)
I. INTRODUCTION Figure 1. Illustration of complementary energy of HPFRCs

High-performance fiber-reinforced concretes (HPFRCs) under tension
have been classified as one of the sustainable construction
materials, which can enhance structure resistance and avoid II. EQUATIONS TO OBTAIN ENERGY
ruinous disasters [1-3]. The term “high-performance” of PARAMETERS OF HPFRCS
HPFRCs refers to strain-hardening behavior accompanied by Based on the definitions of four energy parameters
multiple micro-cracks. This property can be identified with an described in Fig. 1, the equations for obtaining them were built
increase of stress after the first crack due to the crack-bridging to evaluate in the discussion part.
mechanism of reinforcing fibers, and consequently leads to
high strain capacity and large energy absorption capacity of - The elastic energy (named Gel ) is the zone under the
HPFRCs. HPFRCs only produce strain-hardening behavior stress versus strain relationship curve within the elastic stage
with suitable fiber type/content mixed in the plain concrete [1- and calculated using Eq. (1). It characterizes the energy
5]. The interfacial bond between the fiber and HPFRC matrix absorption capacity per unit volume during extending a tensile
is strongly dependent on fiber type/content, which will govern specimen from 0 to e cc .
the mechanical properties of HPFRCs
On the other hand, a strain hardening HPFRC with lower - The hardening energy (named Ghd ) is the zone below the
crack spacing, is characterized as a more ductile material. This stress versus strain relationship curve with the strain range
is because the fracture strain beyond the first crack is mainly from 0 to e pc . The Ghd indicates the energy absorption
due to microcracks, which were generated during strain-
hardening [1-3]. Complementary energy is defined as the area capacity per unit volume during extending a tensile specimen
above the stress versus strain response curve of a material. The from 0 to e pc , it can be derived using Eq. (2).
ratio of complementary energy to the crack tip toughness of
the mortar concrete will represent the capability of producing - The crack tip toughness (named CT ) is the zone above
many microcracks, as illustrated in Fig. 1 [6-10]. It is clear that the stress versus strain relationship curve in the elastic stage
complementary energy is a key engineering property of and given using Eq. (3). It is noted that this expression is
HPFRCs that should be evaluated. Based on the test data of equivalent to the result obtained by Li et al. [13]. The CT
the previous studies of the first author [11,12], this paper reveals the energy per unit volume can not be absorbed during
focuses on the effect of fiber content on the complementary
energy of HPFRCs, which has been not assessed yet. Better extending a tensile specimen from 0 to e cc .

- The complementary energy (named CE ) is the area The HPFRCs used an identical mortar matrix, whose
above the stress versus strain relationship curve with the strain composition is provided in Table 1 according to weight ratio.
range from 0 to e pc [13,14]. The CE represents the energy The partial materials are sand, fly ash, cement and silica fume,
water and superplasticizer. The amount of superplasticizer
per unit volume can not be absorbed during extending a tensile was 0.07 by weight ratio of the cementitious materials. It is
specimen from 0 to e pc , it can be derived using Eq. (4). noted that the sand in the mixture plays an engineering role as
coarse filler material whereas the fly ash and silica fume play
 e e cc  (1) a role as fine filler materials to densify the concrete mixture.
Gel    s (e )d e  The compressive strength of the mortar matrix using a
 
 e 0  cylindrical specimen with a size of 100×200 mm was 89 MPa
 e e pc  at 18-day age. Table 2 provides the properties of long twisted
Ghd    s (e )d e  (2) fiber (T30/0.3) used in this research. Photos of the fiber were
 e 0 
  presented in Fig. 2
 e e cc  All constituent materials of HPFRC were blended using a
CT  s cce cc  Gel  s cce cc    s (e )d e 
(3)
  20-L laboratory mixer (Hobart type). Sand, fly ash, cement
 e 0  and silica fume were dry-mixed for about 10 min. at first.
After dry-mixing, water and then superplasticizer were added
 e e pc  (4)
CE  s pce pc  Ghd  s pce pc    s (e )d e  to the mixture and wet-mixed for about 10 min. Because the
 e 0  plain mortar showed appropriate flow capability and, the steel
 
fibers were gradually added to the mortar mixture by hand and
It is noted that the Ghd and Gel are the true energies further blended for 5 min. After demolding, all specimens
absorbed by a tensile specimen under loading whereas the CE were cured in water at the temperature of 25oC for 14 days.
and CT are the virtual energies only. The detail of mixing work can be referred to in previous
studies [11,12].
III. EXPERIMENTAL TEST B. Experimental setup
A. Materials and preparation of specimens 25
125
cm
162.5
0 1 2 3
L f  30 mm; d f  0.3 mm
Gauge length of 100 mm

50
200
12.5
Wire
162.5
mesh
a) Tensile specimen
Hinge mechanism
Figure 2. Photo of the twisted fiber type in this research

LVDT 1 LVDT 2
Table 1. Composition and compressive strength of Mortar Matrix
Materials Weight ratio
Gauge length
Cement (Type III) 0.8

100
Silica fume 0.07

Silica sand 1
Fly ash 0.2
Superplasticizer 0.04
Hinge mechanism
Water 0.26
Compressive strength (MPa) 89
Table 2. properties of six fiber types

Tensile
Fiber type Length (mm)/ Density b) Experiment setup for direct tensile test
strength
(Notation) Diameter (mm) (g/cm3)
(MPa)
Long twisted Figure 3. Tensile specimen detail and experimental setup
2428 30/0.3 7.9
(T30/0.3)
16
Dog-bone-shaped specimens were used in the direct specimens. Response curves of tensile stress versus strain of
tensile test. The working section over the gauge length of 100 the HPFRCs were built and their complementary energies
mm is rectangular with a dimension of 50 × 25 mm2. To were derived and evaluated.
prevent failure beyond the gauge length, both specimen ends
were reinforced by employing steel wire meshes, as described IV. TEST RESULT AND DISCUSSION
in Fig. 3a. All the prepared specimens were tested through a A. Tensile responses of HPFRCs with different fiber
universal testing machine (UTM) with displacement- contents
controlled loading. The standard displacement rate of 1.0
mm/min. was applied to the movable crosshead of the UTM. Fig. 4 presents the direct tensile responses of the HPFRCs
The data collecting rate during loading was 1 Hz. The using the same fiber type T30/0.3 but fiber content varying
specimen elongation under tension was measured by two from 0.5 to 2.0 vol.%. As can be seen in Fig. 5, all these
linear variable differential transformers (LVDTs), as shown in HPFRCs also exhibit strain-hardening behaviors, although the
Fig. 3b. Besides, the hinge mechanism was designed at both fiber contents were significantly different.
CE
CE
Ghd
Ghd
b) 1.0 vol.% T30/0.3

a) 0.5 vol.% T30/0.3
CE
CE
Ghd
Ghd
c) 1.5 vol.% T30/0.3 d) 2.0 vol.% T30/0.3
Figure 4. Tensile responses of HPFRCs with different fiber contents
Table 3. Tensile parameters with different fiber contents

Table 3 shows the tensile parameters of the HPFRCs using
T30/0.3 type with different fiber contents, varying from 0.5
vol.% to 2.0 vol%. The tensile parameters in Table 4 were
Fiber
content
e cc s cc e pc s pc averaged from three tested specimens for each series, the
(vol.%) (%) (MPa) (%) (MPa) values in the brackets were standard deviations. As presented
in Table 4, all the tensile parameters increased with the
0.013 2.54 0.28 4.86 increase in fiber content. The content of 2.0 vol.% generated
0.5
(0.003) (0.46) (0.077) (0.69)
0.016 2.56 0.52 7.48
the best performance in terms of s cc , e cc , s pc and e pc .
1.0
(0.003) (0.47) (0.092) (0.78)
0.020 3.41 0.49 9.99 B. Energy parameters of HPFRCs
1.5
(0.002) (0.03) (0.035) (1.22) Table 4 gives the results of four energy parameters of
2.0
0.025 3.41 0.55 12.53 HPFRCs, including CE , CT , Ghd and Gel of HPFRCs.
(0.005) (0.27) (0.013) (0.36)
These results were averaged from three specimens tested for
Note: The standard deviations are delivered within parentheses each series. Fig. 5 graphically displays the comparisons of the
17
energy parameters of HPFRCs using different fiber contents. Table 4. Energy parameters of HPFRCs
As can be seen in Fig. 5, all the four energy parameters of
HPFRCs generally increased as fiber volume content Fiber
CE Ghd CT Gel Ghd Gel
increased. As the fiber volume content was changed from 0.5 content
(vol.%) (kJ/m2) (kJ/m2) (kJ/m2) (kJ/m2) CE CT
to 2.0 vol.%, the range values of the energy parameters were
observed as follows: from 0.381 to 1.609 kJ/m2 for the CE ,
0.5 0.381 0.981 0.014 0.019 2.6 1.3
from 0.014 to 0.027 kJ/m2 for the CT , from 0.981 to 5.419
kJ/m2 for the Ghd , and from 0.019 to 0.051 kJ/m2 for the Gel . 1.0 0.579 3.324 0.014 0.027 5.7 1.9
Regardless of fiber content, the CT was observed to be lower
than the Gel at the first crack. At the post crack, the CE was 1.5 0.862 4.086 0.018 0.049 4.7 2.8
significantly lower than the Ghd . This means the material 2.0 1.609 5.419 0.027 0.051 3.4 1.8
capability of HPFRCs was utilized significantly. The ratio Note: Definition of energy parameters can be referred to Fig.1
Gel / CT was 1.3-2.8 while the ratio Ghd / CE was 2.6-5.7, as
provided in Table 4. C. Correlation between complementary energy and number
of microcracks in hardening stage of HPFRCs
As illustrated in Fig. 1, multiple microcracks ( N cr ) are
generated during the hardening stage. According to Naaman
[14], the average crack spacing ( Lav ) of a strain hardening
composite can be predicted using Eq. (5). Eq. (5) relates to a
number of fibers within cross-section ( N f ), which is possibly
estimated using Eq. (6). Finally, N cr within the gauge length
of a tensile specimen can be obtained using Eq. (7).
Mathematically, Eq. (7) describes a linear relationship
a) Complementary energy ( CE ) between N cr and fiber volume content ( V f ).
Ams m
Lav   (5)
( N f  d f ) eq
Vf
N f  2 Ag (6)
af
L 4 LAg eq
N cr   2 V
Lav  Ams m d f f (7)
b) Crack tip toughness ( CT )

where, d f , a f   d 2f / 4 are the fiber diameter and area of

fiber section, respectively; eq is the equivalent bond
strength obtained from a single pullout test; Am and s m are
area and tensile strength of matrix, respectively;  2 is factor
reflecting the orientation of fibers, the  2 value is 1 for case
of 1D, 2  for case of 2D, and 0.5 for the case of 1, 2 and 3D
fiber orientation; Ag is the cross-section area of tensile
specimens; L is the gauge length of tensile specimen;  is
c) Hardening energy ( Ghd ) the crack spacing factor, its value ranging from 1 to 2. It is
highlighted that the author investigated the number of cracks
using experiment only and did not calculate the number of
cracks through theoretical equations proposed by Naaman.
Since the ratio of the complementary energy to the crack
tip toughness ( CE / CT ) of a composite represents capability
of producing multiple microcracks ( N cr ), in this section, the
ratio CE / CT was correlated to N cr of HPFRCs, which were.
Generally, as the ratio CE / CT is higher than 1, the multiple
microcracks possibly occur in the composite. Kanda and Li
[15] stated that the conditions for surely producing multiple
d) Elastic energy ( Gel ) microcracks are ratio CE / CT >2.7 and ratio s pc / s cc >1.3.
Figure 5. Comparison of energy parameters of HPFRCs
18
Table 5 supplies the ratio CE / CT and N cr . As shown in A linear relationship was supposed in this study for ratio
Table V, the ratio CE / CT increased with the increase in fiber CE / CT versus N cr and V f versus N cr of HPFRCs. Using
content. The value of CE / CT was from 26.3 to 58.8 while the test results and linear regression analysis, the CE / CT
the value of s pc / s cc was from 1.9 to 3.7. These ratios versus N cr response curve was performed in Fig. 7 with the
completely satisfy the conditions according to Kanda and Li regression function Eq. (8). Similarly, the fiber volume
[14]. The value of N cr was observed from 2.00 to 16.67. content versus N cr response curve was shown in Fig. 8 with
the regression function Eq. (9). The regression functions are
Table 5. Capability in producing multiple microcracks of HPFRCs useful for the prediction work of microcrack number and ratio
CE / CT with a changed fiber volume content.
Fiber CE
content
CE CT N cr Ncr  0.4456(CE / CT )  10.409 (8)
(vol.%) (kJ/m2) (kJ/m2) CT (ea)
N cr  9.47 V f  2.67 (9)
0.5 0.381 0.014 26.3 2.00
V. CONCLUSION
1.0 0.579 0.014 42.0 7.33
This investigation focused on the complementary energy and
number of microcracks generated during the strain hardening
1.5 0.862 0.018 48.7 10.67
stage. Based on the test and analysis results, several
2.0 1.609 0.027 58.8 16.67 conclusions are summarized from this investigation as follows:
- The complementary energy, crack tip toughness, and the
ratio between them of HPFRCs clearly increased with
increasing fiber volume content.
- The hardening energy of HPFRCs was much higher than
the complementary energy, regardless of fiber content.
This means the material capability in terms of energy
absorption was utilized significantly. Also, the elastic
energy was higher than the crack tip toughness of
a) 0.5 vol.% b) 1.0 vol.% HPFRCs.
- Relationship of microcrack number versus ratio
complementary energy to crack tip toughness as well as
the relationship of microcrack number versus fiber
volume content were analytically modelled based on the
test result and regression analysis. The regression
c) 1.5 vol.% d) 2.0 vol.% functions are useful for the prediction work of
Figure 6. Photos of multiple microcracks of HPFRCs within microcrack number and complementary energy as the
the gauge length fiber volume content in HPFRCs is changed.
ACKNOWLEDGMENT
This research was supported by Ho Chi Minh City
University of Technology and Education, the authors are
grateful to the sponsor. The opinions expressed in this paper
are those of the authors and do not necessarily reflect the
views of the sponsor.
REFERENCES
[1] D. L. Nguyen, J. Song, C. Manathamsombat, D. J. Kim,

Figure 7. A linear relationship between ratio CE / CT and “Comparative electromechanical damage-sensing behavior
number of multiple microcracks of HPFRCs of six strain-hardening steel-fiber-reinforced cementitious
composites under direct tension”, Compos. Part B, 69
(2015), pp. 159–168.
[2] D. L. Nguyen, G. S. Ryu, K. T. Koh, D. J. Kim, “Size and
geometry dependent tensile behavior of ultra-high-
performance fiber-reinforced concrete”, Composites: Part
B, 58 (2014), pp. 279-292.
[3] S. H. Park, D. J. Kim, G. S. Ryu, K. T. Koh, “Tensile
behavior of Ultra-high Performance Hybrid Fiber
Reinforced Concrete,” Construction and Building Materials,
2012, 34(2), pp. 172-184.
[4] K. Wille, D. J. Kim, A. E. Naaman, “Strain hardening UHP-
Figure 8. A linear relationship between fiber content and FRC with low fiber contents,” Mater Struct, 44 (2011), pp.
number of multiple microcracks of HPFRCs 583-598.
19
[5] N. T. Tran, D. L. Nguyen, D. J. Kim, T. T. Ngo, “Sensitivity [11] J. Song, D. L. Nguyen, C. Manathamsombat, D. J. Kim,
of various fiber features on shear capacities of ultra-high- “Effect of fiber volume content on electromechanical
performance fiber-reinforced concrete”, Magazine of behavior of strain-hardening steel-fiber-reinforced
Concrete Research, Vol. 74, No. 4, 2022, pp. 190-206. cementitious composites”, J. Compos. Mater. 2015,
[6] V. C. Li, and H. C. Wu, “Conditions for Pseudo Strain doi:10.1177 /0 021998314568169.
Hardening in Fiber Reinforced Brittle Matrix Composites”, [12] D. L. Nguyen, M. N. -T. Lam, D. J. Kim and J. Song, “Direct
Journal of Applied Mechanics Review, Vol. 45, No. 8, 1992, tensile self-sensing and fracture energy of steel-fiber-
pp. 390-398. reinforced concretes”, Composites: Part B 183 (2020)
[7] V. C. Li, and C. K. Y. Leung, “Theory of Steady State and 107714, pp. 1-19, doi:10.1016/j.compositesb.2019.107714.
Multiple Cracking of Random Discontinuous Fiber- [13] V.C. Li, S. Wang, C. Wu, Tensile strain-hardening behavior
Reinforced Brittle Matrix Composites”, ASCE Journal of of polyvinyl alcohol engineered cementitious composite
Engineering Mechanics, Vol. 118, No. 11, 1992, pp. 2246- (PVA-ECC), ACI Mater. J., 98(6) (2001), pp. 483-492
2264. [14] Yoo, D.-Y., Kim, S.-W., & Park, J.-J. (2017). Comparative
[8] V. C. Li, “On Engineered Cementitious Composites (ECC): flexural behavior of ultra-high-performance concrete
A Review of the Material and its Implications”, Journal of reinforced with hybrid straight steel fibers. Construction and
Concrete Technology, Japan Concrete Institute, Vol. 1, No. Building Materials, 132, pp. 219-229.
3, 2003, pp. 215-230. doi:10.1016/j.conbuildmat.2016.11.104
[9] D. B. Marshall and B. N. Cox, “A J-Integral Method for [15] A. E. Naaman, “Ferrocement & laminated cementitious
Calculating Steady-State Matrix cracking Stresses in composites,” Techno Press 3000, Ann Arbor, Michigan,
Composites”, Mechanics of Materials, North-Holland, No. 2000.
7, 1988, pp. 127-133. [16] T. Kanda, V. C. Li, “Practical design criteria for saturated
[10] A. E. Naaman, “Tensile strain-hardening FRC composites: pseudo strain hardening behavior in ECC,” J. Adv. Concr.
Historical evolution since the 1960”, Advances in Technol. 2006, 4, p. 59.
Construction Materials 2007, Springer Berlin Heidelberg,
pp. 181-202, Stuttgart, Germany.
20
A Study of the Scavenging Process in a Two-stroke

Free Piston Linear Engine at Low Velocity Using CFD
and DPM
Nguyen Huynh Thi Nguyen Van Trang Huynh Thanh Cong
Faculty of Vehicle and Energy Faculty of Vehicle and Energy Vietnam National University - Ho Chi
Engineering Engineering Minh City
Ho Chi Minh City University of Ho Chi Minh City University of Ho Chi Minh City, Vietnam
Technology and Education Technology and Education htcong@hcmut.edu.vn
nguyenhuynhthi@tgu.edu.vn trangnv@hcmute.edu.vn
Dao Huu Huy Huynh Van Loc Truong Hoa Hiep

Faculty of Vehicle and Energy Faculty of Industrial Engineering Faculty of Industrial Engineering
Engineering Tien Giang University Tien Giang University
Ho Chi Minh City University of My Tho City, Vietnam My Tho City, Vietnam
Technology and Education
dhhuy2310@gmail.com
Ngo Duc Huy Vo Bao Toan

Faculty of Vehicle and Energy Engineering Faculty of Vehicle and Energy Engineering
Education Education
Abstract: A free-piston linear engine (FPLE) does not should be suitable for many different fuels [2], reducing
use a crankshaft to control piston motion. The gas exchange friction compared to conventional heat engines [3]. Visibly,
and piston movement are closely related to each other and FPLE is a research trend to improve the performance of
directly affecting to the engine performance. In this study, a hybrid vehicles. The FPLE considered in the present
numerical simulation using Computational Fluid Dynamics research is the two-stroke dual-piston type. The piston
(CFD) and the Discrete Phase Model (DPM) has been assembly has a free linear motion between the top dead
performed to investigate the scavenging process in a two- center (TDC) and bottoms dead center (BDC) and the piston
stroke free piston linear engine during warm up. The assembly motion is controlled by gas and load forces acting
upon it. Cancelling the crank mechanism allows the piston
characteristics of the reciprocating motion can be
to move freely between the two cylinders, giving the engine
determined based on the dynamics model and the piston
the advantage of flexible characteristics, but also poses a
motion profiles are imported into commercial CFD software challenge to the stability of the engine. The destabilization
(Ansys Fluent V.21). Both models have been applied the here is controlling the position of TDC before the
initial conditions such as the velocity of the piston, piston combustion process; different positions of TDC produce
displacement, and inlet pressure to assess the effect of the different end-compression pressure that affects the power of
trapping efficiency. The results show that the trapping the engine. There are so many related researches on the
efficiency ±5% difference in the two models was able to stability control of FPLE, and proposed various control
achieve at piston of 0.32 m/s – 1.25 m/s. In addition, the methods [4-7]. These studies focus only on controlling the
change in gas at each displacement position of the piston motion of the piston. Although considerable progress has
and different inlet pressures is also shown through the been made, the way FPLE is controlled remains one of the
simulation results. problems preventing its commercialization. Another factor
is the mass of air and fuel trapped inside the combustion
Keywords: free piston linear engine, scavenge chamber. The trapped mixtures that are affected by the gas
efficiency, piston motion, Computational Fluid Dynamics exchange process in there have scavenging process. That
(CFD), Discrete Phase Model (DPM) can be key to realizing the engine's combustion efficiency
and emissions potential [8]. But the stability of gas
I. INTRODUCTION
exchange is unpredictable and difficult for measurement
The free-piston engine generator (FPLE) is a linear techniques.
internal combustion engine that is removed the crank and
piston moves freely in the cylinder. With the advantages of CFD (Computational Fluid Dynamics) is a very useful
compactness, high efficiency [1], variable compression ratio tool to analyze the exchange gas through the motion process
of FPLE as well as the scavenging process. Blarigan et al.,

at Sandia National Laboratories, studied the effects of BDC to TDC form is determined by:
design parameters such as gas pressure in box, inlet 𝐹.𝑡
position, and tilt angle on scavenging efficiency and V = V0 − (5)
𝑚
trapping efficiency of FPLE through CFD [9]. Their results
indicate that during the early stages of gas exchange, some
gas travels from the scavenging port to the exhaust port,
causing direct trapping losses. Blarigan et al., at Sandia
National Laboratories They introduced a non-dimensionally
modeled piston movement configuration into the KIVA-3V,
simulating it to obtain the basic FPLE scavenging
performance with constant intake pressure using the model
CFD [10]. The results show that a gas exchange system with
continuously low-rise pressure maximizes efficiency and
reduces unclean emissions. In addition, several other studies
show that FPLE has a faster piston speed than a
conventional engine, which results in shorter gas exchange
duration of FPLE, lower scavenge efficiency but higher trap
efficiency [11-12].
The studies above show that the gas exchange process
has been simulated and optimized the effects of design
parameters to FPLE. However, these studies omitted the gas
Figure 1. A prototype of two-stroke free piston engine [15]
exchange in the start process of FPLE, this process usually
operates at low speed but must ensure the trapping 1- Cylinder 1; 2- Activation mechanism; 3- Compressor; 4-
efficiency. This study presents a CFD simulation to Carburetor, 5- Compressed air chamber; 6-Cylinder 2
investigate scavenging structure and parameters on the
FPLE gas exchange process. The dimensional parameters B. CFD model
are chosen based on a prototype FPLE, the motion of the The piston model and main parameters of FPLE are
free piston is built based on a dynamic model to calculate shown in Fig. 2 and Table 1. The parameter of the FPLE is
the piston’s motion profiles. In this study, the key imported into commercial CFD software Ansys Fluent v.21
parameters such as displacement piston, operating low- to define the calculating surfaces and volumes as well as to
frequency and inlet pressure are inlet parameters. Two generate the mesh. Then there is the meshing process, the
methods CFD and the Eulerian-Lagrangian discrete phase cylinder mesh is generated in a multi-zone method, and the
model (DPM) are used to track the ratio of particles trapped others have generated the mesh in the sweep method. The
and escape from the cylinder then find out trapping velocity of the piston is calculated by a program written in
efficiency and compare the accuracy with each other. Microsoft Visual C++ before importing it into the Ansys
software as an input database, the input database is a user-
II. MODELING AND SIMULATION
defined function (UDF). In the dynamic mesh, the layering
A. Dynamic model is used to add or remove layers of cells according to the
piston position. In addition, the dynamic mesh changes
The mechanical forces acting on the piston include the according to the piston position to change the compression
gas force in the left and right cylinder, mechanical friction ratio, pressure... The position of the piston can be changed
and inertia force, and traction of the actuator. Piston step by step with a step of 0.5mm, from which it is possible
dynamics can be determined by Newton II's law. to control the pressure at the end of the compression stroke
d2 x in different locations.
Fst + Fcyl – Fcyr – Ff = m (1)
dt2
In order to investigate the effects of intake pressure on
Fcyl, Fcyr are the gas force from the left and right cylinder the scavenging process and the pressure at the end
compression, the inlet pressures are adjusted from 1.05bar to
Fcyl = Pcyl  S (2) 1.2bar; initial speeds are 0.32m/s and 1.25m/s. The inlet and
Fcyr = Pcyr  S (3) outlet temperatures are set at 300 K during the simulation
process. The inlet mixture is assumed to be an ideal gas. In
Pcyl, Pcyr is the pressure in the left and right cylinders; S addition, the 𝑘 − 𝜀 turbulent model is employed to capture
is the area of the piston crown. Ff is the friction force, m is turbulence. Input parameters are provided for the
the mass of the translator, d2x/dt2 is the acceleration of the computational fluid dynamic (CFD) model to calculate the
piston, Fsl, Fsr are respectively the spring forces in the left trapping efficiency. In two-stroke, the trapping efficiency is
and right. Fst the force is received from the starting device. defined as follows [13]:
The velocity of piston is calculated by the equation: mass of delivered fresh charge retained in cylinder
t=
𝑑x 𝑑x d2 x mass of delivered fresh charge through all scavenging ports
= ( ) + 𝑡 (4)
𝑑𝑡 𝑑𝑡 0 dt2
In the actual model, the initial speed V0 is based on

the mechanical starter Fig. 1 This speed is hindered by the
spring forces on the left or right. The piston velocity of the
22
particles. The trapping efficiency is defined: as the number

of particles entering the cylinder divided by the number of
particles supplied from the scavenging ports.
Figure 2. a) Parameter of the FPLE, b) Computational mesh of the

FPLE
Table 1. Engine specifications and operating conditions
Parameters Values
Stroke (S) 28 mm
Bore (D) 34 mm
intake port di 0 - 6 mm Figure 3. DPM calculation model
III. RESULTS AND DISCUSSION

Exhaust port de 0 - 13mm
A. CFD simulation results
Transfer port Dt 10 mm
Pressure changes at the inlet port and open/closed
position affect the scavenging process. At an initial speed of
Inlet pressure Pin (bar) 1.05 - 1.25bar
0.32m/s, the initial pressure gradually increases from
1.05bar to 1.2bar. Trapping efficiency is increased but not
Inlet temperature Tin (K) 300K significantly since the mass flow in/out of the cylinder is
almost stable Fig 4. When the piston is close to fully closing
C. DPM model the inlet, the mass flow drops suddenly at the 3.5 mm
The discrete Phase Model (DPM) is used to co- position and at the 4mm position the scavenge port fully
track particles’ motion. DPM is based on a Euler- close. At 1.05bar pressure due to high mass flow at the
Lagrangian framework, the forces such as viscous drag, lift scavenge port but at the exhaust mass flow no significant
force...acting on the particles along their trajectories and difference with the case other so trapping efficiency still
stochastic behaviour of the turbulent flow are taken into tends to increase; at 1.1bar pressure due to mass flow at the
account. Based on tracking the particles throw discrete scavenge port low but high mass flow at exhaust port
phases in boundary areas to escape, trap and reflect. Escape making the exhaust port high so trapping efficiency still
condition is assumed in the inlet and outlet ports, the trap tends to reduce. The flow velocity at the exhaust port is
condition is considered the cylinder and reflects condition is based on the flow velocity at inlet port, the flow velocity at
being a wall of intake and exhaust pipes Fig. 3. To shorten the higher inlet the higher exhaust. The results of the flow
the computational time and the approach become much velocity simulation De = 0mm; De =2mm; De =3.5mm at
simpler, assumptions in particle motion calculations will be 1.05bar Fig. 6 and the rest of the cases are shown in Fig. 7.
simplified: all particles have the same size, and the flow 60
around the particles and particle-particle interaction is
Trapped effiency
neglected in the simulation. In the Lagrangian reference 55

frame, the trajectory is predicted by integrating the force
balance on each individual particle and can be written as 50
(%)
[14]:
45 0.32 m/s
𝑑𝑢𝑝 𝑢−𝑢𝑝 𝑔(𝑝 −)
= + +𝐹 (6) 1.25 m/s
𝑑𝑡 𝑟 𝑝 40
1.05 bar 1.1 bar 1.15 bar 1.2 bar
where F is an additional acceleration (force/unit particle Inlet Pressure (bar)
mass) term; is the particle density and p is the continuum
phase density; u is the local fluid phase velocity and up is the Figure 4. Trapping efficiency from 1.05bar to 1.25bar
velocity of the particle phase, respectively; g is the gravity
of the particle. The first item on the right is the drag force
per unit particle mass, r is the droplet or particle relaxation
time. The initial parameters of the particle such as velocity,
mass flow inlet… similar to the CFD model; the trap
efficiency in the DPM is replaced by the number of
23
Figure 5. Mass flow versus piston displacement at piston speed

0.32m/s Figure 8. Mass flow versus piston displacement at piston speed
1.25m/s
Figure 6. Velocity Streamline at A) De = 0mm; B) De = 2mm; C)

De = 3.5mm with initial pressure = 1.05 bar and initial velocity of
piston = 0.32m/s Figure 9. Velocity Streamline at A) De = 0mm; B) De = 2mm; C)
De = 3.5mm with initial pressure = 1.05bar and initial velocity of
piston = 1.25m/s
Figure 7. Flow velocity versus piston displacement at piston speed

of 0.32m/s
Figure 10. Flow velocity versus piston displacement at piston
When increasing the speed to 1.25m/s the trapping speed of 1.25m/s
efficiency increases at 1.05bar pressure, more mass flows
into the cylinder and out to the cylinder roughly equal to the B. DPM simulation results
other pressure curves. Through the two results, the speed of With the same input conditions as the CFD model, the
0.32m/s still gives trapping efficiency higher than 1.25m/s, trapping performance has a difference in Fig. 11, due to
the reason is a lot of new charge going out of the exhaust particle conditions such as the same size and collision
port but new charge into cylinder low, as shown in the mass between particles in this model. It is possible to observe the
flow Fig. 8 and the velocity simulation De = 0mm; De movement of particles and the number of particles trapped
=2mm; De =3.5mm at 1.05bar Fig. 9, the rest of the cases in the cylinder, the number of trapped particles corresponds
are shown in Fig. 10. to the amount that is lost through the exhaust port and
trapped due to the obstruction of the piston wall at the
scavenging ports Fig. 12.
24
60
Trapped effiency (%)
55
50
45 0.32m/s
1.25m/s
40
1.05 bar 1.1 bar 1.15 bar 1.2 bar
Inlet Pressure (bar) Figure 14. Amount of particle versus piston displacement at piston
speed of 1.25m/s
Figure 11. Trapping efficiency from 1.05 bar to 1.25 bar (DPM)
IV. CONCLUSION
Based on the combination of the dynamic model, CFD
model, and DPM model, the scavenging process in a two-
stroke free-piston linear engine has been modelled and
simulated. Although there is a ±5% difference in the
trapping efficiency between the two models, there are still
similarities in terms of rules. The influence of key
parameters such as operating frequency, piston displacement
changing intake/exhaust port, inlet pressure on scavenging
and trapping efficiency were investigated. Simulation results
show that the trapping efficiency is 0.32m/s higher than 1.25
m/s but not significant. Although the trapping efficiency at
1.25m/s speed is less than 0.32 m/s, there is still an
advantage in terms of air loss between the piston and
cylinder, which affects the starting process. In addition,
simulation results also show that increasing exhaust port
Figure 12. A) Particle residence time at De = 0mm, B) Particle spacing, as well as reducing inlet pressure, can free the
residence time at De = 4mm, C) Particle residence time at De = engine two-stroke piston to increase trapping efficiency. In
6mm the next works, velocity piston, and inlet pressure are
changed more values to find the optimal parameters for the
With a low piston speed of 0.32m/s, the amount of engine. In addition, the inlet temperature, the angle of
particles trapped in the cylinder is high and stable. When the inclination of the intake and exhaust manifolds and the
scavenging ports are completely closed, the number of geometry will be considered.
trapped particles is reduced, causing large air escape
through the exhaust port Fig. 13. With a piston speed of ACKNOWLEDGMENT
1.25m/s, the fast-closing time of the scavenging ports causes
This work belongs to the project grant No: B2019-SPK-
the number of particles not to catch up to the cylinder, but
08 funded by Ministry of Education and Training, and
the number of particles retained is higher due to the fast-
hosted by Ho Chi Minh City University of Technology and
closing time of the exhaust port Fig. 14. However, at
Education, Vietnam.
0.32m/s the trapping efficiency is still higher.
REFERENCES
[1] J. Hansson and M. Leksell., “Performance of a Series Hybrid
Electric Vehicle with a Free-Piston Energy Converter,” IEEE
Xplore, 2007.
[2] R. Mikalsen and A. P. Roskilly, “A review of free-piston
engine history and applications,” Applied Thermal
Engineering, vol. 27, pp. 2339- 2352, Oct 2007.
[3] B. Jia, R. Mikalsen, A. Smallbone, and A. Paul Roskilly, “A
study and comparison of frictional losses in free-piston
engine and crankshaft engines,” Applied Thermal
Engineering, 2018.
[4] R. Mikalsen, A.P. Roskilly, The control of a free-piston
Figure 13. Amount of particle versus piston displacement at piston engine generator. Part 1: Fundamental analyses, Appl.
speed of 0.32m/s Energy, 87 (2010), 1273–1280.
[5] R. Mikalsen, A.P. Roskilly, The control of a free-piston
engine generator. Part 2: Fig. 15. In-cylinder gas pressure
and temperature of each combustion duration. C. Yuan, et al.
Applied Thermal Engineering 173 (2020) 115201 10 Engine
dynamics and piston motion control, Appl. Energy, 87
25
(2010), 1281–1287. [11] C. Yuan, H. Ren, X. Jing, Comparison of the gas exchange
[6] B.R. Jia, Z.X. Zuo, H.H. Feng, G.H. Tian, A. Smallbone, of a loop scavenged free piston engine alternator and the
A.P. Roskilly, Effect of closed-loop controlled resonance conventional engine, Appl. Therm. Eng., 127 (2017), 638–
based mechanism to start free piston engine generator: 649.
simulation and test results, Appl. Energy, 164 (2016), 532– [12] Nguyen Ba Hung, Sung Jaewon, Ocktaeck Lim. (2017). A
539. study of a scavenging process in a two-stroke free piston
[7] B.R. Jia, G.H. Tian, H.H. Feng, Z.X. Zuo, A.P. Roskilly, An linear engine using CFD. 9th International Conference on
experimental investigation into the starting process of free- Applied Energy (pp. 1354-1360). Cardiff, UK: Elsevier Ltd.
piston engine generator, Appl. Energy 157 (2015), 798–804. [13] Grljušić, M.; Tolj, I.; Radica, G.; Sciubba, E. An
[8] Goldsborough SS, Blarigan P V. Optimizing the scavenging Investigation of the Composition of the Flow in and out of a
system for a two-stroke cycle, free piston engine for high Two-Stroke Diesel Engine and Air Consumption Ratio.
efficiency and low emission: A computational approach. Energies 2017, 10, 1.
SAE Paper 2003-01-0001, 2003. [14] A. Vakhrushev, M. Wu, A. Ludwig, G. Nitzl, Y. Tang, & G.
[9] A. Sofianopoulos, Y.C. Zhou, B. Lawler, S. Mamalis, Gas Hackl, “Verification of a Discrete Phase Model with Water-
exchange processes of a small HCCI free piston engine - A Particle Flow Experiments in a Tundish,” in 5th. Int. Conf.
computational study, Appl. Therm. Eng., 127 (2017), 1582– on Simulation; Modeling of Metall, Processes in Steelmaking
1597. (STEELSIM) (2013).
[10] S. Goldsborough, P. Blarigan, Optimizing the scavenging [15] [N. H. Thi et al., "An Investigation on Power Generation
system for a two-stroke cycle, free-piston engine for high Characteristics of Linear Generator Driven by a Free-piston
efficiency and low emissions: a computational approach, Engine," 2021 International Conference on System Science
SAE Paper, No.2003-01-0001, 2008. and Engineering (ICSSE), 2021, pp. 495-499, doi:
10.1109/ICSSE52999.2021.9538476.
26
Investigating the Relationship between Workers’

Needs and Commitments to Garment Enterprise
Tu Tran
Faculty of Fashion and Tourism
Ho Chi Minh City University of Technology and Education
camtuspkt@hcmute.edu.vn
Abstract: This paper analyzes the relationship between and level of engagement with their company. With the data
workers’ needs and commitments to garment enterprises. obtained from the survey and through processing and
Data are collected from the direct survey through the analyzing statistical data, it is hoped that the research results
questionnaire will determine the level of satisfaction of will provide managers with a deeper insight into the factors
garment workers’ needs based on the Hierarchy of Needs that connect employees and businesses. Thereby drawing out
(Maslow, 1943), and the commitment between the worker several measures and recommendations to help the manager
and the garment enterprise. Then, through statistical analysis have better human resource management policies to connect
to review and test the influence of each factor on worker workers with their company, and improve labor fluctuations
in the garment enterprises.
commitment to the enterprise. The research results show that
factors affecting workers’ commitment to garment II. METHODOLOGY
enterprises include: social needs; physiological needs related
to business policies and the lives of workers; safety and A. Research design
respect needs. The research method is quantitative research. Data is
collected through a survey of workers at a garment enterprise
Keywords: needs of garment workers, commitments with in Binh Duong province.
garment enterprises, social needs, physiological needs, safety
B. The scale
and respect needs
1. The scale of satisfaction of needs.
I. INTRODUCTION
Maslow's hierarchy of needs (Maslow, 1943) [2] is a
From the beginning of 2021 until now, textile and garment motivational theory in psychology comprising a five-tier
enterprises have prospered in terms of orders and export model of human needs, often depicted as hierarchical levels
markets, so the need to recruit more workers is quite large. within a pyramid.
However, many businesses are currently facing difficulties in From the bottom of the hierarchy upwards, the needs are:
recruiting and facing a shortage of human resources... [1].
physiological (food and clothing), safety (job security), love
In the report on the recruitment demand for mid-and and belonging needs (friendship), esteem, and self-
senior-level personnel in the Vietnamese market in the first actualization (presented in figure 1).
quarter of this year, Navigos Search Company stated that the
effectiveness of Free Trade Agreements (FTAs), especially
EVFTA (FTAs) Vietnam - EU), although it has not been able
to offset the decrease in export turnover caused by Covid-19,
it has brought significant effects in increasing export orders
for textile and garment enterprises in Vietnam. Besides, the
tense situation in Myanmar has caused more orders to flow to
Vietnam, leading to an increase in demand for positions in the
textile and garment industry. These advantages have helped
recruitment demand in the textile industry increase by 50-60%
over the same period in 2020, especially for middle and senior
management positions. However, according to the Vietnam
Textile and Apparel Association (Vitas), after the 2020 labor
cuts, the textile and garment industries are now struggling to
re-hire because of workers who have returned to their
hometowns or online selling [1].
In the labor structure of the garment industry, workers
account for the highest proportion and the job-hopping rate is
also high. This study is conducted to help understand the
factors affecting the commitment between workers and Figure 1. Maslow's hierarchy of needs (1943)
garment enterprises by surveying workers about their needs

employee engagement with garment enterprises includes 5

The scale of satisfaction of needs includes 5 observed observed variables (presented in table 2). Observed variables
variables (presented in table 1). Observed variables were were measured using a 5-point Likert scale (1: strongly agree,
measured using a 5-point Likert scale degree (1: completely 2: agree, 3: neutral, 4: disagree, 5: strongly disagree).
agree, 2: agree, 3: neutral, 4: disagree, 5: completely C. Data processing methods.
disagree).
The data processing is done with PASW 18.0 software
Table 1. Description of Factors
through the following stages:
Stage 1: Check the reliability of the scale.
Observable Variable Factor
The salary is suitable for the workers’ Table 2. the scale of employee engagement with garment
capacity and contribution enterprises
Overtime money
Observable Variable Factor
Reward work achievements
Try to complete the assigned tasks
Salary, bonus Contribute ideas to develop the team
Allowances (retirement, maternity leave,…) The worker Intend to work long-term for the company
The company’s concern for disadvantaged commitment to
individuals. garment Feeling proud to introduce the company as a
Physiological good environment and the best place where
needs enterprises
Food rations everyone should work
Feeling that the team’s problem is also my
Uniforms problem
Housing and travel allowance policy Testing the reliability of the scale of influencing factors
Seniority allowance and the scale of coherence through Cronbach's Alpha with the
following criteria:
Lunch break time
Work overtime - Total correlation coefficient (Corrected Item – total
Correlation) >= 0.5
The amount of work undertaken
The job is suitable for skills and abilities - Cronbach's Alpha coefficient >= 0.6. Variables that do
Work pressure
not guarantee reliability (Cronbach's Alpha < 0.6) will be
excluded from the scale because the survey is completely new
How does work affect health?
to the respondents (Hoang Trong-Chu Nguyen Mong Ngoc,
Safety needs Company medical service 2008) [4].
Leave and sick leave
- Stage 2: Exploratory factor analysis (EFA), to redefine a
Stability of work set of observations in the research model. Observable
Sanitary condition variables with Factor Loading less than 0.5 will be removed
Guidance on occupational safety (Othman & Owen, 2002) [5].
Communicating with superiors - Stage 3: Correlation coefficient and linear regression
Support and encouragement from superiors
analysis.
Get attention from superiors
Love and Workers are treated fairly Firstly, consider the correlation coefficient between the
belonging needs Colleague support employee engagement with the garment enterprise and the
Friendliness and sociability level of satisfaction of needs. Next, perform multivariable
Cultural and sports activities linear regression analysis by the method of ordinary least
Trade union activities squares (Ordinal Least Squares - OLS), in which the
Be respected and acknowledged for your dependent variable is the association of workers with garment
comments enterprises, and the independent variable expectations will be
Feedback and suggestions from superiors
Esteem needs
The reward for work results
physiological needs, safety needs, social needs, esteem needs,
Competition to improve skills and self-actualization needs.
The reputation of company leader Selecting Enter variable methods is carried out. The
Trained in working skills adjusted coefficient of determination R is used to determine
Self-
actualization Conditions to improve skills the fit of the model. The F-test is used to confirm the
needs Having conditions to show the personal extensibility of this model to apply to the population. And the
ability t-test is used to reject the hypothesis about the regression of
the population is 0.
2. Scale of the worker commitment with garment
III. RESULTS
enterprises
D. Modified models
Commitment to an organization involves three attitudes:
(1) a sense of identification with the organization’s goals, (2) Through testing the scale of independent and dependent
a feeling of involvement in organizational duties, and (3) a variables, the research model is calibrated as follows. The
feeling of loyalty to the organization [3]. The scale of dependent variable represents the level of employee
28
engagement with the enterprise, including two components: X3: Physiological needs related to worker's life
the workers' actions towards the enterprise, and the workers' X4: Safety needs
attitudes towards the enterprise. Independent variables X5: The need to be respected
include social needs (social research), physiological needs e: estimation error.
related to business policy, physiological needs related to
E. Explain the importance of the variables in the model
worker's life, safety needs, need to be respected. The
modified research model is presented in detail in Figure 2. The linear regression equation of the above two models
shows that the degree of employee engagement with the
1. The linear regression equation of the worker's actions enterprise depends on 05 factors: social needs, physiological
variable for the garment enterprise: needs related to corporate policy, physiological needs related
Y = -0.004 + 0.098X1+ 0.417X2 + 0.02X3 + 0.158 X4 + to worker's life, safety needs, esteem needs.
0.056X5 + e (Model 1) Because all the independent variables are measured by the
Which includes: Likert scale, we can also see the importance of each factor for
Y: The degree of employee engagement with the the degree of commitment to the garment enterprises. In
enterprise through actions which, physiological needs related to enterprise policies have
X1: Social needs the strongest influence on workers' actions towards garment
X2: Physiological needs related to corporate policy enterprises. If the level of satisfaction of physiological needs
X3: Physiological needs related to worker's life related to the policy of the enterprise increases by one level,
X4: Safety needs the degree of employee engagement with the enterprise
X5: The need to be respected through actions will increase an average of 0.417 steps.
e: estimation error. Similarly, if the level of satisfaction of the need for safety
increases by one level, the level of employee engagement with
2. Linear regression equation of the variable the workers' the enterprise through action increases an average of 0.158
attitudes towards garment enterprises: steps.
Y = -0.002 + 0.327X1 + 0.131X2+ 0.186X3 + 0.063X4+
0.222X5 + e (Model 2)
Which includes:
Y: The degree of employee engagement with the
enterprise through attitudes
X1: Social needs
X2: Physiological needs related to corporate policy
Figure 2. Modified research model
29
Table 3. Results of Multivariable Regression Analysis
Model Unstandardized Standardized

Coefficients Coefficients
B Std. Error Beta t Sig. F R2

Worker's actions for the garment enterprise
1 -.004 .067 -.066 .948 8.280 0.187
(Constant)
Social needs
.098 .087 .098 1.129 .261
Physiological needs related to corporate policy
.417 .082 .418 5.107 .000
Physiological needs related to worker's life
.002 .085 .002 .028 .978
Safety needs
.158 .078 .157 2.020 .045
The need to be respected
.056 .078 .056 .711 .478
Workers' attitudes towards garment enterprises
-.002 .062 -.029 .977 15.813 0.305
(Constant)
Social needs
.327 .080 .326 4.071 .000
Physiological needs related to corporate policy
.131 .076 .131 1.730 .085
Physiological needs related to worker's life
.186 .078 .186 2.371 .019
Safety needs
.063 .072 .062 .866 .388
The need to be respected
.222 .072 .222 3.070 .002
Regarding the level of employee engagement with Through the research results, we also have more
enterprises through attitudes, social needs have the strongest comments about Maslow's theory of human needs. According
influence. If the level of satisfaction of social needs increases to Maslow (1943), “Individuals must satisfy lower level
by one level, it will increase the level of employee deficit needs before progressing on to meet higher level
engagement with the enterprise through the average attitude growth needs.” However, research shows that workers pay
of 0.327 steps. Similarly, if the level of satisfaction of the attention to both lower-order needs and higher-order needs,
need for respect and physiological needs related to life such as physiological needs, safety needs, social needs, and
increases by one level, it will make the level of worker esteem needs, with different levels of interest. That proves at
attachment to the garment enterprise an average increase of the same time that people have many different needs and may
not be satisfied with all those needs, it is not necessarily that
0.222 and 0.186 respectively. Table 3 presents the results of
they are only interested in one need and when satisfying a
the multivariate regression analysis. certain need is new think of the higher needs as identified by
IV. CONCLUSION Maslow.
Based on the statistical results, the factors affecting V. LIMITATIONS AND FURTHER RESEARCH
employee engagement with garment enterprises include social Firstly, the sampling method is convenient, so the error
needs, physiological needs related to the enterprise's policies, due to sampling is relatively high. Second, the survey
physiological needs related to garment enterprises to business respondents are workers, they do not have much time and do
life, safety needs, and esteem needs. We also identify the not care much about the issues raised by this topic. Therefore,
worker's commitment to the garment enterprise, which the representativeness of the sample is not high.
consists of two components: the workers' actions towards the
enterprise and the workers' attitudes towards the enterprise. To The next research should expand the survey to more
test the hypothesis about the relationship between these above garment companies in different regions of Vietnam with the
factors and the level of employee engagement with the random sampling method. At the same time, it compares the
enterprise, a multivariate regression analysis was performed. level of employee engagement with the enterprise through
In which, the dependent variable is the worker's actions factors such as gender, age, marital status, and working time.
towards the enterprise and the worker's attitude towards the Thus, the research results on the factors affecting the
enterprise; and 05 independent variables include social needs, employee's commitment to the garment enterprise will be
physiological needs related to business policies, physiological more objective and the applicability will be higher. Besides,
needs related to business life, safety needs, and esteem needs. the next research should build a model to calculate the costs
The results show that the physiological needs related to the and losses of the company when workers are not engaged with
policies of enterprises have the strongest influence on workers' their enterprises.
actions and the social needs have the strongest influence on
the attitudes of workers. The other factors have less influence.
30
REFERENCES [3] James L. Gibson, John M. Ivancevich, James H. Donnelly, Jr.,

Robert Konopaske “Organizations: Behavior, Structure,
[1] Kinh te Viet Nam Online, “Nhan luc nganh det may thieu hut Processes”, Fifteen Edition, Published by McGraw-Hill, 2011,
tram trong, doanh nghiep can co giai phap can co” (There is a page 182.
serious shortage of human resources in the textile and garment [4] Hoang Trong, Chu Ngu yen Mong Ngoc (2008), Phan tich du
industry, businesses need a radical solution), Accessed online
lieu nghien cuu vơi SPSS Volume 1, 2 (Analysis of research
at https://vneconomy.vn/nhan-luc-nganh-det-may-thieu-hut-
tram-trong-doanh-nghiep-can-co-giai-phap-can-co.htm, data with SPSS volume 1, volume 2), Hong Duc Publishing
Accessed date 09/05/2021. Company.
[2] Maslow, A. H. (1943). A theory of human motivation. [5] Othman, A. & Owen, L. (2002). The multidimensionality of
Psychological Review, 50(4), 370-96. CARTER models to measure customer service quality in
Islamic banking industry: a study in Kuwait Finance House,
International Journal of Islamic Financial Services, l3(4),1-12.
31
Optimization of Aggregates in Concrete Brick with

Recycled Materials from Stone
Nguyen Thang Xiem Ho Minh Chau Tran Doan Hung Truong Thanh Chung
Faculty of Civil Asia 96 New Material Faculty of Civil Faculty of Civil
Engineering Joint Stock Company Engineering Engineering
Nha Trang University Nha Trang, Vietnam Nha Trang University Nha Trang University
Nha Trang, Vietnam Nha Trang, Vietnam Nha Trang, Vietnam
xiemnt@ntu.edu.vn hungtd@ntu.edu.vn chungtt@ntu.edu.vn
Abstract: Currently, conventional burnt clay brick makes behavior of concrete in the form of a brick.
up 95% of the total use of construction brick in Vietnam.
This paper presents an experimental study with the aim to
There is an urgent need to use alternative materials for produce the concrete brick with recycled materials from
sustainable development. In this paper, we experimentally stone mining with the compressive strength that satisfies the
optimize the proportion and sizes of aggregates to produce building code TCVN 6016-2011 of Vietnam. Experiments
concrete brick with recycled materials from stone mining. were conducted to optimize the sizes of the aggregates and
Experimental results showed that the optimum proportion of their proportions.
the aggregates of the concrete brick are: 50% chippings
stones, 30% young stones, 8% cement, 12% fly ash and This paper is organized as follows. In the next section,
silica fumes; and the optimum sizes of the aggregates are the properties of each individual aggregate in the concrete
brick are investigated. Next, the procedure to make the brick
3.0 mm chippings stones and 2.5 mm young stones. The
is detailed. Next, we show the experiment to optimize the
compressive strength of the concrete brick with the optimum
sizes and the proportion of the aggregates in the concrete
mixture is 9.55 MPa, well above the requirement of the brick. The final section is the conclusions.
building code TCVN 6016-2011 of Vietnam. The use of
local, cheap and abundantly available recycled materials II. MATERIALS
significantly lowers the cost of the concrete brick. The
results of this study were applied to the mass production of A. Young stones
concrete brick by Asia 96 New Material Joint Stock Young stones (Fig. 1a) are formed from the stones that
Company in Vietnam. induced mechanical and chemical weathering. Agents such
as ice, water, cracks or heat cause an expanding effect,
Keywords: concrete brick, compressive strength, widening the cracks in the stone and breaking the stones into
chipping stones, young stones, upland soil, fly ash, silica pieces. The effect of mechanical weathering increases the
fumes exposed surface area of stone, which speed up the chemical
weathering process. The original stones break into smaller
I. INTRODUCTION pieces with lower compressive strength than the original
stones.
According to a report from the Vietnamese government,
burnt clay brick is currently the most used in Vietnam, about B. Upland soil
95% of the total use of construction brick [1]. This is a big Upland soil (Fig. 1b) is a type of terrain that is formed by
issue since the production of burnt clay brick produces a weathering, denuding from mountains. The hills have small
huge amount of CO2. Much effort was made by the slopes and high organic residues due to the presence of
government to promote the use of other types of green plants. The height of hills is usually less than 200 meters.
alternatives. This research is stemmed from a national This type of soil has low economic value, therefore using it
research funding to find the optimum proportion and sizes of as raw material for brick production will lower the cost of the
aggregates to produce the concrete brick with recycled product. Once the upland soil reaches the standard of
materials from stone mining process. cleanliness, it is shoveled into the trough through the
Many types of recycled aggregates were studied as conveyor system. Upland soil materials used for the
replacements for traditional aggregates in concrete over the production of concrete brick must be carefully selected to be
past few decades. Recycled aggregate concrete (RAC) has free of impurities.
been shown to have a great impact on the environment and C. Chippings stones
economy [2-3]. However, the use of RAC in the construction
industry remains limited [4]. A number of studies have Chippings stones (Fig. 1c) used for the production of
worked on the topics to establish the compressive strength concrete bricks are waste products from stone mining. Based
and other mechanical properties with different replacements on the results of the experiments to determine impurity
ratios of various recycled aggregates [5-12]. Sim et al. [13] content in chippings stones, we found that the content of
showed that the particle size of coarse aggregate is an impurities is quite large, about 6.9%. It is needed to lower
important parameter to the behavior of aggregate concretes. the number of impurities in the chippings stones to less than
However, there are very few studies investigating the effect 3% in order to reach the strength requirement of the concrete
of particle sizes and proportion of coarse aggregates on the

brick. The chemical composition of the chippings stones is

shown in Table 1.
Table 1. Chemical composition of the chippings stones
Al2O3 SiO2 Fe2O3 SO3 CaO LOI K2O Na2O MgO
% 18.26 60.12 4.02 1.10 3.48 1.30 4.86 2.44 2.05
Figure 1. Image of (a) young stones, (b) upland soil and (c)
chippings stones after grinding Figure 2. SEM image of fly ash
cps/eV
3
D. Adhesives 2
6 1
PCB40 Portland cement is used as adhesive in this study.

5
Portland cement is an inorganic binder in water, produced by Zr
Ca As
calcining minced limestone (75% to 80%) and clay (20% to 4 K O Cu Mg
S Ti Fe Na
Zr
Al Si P S K Ca Ti Fe Cu As
25%) by heating block (about 1450oC), fast cooling to form 3
clinker; then grinding fine clinker with plaster (3% to 5%)

2
and some other essential additives such as inorganic
additives. Table 2 describes the mineral composition of 1
Portland cement. 0
0 1 2 3 4 5 6 7 8
keV
Table 2. Mineral composition of Portland cement Figure 3. Spectral power distribution enlarged 2000 times
Name of the mineral Chemical formula Abbreviation %
Silicate tricalcite 3CaO.SiO2 C3S 45 - 60
Silicate bicalcite 2CaO.SiO2 C2S 20 - 30
Aluminate tricalcite 3CaO.Al2O3 C3A 4 - 12
Fero aluminate tetra
4CaO.Al2O3.Fe2O3 C4AF 10 - 12
calcite
Aluminate calci 5CaO.Al2O3 - <1
Fero aluminate calci 8CaO.3Al2O3.Fe2O3 - <1
Ferit calci 2CaO.Fe2O3 - <1 Figure 4. Graph particle size distribution of fly ash
E. Fly ash
Fly ash is spherical and sharp, as shown in the scanning
electron microscope (SEM) in Fig. 2. The mean particle size
is about 3.74 μm, as described by the particle size
distribution chart in Fig. 3 and Fig. 4. Adding fly ash to the
mixture of the concrete brick helps to increase the strength
while reducing the cost of the product. Thanks to its small
size, the fly ash fills in the micro-holes of the cement
particles, making the concrete brick denser, waterproof and
higher in strength. The quantitative analysis of the
composition of the fly ash is as follows: O (53.80), Na
(1.84), Mg (1.06), Al (14.27), Si (20.25), P (0.09), S (0.19),
K (0.25), Ca (0.94), Ti (1.95), Fe (4.72), Cu (0.16), As Figure 5. Image of silica fumes grains photographed using
(0.15), Zr (0.34). The elements Si, Al and Ca have relatively microscope
high percentages, which makes the cement easy to hydrate, Table 3. Chemical composition of silica fumes
increasing the rate of cemented grains.
No Oxide %
F. Silica fumes 1 SiO2 96.25
Silica fumes are a type of binder that improves the 2 Al2O3 0.23
3 Fe2O3 0.07
mechanical properties of the mixture. They are mostly 4 CaO 0.31
spherical, contained 95 wt.% of SiO2, with an average 5 MgO 0.04
diameter of about 1.0 mm, about one-hundredth the size of 6 K2O 0.56
cement particles (Fig. 5). The chemical composition of silica 7 Na2O 0.15
fumes is shown in Table 3. 8 LOI 1.92
33
III. MANUFACTURING PROCESS

The process of manufacturing the samples with size 40
mm × 40 mm × 160 mm is described in Fig. 6.
Figure 6. The manufacturing process of sample concrete brick
IV. EXPERIMENTAL STUDIES Experimental studies were carried out on four groups of
samples from M1 to M4 with different proportions of
Fig. 7a shows the mean compressive strength of the aggregates. For each group, a total of nine samples was made
chippings stones with particle sizes from 1.0 mm to 5.0 mm and then averaged to find the compressive strength of the
using the fracture cracking test. For each particle’s size, a concrete brick. The result is presented in Table 4. It is shown
total of nine samples were made and averaged. At the size of that the composition of 50% chippings stones, 30% young
3.0 mm, the mean compressive strength of the chippings stones, 8% cement, and 12% additives (including fly ash and
stones is the highest (Fig. 7a). Similarly, it is found that at silica fumes) gives the highest compressive strength (9.55
the size of 2.5 mm, the mean compressive strength of the MPa). It is well above the strength requirement of
young stones is the highest (Fig. 7b). Those optimum sizes of construction brick according to the building code TCVN
chippings stones and young stones are then used for 6016-2011 in Vietnam.
following experiments to find the optimum proportion of
aggregates of the concrete brick.
Figure 7. Mean values of compressive strength of (a) chippings stones and (b) young stones
34
Table 4. The mixing ratio and the compressive strength of the components
Chippings Young stone Upl-and soil Ce-me-nt Compressive strength [MPa]
Addi-tives [%]
No [%] [%] [%] [%] 7 days 14 days 28 days
70 0 10 1 19 3.03 3.98 5.02
70 0 10 3 17 3.32 3.94 4.98
M1
70 0 10 5 15 2.78 3.48 4.39
70 0 10 8 12 2.63 3.33 4.21
0 90 0 1 19 4.19 5.98 7.55
0 90 0 3 17 4.83 6.01 7.59
M2
0 90 0 5 15 3.88 6.43 8.12
0 90 0 8 12 4.82 5.72 7.22
0 50 30 1 19 2.93 4.33 5.26
0 50 30 3 17 3.89 4.81 6.27
M3
0 50 30 5 15 5.15 6.54 7.65
0 50 30 8 12 4.76 5.75 6.82
50 30 0 1 19 5.71 7.16 8.39
50 30 0 3 17 6.32 7.57 9.02
M4
50 30 0 5 15 7.34 8.70 8.83
50 30 0 8 12 6.91 8.33 9.55
[5] Limbachiya, M. C., Leelawat, T., and Dhir, R. K. (2000).

V. CONCLUSION “Use of recycled concrete aggregate in high-strength
In this paper, we optimize the proportion and sizes of concrete.” Mater. Struct., 33(9), 574–580
aggregates to produce concrete brick with recycled [6] Poon, C. S., Kou, S. C., and Lam, L. (2007). “Influence of
materials from stone mining. Experimental results showed recycled aggregate on slump and bleeding of fresh
concrete.” Mater. Struct., 40(9), 981–988
that the optimum size of the chippings stones and young
[7] de Brito, J., Bizinotto, M. B., and Ferreira, L. (2011).
stones are 3.0 mm and 2.5 mm, respectively. The optimum “Influence of the pre-saturation of recycled coarse concrete
proportion of aggregates of the brick are 50% chippings aggregates on concrete properties.” Mag. Concr. Res.,
stones, 30% young stones, 8% cement, 12% fly ash and 63(8), 617–627
silica fumes, which result in the strength of the concrete [8] Duan, Z. H., Kou, S. C., and Poon, C. S. (2013).
brick being 9.55 MPa, satisfying the building code TCVN “Prediction of compressive strength of recycled aggregate
6016-2011 in Vietnam. The use of local, cheap and concrete using artificial neural networks.” Constr. Build.
Mater., 40, 1200–1206
abundantly available recycled materials lowers
[9] Butler, L., West, J. S., and Tighe, S. L. (2013). “Effect of
significantly the cost of concrete brick. The results of this recycled concrete coarse aggregate from multiple sources
study were applied to the mass production of concrete on the hardened properties of concrete with equivalent
brick by Asia 96 New Material Joint Stock Company in compressive strength.” Constr. Build. Mater., 47, 1292–
Vietnam. This study has the application not only in 1301
Vietnam but also in other places with similar recycled [10] Beltrán, M. G., Barbudo, A., Agrela, F., Galvín, A. P., and
resources. Jiménez, J. R. (2014). “Effect of cement addition on the
properties of recycled concretes to reach control concretes
strengths.” J. Clean. Prod., 79, 124–133
ACKNOWLEDGMENT
[11] Dilbas, H., Çakır, Ö., and Şimşek, M. (2016). “Recycled
The authors would like to thank colleagues at Asia 96 aggregate concretes (RACs) for structural use: An
New Material Joint Stock Company for their support of evaluation on elasticity modulus and energy capacities.”
this study. Int. J. Civ. Eng., 15(2), 247–261
[12] Gholampour, A., Gandomi, A. H., and Ozbakkaloglu, T.
REFERENCES (2017). “New formulations for mechanical properties of
recycled aggregate concrete using gene expression
[1] Chau, H. M (2015). “Research on the situation of using programming.” Constr. Build. Mater., 130, 122–145
adobe brick materials in Khanh Hoa province”
[13] Sim, J. I., Yang, K. H., and Jeon, J. K. (2013). “Influence
[2] Meyer, C. (2009). “The greening of the concrete industry.” of aggregate size on the compressive size effect according
Cem. Concr. Compos., 31(8), 601–605 to different concrete types.” Constr. Build. Mater., 44, 716–
[3] Bravo, M., de Brito, J., Pontes, J., and Evangelista, L. 725.
(2015). “Mechanical performance of concrete made with
aggregates from construction and demolition waste
recycling plants.” J. Clean. Prod., 99
[4] Topcu, I. B. (1997). “Physical and mechanical properties of
concretes produced with waste concrete.” Cem. Concr.
Res., 27(12), 1817–1823 59–74
35
A Comparison on the Flexural Capacity and

Dynamic Performance of a Reinforced Concrete
Beam and a Steel Beam in a High-rise Building
Tham Hong Duong
Ho Chi Minh City Open University
Ho Chi Minh City, Viet Nam
tham.dh@ou.edu.vn
Abstract: Standards and Regulations relating to the usage From a technical viewpoint, especially in the recent trend
of materials for high-rise buildings are now urgent demands of green architecture and sustainable development, engineers
in every urban region and city. The major tasks in such and construction contractors should have a more tentative
constructions are the choice of material used to meet the mind in material, guaranteeing the multi-purpose criteria in
green standards and comply with the regulations. This technical aspects, feasibly constructing with fast bringing into
requires careful analysis, not only in the present time but still service. High-rise buildings have used mainly high-
in the future. From structural viewpoints, a material that has performance concrete to support a very high level of
the higher strength, stiffness, and ductility is the best choice; compression stress. Unfortunately, this causes an increase in
the self-weight of the structure, and under the action of the
however, it is necessary to study more in-depth to understand
wind and lateral loads such as earthquakes, gusts, etc., the
the pros and cons of such material, especially in structures of
massive body of the structure may impose more load for the
high-rise buildings. By analyzing two beams having different foundation. Then if the structure is a skyscraper, the deep
materials but the same in self-weight per meter of length, one foundation would be applied. People often want to reduce the
is made of reinforced concrete having 1% of steel weight of the buildings by using steel beams as an alternative
reinforcement, and the other, a steel beam, some results to heavily supported concrete beams. Steel beams usually
indicate that the steel beam proves to have the higher dynamic prove to be a good replacement for concrete.
stiffness, in terms of natural frequency, and flexural capacity
in terms of the ultimate moment. Nevertheless, the difference Since steel has a modulus of elasticity 7 to 8 times higher
is clear in the aspects of fire protection, ductility, and loads than concrete, the moment of inertia of a steel beam could be
transmitted to the foundation. Aging is also another aspect to reduced to at least one-seventh of that of a concrete beam
having the same stiffness. A small depth of steel beam could
be examined. The motivation for this conference paper is to
then be used, and the total height of the building could be
establish a quantitative comparison between the two kinds of
reduced. With the same stiffness, bending capacity could be
beam elements as an individual structure, and a component of the same, but the natural frequency is varied due to its mass,
a frame as well. Some suggestions are recommended to resulting in a variation in dynamic response. As such, one
engineers and practitioners in real construction works and controversal point to be considered is what the pros and cons
compared to existing valid Standards and Codes of Practice. of each material are, concerning their structural performance
in terms of many different factors involved such as natural
Keywords: dynamic response, flexural capacity, frequency, dynamic stiffness, ductility vs the self-weight, etc.
reinforced concrete beam, steel beam, stiffness In the other words, how to quantify the effectiveness of each
kind of beam, so that a comparison between a concrete beam
I. INTRODUCTION and a steel beam could be conducted. This paper aims to
clarify the applicability of beams, concerning the working
Selecting the materials for construction work is an art of
condition in both static and dynamic loadings.
using materials, balancing between many criteria of
technology, economy, and serviceability. There are so many
criteria, that depend on the designer, the owner, and/or II. METHODOLOGY
different sides involved. People in the engineering sector A. Simply supported reinforced concrete beam
focus on strength and the rigidity of the construction, which is
A beam is a bending element. In the static equilibrium, the
contributed by a connection of many structural elements,
flexural capacity is determined by the maximum value of the
restraints, and boundary conditions. To choose the best
bending moment that the structure could support. Then, for
relevant material, the material and the configuration of the
guaranteeing the serviceability, the structure would deflect
element have to be calculated by applying mechanical
due to that loading, then its deflection would not exceed an
computation, regarding compliance with the design code. An
allowable limit. As such, for the static loading case, a beam or
architect would want to have an artistic external shape, and
beam is called to experience two limit states of equilibrium,
most convenient distribution of a function of the space;
i.e., the former is of ultimate limit state (ULS), and the latter,
sometimes, a green architect is an urgent demand for
the serviceability limit state (SLS). By assigning a
sustainable construction.
predetermined percentage of reinforcement, the flexural
capacity of a reinforced concrete (RC) beam is easily

determined. With a given specification of concrete and B. Steel beam

armature, i.e., compression stress fcu and yield stress fs, by A steel beam has the ultimate bearing capacity considering
taking the equilibrium of all the horizontal forces applied in a the plastic working condition at which the allowable stress is
cross-section (Fig. 1) , as below: allowed to increase by 12 to 15%. In this paper, a hot rolled
steel is assigned in comparison, with a cross-section in I shape.
Unlike the RC beam, working in a fully plastic working
condition, the stress distribution on a cross-section comes
from 2 parts: compression in the upper part of the beam and
tension in the lower part of the beam (Fig. 2).
Figure 1. Determining the ultimate bending moment in RC beam
F x  0  0.8(0.85 fcu )c.b  A 's f 's  As f s  

in which, c is the height of compressive concrete block, f’c ,
equals 0.85 fcu is the cylindrical compressive strength of Figure 2. Stress blocks in a steel beam, working in plastic
concrete, A’s As is the total area of steel in compression zone condition [4]
and tension zone, respectively; b is the beam width. According
to ACI 318-14 (ACI 2014) [1], the ultimate bending moment Full plastic bending capacity, the neutral axis is
 F .y
with both armatures is determined by the following equation:
determined first, y  i i
, then a=2 y , and the fully
 a F
M u  ( As  As, )df y (1  )  As, df y (1  )  a i
1,7 f c d plastic section modulus will be Z=a𝚺Fi , and the fully plastic
in which fc is the compression stress in concrete, ρ is the moment could be Mu=fy.Z
tension reinforcement ratio, a is the thickness of reinforcement There is an essential difference between the ultimate
cover. By examining (2), According to ACI 318-14 (ACI-14), flexural capacity of the RC beam and the steel one. As the RC
the ultimate state attains when the stress armature yields. In beam largely depends on the percentage of the armature to be
case of omitting the compression rebar, the flexural failure reinforced inside the concrete, there are two separate parts of
would occur at the bending moment equaling: stress block, i.e., the compression block for concrete and the
armature for reinforcement; and crack moment could be
As f y viewed as the first indicator of the failure of the structure due
 M u  As f y (d  ) b to the location of the neutral axis (NA). People cannot use the
1.7 f c'b section modulus for an RC beam, even if a conversion for
In case more tentative concern that the cracking condition reinforcing steel into an equivalent concrete area is applied.
is selected to be a criterion of failure, the bending moment of Nevertheless, in the serviceability limit state, people still use
a RC beam at the first crack appearance is the equivalent stiffness for calculating the deflection for an RC
beam. On the other hand, the steel beam working in the
ft Iuncrack
 M cr   
plastics state has only a single location of the NA, and no
yt threshold of the first crack appearance is found. This affects
the design for different stages of load between the two kinds
in which ft=7,5√f’c (in psi unit), Iuncrack is the moment of inertia
of structural material, and the most difficult step in particular
of the uncracked section, and yt equals the effective height d
steel construction is not in the side of a structure, but the side
of the beam, minus the depth of compression concrete block
of a connection, and the ductility instead.
c. In most cases, equations (2a) and (2b) are considered to be
the ultimate limit of a concrete beam. Besides the aspect of dynamic analysis, during conceptual
design, this should be brought into analysis before selecting
Crushing concrete also is another kind of failure. In this
the usage for either material, RC beam, and steel I beam.
situation, the RC beam is at flexural failure due to concrete
crushes.
III. STRATEGIC PLAN FOR THE COMPARISON
a
 M ult  f y As (d  )   The performance of a structure during the process of load-
2 supporting is chosen as the first criterion to compare. Its
In which a=Asfult/(0.85f’cb) with fult=1,2fy is the height of contents are the strength, stiffness, and robustness of the
compression stress zone, replacing an actual non-linear stress structure, subject to both static and dynamic loads and effects.
block of compressing concrete. Because there are different characteristics in inertia as
Excluding the shear failure, the ultimate value would be compared to the mass and the restraint, the response of the
the smaller one computed by (2) for yielding in the armature, structure in an individual material would be different. As such
and (4) for crushing concrete. Particularly according to ACI the strategy should aim at not only the static strength and
440.2R (2017) [2] for beam retrofitting Fiber Reinforced stiffness of the structure, but at dynamic loadings and effects,
Polymer (FRP), the strain at crushing concrete is 0.003 in the together with the response of the structure, both in time-
compression fiber of the beam, following after the yield of domain and frequency-domain analysis. In the case of a high-
steel [2][3]. rise building, and/or soil-structure interaction is taken into
37
account, such a structure as a beam, although being a simple Table 1. Different factors to be checked
element subjected to bending, the response of an RC beam and Factors and Levels
a steel beam is quite a matter. Run
A= Span B=Gravity load C=Stiffness
1 1 1 1
For ensuring the universality of the study, many factors 2 1 2 2
which have a sufficiently wide range of values, and an 3 1 3 3
appropriate method of data collection should be used. 4 2 1 2
Information of a flexural element, even made of concrete or 5 2 2 3
steel, should be sufficiently general and dimensionless. 6 2 3 1
7 3 1 3
Information about a flexural element, even made of concrete
8 3 2 1
or steel, should be sufficiently general and dimensionless. In 9 3 3 2
the light of the abovementioned concept, some key factors are
considered: By assigning a set of properties (i.e., one row in Table 1)
Input data: into a model, the bending moment, deflection, and natural
frequency would be obtained as the outcomes. These
- Geometry: construction height of the beam, cross- outcomes would play as predictors or dependent variables,
section as a fraction of the beam/beam to the span; and the factors would be independent variables, all will
- Gravity Load: the percentage of contribution of the involve in a multi-variable linear regression. This equation
self-weight as compared to total imposed load, for would be modified suitably according to the material input.
bending moment, loading to the foundation. Then the comparison is generally implemented.
- Stiffness: the deflection due to self-weight to that of
total deflection. IV. MODEL FOR COLLECTING DATA
Outcomes D. A pilot tested beam
- The ultimate flexural capacity and deflection. An RC beam is studied. Figure 3 shows a 6-meter span RC
beam having the mass source from loading (including the self-
- Natural responsive frequency for the 1st mode, in weight) equal to 22.19 kN/m. Its natural frequency in the
terms of the total mass of the beam. Dynamic analysis theoretical analysis [6] is
includes modal analysis and static non-linear
pushover analysis, conducted for a given limited
deflection (i.e. displacement controlled pushover). 0.3  0.63
3e10  ( )
 EJ 
2 2
- Other construction costs if any. The cost is assumed  1   12  
to be based on the self-weight of the structure. l 2 m 5.92 2219
C. Factors to be studied  76.53 s 1
For residential buildings, three levels of span length are
considered, i.e. 4m for a short span, 6m for a medium span, or f1 = 12.2 Hz. This value agrees with that of SAP2000
and 10 meters for a long span. For the RC beam, the cross- modal analysis, 12.57 Hz (Fig. 3c)
section is selected by applying an empirical formula; for
instance, the depth equals span divided by 10, and the width
is taken around one-half of the depth. For instance, a 300x600
mm cross-section RC beam would apply to 6m-span.
Percentage of rebar is 1% for all cross-section to be studied.
Unfortunately, for steel beam, it has not an empirical formula
like RC beam. By computing the bending moment, a section
modulus could be determined, and cross-section could be a)
selected in many readily built tables of cold-formed I beam. In
this paper, A36 construction steel as per AISC 1963 (hot rolled
section for construction steel), and Limit State analysis are
applied to this comparison. They are Ultimate Limit State
(ULS) and Serviceability Limit State (SLS). Some additional b)
recommendations are that, although the short span might be
chosen to be less than 30 times the wide flange, wf [5], the
bending moments in RC beams would be used in selecting the
cross-section for the A36 construction steel I beam. As such,
all these abovementioned proportions for both RC beam and
steel beam are kept to be fixed for comparison purposes. The
second factor to be studied is the imposed total load; for
practical construction works, three ranges of loadings are in
order of less than 11 kN/m, 15 to 20 kN/m for medium c)
loading, and 30 to 50 kN/m for heavily imposed loading. This
range of imposed loads would be tentatively used for both RC Figure 3. 6 m-span RC 300x600 beam with load q=17.69 kN/m,
beam and steel beam in this study. bending moment M=96.55 kNm, strain in most tensile zone, and
the natural frequency f1=26.54 Hz (with mass source is the self-
weight)
38
By re-defining the mass source being the self-weight, the

natural frequency for the structure itself is f1=26. 54 Hz. The
mass source would use the specified load patterns instead of
element self-mass and additional mass.
At the moment M = 96.55 kNm, as the strain attains
0.00018 then the stress in concrete exceeds the tensile strength
of concrete (by ft=7,5√f’c in MPa units, nearly equals 1 MPa
on a conservative side) and cracks appear as the calculation
that was conducted previously yields Mcr=107 kNm [3] (c.f
Fig. 4) as below:
Figure 6. 10 meter- steel beam a) dimension of WF24; b) the 1st

mode shape in a scenario of the heaviest load as source mass,
f1=5,39 Hz
E. Collecting data for the comparison

Figure 4. The sagged part in the middle of the beam is considered Planned tasks for collecting data are designed as in Table
to be a 6 m- RC 300x600 beam, with cracking positive moment 1. Level 1 is the smallest value of each factor, and level 3 is
Mcr=107 kNm the largest. Results of the outcomes or response, both in the
static and dynamic modal analysis are tabulated in Table 2.
In a model with rebar as a link element, the results indicate
that rebar changes slightly the natural frequency of the beam, Outcome data are collected by running 9 tasks standing for
about 7%, i.e., from 12.57 Hz to 13.47 Hz (Fig. 5); the strain 9 possible cases or scenarios. In each task, the response would
is also reduced slightly, from 0.00018 down to 0.000163 this be compared in a dimensionless form.
indicates that the rebar restricts the strain at the tensile edge of
the beam. As such, for a more convenient process of data Table 2. Results on responses w.r.t various conditions
collecting, rebar is temporarily not taken into account in Response (*)
modal analysis. Bending/
Task cracking Deflection
f1 (Hz) Weight/total load
moment (m)
(kNm)
a) 22 0.0014 17.08 2/11 (18.2 %)
111
21.85 0.0019 12.98 0.95/10 (10.5 %)
b) 39.37 0.0024 12.19 2/22.2 (9 %)
122
Figure 5. 6m-RC beam with rebar modeling, a) stress distribution 41.2 0.0034 9.56 0.95/18.64 (5,1 %)
in rebar (half of the beam); b) strain in 6 m-RC beam, at the 62.14 0.0036 9.53 2/28.1 (7.1 %)
smallest value, M =126,5 kNm, f1=13,45 Hz 133
66.45 0.0055 7.53 0.95/30.02 (3,2 %)
The cracking moment in a simply supported RC beam 67.37 0.001 20.02 4.5/15.48 (29%)
agrees with that of a selected beam in a 2D frame of a high- 212
rise building, subjected to uniformly distributed loading. This 59.4 0.0025 11.25 1.48/12.46 (11.8 %)
permits a prediction for the ultimate flexural capacity, and the 96.57 0.0015 20.02 4.5/22.19 (20 %)
response of the structure, making use of the approach in the 223
previous studies [3]. As for simply supported steel beams, A36 91.5 0.0038 11.25 1.48/19.17 (7,7 %)
steel and Allowable Stress Design (ASD) are studied [5]. Data 126.5 0.002 20.02 4.5/29.06 (15 %)
to collect are the deflection due to the same load as an RC 231
beam, bending moment, and the natural frequency; some 124.85 0.0056 11.25 1.48/26.04 (5,7 %)
percentage of the self-weight as compared to the total load is 152.5 0.0031 10.65 8/18.98 (42 %)
noted, for a comparison of the mass source that contributes to 313
the free vibration. 182.34 0.0045 8.31 2.41/13.39 (18 %)
320 0.0047 8.4 8/25.64 (31.2 %)
All the data about responses for both RC beam and steel 321
beam are collected, as described in Table 2. Fig. 6 shows a WF 273.45 0.0073 6.52 2.41/20.05 (12 %)
steel beam with its dimensions and the first mode shape of 463 0.0074 6.55 8/37.07 (21.6 %)
vibration. 332
433.47 0.0107 5.39 2.41/29.07 (8.3 %)
(*) Inclined values are of steel beams.
39
The data set ABC=111 means the case in which the Table 4. Results of steel beams in various conditions
shortest beam (A=1) is studied, its span equals L=4m,
supporting the lightest loading (B=1), and the highest Input factors Outcomes Ration S/N
compliance (i.e. the lowest rigidity, or C=1); and the data set
ABC=332 means the case in which the longest beam (A=3) is Nat. Bending Nat.
Run A B C Bending
Frequency Moment Freq.
studied, its span equals L=10m, supporting the heaviest
loading (B=3), and the medium rigidity (medium rigidity has 1 1 1 1 12.98 21.85 22.26 -26.79
the modulus of elasticity laying between the maximum value 2 1 2 2 9.56 41.2 19.61 -32.29
and the minimum value, C=2).
3 1 3 3 7.53 66.45 17.53 -36.45
F. Ranking the importance of the factors involved
Factors A, B, and C may affect the outcomes to different 4 2 1 2 11.25 59.4 21.02 -35.47
levels of influence. it is necessary to compute the difference 5 2 2 3 9.05 91.5 19.13 -39.22
in the maximal and minimal values of the signal-to-noise
(S/N) ratios, in which the largest difference Δ=max S/N-min 6 2 3 1 7.47 125 17.47 -41.94
S/N indicates the most effective factor. The higher Δ, the
7 3 1 3 8.31 182.34 18.39 -45.21
more important factor is.
8 3 2 1 6.52 273.5 16.28 -48.74
For evaluation in the static analysis, the response such as
bending moment, and deflection, is selected to be the 9 3 3 2 5.39 433.5 14.63 -52.74
predictor, and this outcome should be the lowest is the best.
According to Taguchi’s method [7], the signal-to-noise would
be as below: Same as the RC beams, for the WF steel beam, the order
is the span, and the load, and the stiffness, respectively for
1 n 2
( S / N ) s  10log(  yi )
n i 1
(6) static analysis; but in dynamic analysis, the order of
importance is that: the load, span, and the stiffness.
But in dynamic analysis, it is necessary to ensure the G. Ranking the importance for the factors involved
natural frequency is not too small. In dynamic analysis, the For simplicity, the steps of calculating to rank the factors
higher frequency, the stiffer structure is. Optimal performance based on the S/N are not recalled in this paper.
is what having the maximal frequency, implying the stiffness-
to-mass ratio, or a higher frequency. As such, to avoid being For static analysis, the most sensitive is the factor A, or the
seriously sagged, the Signal-to-Noise ratios should be “the span (Δ=max S/N-min S/N=17.47), then comes the factor B,
Larger is the better” or load intensity (Δ=max S/N-min S/N=8.04), and the factor
C, or rigidity seems to be less sensitive with the bending
1 n 1
 )
moment (Δ=max S/N-min S/N=3.66).
( S / N ) L  10 log( (7)
n i 1 yi2 For dynamic analysis, the order of importance is slightly
As such, the comparison should be conducted in both ways different from that in static analysis. The most important
static and dynamic working conditions, for both kinds of factor is the loading intensity, or factor B (Δ=max S/N-min
material, i.e. RC beam and WF steel beam. S/N= 4.24), then follow the span factor A, (Δ=max S/N-min
S/N= 3.53), and the bending stiffness seems to be the least
V. RESULTS influenced factor.
Data of the structural analysis for both kinds of beam in H. Comparison on the performance of the structure
Table 3 (RC beam) and 4 (steel beam). Nine tasks are Table 2 is referred to. The overall performance to be
sufficient for collecting a representative dataset, including 3- evaluated is based on [8]. There are many criteria for an
level factors, A, B, and C. evaluation of performance, such as self-weight and loadings,
span/stiffness/inertia properties, yield stress or deflection, and
Table 3. Results of RC beams in various conditions natural frequency. Criteria of the separate performance are the
Linear static procedure (LSP) and Non-linear static procedure
Input factors Outcomes Ration S/N (NSP) instead of the dynamic procedure [9]. In general,
demand and the capacity of the structure would be compared
Nat. Bending Nat.
Run A B C
Frequency Moment Freq.
Bending relatively to its self-weight. The best performance is that the
1 1 1 1 17.08 22 24.65 -26.85
capacity, together with the natural frequency both attain the
highest by the least possible self-weight.
2 1 2 2 12.19 39.37 21.72 -31.90
3 1 3 3 9.53 62.14 19.58 -35.86 In general, the limit of deflection for a beam is 1/360 times
the span [8] unless otherwise specified (i.e., the bending
4 2 1 2 20.02 67.37 26.03 -36.57
moment at crack appearance). A36 steel beam which is
5 2 2 3 15.82 36.57 23.98 -31.26 designed as per ASD would deflect to a maximal value of
6 2 3 1 13.47 126.5 22.58 -42.04 L/360. This value will be selected as the controlled
7 3 1 3 10.65 152.5 20.55 -43.66 displacement (i.e., limited deflection equals span/360) in a
8 3 2 1 8.4 320 18.48 -50.10
static non-linear pushover analysis.
9 3 3 2 6.55 463 16.32 -53.31
40
data, and the independent variables in the multivariable linear

regression equation could explain well the model. The
performance factor, (abbreviation PF), defined by the ratio of
self-weight per total load divided by the flexural capacity
should be formulated as follows:
RC beam
Figure 7. 4 meter- steel beam of WF8, bending moment at M [M ]
controlled displacement span/360 in pushover static NL analysis ( PF ) RC  0.045  0.241( )  0.397( Dyn ) 
[M ] [M ]
Table 5. Results of Static non-linear pushover analysis  7.297 104 ( sw / q)  0.011( f ) (8)
WF steel beam
Response
M [M ]
Task Bending
Pushover
Flexural Self-
Freq. ( PF ) Steel  0.005  0.66 104 ( )  0.346( Dyn ) 
moment
Load case
capacity weight/
(Hz) [M ] [M ]
(kNm) (kNm) load (%)
5.4 104 ( sw / q)  22.93 104 ( f ) (9)
22 249.47 86.87 18.2 17.08
111
21.85 135.59 2665 9.5 12.98
39.37 249.47 86.87 9 12.19 If the same demand of bearing flexural capacity, the steel
122 structure has the smaller values of the performance factor, or
41.2 135.59 2665 5,1 9.56
the smaller weight for a larger bearing capacity, resulting in a
62.14 249.47 86.87 7.1 9.53
133 more reasonable usage for both static and dynamic responses.
66.45 135.59 2665 3,2 7.53
For example, if some numerical values for a prescribed
212
67.37 832.33 208.7 29 20.02 demand are given, M/[M]=0.25, [Mdyn]/M=10, sw/q=0.05,
59.4 409.87 7005 11.8 11.25 and f=5.39Hz, the PF for RC beam equals 4.014, for steel
96.57 832.33 208.7 20 20.02 beam, -3.44. The steel beam would support a higher load by
223 utilizing a smaller weight, and in the dynamic aspect, the
91.5 409.87 7005 7,7 11.25
response of a steel beam could support a dynamic loading for
126.5 832.33 208.7 15 20.02 a given limit of deflection. A lighter steel beam could provide
231
124.85 409.87 7005 5,7 11.25
a wider range of flexural capacity (i.e. [MDyn]/[M]) than that
152.5 1523 383.00 42 10.65 of a heavier RC beam.
313
182.34 1160.4 20201 18 8.31
The negative coefficients for independent variables
320 1523 383.00 31.2 8.4 indicate an inverse correlation with the predictor (dependent
321
273.45 1160.4 20201 12 6.52 variable). The higher percentage of self-weight and value of
463 1523 383.00 21.6 6.55 the natural frequency, that the beam is someway stiffer, or
332 contributing more dead load for the beam, then the dynamic
433.47 1160.4 20201 8.3 5.39
(*) In plastic design, capacity for steel beam M = 1.12Zfy; for RC beam, the bearing capacity Mpushover bending Moment would be
single reinforcement flexural capacity is calculated by formula (2b). Inclined smaller, resulting in a higher flexural capacity. Steel beam
values are of steel beams. could be predicted better than RC beam, thanks to the very
Results indicate that the WF steel beam, the ultimate small significance and adjusted R-square.
flexural capacity is many times greater than the maximum I. Comparison of the dynamic response of the structure
bending moment that causes a limited deflection, L/360. As
for the RC beam, there is a reversal trend where the maximal Another comparison of the response of the structure made
bending moment in static NL pushover analysis is much by the two materials is plotted in Fig. 8.
greater than the ultimate value, i.e., the value determined by
(2b). This implies that the steel beam could support the Comparison between moment in static and static
dynamic loading better than the RC beam. NL analysis
M/Mpushover
By investigating the correlation between an arbitrary index 0,8
that stands for the ratio economy (i.e., self-weight over 0,6
technical quantity (for example, capacity), good performance
should give this ratio as small as possible. This is to conduct a 0,4
multi-variable linear regression. The index chosen is the self-
weight divided by the maximum bending moment. This 0,2
should be the predictor, standing for the effectiveness of the
0
performance. If a structure has the small self-weight,
associated with a great bearing capacity, both in static and 1 2 3 4 5 6 7 8 9
dynamic ones, that is an outstanding performance. All are RC beam WF steel beam
dimensionless numbers to avoid being dependent on the
system of the unit. Some results are given by the software as
shown in Fig. 8. It is seen that for both kinds of beams, the Figure 8. Comparison between the two kinds of beam material, on
significance F is less than 0.005. The values of adjusted R- static and dynamic pseudo-static pushover analysis
square, the adj. R2=0.94 for steel beam is great, meaning that
the model could predict well the performance by prescribed
41
From Fig. 9 in the case of a 6-meter and 10-meter long natural frequency and the strain. Three factors including the
steel beam (the right part of the figure), there is a large span, vertical load, and stiffness of bending are considered.
difference in the pushover bending moment causing span/360 There are 9 scenarios of possibly occurring cases to be studied,
deflection, and bending moment. For short span RC beam, the and the response is the bending moment, deflection, natural
ratio is close to 1, implying that the short span beam has not frequency for the first mode of free vibration, and the self-
resulted in the difference between values of bending moment weight. For comparison, all the responses are normalized and
in ULS (crack appearance) and SLS (deflection attains limit). converted to dimensionless numbers. Comparisons are made
Scenarios 4 to 9 are for the medium to large span beams. to different criteria including a) the order of importance of the
contributing factors (Signal-to-Noise of the outcomes); b) the
VI. DISCUSSION difference in flexural capacity in static (equation 2b) and
dynamic bearing capacity (using static non-linear pushover
Comparisons are studied from many different aspects analysis); and c) the regression model for predicting the
such as flexural capacity, dynamic response, the controlled- performance of beam (ratio between the percentage of self-
displacement limit of bending moment, etc. Nine scenarios for weight to the total load, and the flexural capacity, namely PF
data collection to replace 33=81 possible cases are studied. values). The main factors are the span and the load in both
Taguchi’s L9(3^3) orthogonal array [10] is used. This array static and dynamic analysis. The steel beam has a good
with 9 tasks gives 9 data of the outcome, including bending performance factor (PF) in terms of self-weight ratio, flexural
moment and natural frequency. This array with 9 tasks gives capacity ratio, and the dynamic response associated with a
9 data of the outcome, including bending moment and natural limit deflection. The model for steel beam proves greater
frequency. Some key points are: reliability, in terms of the high value of adjusted R-square,
 Three-level factors, namely A, B, and C are relative very small significance F value, and small value of
quantities. For instance, a short span or the level of A coefficients for the independent variables. This also indicates
equals unity is a 4-meter span. For factor B or the load, that the RC beam has more complicated characteristics in the
with arbitrarily prescribed values, has the light, correlation between the flexural capacity and the structural
medium, and heavy, or less than 11 kN/m, from 11 to configuration or size. This might be explained that the RC
20 kN/m, and more than 30 kN/m, respectively. This beam is a plastic-elastic structure, and the structure
is slightly not distinguished as seen in Table 2, the last experiences many stages of load support.
column. From some practical points of view, steel material steel
 RC beam is modeled by solid element for controlling material should be relevant for beams in a high-rise building
the strain at cracking moment, while the steel beam is for its controllable effectiveness in performance and cost
easily modeled with the built-in tool in SAP2000. based on self-weight.
Nevertheless, the pilot test on the finite element model
confirms partly the minor role of rebar in calculating REFERENCES
the limit state and free vibration. [1] An ACI Standard and Report, “Building Code Requirements
 Values of the cracking moment for RC beam in various for Structural Concrete (ACI 318-14)”, ISBN: 978-0-87031-
tasks might be slightly lower than those in Table 2. It 930-3, Farmington Hills, MI, USA, 2014.
is determined by simply calculating qL2/8 when the [2] ACI 440.2R-08 (American Concrete Institute), Guide for the
strain in concrete exceeds 3.5e-5. The effect of cracks Design and Construction of Externally Bonded FRP Systems
on the dynamic response of the RC beam is not taken for Strengthening Concrete Structures, Farmington Hills, MI,
into account. USA, 2008.
[3] A.A Kasimzade, S. Tuhta.: Analytical, Numerical and
 In general, the flexural capacity could be evaluated via Experimental Examination of Reinforced Composite Beams
the flexural toughness, which is defined as the area Covered with Carbon Fiber Reinforced Plastics. Journal of
under the load-deflection curve of the structure subject Theoretical and Applied Mechanics, Sofia, Vol. 42, No. 1, pp.
to flexion [11]. But this could be alternatively replaced 55-70.
by a random range of pushover bending capacity for [4] Ezzeldin Yazeed Sayed-Ahmed, “Design of Laterally
different levels of loading in Table 2. By converting unsupported steel I-beam according to the Egyptian code of
the bending moment and the flexural capacity into practice: A proposed simple equation”, Al Azhar University
dimensionless numbers, the model has a more general Engineering Journal, Vol. 7, No. 5, December 2004, pp. 1043-
meaning. 1063.
 The plastic flexural capacity of the wide flange steel [5] AISI 1963 (AISI-S100-16w/S1-18, “North American
Specification for the Design of Cold-Formed Steel Structural
beam is much higher than the bending moment which
Members”, 2016 Edition with Supplement 1, p. 509.
causes an acceptable deflection L/360. This is defined
by the flexural capacity [M]. [6] A. Chopra, “Dynamics of structures”, Prentice-Hall Intl Series,
publisher Pearson Education (US), 2019.
VII. CONCLUSION [7] Duong Hong Tham, “Some statistical techniques applied to
Engineering mechanics problems”, VietNam Journal of
This paper compares the performance of an RC beam and Science and Technology, ISSN 2525-2518, Vol. 57, 6A, 2019,
a steel one. The sizes of RC beams have a fixed ratio of pp. 10-20.
width/height equal to 0.5. An RC beam is modeled with [8] Grant T. Iwamoto, “Performance Evaluation of Spanning
SOLID elements in SAP2000 [12] to detect some strain at Systems”, Master's thesis, Massachusetts Institute of
crack appearance, then determine the cracking moment Mcr. A Technology, USA, 2014.
pilot test on the RC beam reveals the rebars affect weakly the
42
[9] H. S. Lew and Sashi K Kunnath, “Evaluation of analysis and Underground Space in Engineering, Vol. 20, 2005, pp.
procedures for performance-based seismic design of 167-173.
buildings”, 12WCEE2000 Intl Conf. on Earthquake [11] M.N Soutsos, T.T Le, A.P. Lampropoulos, “Flexural
Engineering, Available: performance of FRC made with steel and synthetic fibres”, J.
https://www.iitk.ac.in/nicee/wcee/article/1005.pdf, 8 pages. Construction and Building Materials, Vol. 36. November 2012,
[10] Ozcan Tan, A. Sahin Zaimoglu, Sinan Hinislioglu, Selim Pages 704-710.
Altun, “Taguchi approach for optimization of the bleeding on [12] Sap2000 User Manual, Availabe:
the cement-based grouts”, International Journal for Tunnelling https://www.civilax.org/sap2000-manual, October, 2014.
43
Optimization Model for Biomass Supply Chain

Planning: A Case Study in Mekong River Delta –
Vietnam
Thi-Be-Oanh Cao Duc Duy Nguyen
Department of Industrial Engineering and Management School of Business and Management
National Kaohsiung University of Science and Technology LUT University
Kaohsiung 80778, Taiwan 53850 Lappeenranta, Finland
0000-0002-5451-7879 0000-0001-7553-5913
Thanh-Tuan Dang Chia-Nan Wang

Department of Industrial Engineering and Management Department of Industrial Engineering and Management
National Kaohsiung University of Science and Technology National Kaohsiung University of Science and Technology
Kaohsiung 80778, Taiwan Kaohsiung 80778, Taiwan
0000-0002-8559-0868 0000-0002-2374-3830
Abstract: Renewable energy is an alternative resource to environment [10]: Utilizing biomass for heat and electricity
ensure energy security in many countries and in reducing generation can decrease the use of fossil fuels. Therefore,
environmental problems. An efficient supply chain network reducing the pressure to pressure of natural resources. In
from renewable resources is one of the critical factors in the addition, the possibility of deforestation and desertification
success of renewable energy systems. As the ASEAN country also slowed down. Moreover, biomass energy opens the
with the highest economic development rate and one of the opportunities and jobs offered to the community around the
most vulnerable to climate change, Vietnam Government plants. An optimal supply chain network can help the energy
understands the need of changing the energy source structure plant efficiently operate with low costs and stable material
in order to increase the share of energy in the world. supply capacities [1]. Therefore, it is essential to develop a
model for designing a biomass supply chain.
Renewable energy is linked to energy consumption that is
cost-effective, efficient, and for long-term purposes. Most The growth in the economy and increasing population
importantly, the potential of biomass energy in Vietnam is cause the rapidly growing energy demand in Vietnam. In the
very diverse and has quite large reserves. This paper “Government of Viet Nam. 2015. Revised Power
introduces an optimization model to design a biomass supply Development Plan 2011–2020. Ha Noi”, the annual electricity
chain for producing electricity. Mixed-integer linear demand in Vietnam reaches 506.0 TWh (It was 45.6 TWh in
programming is developed for the biomass factory site, 2005). To satisfy the future demand without dependence on
capacity selection, and biomass allocation to each factory. fossil fuels, The Vietnamese Government has declared
The objective is to minimize the biomass factory opening importation plans to encourage clean energy development.
fixed and operation costs such as purchasing and Based on decision No. 2068/QD-TTg dated 25/11/2015
approving the RE development Strategy of Vietnam until 2030
transportation costs. The proposed model is applied to a
with the vision to 2050, renewable energy will increase to 186
realistic case study in the central Mekong River Delta -
TWh in 2030 and 452 TWh in 2050. Of this, biomass for power
Vietnam. production is 9.0 TWh in 2030 and 20.0 TWh in 2050 (Only
0.3 TWh in 2015). Accordingly, research on biomass in
Keywords: biomass supply chain, renewable energy, Vietnam aligns with the Vietnamese government's energy
sustainability, location-allocation, supply chain management strategy.
I. INTRODUCTION This research focuses on designing biomass-to-energy
Nowadays, the increasing energy demand has raised networks with the motivations above. Mixed-integer linear
concerns for the government in many countries [1-7]. The programming is formulated to optimize the electricity plant
global electricity demand targets 30,000 TWh by 2030 and locations and biomass allocation. Besides the network's
42,000 TWh by 2050 [8]. Although the global electricity constraints, a constraint related to the monopoly sales policy
demand declined by 2% in 2020 compared to 2019 [9], it is between plants and supplies is established in section III. The
expected to be resilient after the COVID-19 pandemic in 2022 model objective minimizes the total cost from factory fixed
[8]. To satisfy the high energy consumption toward cost, transportation cost, and purchasing cost. Based on the
sustainability development, renewable energy, and nuclear results from sensitivity analyses on demand and supplier
power were objective to meet 50% of demand [9]. capacity, decision-makers can understand system behaviors.
The model is illustrated and verified by using the data from
As a renewable resource, biomass energy ensures energy Can Tho, which is central of the Mekong River Delta –
security and benefits the economy, society, and the Vietnam.
44
Copyright © authors. This work is licensed under a Creative Commons
The paper is structured as bellows: A literature review on the locations for a scale of factories’ capacity under demand
biomass supply chain planning is introduced in Section II. and supplier capacity uncertainties. Also focusing on selecting
After that, Section III presents the mathematical model for biomass power plants, Wang et al. [18] adopted a fuzzy
designing the supply chain. Section IV provides a case study multicriteria decision-making model to optimize site
in Vietnam. Finally, the conclusion and future research are selections. Regarding reducing gas emissions from electricity
presented in Section V. generation, Aviso et al. [19] investigated the effectiveness of
indirect and direct utilized residual biomass at power plants in
II. LITERATURE REVIEW Vietnam. A fuzzy linear programming model was developed
This section provides a literature review on supply chain for supporting analyzing and making decisions. Truong et al.
management and biomass for energy generation in Vietnam. [6] addressed the coal-fired power plants development
problem in Vietnam. The power plants used mixed biomass
There have been several studies in supply chain design and and coal as materials for electricity generation with objectives
management in the past decades. Jafarnejad and Aliabadi [11] related to economics, environmental and social implications.
formulated a bi-objective model for designing a sustainable Song et al. [20] focused on determining power generation
supply chain. The economic and environmental objectives plants from rice husks in the Mekong Delta. Plant with a
were simultaneously considered. Bioethanol from the capacity is 11 MW are considered along with CO2 reductions
agricultural waste supply chain in Mexico was addressed by and economic criteria.
Minor-Popocatl et al. [12]. The authors introduced a mixed-
integer mathematical programming model for four-echelon This research aims to bring the potential and essential
network design. Saghaei et al. [13] focused on developing the contributions to the existing biomass supply chain research and
mixed-integer non-linear programming model for woody support for decision-making. Specifically, the contributions
biomass to energy. A hybrid metaheuristic approach was can be summarised as follows:
developed to generate solutions for large-scale problems.
 This research presents a mathematical model for supply
Rahimi et al. [5] addressed the electricity production supply
chain network design. The model develops a constraint
chain from animal manure with a case study in Iran. The model
to consider monopoly sales conditions for rice husks in
decides the optimal facility location's capacity level and
a win-win relationship. The biomass will be stably
material flow throughout the network. Fernández-Puratich et
provided to electricity plants at higher price. Therefore,
al. [14] developed an optimization model for biomass heat and
the policy is beneficial for both parties.
power generation systems. The network considered all the
public, private and residential facilities in Maule Region,  Sensitivity analysis was conducted in demand and
Chile. The model allows decision-makers to select types of supplier capacity to analyze the system behaviors and
biomass, size of plants, and services radius for cost and provide marginal insight.
environmental objectives. Moretti et al. [4] formulated a
mixed-integer linear programming model for an optimal  This study addressed the biomass supply network in
design of a biomass supply chain. The model accounted for a Vietnam, an abundant biomass country yet a limited
multiple echelons network, multi-feedstock, multiple number of renewable supply chain research. Therefore,
transportation modes, and weekly periods. A biodiesel supply the research contributes to biomass supply chain
chain in Iran was investigated by Mohtashami et al. [15]. A research and supports the national energy strategy. It
multi-objective model was developed simultaneously to also contributes to a biomass energy map in Vietnam.
optimize cost, environmental impacts, and social benefits. The  Besides, a large biomass quantity in Vietnam is burned
augmented ε-constraint method was adopted to provide the out in an open environment or traditionally used, thus
trade-off among objectives. Uncertainty is an essential factor causing environmental problems. Therefore, this
that strongly affects any biomass supply chain planning; research has an essential meaning in promoting
therefore, capturing uncertainty can help the design be more biomass for energy production and reducing
practical. Accordingly, Guo et al. [16] formulated multiperiod environmental pollution.
stochastic programming to handle uncertainty to design
biomass for the biofuel supply chain. The results from the III. METHODOLOGY
stochastic model were compared with the results from
This section presents mixed-integer linear programming
deterministic models to provide marginal insight. In line with
for the economic optimization of biomass supply chain design.
tackling uncertainty, Li et al. [17] addressed a production
A model from Ilbahar et al. [3] was adopted as a basic work.
planning problem of a bioethanol plant while demand and
The following notations are used in the model formulation.
yield were uncertain. A heuristic algorithm was developed to
reduce the computational efforts. As an important biomass Indices:
source in Shandong – China, agri-biomass supply chain was
taken into account by Wu et al. [7]. A multidisciplinary
m Index for locations of suppliers, m  {1, 2, …, M}
approach was developed and used to define the optimal f Index for locations of potential factories, f  {1, 2,
number and capacity of facilities in a distribution network, in …, F}
which mixed-integer linear programming was used as an
optimization approach. k Index for type of factories, k  {1, 2, …, K}
Renewable energy research in Vietnam has gained Parameters:
attention from scholars; however, a limited number of studies φ The heat value coefficient (kwh/ton)
focus on biomass supply chain management. Duc and α The unit cost of biomass (USD/ton)
Nananukul [2] considered a location-allocation problem for β The unit transportation cost (USD/ (ton* km)
the electricity supply chain in Vietnam. The study determined
45
k The opening fixed cost of electricity plant f=F k=K

(8)
type k (USD) ∑ ∑ yfk ≤ N
Dmf The distance from the biomass supplier m to f=1 k=1
the electricity plant f (km) k=K
MSm The maximum annual supply capacity of wmf ≤ ∑ yfk Ɐm, f (9)
supplier m (ton/year) k=1
ED The annual electricity demand (kwh)
[max,min] The maximum and minimum of efficiency of k=K m=M k=K (10)
the electricity plant must be met (%) ∑ CFk ×γmin ×yfk ≤ ∑ φ×xmf ≤ ∑ CFk ×γmax ×yfk Ɐf
CFk The capacity of electricity plants type k per k=1 m=1 k=1
year (kwh)
N Number of power plants that can be opened xmf ≥0, wmf{0,1}, yfk{0,1}
(11)
L The minimum quantity of biomass per year
The objective function is given in Eq. (1) total cost:
to establish the contract. factories opening fixed cost, purchasing cost, and
BigM given a large enough number transportation cost.
Decision variables:
Eq. (2) ensure the electricity demand must be satisfied. Eq.
xmf Quantity of biomass supplied from supplier m to power
(3) the supplier capacity must be satisfied. Eq. (4) implies that
plant f (tons)
any power plants must be bought at least L ton biomass per
wmf A binary variable equals one if supplier m ships biomass year to establish the contract. Eq. (5) ensures that no biomass
to power plant f, and 0 o/w. is transported if there is no contract between the supplier and
yfk A binary variable equals one if electricity plant k installs
the power plant. Eq. (6) allows suppliers to supply biomass to
type k, and 0 o/w. only one factory with a contract. Eq. (7) assures that only one
type of power can be opened in a location. Eq. (8) limits the
Objective functions. number of factories that can be operated. Eq. (9) assures no
Minimize: biomass transport to the power plant if it is not installed. Eq.
(10) is lower bound and upper bound about the capacity of
f=F k=K m=M f=F each factory. Eq. (11) presents the variable constraints.
∑ ∑ k ×yfk + ∑ ∑[(β×Dmf ×xmf )+(α×xmf )]
(1) IV. A CASE STUDY
f=1 k=1 m=1 f=1
In this case study, rice husk was considered the primary
In which: biomass supply source due to their availability in Can Tho,
Opening fixed cost : f=F k=K Central of Mekong River Delta in Vietnam (Fig. 1). Can Tho
∑ ∑ k ×yfk is the central economic region in the Mekong River Delta and
f=1 k=1
one of four critical economic areas of Vietnam. As a result, the
economy's development causes an increase in electricity. The
Transportation cost : m=M f=F
electricity demand for our model is furnished by Can Tho
∑ ∑(β×Dmf ×xmf ) power company. We aim to satisfy 10% of demand by using
m=1 f=1 biomass energy, the remainder can be satisfied from other
Purchasing cost : m=M f=F sources.
∑ ∑(α×xmf )
m=1 f=1
s.t.
f=F m=M
(2)
∑ ∑ φ×xmf ≥ E D
f=1 m=1
f=F
(3)
∑ xmf ≤ MSm Ɐm
f=1
xmf ≥ L*wmf Ɐm, f

(4)
xmf ≤ BigM*wmf Ɐm, f Figure 1. A map of Can Tho city in Mekong River Delta in Vietnam
(5)
f=F (Source: Internet)
(6)
∑ wmf ≤1 Ɐm
The research focuses on using rice husk as the main
f=1
material for running turbines in power plants. The data about
k=K rice husk quantity and price were collected from the 41 main
(7)
∑ yfk ≤1 Ɐf rice milling factories in Can Tho. In this research, several small
k=1 milling factories with a low capacity were omitted. In order to
establish the contract with suppliers, the power plants are
required to buy at least 500 tons of rice husk per year, with the
46
price per ton is 26.36 USD on average. It also means that the matrix between supplier - and factory is identified based on the
rice milling factories commit to supplying at least 500 tons to coordinates via Google Maps.
the power plants.
To transform ton biomass to electricity, the coefficient
Rice husk is transported to power plants by a 9-ton truck from Ueda et al. [21] was used to calculate the heat values φ:
with an average price is 0.073 USD per ton per kilometer. The 1.276 tons of rice husk could generate 1000 kWh of electricity
BigM is assumed to equal to the maximum biomass quantity (0.78 kwh/ton).
from a set of suppliers.
Fig. 2 presents the rice milling factories and industrial
Three types of power plants could be installed in Can Tho: parks in Can Tho. The circles present for rice milling factories
7.5MW, 15MW, and 30MW. The investment cost for those which are suppliers. Meanwhile, the blue square symbols
plants was collected from the investment cost (IVC) from present industrial parks where power plants will be installed.
similar capacity factories in Vietnam. Then they are converted
to annual cost by considering the 22 years project lifetime (T) V. RESULTS AND SENSITIVITY ANALYSIS
and the interest rate (i) is 15% per year.
Table 2. The relationship between biomass networks and
r(1+r)T electricity demand
Annuity=IVC×
(1+r)T -1 Demand No. factory Name and type of factory
Each installed power should maintain 80% to 100 % of the 5.0% 1 F4 -15MW
designed capacity. The power plants are planned to be installed 10.0% 1 F4 -30MW
in 13 industrial parks in Can Tho (Table 1). The distance 15.0% 2 F4 -30MW & F8 -15MW
20.0% 2 F4 -30MW & F9 -30MW
25.0% 3 F4, F9 -30MW & F8 -15MW
30.0% - -
Table 3. Costs breakdown from sensitivity analysis in

demand(USD)
Demand Total cost Factory fixed Transport Purchasing

cost ation cost cost
5.0% 13,613,544.36 9,435,946.28 23,085.24 4,154,512.84
10.0% 22,142,002.28 13,776,481.56 56,495.04 8,309,025.68
15.0% 35,759,584.21 23,212,427.84 83,617.86 12,463,538.52
20.0% 44,448,659.44 27,552,963.12 277,644.96 16,618,051.36
25.0% 58,085,189.57 36,988,909.40 323,715.98 20,772,564.20
Figure 2. Rice milling factories and industrial parks in Can Tho

Table 4. The relationship between biomass networks and
(Source: Authors) supplier’s capacity
Supplier capacity No. factory Name and type of factory
Table 1. List of industrial parks in Can Tho 50% 1 F4 -30MW
100% 1 F4 -30MW
Names Notations
150% 1 F4 -30MW
Tra Noc II Industrial Park F1
200% 1 F4 -30MW
Hung Phu I Industrial Park F2
Hung Phu IIA Industrial Park F3 Table 5. Cost break down from sensitivity analysis in supplier
capacity (USD)
Thot Not Industrial Park F4
O Mon Industrial Park F5
Supplier Total cost Factory fixed Transport Purchasing
capacity cost ation cost cost
North O Mon Industrial Park F6
50% 22,224,329.72 13,776,481.56 138,822.48 8,309,025.68
Phu Hung IIB Industrial Park F7
100% 22,142,002.28 13,776,481.56 56,495.04 8,309,025.68
High-tech Industrial Park F8
150% 22,135,621.94 13,776,481.56 50,114.69 8,309,025.68
Thot Not I Industrial Park F9
200% 22,131,677.73 13,776,481.56 46,170.49 8,309,025.68
Thot Not II Industrial Park F10
Cai Rang Industrial Park F11
To obtain marginal insight on the biomass supply chain,
Co Do Industrial Park F12 we conduct several sensitivity analyses on demand and
Phong Dien Industrial Park F13 supplier capacity that gained attention from scholars (e.g.,
[11], [13], and [16]). The MILP model is modeled by CPLEX
47
12.10 to adopt a brand and bound algorithm. The solving time S=50% F4-15MW
is limited to 2 hours, and the MIP gap was set equal to default
from the optimizer. S=100% F4-15MW
D=5%
S=150% F8-15MW
Table 2 presents results from sensitivity analysis on
demand. The satisfied demand is investigated from 5% to 30% S=200% F8-15MW
of annual demand. This is motivated by the energy plan from
the region. It has been shown that the changing in the network S=50% F4-30MW
and capacities of each factory is strongly related to the demand. S=100% F4-30MW
When the demand increases to 30%, the model turns infeasible D=10%
due to a lack of biomass. From data, it was estimated that only S=150% F4-30MW
29.59% of demand would be satisfied by using rice husk in S=200% F4-30MW
Can Tho.
S=50% Infeasible
As shown in Table 2 and Table 3, when the objective is
satisfied with 5% and 10% of annual demand, the model opens S=100% F4-30MW, F8-15MW
only one factory in Thot Not Industrial Park with a capacity of Scenarios D=15%
S=150% F4-30MW, F8-15MW
15MW and 30MW, respectively. Due to the rising demand to
15%, two factories are opened in Thot Not Industrial Park S=200% F4-30MW, F8-15MW
(30MW) and High-tech Industrial Park (15MW). To satisfy
S=50% Infeasible
20% annual demand, the model opens two 30MW factories in
Thot Not Industrial Park and Thot Not I Industrial Park. Model S=100% F4, F9-30MW
opens 3 factories: 2 – 30MW factories in 2 industries parks that D=20%
S=150% F4, F9-30MW
are Thot Not Industrial Park and Thot Not I Industrial Park,
another 15MW factory is opened in High tech Industrial Park. S=200% F4, F8-30MW
The increasing demand causes the increasing in the S=50% Infeasible

required material; therefore, the transportation cost and
purchasing cost are increased when demand changes from 5% S=100% F4, F9-30MW, F8 -15MW
D=25%
to 25% annual demand. S=150% F4, F9-30MW, F8 -15MW
Table 3 presents the changes in the supplier capacity to the S=200% F4, F8-30MW, F9 -15MW
networks. The increase in supplier capacity was investigated
rather drastically, from 50% to 200%. The supply capacity is
based on the practical situation in Can Tho. In the peak Figure 3. Scenario generation and results
seasons, many rice milling factories in Can Tho serve local
milling demand and carry the demand for other provinces. When demand is 25% annual, supplier capacity changes
Therefore, rice husk quantity can be increased. Many rice from 150% to 200%. The model always chooses Thot Not
milling factories maintain a half capacity in the contract due to Industrial Park, Thot Not I Industrial Park, and High-tech
lack of material or restriction regulation, such as during the Industrial Park for installing the factory. The difference
COVID-19 outbreak. As a result, rice husk supply capacity between the two solutions is the capacity of factories F8 and
from the suppliers can be decreased. F9.
From Table 4, when the supplier capacity is 50% to 200%, The result from the scenarios reveal that Thot Not Industrial
the model opens only one 30MW at Thot Not Industrial Park. Park is a good location for opening the biomass factory based
The increase in supplier capacity allows suppliers to join the on our objective constraints and available data. Further
network due to the satisfying - minimum shipped biomass analysis can be conducted by a focus on Thot Not Industrial
quantity (L). Therefore, the transportation cost is decreased Park.
from 138,822.48 USD to 46,170.49 USD per year. VI. CONCLUSION
In this work, Monte Carlo simulation is used to generate This study adopted mixed-integer linear programming for
the scenarios from demand levels and supplier capacity levels. agricultural residue-based biomass to the energy supply chain.
There are 20 generated scenarios in total. The objective is to minimize the factory fixed cost,
As we can see from Fig. 3, when the demand increases to transportation cost, and purchasing cost. After that, sensitivity
15%, 20%, and 25% annual demand, the model becomes analyses were conducted on demand and supplier capacity for
infeasible when the supplier capacity remains 50% due to the future analysis in the supply chain. In this research, the model
conditions of establishing the contract. The minimum shipped was set up to satisfy 5% to 25% annual demand, and the supply
biomass quantity (L) from suppliers to the factory was higher capacity was set to 50% to 200% recently capacity.
than the supplier capacity. It can happen in a real-world From the sensitivity analysis results, the minimum shipped
situation when the supplier capacity is significantly reduced. biomass quantity (L) from suppliers to the factory was robust
The suppliers need to negotiate with factories about the to the solutions. When the supplier capacity falls into a low-
contracts. Or a group of suppliers can gather to have a large capacity situation, many suppliers are out of the condition for
biomass quantity to make a contract. The reducing value of L signing a contract. Therefore, the investigation of this
can relax the biomass shortage situation. However, the model parameter can be future research.
is also infeasible when the needed biomass quantity exceeds
the total supplier capacity. (For example, when the demand is This research may be extended by considering multiple
higher than 29.59% of annual demand). objectives, such as economic and environmental objectives
48
[11], optimizing cost, environmental impacts, and social [12] H. Minor-Popocatl, O. Aguilar-Mejía, E. León-Olivares, and R.
benefits [15], to mention a few. Furthermore, the model can be Tapia-Olvera, "Supply chain design for the generation of
developed to consider multiple echelons networks, multi- bioethanol from agricultural biomass", in Proceedings 2019
feedstock, multiple transportation modes, and weekly periods. IEEE International Autumn Meeting on Power, Electronics and
In addition, uncertainty handling in future research can help Computing (ROPEC), Ixtapa, Mexico, 2019, pp. 1-6: IEEE.
the supply chain design be more realistic. [13] M. Saghaei, H. Ghaderi, and H. Soleimani, "Design and
optimization of biomass electricity supply chain with
REFERENCES uncertainty in material quality, availability and market
demand," Energy, vol. 197, p. 117165, 2020.
[1] X.-g. Zhao and A. Li, "A multi-objective sustainable location
[14] H. Fernández-Puratich, R. Rebolledo-Leiva, D. Hernández, J.
model for biomass power plants: Case of China", Energy, vol.
E. Gómez-Lagos, B. Armengot-Carbo, and J. V. Oliver-
112, pp. 1184-1193, 2016.
Villanueva, "Bi-objective optimization of multiple agro-
[2] D. N. Duc and N. Nananukul, "Supply chain model for
industrial wastes supply to a cogeneration system promoting
renewable energy resource from biomass in Vietnam", in
local circular bioeconomy", Applied Energy, vol. 300, p.
Proceedings 2018 5th International Conference on Industrial
117333, 2021.
Engineering and Applications (ICIEA), Singapore, 2018, pp.
[15] Z. Mohtashami, A. Bozorgi-Amiri, and R. Tavakkoli-
520-525: IEEE.
Moghaddam, "A two-stage multi-objective second generation
[3] E. Ilbahar, C. Kahraman, and S. Cebi, "Location selection for
biodiesel supply chain design considering social sustainability:
waste-to-energy plants by using fuzzy linear programming",
A case study," Energy, vol. 233, p. 121020, 2021.
Energy, vol. 234, p. 121189, 2021.
[16] C. Guo, H. Hu, S. Wang, L. F. Rodriguez, K. Ting, and T. Lin,
[4] L. Moretti, M. Milani, G. G. Lozza, and G. Manzolini, "A
"Multiperiod stochastic programming for biomass supply chain
detailed MILP formulation for the optimal design of advanced
design under spatiotemporal variability of feedstock supply",
biofuel supply chains", Renewable Energy, vol. 171, pp. 159-
Renewable Energy, 2022.
175, 2021.
[17] X. Li, S. Lu, Z. Li, Y. Wang, and L. Zhu, "Modeling and
[5] T. Rahimi, R. Babazadeh, and A. Doniavi, "Designing and
optimization of bioethanol production planning under hybrid
planning the animal waste-to-energy supply chains: A case
uncertainty: A heuristic multi-stage stochastic programming
study," Renewable Energy Focus, vol. 39, pp. 37-48, 2021.
approach", Energy, p. 123285, 2022.
[6] A. H. Truong, M. Ha-Duong, and H. A. Tran, "Economics of
[18] C.-N. Wang, T.-T. Tsai, and Y.-F. Huang, "A model for
co-firing rice straw in coal power plants in Vietnam",
optimizing location selection for biomass energy power plants",
Renewable Sustainable Energy Reviews, vol. 154, p. 111742,
Processes, vol. 7, no. 6, p. 353, 2019.
2022.
[19] K. B. Aviso, C. Sy, R. Tan, and A. T. Ubando, "Fuzzy
[7] J. Wu, J. Zhang, W. Yi, H. Cai, Y. Li, and Z. Su, "Agri-biomass
optimization of carbon management networks based on direct
supply chain optimization in north China: Model development
and indirect biomass co-firing", Renewable Sustainable Energy
and application", Energy, vol. 239, p. 122374, 2022.
Reviews, vol. 132, p. 110035, 2020.
[8] IEA, "World Energy Outlook 2021" IEA, 2021.
[20] N.V. Song, T.V. Ha, T.D. Thuan, N.V. Hanh, D.V. Tien, N.C.
[9] IEA, "World Energy Outlook 2020" IEA, 2020.
Tiep, N.T.M Phuong, P.A. Tu, T.B. Uan, "Development of Rice
[10] R. Mamat, M. S. M. Sani, F. Khoerunnisa, and A. Kadarohman,
Husk Power Plants Based on Clean Development Mechanism:
"Target and demand for renewable energy across 10 ASEAN
A Case Study in Mekong River Delta, Vietnam", Sustainability,
countries by 2040", The Electricity Journal, vol. 32, no. 10, p.
vol. 13, no. 12, p. 6950, 2021.
106670, 2019.
[21] T. Ueda, Y. Kunimitsu, and Y. Shinogi, "Potential conflicts for
[11] E. Jafarnejad and J. Aliabadi, "Multi-objective optimization of
the reuse of rice husk in Thailand", Paddy and Water
costs and pollutants in order to manage the sustainable supply
Environment, vol. 5, no. 2, pp. 123-129, 2007.
chain of bio-fuels", in Proceedings 2017 IEEE International
Conference on Industrial Engineering and Engineering
Management (IEEM), Singapore, 2017, pp. 1990-1994: IEEE.
49
Developing a Digital Competence Performance

Assessment Platform for University Students Based
on the DigComp Framework
Anh Tho Mai Thi Kim Oanh Duong Anh Tuan Ngo
Faculty of Information Technology Institute of Technical Education Institute of Technical Education
Ho Chi Minh City University of Ho Chi Minh City University of Ho Chi Minh City University of
Technology and Education Technology and Education Technology and Education
Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam
https://orcid.org/0000-0001-9159- https://orcid.org/0000-0002-4856- https://orcid.org/0000-0002-2264-
2379 1611 4187
Abstract: Following the rapid development of confirmed the necessity of enhancing young people’s digital
information and communication technologies and digital competence to meet changes in the human resource needs of
media, digital competence has become crucial and played a organizations and businesses [1] and to be able to participate
vital role in the learning process of university students. in modern society and progress professionally [2].
However, measuring and identifying students’ digital A report on the abilities of young people in the digital age
competence remains a challenge to higher education revealed that today’s young people are not ready for changes
institutions with attempts to develop solutions to enhance despite having frequent access to digital tools and media [3].
students’ digital competence. Previous research on digital According to the Program for International Student
competence assessment tools in higher education settings Assessment (PISA) developed and coordinated by the
showed that self-assessment tools account for almost 80% Organization for Economic Cooperation and Development
and yield less reliable data. Therefore, the present study aims (OECD), more than one in every four Australian students aged
to propose a web-based performance assessment tool, 15 years old (27%) have low digital competence proficiency
following the DigComp digital competence framework and [4]. A report at the Tallinn Digital Summit 2017, a platform
inheriting findings of the ERASMUS+ project (2016) with for discussions on plans for digital innovation in Europe, on
modifications to fit local social and higher education September 29, 2017, showed that 44% of the European
contexts. The proposed platform had received comments population did not have basic digital competence although
from eight experts, including four IT experts, three nine out of ten jobs would soon require digital competence.
educational experts, and one in quality assurance, to check Students were also found to lack digital competence [5], not
the content validity before being piloted with the participation to possess high levels of digital competence [6], [7], and not
to have the necessary competencies to meet learning
of 350 IT students of Ho Chi Minh City University of
requirements in the current educational contexts [8]–[11].
Technology and Education (HCMUTE) to check the internal
This means enhancing students’ digital competence needs to
consistency reliability using Cronbach’s alpha analysis. receive special attention from higher education institutions if
Experts and students’ feedback has also been taken into they want their students to have higher competitiveness and
consideration for platform improvement for future surveys on better job prospects after graduation.
a large scale. The results obtained allowed us to confirm that
the proposed web-based digital competence performance To enhance students’ digital competence, higher
assessment platform was found to deliver more reliable institutions need to, first of all, know their students’ current
assessments and could be adapted by Vietnamese higher proficiency levels. Many assessment models and tools have
education institutions for more generalized and accurate been developed by countries, regions, and international
organizations [12], [13]. However, in Vietnam, there have
assessment.
been only a few publications on digital competence for
Keywords: digital competence, higher education, students but none on assessing university students’ digital
competence.
student's digital competence, a web-based performance
assessment tool, DigComp digital competence framework In such a context, it is important to develop an instrument
that assesses students’ digital competence, enabling higher
I. INTRODUCTION education institutions to survey their students’ proficiency
In the age of digital transformation, changes are associated levels and subsequently find specific solutions to help students
with the application of digital technology in all aspects of life. develop digital competence.
Research on digital transformation in ASEAN countries Digital competence and DigComp Framework
showed that employers’ demands for digital competence have
increased by 200% in the past three years and are projected to The concept of digital competence was first mentioned by
continue rising in the next five years. Previous studies Gilster in 1997 [14] as the ability to understand and use

information in various formats from different sources and digital tools to solve problems, communicate, manage
displayed on computers. Following this, the concept has information, collaborate, create and share content, and build
drawn the interest of many authors, and different definitions knowledge effectively, selectively, appropriately, flexibly,
have been provided for different contexts. creatively, ethically, reflectively for learning and research at
universities.
Reviews in higher education settings showed digital
competence has been mainly introduced from two main The assessment questions in the study were built on the
approaches: policy document research and scientific research DigComp framework, covering five areas and the competence
[15]–[17]. indicators as shown in Table 1.
Among the definitions of digital competence from the Table 1. The DigComp framework
policy approach, the definition considered to be the most
complete is proposed by Ferrari in 2012 [18] and defined as Competence
“the set of knowledge, skills, attitudes, abilities, strategies, and Descriptions Competences
areas
awareness that are required when using ICT and digital media The abilities to
to perform tasks; solve problems; communicate; manage
information; collaborate; create and share content; and build identify
knowledge effectively, efficiently, appropriately, critically, information needs, 1.1 Browsing,
creatively, autonomously, flexibly, ethically, reflectively for
search for searching, and
work, leisure, participation, learning, and socializing” (p. 3).
This concept was then used by the European Commission and information and filtering data,
UNESCO to develop the digital competence framework for information, and
resources in digital
European citizens and the global framework. 1.
environments; digital content
From the scientific approach, digital competence is Information
considered to cover (1) information and the ability to find, organize, process, 1.2 Evaluating data,
and data
assess, store and understand it [19], [20]; (2) the ability to analyze and explain information, and
communicate, collaborate and connect via digital tools [21]; literacy
(3) the ability to create digital content in different formats information; digital content
[22]; (4) the ability to use digital technologies for working, compare, and 1.3 Managing data,
studying and in everyday life [23], [24]; (5) the ability to solve
assess the information, and
problems [22]; (6) ethical aspects related when using
technologies in everyday life [19], [25]; and (7) the ability to reliability and digital content
apply the technologies for personal and professional sources of
achievements [26], [27]. A review by Sánchez-Caballé et al.
(2020) [17] showed that these concepts match the digital information.
competence components of the European DigComp
framework. 2.1 Interacting
through digital
Inheriting research findings of the two aforementioned The ability to use
approaches, one notable example is the European DigComp technologies
technologies
framework with three versions, DigComp 1.0 [28], DigComp
effectively and 2.2 Sharing through
2.0 [29], and DigComp 2.1 [30]. These versions are based on
[18] digital competence concept with 21 competence responsibly to digital technologies
components describing knowledge, skills, and attitudes. They 2.
communicate, 2.3 Engaging in
are categorized according to five areas, including (1) Communicati
Information and data literacy; (2) Communication and connect and citizenship through
collaboration; (3) Digital content creation; (4) Safety; and (5) on and
collaborate in digital technologies
Problem solving with various proficiency levels concerning collaboration
essential learning areas to help citizens adapt to life in the 21st academic settings 2.4 Collaborating
century [31]. and everyday life; through digital
UNESCO (2018) evaluated 47 digital competence express oneself technologies
frameworks of economically diverse countries across
continents and compared them to the European DigComp through digital 2.5 Netiquette
framework. They concluded that all competencies in the 47 media.
2.6 Managing
frameworks can be mapped onto DigComp [13]. DigComp
has been recognized by UNESCO as the most updated digital identity
framework and developed as a foundation for the digital
competence enhancement of people of all ages [29]. It is, The abilities to 3.1 Developing
therefore, suitable to be used to enhance university students’ 3. Digital create and edit digital content
digital competence.
content digital content in 3.2 Integrating and
As a result, the present study employs the DigComp 2.1
creation different formats; re-elaborating
version as a conceptual framework. In this study, the
definition of digital competence refers to the set of knowledge, to understand how digital content
skills, and attitudes students need to have when using ICT and
51
Competence Digital competence Assessment Tool based on DigComp

Descriptions Competences framework
areas
copyright and 3.3 Copyright and Studies on the digital competence measurement using
licenses are applied licenses DigComp as the reference framework was reviewed in the
DigComp in Action report [2] with practices from educational
to digital content; to 3.4 Programming institutions. According to the report, the three common digital
program. competence assessment approaches are:
The abilities to  Self-assessment: individuals are asked to assess their

knowledge, abilities, confidence, and usage. Self-
understand risks assessments tend to use Likert-scale, multiple-choice, and
and threats to true-false questions. The pros of this approach include its
ease and low costs of implementation [32]. However,
physical and mental
individuals are often unable to accurately assess their
health in digital abilities [10].
environments;  Knowledge-based assessment: individuals are assessed
understand safety with questions of factual knowledge and procedural
4.1 Protecting knowledge [2]. Although this approach can give a more
and security
devices accurate picture compared to self-assessment [2], it still
measures to protect fails to accurately measure real-life problems.
4.2 Protecting
personal data and
personal data and
 Performance-based assessment: individuals are presented
privacy; understand with real-life situations such as using browsers, word
privacy processors, and spreadsheets [2]. This approach is more
4. Safety how to use and complex technically and costlier, but can produce the most
4.3 Protecting
share information accurate results and be used for certifying purposes.
health and well- Nowadays, much higher education institutions have
in the protection of
being regarded digital competence as one of the graduate
one’s and other attributes [33], [34] and tend to use this assessment
4.4 Protecting the approach.
people’s privacy;
environment As revealed in the study on the quality of instruments to
be aware of the
measure digital competence in higher education by Saltos-
environmental Rivas et al.(2021) [35], almost 80% of the measurements have
impact of been self-assessments. Besides, the criteria were not
consistent, some measurement instruments excluded some
technologies and
competencies, and many others did not report validity and
how to use digital reliability. This is one of the major drawbacks when it comes
technologies safely to re-using such instruments on a large scale, undermining
overall findings and data accuracy.
and responsibly.
A systematic literature review on digital competence in the
The abilities to higher education context by Zhao et al. (2021) [15] also
identify and solve
pinpointed a similar gap, claiming that self-assessments could
not provide accurate results concerning digital competence
technical problems 5.1 Solving
and suggesting future research to use assessment tools with
when using tools technical problems realistic situations as in performance-based evaluation.
and digital 5.2 Identifying As a result, the present study aims to develop a
performance-based assessment tool targeting university
environments; needs and
students, using DigComp as a reference and inheriting
identify, evaluate technological findings of the ERASMUS+ 2016 project with modifications
5. Problem- to fit Vietnam’s social and higher education contexts.
and use suitable responses
solving
digital technologies 5.3 Creatively using II. METHOD
to solve a task or digital technologies A. Developing students’ digital competence assessment tool
problem creatively 5.4 Identifying The development of a tool to assess students’ digital
to create
competence took into account performance-based instruments
digital competence in the “DigComp in Action” report [2] and other projects,
knowledge; keep gaps including IKANOS (https://ikanos.eus/en/ikanos-model/),
competencies up- TASK (http://www.taskeuproject.com/the-tool/), and
ERASMUS+ 2016 [36], [37]. Using DigComp as a reference,
to-date. the tool presented in this study is a web-based performance
assessment platform called MATPlatform. The MATPlatform
adapted the assessment aimed at office workers of the
52
ERASMUS+ 2016 project and followed the procedures as C. Structure, content, and web-based architecture of the
shown in Figure 1. MATPlatform
C1. Structure of the MATPlatform
The assessment tool’s structure inheriting research
findings of [36], [37] has five evaluations for five competence
areas. Each area is assessed with five questions of knowledge,
skills, and attitudes adapted to suit Vietnam’s grading scale of
10.
 three knowledge questions with three levels: basic (1
point), intermediate (2 points), and advance (3 points)
 one skills question (4 points)
 one attitude question (no point) was used to compare
the reliability of respondents ’ responses and
performance
Figure 1. Developmental process of the digital competence To accurately assess respondents' knowledge and skills,
assessment tool [36]
questions were designed in different types such as multiple-
B. Assessment structure and content of ERASMUS+ 2016 choice, true or false, gap fill after handling technical
project requirements, and simulation questions.
In terms of structure, the ERASMUS+ 2016 project’s Students’ levels of digital competence were based on three
assessment was based on the European digital competence proficiency levels of DigComp, including basic, intermediate,
framework, but after consultations with experts concerning and advance, and developed according to Vietnam’s grading
essential competences, the 21 competences of the DigComp scale with basic (1-3 points), intermediate (4-6 points), and
framework were categorized into four areas, including (1) advance (7-10 points). Levels of competence components are
Essential competences consisting of the most important in percentages as shown in Figure 3.
competences (1.1, 1.2, 1.3, 2.2, 2.6, 4.2), (2) Transversal
competences (2.1, 2.4, 2.5, 3.1, 3.3, 4.1, 4.3, 5.3, 5.4), (3)
Complementary competences (2.3, 3.2, 4.4, 5.2), and (4)
Exclude competences (3.4, 5.1). Competence 3.4 has been
proposed to be excluded from the assessment.
Figure 3. Digital competence levels

Percentages are based on the total scores one particular
Figure 2. Relevance and levels of competences [36] student gets according to competence components and the
Regarding content, the set of assessment questions of the maximum score. If a student gets below 10%, the level of
ERASMUS+ 2016 project was developed with consultations digital competence is then beginner. Percentages from 10% to
of 145 experts, and the relationship between the contents and 30%, from 40-69%, and 70% or higher are equivalent to basic,
the assessment goals was considered carefully by 21 experts intermediate, and advance respectively.
from five European countries and end-users. Besides, this C2. Content of the MATPlatform
assessment has been improved and piloted with end-users
across Europe (n=525). Questions in the ERASMUS+ 2016 Due to differences in cultures, social contexts, and
assessment have been replaced or modified for increased languages, in the process of developing the MATPlatform,
reliability and validity, following feedback from experts and questions in the original assessment tool were modified into
end-users, reliability and validity based on Classical Test propositions to be used in the Vietnamese context. Questions
Theory, Cronbach’s alpha for five tests with different indexes of the MATPlatform (version MAT01) correspond to
like difficulty indexes, discrimination indexes, means, and competence components of Vietnamese students and are
standard deviations. suitable for the social and higher education contexts in
Vietnam.
53
Table 2. The initial bank of questions 100 items, covering five competence areas as in the DigComp
framework, and is available at:
Number of
Competence areas Competences
questions https://nanglucso.hcmute.edu.vn
Information and data 1.1, 1.2, 1.3 15 F. The content validity and reliability of the tool
Literacy (Area 1)
 The content validity
Communication and 2.1, 2.2, 2.3, 30
collaboration (Area 2) 2.4, 2.5, 2.6 The content validity was carried out by eight experts,
Digital content creation 3.1, 3.2, 3.3 15 including four IT experts (E1, E2, E3, E4), three educational
(Area 3) experts (E5, E6, E7), and one expert in quality assurance (E8).
All experts gave positive feedback on the assessment tool. In
Safety (Area 4) 4.1, 4.2, 4.3, 4.4 20 terms of question quality and logic, six out of eight experts
Problem-solving (Area 5) 5.1, 5.2, 5.3, 5.4 20 gave good and very good feedback, whereas experts E1 and
Total number of questions 100 E8 expressed concerns with some questions in the second and
third competence areas.
C3. Web-based architecture of the MATPlatform:
The question contents were recommended to be shortened,
and the quality of illustrative pictures should be enhanced (see
Table 3).
Table 3. Experts’ comments on the MATPlatform

Competence Competences –
Comments
areas Questions
Engaging in The option at the end of
citizenship through the question could be
digital technologies confusing to
Area 2. – Question 3) respondents.
Communicati
Figure 4. Architecture of MATPlatform on and The technology
Collaborating
collaboration mentioned in the
The web-based microservice assessment platform enables through digital
question should be
flexible and simple upgrades or maintenance, depending on technologies –
replaced with a newer,
the number of users. It can thus be used for surveys of different Question 4
more common one.
sizes. Developing digital
Image display error
The backend system using the Spring framework and content – Question 2
connected to the MySQL database through Hibernate JPA Copyright and
facilitates quick, accurate, and reliable interaction with the licenses – Question Options are repeated.
database. Data can be easily exported to formats that can be Area3. Digital
1
content
used by statistical software for quantitative analysis of users’ The illustrative image
creation
digital competence. Copyright and should be consistent
The frontend system using NodeJS (React library), a licenses – Question with the image needed
library maintained by Facebook to work in the backend, helps 5 for copyright check to
render the interface in the backend and reduce server costs, avoid confusion.
hence making the system operate fast and smoothly. This
enables assessments of many users at one time. All comments had been taken into consideration before the
pilot test was carried out with 350 IT students to check the
D. Validity and reliability of the MATPlatform tool’s reliability.
To determine the content validity of an instrument,  The reliability based on Cronbach’s alpha
questionnaires are the most commonly used method for
obtaining information to facilitate the process of expert The tool’s reliability using Cronbach’s alpha was
assessment [38]. Adapting the questions of the ERASMUS+ determined in the pilot study with 350 IT students at
2016 project [36], which were carefully reviewed by 21 HCMUTE, 68% of whom were third-year students or older
European experts, the content validity of this tool was also (see Table 4).
reviewed by eight experts. The expert reviewers are university
lecturers experienced in digital competence. Table 4. Description of the participants
In addition, Cronbach’s alpha reliability was determined First- Second- Third- Fourth-year
based on the pilot study with information technology students year year year onwards
(n=350) at HCMUTE. Number of
27 85 97 141
students
III. RESULT
Percentages 8% 24.% 28% 40%
E. The platform
The performance-based digital competence assessment We received emails from students sharing their challenges
tool for university students in Vietnam has an initial bank of and interests in the assessment. They also shared their
54
opinions about questions that confused them and should be adjust and add more areas and competence indicators [2]. The
modified. All comments from the students were taken into web-based architecture of the platform can also be reused
account to improve the assessment tool’s version MAT02 easily with a replacement digital competence framework.
which was used for another study with students at three
Vietnamese universities, including one autonomous However, concerning the tool’s validity and reliability, the
university, one public university, and one private university. initial item bank for each competence is still limited.
Therefore, the inclusion of more items as well as the
Results showed that IT students have lower intermediate implementation of large-scale tests with a high number of
levels of digital competence proficiency in general. Besides, students could improve the reliability of the tool [36]. Besides,
their proficiency in Digital Content Creation (Area 3) is the more simulation questions to assess students’ skills would be
lowest with just getting a basic level of proficiency (see Table desirable.
5).
V. CONCLUSION
Table 5. Digital competence levels of students
The study developed a digital competence assessment
Area 1 Area 2 Area 3 Area 4 Area 5 platform for university students in Vietnam, utilizing
performance-based assessment based on the DigComp
Area
5.57 5.03 3.51 5.41 5.45
framework. It enables more accurate students’ levels of digital
Mean competence compared to self-assessment questions. By
having consultations with experts and conducting a pilot test
Mean 4.99 with 350 IT students at HCMUTE, the tool’s reliability and
The tool’s reliability based on Cronbach’s alpha is shown validity have initially been confirmed. The tool will be
in Table 6 using the SPSS Statistic version 25. Cronbach’s upgraded with another version to be used at other faculties at
alpha coefficients for each competence area revealed high HCMUTE and other universities. Future research could
consider designing additional questions and diversifying
reliability of the tool (> 0.6) and acceptable corrected item-
simulation questions to better assess students’ skills. The
total correlation (≥ 0.3) as in [39].
platform could also be applied to different educational levels
Table 6. Reliability statistic results with new appropriate question banks.
Reliability
Cronbach’s Alpha Number of items REFERENCES
Statistic
[1] J. Change and P. Huynh, “ASEAN in Transformation - The
MATPlatform 0.951 20 Future of Jobs at Risk of Automation,” Bureau for Employers’
Activities. p. 48, 2016.
Area 1 0.639 3 [2] S. Kluzer and L. Pujol Priego, DigComp into action - Get
inspired, make it happen. A user guide to the European Digital
Area 2 0.894 6 Competence Framework. (JRC Science for Policy Report).
Area 3 0.741 3 Publications Office of the European Union. 10.2760/112945,
2018.
Area 4 0.925 4 [3] L. Pangrazio, Young People’s literacies in the digital age
continuities, conflicts, and contradictions. 2019.
Area 5 0.892 4
[4] S. Thomson and L. De Bortoli, “Preparing Australian students
for the digital world: results from the PISA 2009 Digital
IV. DISCUSSION Reading Literacy Assessment Preparing Australian Students
for the Digital World” no. January. 2012.
According to a systematic mapping study on the quality of
[5] G. M. Cabezas and M. S. Casillas, “Are Future Social
quantitative instruments to measure digital competence in Educators Digital Residents?,” Rev. Electron. Investig. Educ.,
higher education by Saltos-Rivas et al.(2021) [35], most vol. 19, no. 4, pp. 61–72, 2017, doi:
digital competence tools use self-assessments with 10.24320/redie.2017.19.4.1369.
inconsistent criteria or lack validity and reliability tests, which [6] M. Núñez-Canal, M. de las M. de Obesso, and C. A. Pérez-
makes them difficult to be reused or upgraded. Furthermore, Rivero, “New challenges in higher education: A study of the
any survey with a large number of self-assessment questions digital competence of educators in Covid times,” Technol.
Forecast. Soc. Change, vol. 174, no. October 2021, 2022, DOI:
may give respondents boredom, causing responses to be less 10.1016/j.techfore.2021.121270.
accurate. [7] F. Mesároš and P. Mesároš, “Digital competencies in process
As a result, using performance-based assessment and the of creating the knowledge company in the construction sector,”
2010 - 27th International Symposium on Automation and
flexible framework of DigComp and inheriting findings of the Robotics in Construction, ISARC 2010. pp. 544–550, 2010,
ERASMUS+ 2016 project [36], [37] enabled us to save time DOI: 10.22260/isarc2010/0058.
and costs when developing an assessment tool for university [8] J. C. Verhoeven, D. Heerwegh, and K. De Wit, ICT learning
students. Although the amount of time needed to do the experience and research orientation as predictors of ICT skills
assessment test is long, most students in the pilot study and the ICT use of university students, vol. 21, no. 1. 2016.
finished the test and shared positive opinions, bringing joy and [9] W. Ng, “Can we teach digital natives digital literacy?,”
motivation to the research team. Comput. Educ., vol. 59, no. 3, pp. 1065–1078, 2012, DOI:
10.1016/j.compedu.2012.04.016.
The evaluation and verification of validity and reliability [10] K. Aesaert, J. Voogt, E. Kuiper, and J. van Braak, “Accuracy
also helped the assessment tool to be further improved [35]. and bias of ICT self-efficacy: An empirical study into students’
We believe that the platform can be adapted by not only over- and underestimation of their ICT competences,” Comput.
Human Behav., vol. 75, pp. 92–102, 2017, DOI:
universities in Vietnam but also by other educational levels. 10.1016/j.chb.2017.05.010.
The development of the tool using the DigComp framework
with a flexible structure enables educational institutions to
55
[11] L. Johnson, S. Adams Becker, M. Cummins, V. Estrada, A. literacy models,” Ital. J. Sociol. Educ., vol. 9, no. 1, pp. 6–30,
Freeman, and C. Hall, Horizon Report - 2016 Higher Education 2017, DOI: 10.14658/pupj-ijse-2017-1-2.
Edition. 2016. [27] M. Senkbeil and J. M. Ihme, “Motivational factors predicting
[12] A. Çebi and I. Reisoglu, “Digital competence: A study from the ICT literacy: First evidence on the structure of an ICT
perspective of pre-service teachers in Turkey,” Journal of New motivation inventory,” Computers and Education, vol. 108. pp.
Approaches in Educational Research, vol. 9, no. 2. pp. 294– 145–158, 2017, DOI: 10.1016/j.compedu.2017.02.003.
308, 2020, DOI: 10.7821/naer.2020.7.583. [28] A. Ferrari, “DIGCOMP: A Framework for Developing and
[13] X. Jashari, B. Fetaji, A. Nussbaumer, and C. Gütl, “Assessing Understanding Digital Competence in Europe .” 2013, DOI:
Digital Skills and Competencies for Different Groups and 10.2788/52966.
Devising a Conceptual Model to Support Teaching and [29] R. Vuorikari, Y. Punie, S. Carretero, and L. Van Den Brande,
Training,” 2021, pp. 982–995. “DigComp 2.0: The Digital Competence Framework for
[14] P. Gilster, Digital literacy. New York: John Wiley, 1997. Citizens. Update Phase 1: The Conceptual Reference Model,”
[15] Y. Zhao, A. M. Pinto Llorente, and M. C. Sánchez Gómez, EU Commission JRC Technical Reports. pp. 1–40, 2016, DOI:
“Digital competence in higher education research: A 10.2791/11517.
systematic literature review,” Comput. Educ., vol. 168, no. [30] S. Carretero, R. Vuorikari, and Y. Punie, “DigComp 2.1: The
August 2020, p. 104212, Jul. 2021, doi: Digital Competence Framework for Citizens. With eight
10.1016/j.compedu.2021.104212. proficiency levels and examples of use,” Publications Office of
[16] M. Spante, S. S. Hashemi, M. Lundin, and A. Algers, “Digital the European Union. p. 48, 2017, DOI: 10.2760/38842.
competence and digital literacy in higher education research: [31] A.-B. Enochsson, “Teenage pupils’ searching for information
Systematic review of concept use,” Cogent Education, vol. 5, on the Internet,” in Proceedings of ISIC, The Information
no. 1. pp. 1–21, 2018, DOI: 10.1080/2331186X.2018.1519143. Behaviour Conference, Krakow, Poland, 9-11 October: Part 2.
[17] A. Sánchez-Caballé, M. Gisbert-Cervera, and F. Esteve-Mon, Information Research, 2019, pp. 24(1), paper isic1822,
“The digital competence of university students: a systematic [Online]. Available: http://informationr.net/ir/24-
literature review,” Aloma, vol. 38, no. 1, pp. 63–74, 2020. 1/isic2018/isic1822.html (Archived by WebCite® at
http://www.webcitation.org/76lUyYjpm).
[18] A. Ferrari, “Digital Competence in Practice: An Analysis of
Frameworks,” Jt. Res. Cent. Eur. Comm., p. 91, 2012, DOI: [32] International Telecommunication Union (ITU), Digital Skills
10.2791/82116. Toolkit. 2018.
[19] P. Cardoso and N. R. Oliveira, “Scholars’ use of digital tools: [33] Leeds Metropolitan University, “Embedding Digital Literacy
open scholarship and digital literacy,” in INTED2015 as a Graduate Attribute,” 2014, [Online]. Available:
Proceedings, 2015, pp. 5756–5763, [Online]. Available: https://www.leedsbeckett.ac.uk/partners/files/UG_Embedding
http://library.iated.org/view/CARDOSO2015SCH. _Digital_Literacy.pdf.
[20] I. Peña-López, “From Laptops to Competences: Bridging the [34] R. Sharpe, “Digital literacy: from a definition to a graduate
Digital Divide in Education,” RUSC. Universities and attribute to a measure of learning gain,” Queen’s Learn. Teach.
Knowledge Society Journal, vol. 7, no. 1. p. 14, 2010. Conf. 2018 Creat. Innov. Teach., no. June, 2018.
[21] J. B. Son, S. S. Park, and M. Park, “Digital literacy of language [35] R. Saltos-Rivas, P. Novoa-Hernańdez, and R. Serrano
learners in two different contexts,” JALT CALL J., vol. 13, no. Rodrı´guez, “On the quality of quantitative instruments to
2, pp. 77–96, 2017, DOI: 10.29140/jaltcall.v13n2.213. measure digital competence in higher education: A systematic
mapping study,” PLoS One, 2021.
[22] M. Morellato, “Digital Competence in Tourism Education:
Cooperative-experiential Learning,” J. Teach. Travel Tour., [36] J. Bartolomé, P. Garaizar, and X. Larrucea, A Pragmatic
vol. 14, no. 2, pp. 184–209, 2014, DOI: Approach for Evaluating and Accrediting Digital Competence
10.1080/15313220.2014.907959. of Digital Profiles: A Case Study of Entrepreneurs and Remote
Workers, no. 0123456789. Springer Netherlands, 2021.
[23] A. Loureiro, I. Messias, and M. Barbas, “Embracing Web 2.0
& 3.0 Tools to Support Lifelong Learning - Let Learners [37] J. Bartolomé et al., “Developing a Digital Competence
Connect,” Procedia - Soc. Behav. Sci., vol. 46, pp. 532–537, Assessment and Accreditation Platform for Digital Profiles,”
2012, DOI: 10.1016/j.sbspro.2012.05.155. no. November 2018, DOI: 10.21125/inted.2018.0888.
[24] L. Ilomäki, S. Paavola, M. Lakkala, and A. Kantosalo, “Digital [38] A. G. V. Muñoz-Repiso, S. C. Martín, and V. B. Gómez-
competence – an emergent boundary concept for policy and Pablos, “Validation of an indicator model (INCODIES) for
educational research,” Educ. Inf. Technol., vol. 21, no. 3, pp. assessing student digital competence in basic education,”
655–679, 2016, DOI: 10.1007/s10639-014-9346-4. Journal of New Approaches in Educational Research, vol. 9,
no. 1. pp. 110–125, 2020, DOI: 10.7821/naer.2020.1.459.
[25] T. Hallaq, “Evaluating Online Media Literacy in Higher
Education: Validity and Reliability of the Digital Online Media [39] J. Nunnally, Psychometric Theory. New York: McGraw-Hill,
Literacy Assessment (DOMLA),” Journal of Media Literacy 1978.
Education, vol. 8, no. 1. pp. 62–84, 2016, [Online]. Available:
www.jmle.org.
[26] C. Iordache, I. Mariën, and D. Baelden, “Developing digital
skills and competences: A quick-scan analysis of 13 digital
56
The Performance of Geonet Reinforced Straw Rolls

as a Flexible Waterbreaker for Riverbank Protection
Minh-Duc Nguyen Le-Nhat-Huy Nguyen Tran-Phuong-Thao Hua
Faculty of Civil Engineering Office of Project Management and Faculty of Foreign Languages
HCMC University of Technology Land Development HCMC University of Technology
and Education Tan Hong District and Education
Ho Chi Minh City, Vietnam Dong Thap Province, Vietnam Ho Chi Minh City, Vietnam
ducnm@hcmute.edu.vn nguyenlenhathuy993@gmail.com huatranphuongthao@hcmute.edu.vn
Abstract: The paper presents a series of experiments to by water flow, forming a new land bank that continues to
investigate the performance of flexible water breakers using erode according to the above mechanism (Figure 1). This
reinforced straw rolls under the mechanical and biological landslide mechanism is consistent with the riverbank erosion
impact of water in the Mekong River. The water breaker process proposed by [1, 2].
includes either single reinforced straw rolls or triple There are several methods for strengthening riverbanks
(triangular shape) reinforced straw rolls. The straw was dried, against local erosion, including hard protection (including
rolled tightly, then covered by Geonet to enhance the stability reinforced concrete embankment and stony bed) and soft
and duration. A single row or two parallel rows (1m of protection (method using biotechnologies and vegetation).
spacing) of the proposed water breaker were placed along the Soft protection methods based on biotechnology and
riverbank to evaluate their performance against the impact of vegetation are currently preferred due to their environmental
the river wave. The experimental results reveal that the water friendliness, ease of application, and low cost. The use of
breaker using triple-reinforced straw rolls was more durable vegetation has been studied and proven to be effective in
than a single straw roll. Underwater waves, the reinforced protecting canal banks and embankments from erosion [3-6].
straw rolls dissolved over 95% of wave energy and reduced Vetiver grass, for example, is widely used in many countries
up to 6 times the water velocity. Due to the mechanical and and produces positive results [7, 8]. However, these methods
biological effects on the Mekong River, after 20 weeks, the have the disadvantage of requiring a backup period for Vetiver
triple-reinforced straw rolls reduced up to 50% of floating grass to grow (about 2-3 months if the weather is warm and 3-
height compared to that in the first week. The reinforced 4 months if the weather is cold), and they must be protected
straw rolls were a good habitat and environment for and cared for to maintain and develop, particularly in high and
steep talus [9].
vegetation and fishes in the Mekong River.
The straw rolls were proposed as a dike edge erosion
Keywords: geonet, reinforced straw rolls, river control material in the National Engineering Handbook, Part
embankment protection, flexible water breaker, erosion 654, on the design of canal erosion protection [10]. Straw rolls
control, wave energy dissipation have numerous advantages, including locally available
materials, simple and inexpensive construction methods, and
I. INTRODUCTION the ability to protect dikes from immediate erosion. Straw rolls
The Mekong River dike is a multi-purpose project that are temporary biodegradable rolled erosion control products.
protects people's safety, infrastructure, production This group is made up of stems and natural plant fibers (such
development, flood control for silt, fishery, and field as jute, straw, and coir fibers) that are covered in a net to
sanitation. Natural factors such as rain, water evaporation, and increase durability in the aquatic environment. [5, 11, 12]
changes in water levels, combined with the effects of human found that straw rolls covered with nets can protect the dike
activities such as riverbed sand mining and building with a maximum flow velocity of 30-90cm/sec. However, a
construction, have increased the load on riverbanks and few studies have demonstrated the ability of straw rolls to
caused dozens of landslides along the river and dike system in block and dissipate wave energy, preventing edge erosion for
recent years. dikes, particularly in terms of durability, resistance to decay,
and the impact of straw rolls on water and flow environment
One of the causes of local dike erosion is the impact of in river and dike systems. This shortcoming makes it difficult
transverse waves caused by wind perpendicular to the dike, to apply the method of using straw rolls to prevent erosion of
which causes erosion, a cleft, and landslides. The following is the area's riverbanks and dikes.
a description of the erosion process: (1) the embankment is
built up to the design height by compacted clay layers; (2) Experiment with five different straw roll arrangement
during the flood season, water begins to enter the cases to determine the durability, environmental impact of
embankment, and under the influence of wind, waves up to 30 straw rolls, and ability to break waves and dissipate wave
cm high appear, acting perpendicular to the embankment; (3) energy of the linked straw rolls. In addition, the study
the saturated soil layer begins to appear eroded, washed away compares and evaluates the economic and technical criteria
by the impact of waves, and appears cleft; (4) the upper soil for five methods to provide the best method of arranging straw
layer (cleft) has subsided after the clay layer has been deeply rolls to protect riverbanks and embankments. Finally, the
eroded; and (5) the landslide clay is eventually washed away study tested the wave breakability of straw roll under limited

water speed conditions to determine its ability to protect Straw is rolled by machine after harvesting and then dried
riverbanks and embankments from erosion caused by waves with a weight of 14.2 kg, a diameter of 40cm, and a length of
and water currents. 70cm. Straw rolls are reinforced by compacting in geonet
(Figure 2a). After that, the reinforced straw rolls were linked
to form three triangular linked rolls (Figure 2c). The straws
(a) are connected using bamboo stiffeners to make a single roll
row (Figure 2b) and a triple roll row (triangular shape) (Figure
2d).
First, the wave barriers are composed of a single row with
single or triple rolls (Figure 3). Furthermore, straw rolls are
arranged in double rows with a 1-meter gap between them to
increase their ability to dissipate wave energy (Figure 4).
(a)
Wave height Wave height
(b) measurement measurement
Wave length
measurement
Reinforced
straw roll
Water level
(c)
Wave height (b) Wave height

measurement measurement
Triple rolls
(triangular shape)
Wave length
measurement
Water level
(d)
Figure 3. Single row layout: (a) single roll; (b) triple rolls
Wave height Wave height
measurement a) measurement
Figure 1. Riverbank erosion under wave impact (a) Initial bank Wave length
Single roll Single roll measurement
position; (b) Toe erosion; (c) Failure of cantilever overhang; (d)
Water level
Repositioned bank line [1, 2]
1m
0.4m
0.4m
II. MATERIALS AND TEST METHODS

A. Straw roll material Wave height
Wave height measurement
measurement b) Triple rolls
(triangular shape) Wave length
Single roll measurement
0.4m
Water level
0.4m
0.4m
1m
a) b) Wave height
Wave height c)
Triple rolls Triple rolls measurement
measurement
(triangular shape) (triangular shape)
Wave length
measurement
0.4m
0.4m
Water level
0.4m
0.4m
1m
c) d)
Figure 4. 2 rows layout: (a) 2 rows single roll; (b) 2 rows of single
Figure 2. Preparation of straw rolls reinforced by geonet (a) single roll and triple roll; (c) 2 rows triple roll
rolls; (b) a single roll row; (c) triple roll; (d) a triple roll row
58
B. Experimental methods The diameter and length of the straw roll were measured
After being compacted and reinforced using Geonet, the at intervals to determine the durability of the straw roll under
straw rolls were soaked in the river water to evaluate (1) the the impact of water and river flow. The volume of the straw
durability of the straw rolls overtime, (2) e ability to dissipate roll at time t was calculated using its measured diameter and
wave energy, protect the river dike and embankment, and (3) length by assuming the cylindrical shape of the straw rolls.
the biology influence of straw rolls on the river environment The volume loss of straw roll after time t would be
including aquatic animals and plants. calculated as:
The durability of straw rolls was examined by recording V0  Vt
the changes in diameter, length, volume, and floating height Vt  100% (1)
of straw rolls overtime when soaking in the Mekong River. To V0
determine the effects of reinforced straw rolls on breaking
waves and dissipating the wave energy, 20m length of In which V0 = initial volume of the straw roll
different reinforced straw roll rows were placed along Vt = volume of the straw roll at time t
riverbanks to determine their ability to block and prevent the
hydraulic impact of waves causing erosion. The incident wave 100
(the wave that hits the straw) and output wave (the wave that Triple reinforced
comes out of the straw) parameters, including the height and 80 roll
Volume loss, Vt (%)

the length of the river wave, were recorded manually by a
vertical and horizontal ruler arranged as shown in Figures 3 60
and 4.
III. TEST RESULTS 40
C. Straw roll durability in water 20
Geonet reinforced straw rolls were tested for 20 weeks in
some Mekong River tributaries to determine their durability 0
under the impact of river water. Experimental test results show 0 2 4 6 8 10 12 14 16 18 20
that the geonet covering the straw rolls increases its durability, Soaking time (weeks)
resists decay, and keeps the straw rolls for a longer time after
soaking in river water. The floating height was reduced by half Figure 6. Correlation of volume loss with a soaking time of
after the two weeks of soaking for the unreinforced straw roll. reinforced straw rolls
It was submerged underwater level entirely after a month
As shown in Figure 6, the volume loss of reinforced triple
(Figure 5). In contrast, for the first four weeks, the floating
roll was higher than that of reinforced single roll after the same
heights of the reinforced single and reinforced triple roll were
time of soaking. In both cases, after 20 weeks (approximately
kept over 39cm and 10cm, respectively. The reinforced rolls
five months), the straw rolls had lost most of their weight (i.e.,
then decayed and sank gradually overtime. The floating height
75.3 and 87.4% for reinforced single roll and reinforced triple
of the reinforced triple roll decreased by about 10cm after 15
roll, respectively). The decomposition rate of a straw roll is
weeks.
relatively slow for the first six weeks of soaking, in which the
In contrast, that of the single reinforced roll was reduced volume loss of the rolls was less than 10%. After 12 weeks
by about 5cm in the same soaking time period. The reinforced (about three months), over 50% of the volume of the
straw rolls then sank faster; after 20 weeks, the floating height reinforced straw rolls was maintained, which ensures the
for the triple roll was only 20cm (half of the original floating protection of riverbanks and embankments during the three
height) and 3cm for a single roll row. In short, after about five months of peak flood in the Mekong Delta from August to
months (20 weeks), the reinforced triple roll was still capable October every year, buoyancy is over 30cm [13].
of blocking waves up to 20cm in height, whereas a single
D. Wave breaking ability of reinforced straw rolls
reinforced roll was almost completely submerged in water
(Figure 5). After about two weeks of soaking in river water and
reaching a stable floating height, the reinforced straw rolls
50 were tested for their ability to break waves and dissipate wave
energy from artificial waves with a wave height of 10cm to
40 25cm, a cycle of 0.7 to 1 second, and a wavelength of 60cm to
Floating height (cm)
100cm. The wave height and wavelength parameters are

30 determined for incoming and outgoing waves by measuring
Triple reinforced roll
length and height (Figure 7).
20
10
0
0 2 4 6 8 10 12 14 16 18 20
Soaking time (weeks) a) b)
Figure 5. Variation of the floating height of unreinforced and Figure 7. Experimental image of reinforced straw rolls (a) 2 rows
reinforced straw rolls overtime when soaking in the Mekong River triple roll; (b) 1-row single roll.
59
Experiment results show that reinforced straw rolls

reduced the height of waves significantly from 10-25cm to
less than 5cm (Figure 8). The wave height was still
significantly reduced to only 4-5cm for the arrangement of
either 1-row single roll or two rows single roll. Water level
6
Height of wave after straw rolls , cm
3 Figure 9. Model of wave energy and wave crest velocity

1 row single roll Figure 10 shows the obtained results of wave crest water
2 1 row triple roll velocity of the incoming wave and the wave after the
2 rows triple roll reinforced straw rolls. The crest velocity of the incoming wave
1 varied from 40cm/sec to 90cm/sec, which was also the
2 rows single roll
allowable water velocity (1-3 feet per second) for dike
2 rows of single roll and triple roll protection with straw wrapped in nets proposed by [5, 11, 12,
0 14]. The experimental results demonstrated that the reinforced
10 12 14 16 18 20 22 24 26 28 straw rolls significantly reduced the wave crest velocity after
Height of incoming wave, cm
the reinforced straw rolls, with the Uc-out being less than
20cm/s. In addition, the 2 rows layout reduces wave velocity
Figure 8. Correlation between the incoming wave height of the more than the 1-row layout. In which, the arrangement of 2
straw roll and the wave height after the reinforced straw rolls rows triple roll provides the most effective reduction in wave
crest velocity (Uc-out is less than 15cm/sec).
The wave-breaking ability of reinforced straw rolls in 2
rows arrangement is superior to that of a 1-row arrangement. The wave energy dissipation was determined by the wave
The arrangement of 2 rows single roll (using 250 straw energy dissipation level, which is calculated as follows:
rolls/100m long) for better wave resistance with a lower wave
Ein  Eout
height of about 1cm for the case of 1 row triple roll (using 375 E  100% (4)
straw rolls/100m long). Ein
To evaluate the ability to break waves and dissipate In which
energy, the wave speed and wave energy entering and leaving
the straw roll are determined using the calculation model Ein = energy of the incoming wave,
shown in Figure 9. Eout = energy of wave after the reinforced straw rolls
The water velocity at the wave crest, Uc is determined by:
25
H
Uc 
Wave crest velocity after the
reinforce straw rolls, cm/sec
T (2) 20
In which: T = Wave period
H = Height of wave 15
The wave energy is determined by:
10
1 row single roll
gH 2 L
E  Ek  E p  (3) 5
1 row triple roll
8 2 rows single roll
2 rows of single roll and triple roll
In which: E = Total energy of the wave 2 rows triple roll
0
Ek= Momentum 30 50 70 90
Wave crest velocity of incoming wave, cm/sec
Ep= Potential energy
ρ = Specific gravity of water 1 T/m3 Figure 10. Effects of the reinforced straw rolls on reducing the
wave crest velocity
g = Gravity acceleration 9.81 m/s2
L = Wavelength
60
 Geonet significantly improved the durability of reinforced

100
straw rolls under the impact of river water. After the first
Wave energy dissipation, E, %
five weeks, the straw roll was almost undissolved, with

98 stable floating heights of 40cm and 10cm for rows of 3 and
1 roll, respectively. After 20 weeks, the straw roll had
96 disintegrated to 75% of its original volume; the floating
height of the triple roll was 18-20cm, while the single roll
2 rows triple roll was almost completely submerged in water. The
94
2 rows of single roll and triple roll arrangement of triple rolls would be able to ensure the
2 rows single roll protection of riverbanks and dikes during the three months
92 1 row triple roll of flood peaks with a breakwater height (floating height)
1 row single roll of over 30cm.
90
 The reinforced straw rolls reduce the height of the
0 20 40 60 80 100 incoming wave (with a height of 10cm to 25cm) to less
Incoming wave energy, Ein, N.m/m than 5cm. The wave-breaking ability of reinforced straw
rolls in 2 rows arrangement was higher than that of a 1-
Figure 11. Wave energy dissipation of reinforced straw rolls row arrangement. The reinforced straw rolls reduce the
In all five experimental cases, wave energy analysis wave crest water velocity, from Uc-in = 90 cm/s to less than
revealed that the straw coil dissipated more than 95 percent of 20 cm/s, ensuring that riverbank erosion control is
the original wave energy (Figure 11). The difference in wave maintained under the limited velocity condition; finally,
energy dissipation between the two cases is negligible. These the reinforced straw rolls dissipated more than 95 percent
findings indicate that straw rolls are especially effective at of wave energy, thereby protecting canals and dikes from
dissipating wave energy and protecting canals and dikes from local dike erosion.
local dike erosion.  Finally, straw rolls were good shelters for aquatic species
such as snails, fish, shrimp, and crabs. After decomposing,
3.2. The effects of soaking straw rolls on the environment
the straw rolls can be used as fertilizer for crops and fruit
of river water
trees, and the geonet can be recycled to reinforce other
The effect of the decomposed straw on the aquatic straw rolls.
environment was investigated after soaking straw rolls into the The proposed studied results encourage the usage of
river water environment. It was observed that straw rolls did geonet reinforced straw rolls to protect river banks and
not harm aquatic species' habitats and provided good shelters embankments against erosion for riverbanks in the Mekong
for various aquatic species, including snails, fish, shrimp, and Delta.
crabs. Furthermore, after five months of soaking, the sprouted
rice seeds in the straw roll thrived (Figure 12a). After REFERENCES
disintegration, straw rolls could be used as fertilizer for crops
and fruit trees, and the geonet can be recycled to reinforce [1] Thorne, C. and Lewin, J. (1979). Bank Processes, Bed
other straw rolls (Figure 12b). Material Movement and Planform Development in a
Meandering River.
a) b) [2] Johnson, A.W. and Stypula, J.M. (1993). Guidelines for Bank
Stabilization Projects in the Riverine Environments of King
County. Seattle, Wash.: King County Dept. of Public Works,
Surface Management Division.
[3] Greenway, D.R. (1987). Vegetation and slope stability. Slope
stability: geotechnical engineering and geomorphology. p.
187-230.
[4] Coppin, N.J., and Richards, I.G. (2007). Use of Vegetation in
Figure 12. The environmental impact of straw rolls (a) rice grown Civil Engineering CIRIA C708. Construction Industry
on straw rolls; (b) plantation after the straw rolls decomposed Research & Information Association, p. 238.
[5] Gray, D.H. and Sotir, R. (1996). Biotechnical and Soil
IV. CONCLUSION Bioengineering Slope Stabilization: A Practical Guide for
The article investigates the wave breakability of five Erosion Control. John Wiley and Sons, p. 378.
different ways of arranging straw rolls to protect canal banks [6] Morgan, R.P.C. and Rickson, R. (1995). Slope Stabilization
and embankments from erosion. The study also determined and Erosion Control: A Bioengineering Approach. Taylor &
the durability and floating height of the straw roll in the Francis, p. 288.
[7] Grimshaw, R.G. (1995). Vetiver grass - Its use for slope and
Mekong River's water environment. Straw rolls have been
structure stabilization under tropical and semi-tropical
shown in studies to provide erosion protection for canal banks, conditions. Vegetation and slopes. Thomas Telford, London:
embankments, and waves with crest speeds of up to 90cm/s. Institution of Civil Engineers, p. 10.
Straw rolls arranged in 2 rows (i.e., two rows single roll, two [8] Nilaweera, N.S. and Hengchaovanich, D. (1996). Assessment
rows single roll and triple roll, and two rows triple roll) are of strength properties of vetiver grass roots concerning slope
recommended to provide the best protection against erosion of stabilization. in Proceeding of the international conference on
the embankment. In addition, the following conclusions are vetiver: A miracle grass to commemorate the 50th anniversary
proposed: (golden jubilee) celebrations of His Majesty the King of
Thailand's accession to the throne, 87
61
[9] Truong, P., Van, T.T., and Pinners, E. (2007). Technical [13] Tuan, L.A., Du, L.V., Can, T.Q., Ngoc, P.T.B., and Thuong,
guidance on the application of Vetiver grass technology for V.T. (2011). The summarized report on some activities for
natural disaster mitigation and environment protection (in climate change in the river region of An Giang province.
Vietnamese). Building capacity on climate change for civil society
[10] NRCS. (2007). Stream restoration design: Part 654. Natural organizations in Vietnam (in Vietnamese). 2011, Can Tho,
Resources Conservation Service, Washington, DC. December 2011, p. 24.
[11] Norman, J.M. (1975). Design of stable channels with flexible [14] Fischenich, C.J. (2001). Stability thresholds for stream
linings. Hydraulic Engineering Circular. 1975, Federal restoration materials. Engineer Research and Development
Highway Administration, Washington, DC. Hydraulics Center Vicksburg MS Environmental Lab, p. 11.
Branch., p. 152.
[12] Paul, N., Jett, M. (1997). Field Performance Testing of
Selected Erosion Control Products. 1997, Austin, Texas:
Texas Department of Transportation (TxDOT), p. 177.
62
SSD21, Educational Toolbox for Static, Stability,

and Dynamic Analysis of Frame
Truong Thanh Chung Le Nguyen Anh Vu Le Cong Lap Nguyen Thang Xiem
Faculty of Civil Faculty of Civil Faculty of Civil Faculty of Civil
Engineering Engineering Engineering Engineering
Nha Trang University Nha Trang University Nha Trang University Nha Trang University
Nha Trang, Vietnam Nha Trang, Vietnam Nha Trang, Vietnam Nha Trang, Vietnam
chungtt@ntu.edu.vn vulna@ntu.edu.vn laplc@ntu.edu.vn xiemnt@ntu.edu.vn
Abstract: Nowadays, commercial structural analysis Several improvements compared to other programs as
software is both powerful and easy to use. That combination mentioned above are as follows. First, the results of the
comes with the danger of being a “black box” where users toolbox SSD21, such as node displacement, internal forces
provide the input and take the output from the software diagrams, modal analysis, stability analysis are shown to be
without understanding of how it works. On the other hand, very close to professional commercial software, such as
hand-calculation methods currently taught in university such SAP2000 [7]. Second, we introduced the ability to model
as force method and displacement method are completely internal hinges in frame structures, which is important for a
out-of-date since they only allow solving relatively simple wide class of structural analysis. Third, we introduced many
problems. The authors have developed an open-source options in the code to allow users to change the appearances
of the outputs: for example, show/hide node and beam
MATLAB-based educational structural analysis program to
numbering, show/hide numerical value and numerical scale
use as a tool for undergraduate and graduate classes in order bar of internal force diagrams. Finally, a non-linear structural
to encourage students to understand the underlying principles analysis (P-delta effect) was developed for advanced users.
of Finite Element Method and apply for static, stability, and The flowchart of the FEM program SSD21 for static and
dynamic analysis of frame structures. dynamic analysis is shown in Fig. 1.
Keywords: educational toolbox, static analysis, stability In the next section, several examples of structural analysis
analysis, dynamic analysis, finite element method are shown to illustrate the capabilities of SSD21 toolbox,
including static analysis of beam structure, static and stability
I. INTRODUCTION analysis of concentrically braced frame, and modal analysis of
Structural analysis due to the effects of static and dynamic frame structure. It is noted that for the purpose of illustration,
loads is an essential part of civil engineering programs. The the information regarding numerical values of dimensions and
modern approach to teach structural analysis is using load amplitudes was neglected for concision. The final section
programming tools where students learn the computational is the conclusion.
methods, such as the finite element method to analyze internal
forces of structures subjected to static loads, or direct
integration method to analyze vibration of structures subjected
to dynamic loads. Several programs on MATLAB platform
were developed for this purpose, such as CALFEM [1],
NUMA-TF [2], FEM Toolbox for Solid Mechanics [3],
MASTAN2 [4], VisualFEA [5]. The authors have looked
closely into each of those programs, unfortunately we found
that they all have serious drawbacks: CALFEM is lack of
capability to model frame structures with internal hinges;
NUMA-TF is coded using Object Oriented Programming
(OOP) method, which is extremely difficult for students to
understand; FEM Toolbox for Solid Mechanics has a bug that
caused it to produce the wrong internal forces diagrams; and Figure 1. Flowchart of FEM program SSD21 for static and
finally, both MASTAN2 and VisualFEA are closed-source dynamic analysis
code. For that reason, the authors have developed an open-
source educational structural analysis toolbox (SSD21) based II. EXAMPLES
on MATLAB platform [6]. Much effort was made to simplify A. Static analysis of beam structure
the code so that it can be understood and easy to use by
students in classes. The source code is completely free to In this problem, the beam structure is subjected to uniform
download, use, and modify in the following address: force, concentrated force, and concentrated moment as shown
https://github.com/chungntu/SSD21. This paper is served as a in Fig. 2a. Both ends of the middle beam (draw using red
short introduction to SSD21. Students are highly color) are hinges. The uniform force in vertical direction is
recommended to download and review the provided source illustrated as a dot line, with the dot line above the beam
code for better understanding. meaning that the force direction is pointed down. The

concentrated force is illustrated as an arrow, and the hinges at both ends, therefore they are effectively act as truss
concentrated moment is illustrated as an arc-arrow. elements. The loads acting on the structures are uniform forces
on each floor, positive wind loads at the left side of the frame,
The deformed shape of the beam, the shear forces diagram, and negative wind loads at the right side of the frame. For the
and the moment diagram are shown in Fig. 2b, Fig. 2c, and uniform load in vertical direction, a line of triangle is used for
Fig. 2d, respectively. These figures are exported automatically illustration.
as both MATLAB files (.fig) and picture files (.jpeg) and
saved in the same directory of the input text file (.s2k). Fig. The deformed shape is shown in Fig. 4b. The normal
3a, b show the shear forces and the moment diagrams from forces diagram, shear forces diagram, and moment diagram
SAP2000® for comparison. Both figures are numerically are shown in Fig. 4c, Fig. 4d, and Fig. 4e, respectively. Table
identical with corresponding diagrams from SSD21 (Fig. 2c, 2 shows the node displacement of the frame structure. It is
d), except that the shear force diagram by SAP200 is upside noted that for complex structures such as in this example, it is
down. Table 1 shows the node displacement of the beam recommended to use a scale line to show the magnitude of the
structure. displacement and forces rather than showing directly as in
Example A.
(a)
(a)
(b)
(c)
(b)
(d)
Figure 2. Top to bottom. (a) Beam structure subjected to static

loads, (b) Deformed shape, (c) Shear forces diagram, (d) Moment
diagram
(a) (c)
(b)
Figure 3. Top to bottom. (a) Shear forces diagram, (b) Moment

diagram of the beam using SAP2000
(d)
Table 1. Node Displacement of Beam Structure Example A
Node # X-direction [m] Z- direction [m] Rotation [rad]
1 0.00000000 0.00000000 -0.00000023
2 0.00000000 0.00000000 0.00000000
3 0.00000000 -0.00000068 -0.00000034
4 0.00000000 -0.00000158 -0.00000008
5 0.00000000 -0.00000113 0.00000049

(e)
6 0.00000000 0.00000000 0.00000015
7 0.00000000 0.00000000 0.00000004

Figure 4. Top to bottom. (a) Concentrically braced frame subject
B. Static analysis of concentrically braced frame to static loads, (b) Deformed shape, (c) Normal forces diagram, (d)
A concentrically braced frame subjected to static loads is Shear forces diagram, (e) Moment diagram
shown in Fig. 4a. The frame elements in the middle bay have
64
Table 2. Node Displacement of Concentrically Braced Frame in

Example B
Node # X-direction [m] Z- direction [m] Rotation [rad]
1 0.00000000 0.00000000 -0.00012326
2 0.00002691 -0.00000864 -0.00034288 (a)
3 0.00004337 -0.00001294 -0.00046463
4 0.00000000 0.00000000 -0.00001778
5 0.00001930 -0.00001195 0.00001625
6 0.00003568 -0.00001987 0.00006220

(b)
7 0.00000000 0.00000000 -0.00000373
9 0.00002700 -0.00002319 -0.00001955
10 0.00003338 -0.00003319 -0.00010830
11 0.00000000 0.00000000 -0.00049051
12 0.00004117 -0.00000846 0.00037734

(c)
C. Modal analysis of frame structure

In this example, the modal analysis of a plane frame is
carried out (Fig. 5). At each node of the frame, there is a
concentrated mass acting in horizontal direction, which is Figure 6. Top to bottom. (a) Modal mode 1, (b) Modal mode 2, (c)
illustrated by a horizontal arrow. The self-weight of the Modal mode 3
structure is also considered in the modal analysis.
D. Stability analysis of concentrically braced frame
As the result of the modal analysis, Fig. 6a, Fig. 6b, and
Fig. 6c show the modal mode shape 1, modal mode shape 2, In this example, the linear stability analysis (or linear
and modal mode shape 3 of the frame structure, respectively. buckling analysis) of a concentrically braced frame is carried
Table 3 shows the solutions of circular natural frequencies ω out. Fig. 7a shows the frame under the static loads, which is
(rad/s) of the frame structure using SSD21 toolbox compared the same as the frame in Example B. The result of buckling
to SAP2000, which are nearly identical. Though not shown analysis is the critical load ratio, which is the parameter that
here, the mode shapes are able to animated and exportable as multiples to the current load that makes the structure buckle.
move files (.avi), which is very useful for students to visualize Fig. 7b, Fig. 7c, and Fig. 7d show the buckling mode 1 to
how the system behaves in each mode. 3 of the frame structure. It is noted that the results presented
here are linear analysis, which does not consider the P-delta
Table 3. Circular Natural Frequencies of The Frame Structure with effect of the structure. We have developed an advanced
Comparison to SAP2000
version of stability analysis based on non-linear algorithm;
Circular Natural Frequencies ω (rad/s) readers who are interested may refer to the source code [8] for
Mode #
SSD21 SAP2000 Differences (%) further information.
1 56.571488 56.571476 2.1E-05 Table 4 compared the critical load ratios of the SSD21
2 132.999642 132.999571 5.3E-05
compared SAP2000, which shows a difference of less than
2%. The buckling mode shapes are also able to be animated
3 144.036821 144.036770 3.5E-05 and exportable as move files (.avi).
4 157.984938 157.984930 5.1E-06
Table 4. Linear Bulking Analysis Critical Load Ratio Compared to
5 264.952482 264.952481 3.8E-07 SAP2000
6 322.450038 322.450013 7.8E-06 Buckling Critical Load Ratio
Mode # SSD21 SAP2000 Differences (%)
1 336.238797 340.723475 1.31
2 494.847193 500.897242 1.21
3 646.964575 654.695405 1.18
4 869.997952 881.29009 1.28
5 931.286847 931.560272 0.03
6 1246.659921 1266.497852 1.57

Figure 5. Concentrated mass structure
65
presented. We have been using SSD21 in actual teaching at

Nha Trang university and it has shown its effectiveness as a
(a) teaching and learning tool. New ideas are continuously being
added to SSD21 so that the toolbox gets more convenient to
use for both students and lecturers.
ACKNOWLEDGMENT
This study is funded by Nha Trang University, contract
number 36/2021/HĐTR.
(b)
REFERENCES
[1] Ristinmaa, M., Göran S., and Karl-Gunnar O. "CALFEM as a
tool for teaching university mechanics." International Journal
of Innovation in Science and Mathematics Education (2000).
[2] Rangel, Rafael Lopez, and Luiz Fernando Martha. "LESM—
An object‐oriented MATLAB program for structural analysis
of linear element models." Computer Applications in
Engineering Education 27.3 (2019): 553-571.
(c)
[3] Zaicenco, A., et al. "Harmonization of seismic hazard and risk
reduction in countries influenced by vrancea earthquakes."
Programme for Security through Science, NATO, (2004).
[4] Ronald A., William M. "MASTAN2, educational analysis
software for the 21st century." Proceedings of the 6th
International Conference on Computation of Shell and Spatial
Structures, Ithaca, New York, (2008).
[5] Lee, J. Y., and S. Y. Ahn. "Finite element implementation for
(d) computer-aided instruction of structural mechanics."
Proceedings of the 6th International Conference on
Computation of Shell and Spatial Structures, Ithaca, New York,
(2008).
[6] MathWorks. (2022). Matlab overview. [Online]. Available:
https://www.mathworks.com/products/matlab.html
Figure 7. (a) Structure subject to static loads, (b) Bulking mode 1, [7] Computers and Structures, Inc. (2022). SAP2000- Structural
(c) Bulking mode 2, (d) Bulking mode 3 Analysis And Design. [Online]. Available:
https://www.csiamerica.com/products/sap2000
[8] GitHub. (2022). SSD21. Online]. Available:
III. CONCLUSION https://github.com/chungntu/SSD21
In this paper, a review of the educational toolbox for static,
stability, and dynamic analysis of frame structure (SSD21) is
66
Digital Competence of University Students: A

Comparative Study at Three Universities in Vietnam
Quynh Trang Mai Anh Tuan Ngo
Anh Tho Mai
Faculty of Fashion and Tourism Institute of Technical Education
Faculty of Information Technology
Ho Chi Minh City University of Ho Chi Minh City University of
Ho Chi Minh City University of
Technology and Education Technology and Education
Email: mqtrang@hcmute.edu.vn https://orcid.org/0000-0002-2264-
https://orcid.org/0000-0001-9159-
4187
2379
Abstract: Digital competence is one of the eight core the Internet, information and communications technology
competencies required for lifelong learning. It is also a (ICT), multimedia, and artificial intelligence (AI) [2].
prerequisite to one’s ability to study and work in a learning In the context of educational digital transformation,
environment that is experiencing an increasing influence of multiple studies have been conducted to examine essential
ICT and digital media. Previous research on digital human and contextual factors to prepare for successful
competence in higher education settings showed that today’s learning in a digitalized environment.
students have been exposed greatly to technologies but failed
to possess high levels of digital competence as expected. Blayone’s review (2018) of 76 studies, including 25
Therefore, universities need to enhance students’ digital studies published between 2015 and 2017 [3], positioned
competence, and such enhancement is essential and urgent digital competence as one of the crucial factors in educational
digital transformation. The study was followed up by
amid the ongoing digital transformation in education. The
evaluating digital competence frameworks to see whether
present study used mixed methods with an explanatory
digital competence is a readiness key factor in online learning
design. Utilizing the quantitative descriptive method, the or not [4]–[6]. Results showed that digital competence
study first examined and compared the digital competence of represents the most important factor needed for digital
3,467 students from three universities: an autonomous public learning readiness.
university (HCMUTE), a public university (HCMNLU), and
a private university (HUFLIT), adopting digital competence Other studies on student readiness for online learning have
performance assessment tool that is based on the DigComp also been carried out to investigate essential digital
framework. Follow-up semi-structured interviews were also competencies students need for successful learning. Two
conducted for further analysis of survey results. Findings studies by the authors at the University of New South Wales
revealed a lower intermediate level of students' digital (Australia) revealed 58 competencies believed to be essential
for digital learning, 22 of which are related to the use of
competence at all three universities. While HCMUTE
technologies [7], [8].
students reported a basic level with two competencies, the
other universities had ten. Two competencies that all students Rawda Ahmed Omer’s study (2016) at Najran University
performed at the lowest level were (2.6) “Managing digital (Saudi Arabia) which aimed to identify the most important
identity” and (3.3) “Copyright and licenses”. Interviews with practical competencies, also highlighted the importance of
students revealed that students did not receive instructions digital competence [9]. Both studies of [10] and [11]
regarding these two competencies. Besides, different digital confirmed that digital competence is the top factor in
competence among the three universities was found to result maintaining a learning and ensuring students’ learning
from varying levels of access to digital learning management outcomes in the current learning environments.
systems (LMS) and digital tools in the learning process at Possessing digital proficiency enables students to have
these institutions. The study’s findings serve as the better interpretability and understanding of online learning
foundation for higher education institutions to work towards [12], helps them to progress in studies, personal and
specific solutions to increase their students’ digital professional lives, and facilitates practice in online learning
competence. [13], hence being more successful when studying and working
in a learning environment with increased technological
Keywords: digital competence, higher education, integration [14], [15].
student's digital competence, digital transformation in In addition, digital competence is considered a prerequisite
education to studying and working in open, global learning
I. INTRODUCTION environments, and a vital factor to be successful in learning,
research, and future career development [16]. Therefore,
Constant changes in digital technologies have greatly possessing digital competence is regarded as a requirement for
affected all aspects of life and become an essential part of the students in the 21st century [17].
learning environment. Educational institutions have been
making use of digital technologies to transform traditional The issue of digital competence development has thus
learning systems into modern, digitized ones [1] by become a major focus on the agendas of scholars,
establishing connections among technologies of computers, practitioners, and policymakers worldwide to ensure citizens

have full capacities to participate in an increasingly digitized DigComp has been recognized by UNESCO as the most
society [18] and competencies needed for lifelong learning updated after testing 47 frameworks in countries with
[19]. Digital competence development is one of the priorities economic diversity across the globe [31], concluding that all
of education in general [20]. competencies covered by the 47 frameworks reflect DigComp
[32].
In the higher education space, according to a review by
Sánchez-Caballé et al. (2020) [17], most researchers found DigComp was developed to facilitate the establishment of
that although students were born in the digital age, they do not policies to improve citizens’ digital competence and serves as
possess high levels of proficiency as expected. As a result, it a tool to create educational and training initiatives [33].
is essential and urgent that universities help students develop Among the three DigComp versions, DigComp 2.1 [30] is the
digital competence in the 21st century [13]. Universities’ latest and further developed by Institute for Prospective
attempts to enhance students’ digital competence not only Technological Studies (IPTS) from DigComp 2.0 [33] and
contribute to their successful learning but also meet new DigComp [21]. Components and descriptors of DigComp 2.1,
requirements of human resources [10]. However, to be able to based on knowledge, skills, and attitudes, comprise five areas
work on enhancing their digital competence, higher education that form digital competence, including “Information and data
institutions need to, first of all, understand their current levels literacy”, “Communication and collaboration”, “Digital
of proficiency. content creation”, “Safety” and “Problem-solving”.
Therefore, this study aims to survey and compare the Covering essential learning areas to help citizens adapt to
current proficiency levels of students of three universities in lives in the 21st century [34], DigComp is used to enhance the
Vietnam, including one autonomous public university (Ho Chi digital competence of all citizens of all ages [33], thereby
Minh City University of Technology and Education - being suitable for university students.
HCMUTE), one public university (Ho Chi Minh City Nong
Lam University - HCMNLU) and one private university (Ho Types of digital competence assessment using DigComp
Chi Minh City University of Foreign Languages & framework
Information Technology - HUFLIT). Based on survey results, The “DigComp in action” report [35] presented three
the study continued to conduct in-depth interviews with common assessment types with specific measurements as
students to have initial judgments which served as a shown in Table 1.
foundation for proposing solutions to improve students’
digital competence. Table 1. Types of assessment and tools using the DigComp
framework
Digital competence of university students
The concept of digital competence often referred to in Assessment
Types of assessment
discussions on research and policies is regarded as a set of tools/studies
abilities to use technologies to optimize daily lives [21], or Ikanos
“the confident, critical and responsible use of the technologies
(https://ikanos.eus/en/ik
from the society of information for work, entertainment and
education” [19, p9]. Self-assessment: individuals’ anos-
knowledge, abilities, confidence, model/audit/ikanos-
The definitions of digital competence used by researchers and usage are self-assessed. test/)
in the higher education space include (1) information and the Common self-assessment tools Digital competence
ability to find, assess, store and understand it [22], [23]; (2) include Likert-scale, multiple- wheel (https://digital-
the ability to communicate, collaborate and connect via digital choice, and true-false questions competence.eu/)
tools [24]; (3) the ability to create digital content in different [36]. While the implementation Digital skills
formats [25]; (4) the ability to use digital technologies for of this type of assessment is easy Accelerator
working, studying and in everyday life [26], [27]; (5) the and low-cost, the results are not (https://www.digitalskil
ability to solve problems [25]; (6) ethical aspects related when always accurate [37]. lsaccelerator.eu/learnin
using technologies in everyday life [22], [28]; and (7) the
g-portal/online-self-
ability to apply the technologies for personal and professional
achievements [18], [29]. A systematic literature review by assessment-tool/)
Sánchez-Caballé et al. (2020) showed that these concepts Knowledge-based assessment:
match the digital competence components of the European questions of factual knowledge Smartivemap
DigComp framework [17]. and procedural knowledge are (https://www.smartive.c
used in knowledge-based ompany/smartivemap)
In the present study using the DigComp 2.1 framework as
a reference [30], the digital competence definition adopted is assessments [35]. More accurate Skillage
the set of knowledge, skills, and attitudes students need to results can be generated from this (https://all-
have when using ICT and digital tools to solve problems, approach compared to self- digital.org/resources/sk
communicate, manage information, collaborate, create and assessments [35], it is unable to illage-ready-get-hired-
share content, and build knowledge effectively, selectively, measure real-life problems test-ict-skills/)
appropriately, flexibly, creatively, ethically, reflectively for accurately.
learning and research at universities.
Theoretical framework
68
Table 2. The participants

Performance-based assessment:
individuals are asked to handle Type of Number of
No. University
real-life situations including Ikanos – BAIT: the new university participants
using web browsers, word digital competence Ho Chi Minh City
processors, and spreadsheets evaluation system University of public,
[35]. Although this approach is (https://ikanos.eus/en/ik 1 1,336
Technology and autonomous
complex and costlier, it can anos-model/) Education (HCMUTE)
generate the most accurate Task
results among the three types of Ho Chi Minh City Nong
(http://www.taskeuproj public, non-
assessment. Digital competence 2 Lam University 1,021
ect.com/the-tool/) autonomous
(HCMNLU)
has been considered one of the
attributes of graduates by ERASMUS + 2016 Ho Chi Minh City
numerous universities [38], [39], (http://pathwaysforempl University of Foreign
and this assessment approach is oy.eu/) 3 Languages & private 1,100
often used for certifying Information Technology
purposes. (HUFLIT)
Total 3,467
A systematic mapping study by Saltos-Rivas et al. (2021)

Table 3. Number of questions of matplatform
on the quality of quantitative instruments to measure digital
competence in higher education [40] showed that almost 80% Number
of the publications on assessment tools in the higher education Competence
Area Competences of
space adopted the self-assessment type, resulting in less name
questions
accurate data. This makes it hard for such tools to be reused or
1.1 Browsing, searching,
adapted for larger applications, and to present general and and filtering data,
accurate findings. Zhao et al. (2021) [41] shared the same information, and digital
view, suggesting the use of performance-based assessments to content
obtain more accurate results on students’ digital competence. Information
1.2 Evaluating data,
1 and Data 15
As a result, to measure the digital competence of students information, and digital
Literacy
at three Vietnamese universities, the present study made use content
of performance-based assessments developed from DigComp 1.3 Managing data,
and aimed to address two questions: information, and digital
content
1. What are the digital competence levels of students at
three Vietnamese universities? 2.1 Interacting through
digital technologies
2. Are there significant differences in digital 2.2 Sharing through
competence among three Vietnamese universities? digital technologies
2.3 Engaging in
II. METHOD Communicati citizenship through digital
2 on and 30
The present study used mixed methods with an technologies
Collaboration
explanatory design. Quantitative data were collected to have 2.4 Collaborating through
an overview of students’ current digital competence, followed digital technologies
by semi-structured interviews with students to clarify survey 2.5 Netiquette
data. 2.6 Managing digital
In the first phase of the study, data on the digital identity
competence of students from three Vietnamese universities 3.1 Developing digital
(see Table 2) were collected with MATPlatform (version 02). content
This performance-based assessment tool was developed based Digital
3.2 Integrating and re-
on the DigComp framework. 3 content 15
elaborating digital content
creation
The web-based MATPlatform (version 02), inherited 3.3 Copyright and
findings of the ERASMUS+ 2016 [42], [43], is the upgraded licenses
version of the MATPlatform (version 01) whose content 4.1 Protecting devices
validity was tested by eight experts, with four in IT, three in
4.2 Protecting personal
education and one in quality assurance. The first version
data and privacy
underwent a pilot test with the participation of 350 IT students
4 Safety 4.3 Protecting health and 20
at HCMUTE and was found to be valid and reliable
well-being
(Cronbach’s alpha α = 0.951). The survey comprised 100
questions in five competence areas as shown in Table 3. 4.4 Protecting the
environment
69
5.1 Solving technical performed at basic level in seven competencies, including 1.2
problems (3.77), 2.4 (3.56), 2.6 (3.34), 3.1 (3.49), 3.2 (3.56), 3.3 (2.69)
5.2 Identifying needs and and 5.4 (3.57).
Problem- technological responses From Table 5, it can be seen that despite the intermediate
5 20
solving 5.3 Creatively using level of all participating students, there existed significant
digital technologies differences in proficiency at all three universities, with the
5.4 Identifying digital highest recorded at HCMUTE (5.40), compared to HCMNLU
competence gaps (4.16) và HUFLIT (4.10).
Total number of questions 100 Table 4. Descriptive statistical results about students’ digital
Students’ digital competence is evaluated with questions competence
concerning knowledge, skills, and attitudes on a grading scale Competence Areas Competences Mean SD
of 10. More specifically, in each competence, there are five 1.1 5.71 3.33
questions, including three knowledge questions (1 point for 1 1.2 3.77 2.78
basic level, 2 points for intermediate level, and 3 points for 1.3 4.08 3.08
advance level), one skill question (4 points), and one attitude 2.1 6.40 3.52
question (0 points). The attitude question is not graded but 2.2 5.18 3.25
used to test the reliability of the data collected. 2 2.3 5.04 3.14
The tool made use of different question types like 2.4 3.56 2.82
multiple-choice, true or false, gap fill after handling technical 2.5 5.12 3.24
requirements, and simulation questions for accurate 2.6 3.34 2.92
assessments of students’ knowledge and skills. 3.1 3.49 3.15
3 3.2 3.56 2.36
To measure students’ levels of digital competence, the
3.3 2.69 2.85
three proficiency levels of DigComp (basic, intermediate, and
4.1 5.00 3.85
advance) were used but adapted to suit the grading scale in
4.2 6.11 3.88
Vietnam, thereby 1-3 points for the basic level, 4-6 points for 4
4.3 4.48 3.45
the intermediate level, and 7-10 points for the advance level.
4.4 4.87 3.44
Proficiency levels are also put in percentages (see Figure 1),
5.1 6.50 3.98
using students’ scores in a particular competence and the
5.2 5.76 3.85
maximum score. In particular, a percentage below 10% 5
5.3 4.15 3.05
indicates beginner level, 10%-30%, 40-69%, and over 70% 5.4 3.57 2.4
show basic, intermediate, and advanced levels respectively.
Average 4.62 3.22
Below 10% 10%-30% 40%-69% Over 70%
BEGINNER BASIC INTERMEDIATE ADVANCE Table 5. Comparison of digital competence of students at three
universities
Figure 1. Digital competence levels HCMUTE HCMNLU HUFLIT

Universities
(1,336) (1,021) (1,110)
Descriptive analysis of students’ digital competence data Are Comp
was performed in the first phase of the study, using SPSS 25.0. Mean SD Mean SD Mean SD
as etence
In the second phase, follow-up interviews were carried out 1.1 6.12 3.03 5.40 3.48 5.50 3.49
with eight students from different faculties at each university 1.2 4.03 2.64 3.36 2.72 3.83 2.94
1
(n=24), coded from S1-S8 (HCMUTE), S9-S16 (HCMNLU),
and S17-S24 (HUFLIT), and met the suggested sample size 1.3 4.43 3.04 3.71 3.02 3.99 3.15
range of Creswell (2018) [44]. The interview protocol was 2.1 7.31 2.69 5.89 3.77 5.78 3.92
also inspired by DigComp and included questions focusing on 2.2 5.92 2.76 4.75 3.35 4.69 3.53
the competences that students performed at a basic level. 2 2.3 5.79 2.71 4.53 3.20 4.60 3.40
Thirty-minute online interviews were done with each student 2.4 4.14 2.60 3.31 2.91 3.09 2.88
before data were analyzed with the support of MAXQDA.
2.5 6.00 2.72 4.62 3.36 4.51 3.47
III. RESULT 2.6 3.85 2.84 3.11 2.88 2.94 2.96
3.1 4.21 2.99 3.19 3.17 2.89 3.16
In this section of the paper, we present the results obtained
3 3.2 4.35 1.81 3.15 2.51 2.97 2.53
from the quantitative and qualitative analyses carried out. This
study focused on the students’ perception of digital 3.3 3.29 2.80 2.22 2.71 2.38 2.91
competence and competence differences among three 4.1 6.14 3.40 4.39 3.87 4.18 3.99
Vietnamese universities. 4.2 7.41 2.85 5.44 4.13 5.16 4.26
4
4.3 5.41 3.07 3.88 3.46 3.90 3.64
Descriptive Analysis
4.4 5.87 2.84 4.30 3.52 4.19 3.72
First of all, the focus was on assessing students’ digital
5.1 7.68 3.06 5.84 4.24 5.69 4.37
competence in five areas of DigComp (see Table 3 for
5.2 6.82 3.16 5.13 4.03 5.06 4.14
competence areas and competencies). Results obtained in 5
Table 4 indicated the intermediate level of proficiency at all 5.3 4.98 2.66 3.67 3.14 3.59 3.20
universities (4.62). Competence 5.1 was the competence 5.4 4.22 1.94 3.25 2.53 3.07 2.60
students were the most proficient (6.50), but they are the least Average 5.40 2.78 4.16 3.30 4.10 3.41
proficient in competence 3.3 (2.69). Besides, students
70
HCMUTE students’ digital competence was found to be Results showed that there have yet to be specific plans to
upper-intermediate (5.40). While there were two promote digital competence among students at all three
competencies where students performed at the basic level, universities. However, differences in facilities including
competence 2.6 (3.85) and competence 3.3 (3.29), they teaching environments and applying digital tools appeared to
performed well with competencies 2.1 (7.31), 4.2 (7.41), and result in different levels of proficiency at these universities.
5.1 (7.68). At HCMNLU and HUFLIT, students showed their HCMUTE has been implementing the learning management
basic level in up to ten competencies, including 1.2, 1.3, 2.4, system (LMS) since 2014, whereas the LMS of HCMNLU
2.6, 3.1, 3.2, 3.3, 4.3, 5.3, and 5.4, and no competencies at the and HUFLIT was introduced at the beginning of the 2021-
advanced level. Proficiency levels of HCMNLU were slightly 2022 academic year when the Covid-19 caused universities to
higher than those of HUFLIT students. implement online teaching. Moreover, HCMUTE has
continued to diversify its online learning by developing
Regarding proficiency levels by area, in the first Massive Open Online Courses (MOOCs) on the online
competence area, although students did well with competence learning system platform UTEx since 2018, and at the same
1.1 (Browsing, searching, and filtering data, information, and time allowed students to register for such courses as well as
digital content) but did not with competencies 1.2 (Evaluating other MOOCs on popular open online course platforms, with
data, information, and digital content) and 1.3 (Managing self-study time accounting for 70-100% of the duration of one
data, information, and digital content). course. The constant exposure to digital learning
In the second competence area, the basic level of environments and the support of tutorial clips might be
proficiency of students of all three universities was attributed to higher proficiency levels of HCMUTE students
competence 2.6 (Managing digital identity). Such a low level compared to their peers at HCMNLU and HUFLIT (see Table
was also found at competence 2.4 (Collaborating through 7)
digital technologies) at HCMNLU and HUFLIT. Table 7. Digital competence proficiency of students at three
While students of both HCMNLU and HUFLIT reported universities
the basic level in all competencies in the third competence
Universities Basic Intermediate Advance
area, HCMUTE students scored low only at competence 3.3
(Copyright and licenses). HCMUTE 2 15 3
Although HCMUTE students’ performance in the fourth HCMNLU 10 10 0
competence area ranged from intermediate to advance,
HUFLIT 10 10 0
students of HCMNLU and HUFLIT performed at the basic
level at competence 4.3 (Protecting health and well-being). A further look into students’ varying levels of proficiency
As for the last competence area, HCMUTE students in each competence area provided other interesting findings.
continued to demonstrate their intermediate and advanced Area 1
proficiency in all competencies. On the contrary, students at All interviewees found searching for information on the
HCMNLU and HUFLIT had basic proficiency in Internet through the Google search engine familiar. However,
competencies 5.3 (Creatively using digital technologies) and students evaluating and managing search data are mainly done
5.4 (Identifying digital competence gaps). with their own experience and without instructions from their
The MATPlatform tool (version 02) was found to be valid universities.
and reliable (Cronbach’s alpha α = 0.945) (see Table 6). For example, 3 common statements given by all students
Besides, Cronbach’s alpha in every competence area is higher from S1 to S24 include: “I often google, and if I see a result
than 0.7 and there were no areas with a corrected item-total that seems to help me do the assignments, then I will take it as
correlation smaller than 0.4, as suggested by [45]. a reference”, and “I don’t know how to assess the reliability
Table 6. Reliability of matplatform v02 of data”, and “I look information up directly on web browsers
and barely store searchable data.”
Reliability Number of This helps explain why in Area 1, all students performed
Cronbach’s Alpha
Statistic items well at competence 1.1 (Browsing, searching, and filtering
MATPlatform 0.945 20 data, information, and digital content), but badly in the other
version 02 two competencies, 1.2 (Evaluating data, information, and
digital content) and 1.3 (Managing data, information, and
Area 1 0.703 3 digital content).
Area 2 0.875 6 Area 2
Area 3 0.732 3 HCMUTE students said that they have studied in a
blended learning environment, combining face-to-face and
Area 4 0.904 4
LMS, MOOCs developed by HCMUTE itself, or other
Area 5 0.884 4 MOOCs selected by the university to replace certain courses
in the curriculum. Some lecturers have also made use of web
Interview Analysis tools 2.0 in their teaching process.
In-depth interviews with students regarding digital At HCMNLU and HUFLIT, the primary method of
competence development activities for students were learning has been face-to-face, and students only started to get
conducted to explain general differences in students’ themselves familiarized with LMS in the first semester of the
proficiency levels at the three universities as well as 2021-2022 academic year as a result of the pandemic.
differences in specific competencies in one particular area.
71
The diverse digital learning environment and frequent Different levels of proficiency at the three Vietnamese
access to digital tools have caused HCMUTE students to universities indicated that the widespread and diverse use of
perform better at competence 2.1 (Interacting through digital digital learning platforms, and increased access to digital
technologies) and competence 2.4 (Collaborating through environments and tools in higher education settings are the
digital technologies). starting point to enhance students’ digital competence and
Nevertheless, as shared by all students from S1 to S24, help them progress in online learning [13]. The study’s
they had limited or no instructions on managing digital findings are in line with previous studies [55]–[57] showing
identity through courses, for example: “I have never been that blended learning had a positive relationship with students’
taught about managing digital identity in any courses”, and “I improvement of study results in comparison to face-to-face
mainly learn about managing digital identity from the Internet learning. This is indicated by the fact that without specific
by yourself”). This explains the basic level at all universities policies to promote students’ competence so far at HCMUTE,
at competence 2.6 (Managing digital identity). its students have had opportunities to develop some
competencies like 2.1 (Interacting through digital
Area 3 technologies), 4.2 (Protecting personal data and privacy), and
Students at all three universities said that they study in 5.1 (Solving technical problems) due to the presence of good
traditional, lecture-based classes for many courses. Some digital environments, access to positive teaching methods, and
other courses require them to study by working in groups or web tools.
the project-based learning format, and students then submit However, in Area 3 (Digital content creation), which is
reports, presentation slides, website links, or video clips for regarded as one of the fundamental digital competency
final assessments. Such products are accessible on LMS or components [58], [59], only low-intermediate and basic levels
Google Drive so that every student can view and give were recorded at all three universities. Proficiency levels in
feedback. For example, students from S9 to S24 always stated: Area 4 (Safety) were also found at lower-intermediate and
“I am excited about courses where I can participate and create basic, similar to Gallego-Arrufat et al.’s findings (2019) [60].
knowledge with other classmates, but such courses are not This may indicate students’ need for guidance and instructions
many”, and “I enjoy working in groups and sharing from higher education institutions to acquire competencies
knowledge under the guidance of teachers, but such classes essential for studies and future work [10], [13], [61].
are not many”. Early access to the LMS and MOOCs might Universities work towards digital competence enhancement
be the reason for higher proficiency levels of HCMUTE
by developing “a competency model design of the subjects
students at competencies 3.1 (Developing digital content) and with the practical inclusion of activities mediated by
3.2 (Integrating and re-elaborating digital content). technology to guarantee sufficient adequacy between the
Regarding competence 3.3 (Copyright and licenses) where academic world and the labor market” [13, p82]. In addition,
all students performed at the basic level, most students have with competencies reporting the basic level of all students like
heard but have not learned more about or been instructed 2.6 (Managing digital identity) and 3.3 (Copyright and
about them in any courses. (“I have heard about them a lot, but licenses), universities could produce tutorial clips for specific
I am not sure about applying them in learning”, “I often courses in a bid to gradually increase students’ digital
include pictures, diagrams, and videos on the Internet into my competence.
presentation slides, but I haven’t noticed any comments on Initial findings obtained from analysis and comparison of
this by my lecturers.”) three universities representing three types of institutions
Area 4 revealed a sketch of Vietnamese students’ digital competence.
According to most students, they are aware of issues in The use of mixed methods with explanatory design, with
competencies 4.1 (Protecting devices) and 4.2 (Protecting performance-based assessments to increase reliability
personal data and privacy). However, they are not sure about compared to self-assessments and follow-up interviews to
competencies 4.3 (Protecting health and well-being) and 4.4 clarify results, contributed to a clearer picture of the digital
(Protecting the environment), and lecturers do not seem to competence of students at all three universities.
give explicit instructions on such issues.
V. LIMITATIONS AND FUTURE RESEARCH
Area 5
The study examined the overall digital competence of
All students found issues in competencies 5.1 (Solving
students of the three universities rather than conducting in-
technical problems) and 5.2 (Identifying needs and
depth analysis such as correlations between digital
technological responses) familiar but found the concepts of
competence and number of years at university, genders, and
digital competence and competence gap new. They only got
disciplines. Therefore, future studies could consider
to know these concepts when participating in the survey. The
investigating students’ digital competence on a larger scale
late introduction of online teaching and learning could explain
and analyzing correlations of factors affecting their
why students at HCMNLU and HUFLIT showed low
competence development to have a bigger picture and a basis
proficiency in competencies 5.3 (Creatively using digital
to propose practical solutions to enhancing students’
technologies) and 5.4 (Identifying digital competence gaps).
competence. The implementation of strategies to help students
IV. DISCUSSION AND CONCLUSION develop digital competence also needs to be explored in future
research.
The lower-intermediate proficiency level of students at the
three universities supported findings that current students
were born in the digital age [46], [47], had frequent access to
technologies, did not possess advanced digital competence
[48]–[50], and did not have essential competencies required of
learning in the current context of education [51]–[54].
72
REFERENCES [18] C. Iordache, I. Mariën, and D. Baelden, “Developing digital

skills and competences: A quick-scan analysis of 13 digital
[1] S. R. Hiltz and M. Turoff, “The Evolution of Online Learning literacy models,” Ital. J. Sociol. Educ., vol. 9, no. 1, pp. 6–30,
and the Revolution in Higher Education”, October, vol. 48, no. 2017, doi: 10.14658/pupj-ijse-2017-1-2.
10. pp. 59–64, 2005.
[19] European Commission, “Proposal for a COUNCIL
[2] S. Harwell, S. Gunter, S. Montgomery, C. Shelton, and D. RECOMMENDATION on Key Competences for LifeLong
West, “Technology Integration and the Classroom Learning Learning”, 2018, [Online]. Available:
Environment: Research for Action”, Learn. Environ. Res., vol. http://www.theeuropeanlibrary.org/tel4/record/200000467889
4, no. 3, pp. 259–286, 2001, doi: 10.1023/A:1014412120805. 8.
[3] T. Blayone, “Reexamining digital-learning readiness in higher [20] L. Slavova and K. Garov, “Increasing the digital competences
education: Positioning digital competencies as key factors and of students”, Mat. Informatics, vol. 62, no. 1, pp. 42–51, 2019.
a profile application as a readiness tool”, Int. J. E-Learning
Corp. Gov. Heal. High. Educ., vol. 17, no. 4, pp. 425–451, [21] A. Ferrari, “Digital Competence in Practice: An Analysis of
2018. Frameworks”, Jt. Res. Cent. Eur. Comm., p. 91, 2012, doi:
10.2791/82116.
[4] T. Blayone, O. Mykhailenko, R. VanOostveen, O. Grebeshkov,
O. Hrebeshkova, and O. Vostryakov, “Surveying digital [22] P. Cardoso and N. R. Oliveira, “Scholars’ use of digital tools:
competencies of university students and professors in Ukraine open scholarship and digital literacy”, in INTED2015
for fully online collaborative learning”, Technol. Pedagog. Proceedings, 2015, pp. 5756–5763, [Online]. Available:
Educ., vol. 27, no. 3, pp. 279–296, 2018, doi: http://library.iated.org/view/CARDOSO2015SCH.
10.1080/1475939X.2017.1391871. [23] I. Peña-López, “From Laptops to Competences: Bridging the
[5] T. Blayone, O. Mykhailenko, R. VanOostveen, and W. Barber, Digital Divide in Education”, RUSC. Universities and
“Ready for digital learning? A mixed-methods exploration of Knowledge Society Journal, vol. 7, no. 1. p. 14, 2010.
surveyed technology competencies and authentic performance [24] J. B. Son, S. S. Park, and M. Park, “Digital literacy of language
activity”, Educ. Inf. Technol., vol. 23, no. 3, pp. 1377–1402, learners in two different contexts”, JALT CALL J., vol. 13, no.
2018, doi: 10.1007/s10639-017-9662-6. 2, pp. 77–96, 2017, doi: 10.29140/jaltcall.v13n2.213.
[6] T. Blayone, O. Mykhailenko, M. Kavtaradze, M. Kokhan, R. [25] M. Morellato, “Digital Competence in Tourism Education:
VanOostveen, and W. Barber, “Profiling the digital readiness Cooperative-experiential Learning”, J. Teach. Travel Tour.,
of higher education students for transformative online learning vol. 14, no. 2, pp. 184–209, 2014, doi:
in the post-soviet nations of Georgia and Ukraine”, Int. J. Educ. 10.1080/15313220.2014.907959.
Technol. High. Educ., vol. 15, no. 1, 2018, doi: [26] A. Loureiro, I. Messias, and M. Barbas, “Embracing Web 2.0
10.1186/s41239-018-0119-9. & 3.0 Tools to Support Lifelong Learning - Let Learners
[7] M. Parkes, S. Stein, and C. Reading, “Student preparedness for Connect”, Procedia - Soc. Behav. Sci., vol. 46, pp. 532–537,
university e-learning environments,” Internet High. Educ., vol. 2012, doi: 10.1016/j.sbspro.2012.05.155.
25, pp. 1–10, 2015, doi: 10.1016/j.iheduc.2014.10.002. [27] L. Ilomäki, S. Paavola, M. Lakkala, and A. Kantosalo, “Digital
[8] M. Parkes, C. Reading, and S. Stein, “The competencies competence – an emergent boundary concept for policy and
required for effective performance in a university e-learning educational research”, Educ. Inf. Technol., vol. 21, no. 3, pp.
environment”, Australas. J. Educ. Technol., vol. 29, no. 6, pp. 655–679, 2016, doi: 10.1007/s10639-014-9346-4.
771–791, 2013, doi: 10.14742/ajet.38. [28] T. Hallaq, “Evaluating Online Media Literacy in Higher
[9] S. G. A. Rawda Ahmed Omer, “E-Learning Competencies Education: Validity and Reliability of the Digital Online Media
Practice Level among Faculty Members at Najran University”, Literacy Assessment (DOMLA)”, Journal of Media Literacy
SJETR Journal., vol. Vol. 2, No, no. 1. 2016. Education, vol. 8, no. 1. pp. 62–84, 2016, [Online]. Available:
[10] M. Florence, S. Brandy, and F. Claudia, “Examining student www.jmle.org.
perception of readiness for online learning: Importance and [29] M. Senkbeil and J. M. Ihme, “Motivational factors predicting
confidence”, Online Learn. J., vol. 24, no. 2, pp. 38–58, 2020, ICT literacy: First evidence on the structure of an ICT
doi: 10.24059/olj.v24i2.2053. motivation inventory”, Computers and Education, vol. 108. pp.
[11] T. Yu, “Examining construct validity of the student online 145–158, 2017, doi: 10.1016/j.compedu.2017.02.003.
learning readiness (SOLR) instrument using confirmatory [30] S. Carretero, R. Vuorikari, and Y. Punie, “DigComp 2.1: The
factor analysis”, Online Learning Journal, vol. 22, no. 4. pp. Digital Competence Framework for Citizens. With eight
277–288, 2018, doi: 10.24059/olj.v22i4.1297. proficiency levels and examples of use”, Publications Office of
[12] A. A. Mosa, M. Naz’ri bin Mahrin, and R. Ibrrahim, the European Union. p. 48, 2017, doi: 10.2760/38842.
“Technological Aspects of E-Learning Readiness in Higher [31] UNESCO, “A Global Framework of Reference on Digital
Education: A Review of the Literature”, Computer and Literacy,” Information Paper, vol. 51, no. 51. pp. 1–146, 2018.
Information Science, vol. 9, no. 1. p. 113, 2016, doi: [32] X. Jashari, B. Fetaji, A. Nussbaumer, and C. Gütl, “Assessing
10.5539/cis.v9n1p113. Digital Skills and Competencies for Different Groups and
[13] E. López-Meneses, F. M. Sirignano, E. Vázquez-Cano, and J. Devising a Conceptual Model to Support Teaching and
M. Ramírez-Hurtado, “University students’ digital competence Training”, 2021, pp. 982–995.
in three areas of the DigCom 2.1 model: A comparative study [33] R. Vuorikari, Y. Punie, S. Carretero, and L. Van Den Brande,
at three European universities”, Australas. J. Educ. Technol., “DigComp 2.0: The Digital Competence Framework for
vol. 36, no. 3, pp. 69–88, 2020, doi: 10.14742/AJET.5583. Citizens. Update Phase 1: The Conceptual Reference Model”,
[14] N. Bergdahl, J. Nouri, and U. Fors, “Disengagement, EU Commission JRC Technical Reports. pp. 1–40, 2016, doi:
engagement and digital skills in technology-enhanced 10.2791/11517.
learning”, Educ. Inf. Technol., vol. 25, no. 2, pp. 957–983, [34] A.-B. Enochsson, “Teenage pupils’ searching for information
2020, doi: 10.1007/s10639-019-09998-w. on the Internet”, in Proceedings of ISIC, The Information
[15] T. He, C. Zhu, and F. Questier, “Predicting digital informal Behaviour Conference, Krakow, Poland, 9-11 October: Part 2.
learning: an empirical study among Chinese University Information Research, 2019, pp. 24(1), paper isic1822,
students”, Asia Pacific Educ. Rev., vol. 19, no. 1, pp. 79–90, [Online]. Available: http://informationr.net/ir/24-
2018, doi: 10.1007/s12564-018-9517-x. 1/isic2018/isic1822.html (Archived by WebCite® at
[16] C. Killen, “Collaboration and coaching: powerful strategies for http://www.webcitation.org/76lUyYjpm).
developing digital capabilities,” in Digital Literacy Unpacked, [35] S. Kluzer and L. Pujol Priego, DigComp into action - Get
Facet Publishing, 2018. inspired, make it happen. A user guide to the European Digital
[17] A. Sánchez-Caballé, M. Gisbert-Cervera, and F. Esteve-Mon, Competence Framework. (JRC Science for Policy Report).
“The digital competence of university students: a systematic Publications Office of the European Union. 10.2760/112945,
literature review,” Aloma, vol. 38, no. 1, pp. 63–74, 2020. 2018.
73
[36] International Telecommunication Union (ITU), Digital Skills pre-service education teacher: Face-to-face vs blended
Toolkit. 2018. education and gender”, J. E-Learning Knowl. Soc., vol. 16, no.
[37] E. Litt, “Measuring users’ internet skills: A review of past 3, pp. 1–9, 2020, doi: 10.20368/1971-8829/1135214.
assessments and a look toward the future”, New Media Soc., [51] J. C. Verhoeven, D. Heerwegh, and K. De Wit, ICT learning
vol. 15, no. 4, pp. 612–630, 2013, doi: experience and research orientation as predictors of ICT skills
10.1177/1461444813475424. and the ICT use of university students, vol. 21, no. 1. 2016.
[38] Leeds Metropolitan University, “Embedding Digital Literacy [52] W. Ng, “Can we teach digital natives digital literacy?”,
as a Graduate Attribute”, 2014, [Online]. Available: Comput. Educ., vol. 59, no. 3, pp. 1065–1078, 2012, doi:
https://www.leedsbeckett.ac.uk/partners/files/UG_Embedding 10.1016/j.compedu.2012.04.016.
_Digital_Literacy.pdf. [53] K. Aesaert, J. Voogt, E. Kuiper, and J. van Braak, “Accuracy
[39] R. Sharpe, “Digital literacy: from a definition to a graduate and bias of ICT self-efficacy: An empirical study into students’
attribute to a measure of learning gain”, Queen’s Learn. Teach. over- and underestimation of their ICT competences”, Comput.
Conf. 2018 Creat. Innov. Teach., no. June, 2018. Human Behav., vol. 75, pp. 92–102, 2017, doi:
[40] R. Saltos-Rivas, P. Novoa-Hernańdez, and R. Serrano 10.1016/j.chb.2017.05.010.
Rodrı´guez, “On the quality of quantitative instruments to [54] L. Johnson, S. Adams Becker, M. Cummins, V. Estrada, A.
measure digital competence in higher education: A systematic Freeman, and C. Hall, Horizon Report - 2016 Higher Education
mapping study”, PLoS One, 2021. Edition. 2016.
[41] Y. Zhao, M. C. Sánchez Gómez, A. M. Pinto Llorente, and L. [55] A. Y. Al-Qahtani Awadh and S. Higgins, “Effects of
Zhao, “Digital competence in higher education: Students’ traditional, blended and e-learning on students’ achievement in
perception and personal factors”, Sustain., vol. 13, no. 21, pp. higher education”, J. Comput. Assist. Learn., vol. 29, no. 3, pp.
1–17, 2021, doi: 10.3390/su132112184. 220–234, 2013, doi: Al-Qahtani, A. A. Y., & Higgins, S. E.
[42] J. Bartolomé, P. Garaizar, and X. Larrucea, A Pragmatic (2012). Effects of traditional, blended and e-learning on
Approach for Evaluating and Accrediting Digital Competence students’ achievement in higher education. Journal of
of Digital Profiles : A Case Study of Entrepreneurs and Remote Computer Assisted Learning, 29(3), 220–234.
Workers, no. 0123456789. Springer Netherlands, 2021. doi:10.1111/j.1365-2729.2012.00490.x.
[43] J. Bartolomé et al., “Developing a Digital Competence [56] J. S. Lewis and M. A. Harrison, “Online delivery as a course
Assessment and Accreditation Platform for Digital Profiles”, adjunct promotes active learning and student success”, Teach.
no. November, 2018, doi: 10.21125/inted.2018.0888. Psychol., vol. 39, no. 1, pp. 72–76, 2012.
[44] J. Creswell and V. L. Plano Clark, “Designing and conducting [57] M. A. Harjoto, “Blended versus face-to-face: Evidence from a
mixed methods research”, 2018. graduate corporate finance class”, J. Educ. Bus., vol. 93, no. 3,
pp. 129–137, 2017.
[45] J. Nunnally, Psychometric Theory. New York: McGraw-Hill,
1978. [58] J. Gil-Flores, J. Rodríguez-Santero, and J. J. Torres-Gordillo,
“Factors that explain the use of ICT in secondary-education
[46] B. J. Bowe and D. Y. Wohn, “Are There Generational classrooms: The role of teacher characteristics and school
Differences? Social Media Use and Perceived Shared Reality”, infrastructure”, Comput. Human Behav., vol. 68, pp. 441–449,
in ACM International Conference Proceeding Series, 2015, vol. 2017.
2015-July, doi: 10.1145/2789187.2789200.
[59] J. Tondeur, K. Aesaert, B. Pynoo, J. van Braak, N. Fraeyman,
[47] P. Thompson, “The digital natives as learners: Technology use and O. Erstad, “Developing a validated instrument to measure
patterns and approaches to learning”, Computers and preservice teachers’ ICT competencies: Meeting the demands
Education, vol. 65. pp. 12–33, 2013, doi: of the 21st century”, Br. J. Educ. Technol., vol. 48, no. 2, pp.
10.1016/j.compedu.2012.12.022. 462–472, 2017, doi: 10.1111/bjet.12380.
[48] G. M. Cabezas and M. S. Casillas, “Are Future Social [60] M. J. Gallego-Arrufat, N. Torres-Hernández, and T. Pessoa,
Educators Digital Residents?”, Rev. Electron. Investig. Educ., “Competence of future teachers in the digital security area”,
vol. 19, no. 4, pp. 61–72, 2017, doi: Comunicar, vol. 27, no. 61, pp. 53–62, 2019, doi: 10.3916/C61-
10.24320/redie.2017.19.4.1369. 2019-05.
[49] F. Mesároš and P. Mesároš, “Digital competencies in process [61] M. Pérez-Mateo, M. Romero, and T. Romeu-Fontanillas,
of creating the knowledge company in construction sector”, “Collaborative construction of a project as a methodology for
2010 - 27th International Symposium on Automation and acquiring digital competences”, Comunicar, vol. 21, no. 42, pp.
Robotics in Construction, ISARC 2010. pp. 544–550, 2010, 15–23, 2014, doi: 10.3916/C42-2014-01.
doi: 10.22260/isarc2010/0058.
[50] F. D. Guillen-Gamez, M. J. Mayorga-Fernández, and M. T. Del
Moral, “Comparative research in the digital competence of the
74
Load Capacity Evaluation of Simple Reinforced

Concrete Girder Bridges with Considering the
Corrosion of Reinforcement and Concrete
Tran The Truyen Tran Thu Minh Nguyen Dac Duc
Department of Bridge and Tunnel Department of Bridge and Tunnel Department of Bridge and Tunnel
Engineering Engineering Engineering
University of Transport and University of Transport and University of Transport and
Communications Communications Communications
Hanoi, Viet Nam Hanoi, Viet Nam Hanoi, Viet Nam
tranthetruyen@utc.edu.vn minhtran250999@gmail.com ngdacduc@utc.edu.vn
Tran Duc Manh Nguyen Quoc Cuong

Department of Bridge and Tunnel Engineering Department of Civil Engineering
University of Transport and Communications Vinh University
Hanoi, Viet Nam Vinh, Viet Nam
manhbangs@gmail.com nqcuongbkdn@gmail.com
Abstract: This paper presents the results of the calculation it is necessary to clearly understand the mechanical properties
and evaluation of the resistance of simple span reinforced and critical load capacity of reinforced concrete structures.
concrete girder bridge structure taking into account corrosion Thereby contributing to the maintenance of the bridge in the
of reinforcement and reduction of concrete strength. The future.
degree of material corrosion is considered of the reduction Several studies have determined that reinforcement
diameter of reinforcement, changes in the bonding strength, corrosion is the most common cause affecting the resistance
and reduction of concrete strength. Calculation results allow of reinforced concrete structures. The bearing capacity of the
diagnosing the exploitability of simple span reinforced reinforced concrete bridge exposed to the environment,
concrete girder bridge structure. especially the marine environment, is reduced in load. The
influence of corrosive oxides, causes expansion due to tensile
Keywords: reinforcement, corrosion, bridge girder, stress acting on the surrounding concrete layers and thus leads
concrete, resistance to the appearance of cracks in the concrete because of its poor
bearing capacity. This is a matter of concern because cracks
I. INTRODUCTION can lead to the entry of oxygen or chloride ions that corrode
In practical construction, reinforced concrete beam the reinforcement. A reduction in the effective cross-sectional
modules with simple spans are commonly used. The area of the reinforcement is also inevitable because iron is
phenomenon of cracked concrete beams is inevitable. The consumed during oxidation and this can lead to a significant
bearing capacity of concrete beams is gradually reduced over reduction in the load-carrying capacity. After a period of use,
time under the corrosion of internal reinforcement, and at the the corrosion of reinforcement due to the carbonation of
same time gradually reduces the bearing cross-section and has concrete or the penetration of chloride ions often occurs in
serious consequences such as damage to the structure or the reinforced concrete structures or structures [1–3]. Corrosion
service life of the beam. construction. Because of these serious of reinforcement will reduce the bearing capacity of the
consequences, when we discover the phenomenon of cracked reinforced concrete structure.
concrete beams, we need to find out carefully about the Repeated load and environmental impacts will cause
resistance and the causes of the cracks to find solutions for material failure in reinforced concrete structures. Fatigue
timely and effective recovery. damage and corrosion of reinforcement are the main causes
Bridges are considered an important element in the leading to the decrease in load capacity of reinforced concrete
transport infrastructure system, which should be periodically structures [4,5]. The fatigue failure mechanism of corroded
inspected and monitored to ensure their safety for exploitation reinforced concrete structures is more complex than that of
and use. Not only that, in recent years, the demand for non-corroded reinforced concrete. The cause is that the
prolonging the service life and increasing additional load for deterioration of the reinforced concrete structure caused by the
bridge works has raised concerns about the durability and effects of corrosion and fatigue is not a simple superposition
structural evaluation of the existing infrastructure. It can be of these two failure factors, but a process of combining their
clearly seen that the deterioration of bridge structures can mutual influence. 6]. Therefore, reinforcement corrosion
affect the load-carrying capacity and serviceability. Therefore, poses a significant threat to the usability and serviceability of

reinforced concrete structures in general and bridges in maintenance cost when affecting the load capacity of the
particular [7]. structure. Nogueira et al. [14] performed a study on
determining the onset of corrosion using a probabilistic
Many researchers report that the corrosion mechanism of method by combining Fick's law and reliability theory.
reinforcement can potentially lead to reduced structural Research results show that the influence of uncertainty related
strength and performance of critical bridge components such to concrete coating depth and water-cement ratio on the failure
as decks, piers, and load-bearing members, depending on the probability of the structure has been analyzed. Nogueira and
degree of severity of environmental exposure (Akiyama, Leonel [15] proposed a simplified method to determine the
Frangopol, & Matsuzaki, 2011; Moser et al., 2009; Silano & optimal curing time as a function of the water-cement ratio,
Brinckerhoff, 1993; Strauss, Wendner, Bergmeister, & Costa, coating depth and the aggressiveness of the medium. The
2013). Concrete bridges that meet design requirements for study proposed a theoretical basis for determining the optimal
residential traffic may over time deteriorate as they gradually values of the thickness of the concrete coating according to the
degrade due to chloride contamination. Frangopol, Dong, and predetermined curing time to ensure the durability of the
Sabatino (2017) reported that the condition of reinforced structure against the effects of corrosion.
concrete bridge structures is deteriorating due to multiple
degradation mechanisms including aging, environmental With the rapid development of society and traffic demand,
hazards (e.g. airborne chloride) and mechanical hazards (eg it is necessary to calculate and predict the load capacity of the
traffic loads). Many existing reinforced concrete bridges near bridge in the process of operation. In this research paper, we
coastal areas have had corrosion reinforcement showing will focus on assessing the resistance of reinforced concrete
undesirable durability (Nakamura, Watanabe, Koga, girders for rural bridges using reinforced concrete beams,
Nakamura, & Ikawa, 2008). taking into account the influence of reinforcement corrosion.
The result obtained is the resistance of the beam before and
Corrosion of reinforcement in reinforced concrete after corrosion. From there, it is possible to draw general
structures due to chloride penetration begins when the conclusions to predict the health of the bridge structure to
chloride concentration at the concrete/reinforced surface serve to ensure safety in the operation and maintenance of the
reaches a threshold value. At this point, the amount of bridge during different periods.
hydroxyl entering the concrete pores exceeds the chemical
threshold, causing the separation of the protective chemical II. MECHANICAL BEHAVIOR OF CORRODED
layer around the reinforcement [8]. The chloride concentration SIMPLE RC GIRDER BRIDGES
at the junction increases with time and reaches a threshold
value due to the transport of ions from the outer surface of the Factors affected by reinforcement corrosion: load
structure to the interior. The penetration of chloride into capacity, stiffness and force distribution are considered as
concrete pores is a complex phenomenon regulated by mechanical properties of reinforced concrete structures in
complex physical and chemical mechanisms. However, the general and concrete simple girder bridge structures.
representation of this phenomenon can be simplified by reinforced concrete in particular. Corrosion of the
considering only the diffusion of chloride ions along with the reinforcement leads to a decrease in the cross-sectional area
concrete cover. of the reinforcements, resulting in reduced shear strength and
bending moment, as well as a decrease in the stiffness of the
Currently, several models have been built [9], Samson et structure. Changing the ductility of steel directly affects the
al. present mathematical models to represent the motion of distribution of forces and moments and limits the load-
particles in saturated concrete. Samson and Marchand [10] carrying capacity of the structure and can also severely
studied the effect of temperature on chloride transport into reduce the ability of the structure to withstand seismic loads.
concrete pores and emphasized the effect of temperature on Not only that, the corrosion causing volume expansion of the
chloride ion diffusion causing corrosion in the structure. reinforcement can cause the surrounding concrete to crack
Bastidas-Arteaga et al. [11] presented a study on the and peel off, also in terms of the compressive strength of a
mechanical degradation processes due to corrosion of concrete element, the peeling of the outer concrete layer.
reinforcement, concrete cracking and biological damage, as reduces the inner swing arm, thereby reducing the bending
well as their combined effects on the strength reduction. capacity.
degree of reinforced concrete members over time. Mechanical
degradation and biological damage in reinforced concrete In addition, the concrete layer around the reinforcement
structures significantly affect the structure, causing and the belt reinforcement is corroded, affecting the
deterioration in strength over time and should be included as anchorage and shear strength of the beam. If the concrete in
part of the planning and design of the structure, especially, this area has cracked due to corrosion then it has reached its
when the works are affected by the environment prone to maximum tensile strength. Therefore, any mechanical load
erosion. Zhao et al. [12] presented an analysis of the causing tensile stress will act to make the knots larger.
mechanical degradation in the concrete shell due to corrosion Cracked concrete not only affects the shear strength and bond
in the non-cracked, initial, and later stages of the partial between concrete and reinforcement but also reduces the
cracking. Cracks appear on different structures along the load-carrying capacity of the structure in the long term due to
radial direction of the concrete coating causing expansion less protection of reinforcement and a further aggressive
stress due to corrosion of reinforcement and steel loss when environment. direct access to the reinforcement.
cracking the concrete surface. Frangopol et al. [13], proposed Furthermore, cracks, depending on their orientation, can also
a model to represent the mechanical behaviour of reinforced change the stiffness, thereby altering the force distribution in
concrete beams as a function of time to determine the decrease the structure and ultimately the deterioration of the bridge
in the diameter of reinforcement. The proposed design method structure's health.
shows that corrosion rate affects both material cost and
76
Corrosion of steel
reinforcement
Pitting Generalised
Corrosion corrosion
Corrosion
mechanism
Pitting Generalised
Corrosion corrosion
Mechanical Bond concrete / Local

Cover cracking
properties steel Effects
Moment Shear Anchorage Global

capacity capacity capacity Effects
Load-carrying Redistribution / Structural

capacity redundancy Effects
Figure 1. Effects of corrosion on load-carrying capacity, stiffness
III. EVALUATION OF LOAD CAPACITY FOR In order to save the computational time and volume of
results files, only half of the beam was modeled for other
REINFORCED CONCRETE BRIDGE GIRDER USING
beams due to the symmetry in geometry and loading
NUMERICAL MODELING METHOD
conditions.
A. Model description
In this research, ATENA 3D software is applied for load
capacity analysis with reinforced concrete girder. From these Steel plate
models, it is possible to know the time of occurrence, and the
development of cracks in terms of size relating to a load
resistance, and thereby serve for bridge maintenance.
Figure 2. General view of reinforcement arrangement of T

beam with 15 meters length
T-beams are commonly used in Vietnam with bridge

constructions, especially with small and medium-sized
bridges, and rural traffic bridges. This study calculated and
analyzed a sample of T-beam with length L = 15 m; height H
= 1 meter. Materials used: concrete with a strength of 28.5 Figure 3. Model of 15 meters T-beam and reinforcement
MPa, reinforcement with a yield strength of 560 MPa. modeling
77
B. Element category recommended for concrete. This parameter set is generated

Several element types are used to simulate the FE models. from code and suggestions. This setting is recommended
Such as brick/tetra elements are used to simulate the concrete, unless there is a reasonable and compelling reason to change
steel plate, support, and bearing plate. it. In the current Atena, Eurocode 2 compliant concrete 3D
data is available.
In compliance with the concrete regulations of Eurocode
2, concrete with a strength of 28.5 MPa has been chosen with
the B-liner type of the software with available parameters.
E. Bond – slip modeling
The bond model is selected, and additional parameters
Figure 4. Elements type modeling related to the bond can be entered (bond material,
circumference, fixed end slip). The bar circumference is
C. Material constitutive calculated automatically at the time of area calculation
following CEB – FIP model code 1990.
The mass loss of the corroded reinforcing bar is simulated
by decreasing the sectional area in the FE model. The tensile
capacity curve of the corroded steel bar was obtained from
the material test after the test finished. The density of the steel
bar is 7.8×10−2 (MN/m3) and the coefficient of thermal
expansion is 1.2×10−5 (1/K). At the same time, the yield
strains in ATENA are obtained by all the plastic stains by
subtracting the peak elastic strain value. Figure 4 presents the
stress-strain law of all steel bars in ATENA.
Figure 7. Bond strees – slip diagram in CEB – FIP model

code 1990
F. Results and recommendations

Then, a numerical study was performed to evaluate the
strength of the main reinforcement at the 25% corrosion level.
It is assumed that the diameter of the reinforcement is reduced
on a cross-section with a diameter of 24 mm compared to the
uncorroded case with a diameter of 32 mm.
Figure 5. Stress-strain law for reinforcement in ATENA
D. Concrete type
The equivalent uniaxial stress-strain relationship of
concrete is needed in the FE model. The concrete stress-strain
model is for the design of concrete.
Figure 8. Result of load - displacement relationship graph

of two calculated cases
Compared to the beam with 0% reinforcement corrosion

degree, the load capacity of the beam reduces to 7%. From
numerical analysis, a linear relationship between the decrease
in the capacity of this T beam and the degree of reinforcement
Figure 6. Stress-strain law for concrete
corrosion was also detected. This simulation only considers
the reduction in diameter of the reinforcement, excluding the
By definition, the material type can be selected from a list effects of the environment and other factors.
of build parameters. A set of default parameters is
automatically generated. Select the fracture plastic material
3C Nonlinear Cementitious 2 as an example. This material is
78
temperature on ionic transport in cementitious materials.

Cement and Concrete Research, v. 37, 2007, p. 455-468.
[7] Sarja and Vesikari, Durability Design of Concrete Structures,
Architecture and Civil Engineering 7, 1996.
[8] Val, d. V.; Stewart, M. G. Life-cycle cost analysis of reinforced
concrete structures in marine environments. Structural Safety,
v. 25, 2003, p. 343-362.
[9] Samson, E.; Marchand, J.; Snyder, K. A. Calculation of ionic
diffusion coefficients on the basis of migration test results.
Materials and Structures, v. 36, 2003, p. 156-165.
Figure 9. Cracks distribution in the outer surface of T-beam [10] Samson, E.; Marchand, J. Modeling the effect of temperature
on ionic transport in cementitious materials. Cement and
As a result, the corrosion occurring to the main Concrete Research, v. 37, 2007, p. 455-468.
reinforcement of the beam greatly affects the resistance of the [11] Bastidas-Arteaga, E.; Sánchez-Silva, M.; Chateauneuf, A.;
beam. The decline in resistance is markedly noticeable. This Ribas-Silva, M. Coupled reliability model of biodeterioration,
chloride ingress and cracking for reinforced concrete
demonstrates that actual corrosion can be more severe than structures., Structural Safety v. 30, 2008, p. 110-129
this simulation. Therefore, there is a need for more extensive [12] Zhao, Y.; Yu, J.; Jin, W. Damage analysis of cracking model
studies to accurately assess the corrosion level of the bridge of reinforced concrete structures with rebar corrosion.
girders under environmental conditions, corrosion time, etc. Corrosion Science, v. 53, 2011, p. 3388-3397.
to accurately predict the health of the bridge structure. From [13] Frangopol, D. M.; Lin, K-Y.; Estes, A. C. Reliability of
there, propose appropriate and specific solutions to limit reinforced concrete girders under corrosion attack. Journal of
corrosion. Structural Engineering, ASCE, v. 123, n. 3, 1997, p. 286-297.
[14] Nogueira, C. G.; Leonel, E. D.; Coda, H. B. Reliability
The FE models are verified by comparing the predicted algorithms applied to reinforced concrete structures durability
load-deflection curves and the crack distribution of FE assessment. IBRACON Structures and Materials Journal, v. 5,
models with practical experience. Based on the results of FE n. 4, August, 2012, p. 440-450.
models, the impact of the thickness of the steel plate, the [15] Nogueira, C. G.; Leonel, E. D. Probabilistic models applied to
location of the bearing and the distance of stirrups should be safety assessment of reinforced concrete structures subjected to
chloride ingress. Engineering Failure Analysis, v. 31, 2013, p.
probed. Therefore, several recommendations for strengthened 76-89.
corroded beams are proposed. The following can be [16] CEB-FIP MODEL CODE 1990.
concluded from this study: [17] Jiansheng, Shen & Gao, Xi & Li, Bo & Du, Kun & Jin, Ruoyu
(i) The load-deflection curves of the developed FE models & Chen, Wei & Xu, Yidong. Damage Evolution of RC Beams
Under Simultaneous Reinforcement Corrosion and Sustained
and the cracking distribution are in good acceptance with the Load. Materials, 2019. 10.3390/ma12040627.
crack and deflection theory of simple-span girder. [18] Wei-liang, Jin & Yu-xi, Zhao. (2001). Effect of corrosion on
(ii) It is necessary to develop more studies in the future bond behavior and bending strength of reinforced concrete
beams. Journal of Zhejiang University SCIENCE. 2. 298-308.
considering the simultaneous corrosion of other 10.1007/BF02839464.
reinforcements such as stirrup, deck reinforcement and the [19] Adelaide L, Richard B, Ragueneau F, Cremona C (2012). A
deterioration of concrete strength over time to have a more simplified numerical approach of global behaviour of RC
accurate assessment. beams degraded by corrosion. European Journal of
Environmental and Civil Engineering 16(3–4):414–439
ACKNOWLEDGEMENT [20] Thanh-Hung Nguyen, Anh-Thang Le & Duy-Duan Nguyen
(2020) Bending strength diagnosis for corroded reinforced
This research is funded by the Vietnam Ministry of concrete beams with attendance of deterministic, random and
Education and Training under grant number B2022-GHA-05. fuzzy parameters, Journal of Structural Integrity and
Maintenance, 5:3, 183-189
REFERENCES [21] Al-Kutti, Walid. (2014). Enhancement in Chloride Diffusivity
[1] Kumar, V. Protection of steel reinforcement for concrete—A due to Flexural Damage in Reinforced Concrete Beams.
review. Corros. Rev., 1998, 16, 317–358. Journal of Materials in Civil Engineering. 26. 658-667.
[2] Lee, H.; Lee, H.; Min, S.; Lim, S.; Singh, J. Carbonation- 10.1061/(ASCE)MT.1943-5533.0000836.
induced corrosion initiation probability of rebars in concrete [22] Nguyễn Công Luyến (số 26, 09/2018). Nghiên cứu ảnh hưởng
with/without finishing materials. Sustainability, 2018, 10, của hiện tượng ăn mòn cốt thép đến kết cấu bêtông cốt thép
3814. bằng mô hình 3D-RBSM.
[3] Wu, J.; Diao, B.; Xu, J.; Zhang, R.; Zhang, W. Effects of the [23] Michel, A., Geiker, M. R., Olesen, J. F., & Stang, H. (2013).
reinforcement ratio and chloride corrosion on the fatigue Reinforcement Corrosion: Numerical Simulation and Service
behavior of RC beams. Int. J. Fatigue, 2020, 131, 105299. Life Prediction. Kgs. Lyngby: Technical University of
[4] Song, L.; Yu, Z. Fatigue performance of corroded reinforced Denmark. (B Y G D T U. Rapport).
concrete beams strengthened with CFRP sheets. Constr. Build. [24] Thanh-Hung Nguyen, Manh-Hien Nghiem, Duy-Duan Nguyen
Mater. 2015, 90, 99–109. (2020) Evaluation of loading capacity of corroded reinforced
[5] Blikharskyy, Y.; Selejdak, J.; Kopiika, N. Corrosion fatigue concrete beams using experiment and finite element method
damages of rebars under loading in time. Materials, 2021, 14, [25] Mohammad, Tahershami (2016). Structural Effects of
3416. Reinforcement Corrosion in Concrete Structures
[6] Bigaud, D.; Ali, O. Time-variant flexural reliability of RC [26] Hien, Nguyen Thi, Takumi, Shimomura (2018). Evaluation of
beams with externally bonded CFRP under combined fatigue- critical crack width for reinforcement corrosion in RC member
corrosion actions. Reliab. Eng. Syst. Saf., 2014, 131, 257–270 based on numerical simulation of transport of chloride ions in
SAMSON, E.; MARCHAND, J. Modeling the effect of concrete.
79
Effect of Biodegradable and Metallic Mordants on

Dyeing Cotton Fabric with Spent Coffee Grounds
Tuan Anh Nguyen
Faculty of Fashion and Tourism
Ho Chi Minh City University of Technology
nta@hcmute.edu.vn
Abstract: This paper is to investigate dyeability of cotton coffee extracts in the presence of ferrous sulfate (FeSO4) was
fabric with extracts from coffee grounds under certain highest value among above the metallic mordants.
conditions and various mordants. Three kinds of Mongkholrattanasit et al. (2021) demonstrated that chitosan
biodegradable additives (almond, breadfruit and vinegar) and with low molecular weight in acidic medium played a role as
two kinds of metallic additives (ferrous sulfate and potassium a good mordant due to crosslinking between cotton and brown
alum) were used to improve color fastness of cotton fabric pigments [10].
dyed with coffee grounds extracts. The color space (CIE lab) In this work, five different mordants from natural and
and color histogram on digital images obtained from treated artificial compounds were applied to evaluate color
samples were applied to clarify the color differences in the performances for cotton fabrics dyed with spent coffee
presence of mordants. Besides, the tests of color fastness on grounds under specific conditions.
dyed fabrics to washing process based on the AATCC
standards were also carried out to evaluate more obviously II. EXPERIMENTS
the dyeing performances. The results obtained from this work A. Materials
provided additional information to commercially develop the
Spent coffee grounds (namely, Robusta) collected from
products from spent coffee grounds different coffee shops and dried in 2 days under the sun. The
obtained materials were continuously conditioned within 24
Keywords: cotton fabric, coffee grounds, dyeability, color
hours in the experimental lab to balance moisture (at 23oC and
fastness, mordant 65% RH). Then, coffee grounds were filtered by an espresso
I. INTRODUCTION machine (coffee maker) to obtain a stock solution.
Coffee grounds are known as the byproducts of the Greige plain woven cotton fabric with GSM 135.41, Tex
beverage industry with huge global consumption. Besides 30, TPI 220/190 (warp/weft densities) was desized in water
caffeine, coffee contains a variety of high molecular weight within 30 minutes in order to remove all unnecessary
compounds including tannin, fixed oils, carbohydrates and impurities. Then, the specimens were dried and relaxed at the
proteins [1,2]. Most spent coffee grounds were uncollected room temperature before dyeing with coffee extracts.
and unrecycled, causing negative impacts on our ecological In this work, three kinds of natural mordants including
environment. To solve these issues, many recent ideas and almond (Terminalia Catappa), breadfruit (Artocarpus Altilis)
discoveries of using spent coffee grounds to serve human life leaves and vinegar (mainly acetic acid) and two kinds of
have been reported as follows [3,4,5]. metallic mordants including ferrous sulfate (FeSO4) and kali
Firstly, the authors have successfully extracted antioxidant alum KAl(SO4)2 were used. For almond and breadfruit
phenolic compounds from coffee grounds to widely apply in additives, each liquid was extracted from fresh leaves by
food and pharmaceutical products [5]. Accordingly, some boiling in water at 1:20 ratio for 30 minutes. For other
works have proved that spent coffee grounds could inhibit the chemicals, vinegar, ferrous sulfate and kali alum were
growth of gram-positive bacteria [6]. Also, the studies purchased and stored in ambient condition.
addressed that coffee might fight against some human diseases B. Methods
such as diabetes, cancer, cirrhosis, and depression [1].
Secondly, coffee grounds can be used as a natural finishing
agent for textile materials. Especially, available contents of
pigments extracted from coffee grounds could dye some
common textile fabrics [7,8]. However, two most limitations
for coloring textile materials were probably color range and
color strength that the scientists were attempting to fix them.
Noticeably, brown appearance and color strength of coffee
dyed wool fabrics were significantly improved as mordanting
with tannin acid [7]. In other works, Lee (2006) found that the
best mordants of improving color strength and light fastness
for cotton fabrics dyed with coffee grounds were metallic
mordants such as FeSO4, CuSO4, and SnSO4 [9]. In fact, it was
evidenced that the color strength of dyed cotton fabric with Figure 1. Model of color space in terms of CIE Lab system

All above purified cotton samples were rotated with 3. In particular, a and b clearly increased from 3.4860 to
coffee grounds extracts in a semi-automatic beaker dyeing 12.4343 at liquor ratio of 1:80 to 1:5, respectively.
machine (namely Copopower) to control basic parameters
including dyeing concentration, time and temperature. Color Table 1. Values h and c for various liquor ratios with coffee
performance of dyed fabrics was determined on color analysis grounds extract
among scanned photos from experimental samples through Liquor ratio h c
graphic software (PTS) to recognize color values. A portable 1:80 74.9389 12.7137
X-rite spectrophotometer (I1Pro) was applied to evaluate 1:40 74.5624 12.9518
change in luminosity, hue and saturation (CIE Lab) of samples
1:20 74.2974 14.6576
in terms of reflective surfaces as illustrated in Fig. 1.
1:10 73.7800 16.8997
To determine the color fastness of dyed cotton fabric 1:05 72.5100 17.7626
with coffee grounds and various additives, treated samples
were washed in Miele’s washer/dryer according to the Similarly, surveys on the effects of extract concentration
AATCC 61. on dyeability of fabrics in terms of Lhc also show that at
higher ratios (up to 1:5), hue (h) slightly decreased but chroma
III. RESULTS AND DISCUSSION (c) markedly increased as shown in Table 1. In combination
C. Effect of exhausting time and liquor ratio on dyeability of with the investigation results on exhausting time above, it can
cotton fabric with spent coffee extract conclude that color formation of coffee extract on material
took place very rapidly, depending on dyeing solution
concentrations.
From the above findings, it could be affirmed that these
investigations on dyeing time and extract concentration only
evaluated color performance for just finished samples but not
for dyed fabrics after washing. Clearly, mordants were
necessary for dyed cotton fabric with coffee extract to enhance
color performance that will be discussed in the following.
D. Difference in luminosity between fabrics dyed with
various mordants
Figure 2. Impact of exhausting time on color based on the CIE Lab

system
Fig. 2 describes that increase in exhausting time from 1

to 24 hours for cotton fabric dyed with coffee extract induced
unmeaningful change for values Lab, i.e., less color difference
for both value a (red) and value b (yellow).
Figure 4. Photos of dyed fabrics (1) without mordant and with

mordants of (2) breadfruit, (3) almond, (4) vinegar, (5) potassium
alum, and (6) ferrous sulfate at liquor ratio of 1:40
Figure 3. Comparison of Lab values among dyed fabrics at various

liquor ratios (1:5, 1:10, 1:20, 1:40, 1:80) for 24 hours
Obviously, the dyeability of cotton fabric with coffee

solution seems to not be affected by an exhausting period
more than one hour. However, change in liquor ratios causes
Figure 5. Luminosity on color histogram for dyed samples with
big difference in the values a and value b as illustrated in Fig. coffee-ground extract with/without mordants
81
Based on luminosity on digital images, lightness as well

as darkness might show color difference among dyed samples.
Particularly, the darkest sample was dyed with ferrous sulfate
and the lightest sample was dyed with breadfruit (Figs. 4 and
5). As known that tannin (a kind of polyphenols) is a chemical
compound found in the extract of breadfruit and almond 1st A 2nd A 1st D 2nd D
leaves, which is able to form strong complexes with
celluloses.
E. Color difference of fabrics dyed with various mordants in
terms of CIE lab
1st B 2nd B 1st E 2nd E
A: Almond
B: Breadfruit
C: Vinegar
D: Ferrous sulfate
E: Potassium alum
1st C 2nd C
Figure 7. Color difference between 1st dyed samples and 2nd dyed
samples
Figure 6. Luminosity in terms of CIE Lab of five dyed samples by

coffee grounds and additives including (a) almond, (b) breadfruit,
(c) vinegar, (d) potassium alum, and (e) ferrous sulfate
Table 2. Values h and c for dyed fabric with various mordants

Mordant h c
Almond (A) 77.6535 17.7122
Breadfruit (B) 75.5985 14.7435
Vinegar (C) 74.3751 13.2779
Ferrous sulfate (D) 80.1458* 14.8924
Potassium alum (E) 74.9538 13.1198
To evaluate color difference among dyed samples with

coffee extract in the presence of various mordants more
clearly, a color system of CIE Lab as illustrated in Fig. 6 was
proposed. Based on the theoretical model in Fig. 1 and the
results in Fig. 6, it was obvious that dyed fabric with coffee
extract in the presence of almond had the value b which was
the most different and in presence of ferrous sulfate had the
value a whose was the most different. In fact, the dyed sample
with ferrous sulfate was greyest compared with others (see
Fig. 4). According to the color system of CIE Lch (Table 2),
value h on the dyed sample with ferrous sulfate also was the
highest or the most distinguished.
It can be concluded that the fabric dyed with coffee
grounds extract in the presence of various mordants caused
different color expressions. However, it was not possible to
affirm color fastness as well as other color features that the
following sections will be discussed.
F. Color difference among additional dyed fabrics
In this work, additional dyeing operations on treated
samples were carried out in order to evaluate the color
dyeability of cotton fabric. Figs. 7 and 8 showed that the dyed
sample with metallic mordants caused the largest change in
luminosity compared with the 1st dyeing time and the 2nd
dyeing time. On the other hand, additional dyeing for dyed
samples with coffee grounds extract only was significantly Figure 8. Color transition in terms of luminosity histogram
efficient as using metallic mordant, but biodegradable between 1st dyed and 2nd dyed samples
mordants such as almond, breadfruit and vinegar.
82
Figure 9. Value a and b of the 1st dyed and the 2nd dyed samples
Table 3. Values c and h for dyed fabric after 1 and 2 dyeing cycles
Mordant Dyeing cycle h c
1st time 77.6535 17.7122
Almond (A)
2nd time 77.9484 21.1874
1st time 75.5985 14.7435
Breadfruit (B)
Figure 11. Value a and b of dyed sample with coffee grounds
2nd time 74.2939 16.0983 before and after the 1st washing cycles
1st time 74.3751 13.2779
Vinegar (C) Table 4. Values c and d for dyed fabric before and after the first
2nd time 72.0387 15.0184 washing cycle
1st time 80.1458 14.8924
Ferrous sulfate (D) Mordant Washing cycle c h
2nd time 83.2995 9.9175 0
1st time 21.1874 77.9484
74.9538 13.1198 Almond (A)
Potassium alum (E) 1
2nd time 17.4121 76.3829
72.8909 15.7981 0 16.0983 74.2939
Breadfruit (B)
Similarly, the results of change of value a and b were 1 15.4104 76.9276
illustrated in Fig. 9. It was observed that there was a big 0 15.4104 76.9276
difference in value a and b between sample D and E. Besides, Vinegar (C)
1
as shown in Table 3, the value c and the value h of dyed fabric 15.1826 73.4266
0
for ferrous sulfate mordant were much remarkable from Ferrous sulfate (D)
9.9175 83.2995
14.8924 to 9.9175 and from 8.01458 to 83.2995, respectively. 1
9.7773 76.5744
0 15.7981 72.8909
G. Color fastness of dyed fabrics to washing Potassium alum (E)
1 17.0141 77.1083
Figs. 10, 11 and Table 4 indicated that color change with

washing condition (a standardized cycle) for dyed fabric with
coffee-ground extract in the presence of almond and ferrous
sulfate was noticeable (especially value b and c of sample A
A0 A1 D0 D1 (almond) decreased from 21.0782 to 16.9231 and from
21.1874 to 17.4121, respectively). Based on the above
obtained investigations, it can conclude that color fastness of
dyed sample with coffee extract to washing was relatively
good in the presence of breadfruit and vinegar mordants
although their dyeability was quite low as comparing to
B0 B1 E0 E1
others. In general, as considering both dyeing and using
efficiencies among samples dyed with coffee grounds extract
0: Unwashed sample
in various mordants, ferrous sulfate was supposed to be
1: Washed sample
highest quality.
IV. CONCLUSION
C0 C1
This study investigated the dyeability of cotton fabric dyed
Figure 10. Value a and b of dyed sample with coffee grounds with coffee grounds extract in the presence of various
before and after the 1st washing cycles mordants including biodegradable materials (almond,
breadfruit and vinegar) and metallic compounds (ferrous
sulfate and potassium alum). The experimental results were
reported according to the color model of the CIE Lab and Lch
83
obtained from digital images on PTS software. The [3] C. H. Nam, C. H. Xiang, Natural Dyeing Application of Used
investigations of exhausting time, liquor ratio, dyeing cycle Coffee Grounds as a Potential Resource, International Journal
of Fashion Design, Technology and Education, 12(3), 335-245,
and washing cycle were carried out to clarify the dyeability of 2019
cotton fabric with coffee grounds solutions. There were some [4] J. H. Bae, K.H. Hong, Optimized Dyeing Process for
interesting differences in color among dyed samples related to Enhancing the Functionalities of Spent Coffee Dyed Wool
various mordants. The initial findings in this study might Fabrics Using a Facile Extraction Process, Polymer, 11(4),
provide useful information for dyeing cotton with coffee 2019
grounds extract in the future. The coming works will be [5] K. H. Hong, Effects of Tannin Mordanting on Coloring and
focused on determining dye composition (such as tannin and Functionalities of Wool Fabrics Dyed with Spent Coffee
Grounds, Fashion and Textiles: International Journal of
polyphenols) and enhancing color fastness as well as Interdisciplinary Research, 33(5), 2018
antibacterial and antioxidant abilities. [6] S. I. Mussatto, L. F. Ballesteros, S. Martins, J. A. Teixeira,
Extraction of Antioxidant Phenolic Compounds from Spent
ACKNOWLEDGMENT Coffee Grounds, Separation and Purification Technology, 83,
The author would like to thank Ho Chi Minh City 173-179, 2011
University of Technology and Education for financial support [7] E. Koh, K. H. Hong (2017), Preparation and Properties of
and facilities. Cotton Fabrics Finished with Spent Coffee Extract, Cellulose
24(2)
REFERENCES [8] K. H. Hong, Effects of Tannin Mordanting on Coloring and
Functionalities of Wool Fabrics Dyed with Spent Coffee
Grounds, Fashion and Textiles 5(1), 2018
[1] G. V. De Melo Pereira, D. P. De Carvalho Neto, A. I. M Júnior, [9] Y. H. Lee, Dyeing, Fastness and Deodorizing Properties of
F.G. Do Prado, M.G. B Pagnoncelli, S.G. Karp, C.R. Soccol, Cotton, Silk, and Wool Fabrics Dyed with Coffee Sludge
Chemical Composition and Health Properties of Coffee and (Coffea arabica L.) Extract, Applied Polymer Science, 103(1),
Coffee Byproducts, Advance in Food and Nutrition Research, 251-257, 2006
91, 65-96, 2020 [10] R. Mongkholrattanasit, M. Nakpathom, N. Vuthiganond, Eco-
[2] M. Muzaifa, D. Hasni, Febriani, A. Patria, A. Abubakar, dyeing with Biocolorant from Spent Coffee Ground on Low
Chemical Composition of Green and Roasted Coffee Bean of Molecular Weight Chitosan Crosslinked Cotton, Sustainable
Gayo Arabica Civet Coffee, IOP Conference Series Earth and Chemistry and Pharmacy, 20, 2021
Environmental Science, 425, 2020
84
The Importance of Green Technology for

Sustainable Development Education: A Case
Study at Lac Hong University
Nga Hong Thi Doan Truong Van Nguyen
Office of Educational Testing and Quality Assurance Office of Educational Testing and Quality Assurance
Lac Hong University Lac Hong University
Bien Hoa City, Vietnam Bien Hoa City, Vietnam
hongngatcnh@lhu.edu.vn vantruong@lhu.edu.vn
Abstract: Over the last few years, green technology has as a paper providing useful information for green technology
received a lot of attention since it is considered for and as well as benchmark with the readiness of the LHU in
development and application of products, equipment and practice.
systems to conserve the environment and natural resources to
II. GREEN TECHNOLOGY IN EDUCATION AND
minimize human activities’ negative impact. Accordingly,
green technology is applied in many different fields such as SUSTAINABLE DEVELOPMENT EDUCATION
agriculture, food processing, potable water, sustainable Green technology known as the clean technology is one of
energy, consumer products, automobiles, construction, the elements of environmental science to preserve nature
industrial automation, computer and information surrounding and natural resources to minimize the negative
communication, education, health, aircraft and space travel. effects of human activities. [4]
The purpose of this study is to emphasize the importance of Green technologies encompass various aspects of
green technology for sustainable development education as technology which help us reduce the human impact on the
well as benchmark with the readiness of Lac Hong University environment and create ways of sustainable development.[6]
for green technology adoption in practice.
According to M.Bhardwaj and Neelam [7], green
Keywords: green technology, education, Lac Hong technology is the development and application of products,
University, sustainable development equipment and systems used to conserve the natural
environment and resources, which minimize and reduce the
negative impact of human activities.
I. INTRODUCTION
Nowadays, educational institutions in general and Lac As mentioned, green technology is applied in a lot of
Hong University (LHU) in particular need to connect closely different fields. In this research, I will focus on researching
with the world as well as general trends in order to provide and discussing some issues for the green technology in
learners with a meaningful environment to enhance their education.
awareness and learn relevant competencies for the rapidly According to P.Rao and P.S. Aithal [2], green higher
changing world. education is all about creating knowledge, skills, attitudes and
The mission of the LHU is to provide highly qualified values in terms of the environment. It’s very necessary for
human resources and competent, virtuous talents to serve the higher education due to the relationship of the environment
national industrialization and modernization in the context of with the economy. Higher education has a salient role in
international integration. Therefore, LHU is constantly creating and developing human resources. This resource that
focusing on developing the quality assurance to meet the is created should not just look at the economic point of
society, stakeholders’ requirements and global integration. business but societal aspect as well. Green education also
includes enhancing student knowledge in adopting green
In the context of Covid-19 pandemic, online teaching and technology. Computer and information technologies are
learning has never been more popular. Accordingly, some already considered as green technologies because of their
problems related to solutions for both direct and online contribution to clean environment in a lot of industrial
learning are noted especially some technical problems. Along processes.
with the general trends of the world, not only technology but
also green technology for sustainable development education The study by V.G.Sadh [9], Green education is inculcating
is considered a new trend. The concept of the green technology and cultivating the values, skills, knowledge towards
for sustainable development education is a new concept which sustainability of the environment. Our economy is also linked
receives much attention and approaches. So, all the up with the environment but the products which we are
stakeholders such as lecturers, students, managers, etc. need to manufacturing and creating should be community oriented.
update and rethink their roles and actions related to the above There is a tremendous demand for a Green job. Several sectors
issues. Facing this situation, the researcher had chosen the require competent green managers to organize, plan, control
topic “The importance of green technology for sustainable and implement solar energy, clean nuclear energy, and wind
development education: A case study at Lac Hong University” energy. In this regard, academicians have the major

responsibility to develop and mold the students for this genre. in The Impact Rankings Methodology 2021 including
Technology can help to spread awareness about Green and proportion of graduates with relevant qualification for
make it more feasible and accessible for everyone. teaching, public resources (lifelong learning), public events,
vocational training events, education outreach activities
In terms of sustainable development education, T.A.Aina beyond campus, lifelong learning access policy, proportion of
[1], emphasized the sustainability of the university as a process first generation students. However, with the approach as be
continuously in higher education. Transformation will not be presented above, the study has focused on two main indicators
easy, it will be a long, drawn-out process with many as follows:
crossroads and potholes on the way.
Indicator of public resources (lifelong learning): Provide
Sustainable development education is a high education for access to educational resources for those not studying at the
all sectors including the kindergarten, primary, tertiary university – e.g. computers, library, online courses, and access
education and tertiary education sectors as well as non-formal to lectures.
education. Therefore, schools are the models that play a key
role in promoting changes in education for sustainable Indicator of lifelong learning access policy: A policy that
development. The life experiences of learners and learners ensures that access to these activities is accessible to all,
need to be implemented in schools, including food-related regardless of ethnicity, religion, disability or gender.
issues, consumer education, social learning, energy use and Therefore, some survey questions are added to emphasize
personal resource management. Young people also can test the importance of green technology for sustainable
and develop their own life skills and lifestyle [8]. According development in terms of teaching and learning activity at
to D.S. Rathod [3], sustainable development education aims to LHU.
promote teaching that respects original and traditional
knowledge and also promotes the use of native languages in This research applies both qualitative and quantitative
education. Thus perspectives on sustainability need to be research methods.
integrated into programs at all levels of education. In terms of lecturer: first of all, 10 out of 100 participants
According to R.B. Mustapha [5] stated that TVET also joined the pilot survey and discussed questions in the
plays a role in improving green development. Education and questionnaire and interview. The aspects of the question pay
training to enhance the greening of the economy and attention to evaluate the effect of digital transformation on
companies should be applied to all levels of education. The lecturer’s readiness after 3 years at LHU. Then, some
basic education level should be started with an environmental questions were adjusted for appropriateness with the scale of
education that plays an important role in creating awareness. research concepts. Next, the questionnaire was issued to 100
In addition, the sustainable development goal 4 in The Impact lecturers who are teaching at LHU with the interviewees’
Rankings Methodology 2021 also aims at improving quality permission.
education by exploring how universities support early years, In this study, the author selected non-probability sampling
lifelong learning, and their nations through ensuring equality methods. The survey period starts from 15 February-15
of access to their facilities. So far, that can help equip locals March. The collected questionnaires: 100 votes. The valid
with the tools required to develop innovative solutions to the questionnaires: 100.
world’s greatest problems.
In terms of student: the study conducted a survey for
This study aims to emphasize the importance of green students who are learning at LHU after end-of-semester via an
technology for education, especially provide lecturers and online survey system. The purpose of this survey is to evaluate
students with the knowledge about the green technology. As a opinions of learner about current teaching activities after the
result, it can change and enhance the awareness of lecturers lecturers have applied digital technologies. The valid
and students in terms of the teaching and learning environment questionnaires: 1439.
and activities via applying green technology approaches at Lac
Hong University (LHU). Accordingly, green technology for IV. RESEARCH RESULTS
education can be understood as a tool which is To evaluate the effect of digital transformation on
environmentally friendly, and is used to actively integrate with lecturer’s readiness after 3 years at LHU, the study conducted
other materials and skills to minimize teaching and learning a survey with the following results:
activities’ negative impact. Hand in hand with this one,
lecturers also need to convey, and share knowledge of green
technology and sustainable development with their students. 120%
97%
In addition, they can be provided with a chance to rethink their 100% 83%
learning methods as well as added meaning value for work- 77%
80% 63% 67%
life environment and sustainable development. Therefore,
along with specific actions, an old educational environment 60%
40%
will be replaced by a new green environment like a green 40% 33%
23%
ecosystem where our learners will participate to study together 17% 17%20%
20% 3% 3% 7% 3%
in a more friendly, intriguing and positive way to achieve
results and wishful sustainable development for themselves, 0%
LHU and society.
III. RESEARCH METHODS
In the context of research on sustainable development
education, the methodology in this article is based on
Figure 1. The tools/system used by lecturers
indicators from the Sustainable Development Goal 4 (SGD4)
86
For the first item which stated “In the teaching and 0,60% 4,70%
learning activity, which tools/system have you been 0,10%
Highly
using?”, Figure 1 shows the selections of the lecturers with dissatisfied
97% with Zoom, 83 % with Moodle, 77% with Google Meet, 34,80%
67% with Quiz, 40% with Kahoot, etc. 59,80% Dissatisfied
Neutral
3,00%
Definitely agree
29% Figure 4. Learner’s satisfaction of materials for teaching and
Mostly agree learning
68% Neither agree nor 0,30% 4,20% Highly

0,10%
disagree dissatisfied
34,50% Dissatisfied
60,90%
Figure 2. Overall readiness of lecturers Neutral
For the second item which stated “Lecturers are always Satisfied
ready for digital transformation at LHU in practice”, Figure 2
shows that 68% of the lecturers definitely agreed, 29% mostly
agreed and 3% neither agreed nor disagreed. On the other Figure 5. Learner’s satisfaction of professional knowledge of
hand, in terms of the third item which stated “In general, lecturers
applying for digital transformation can help LHU’s lecturers
enhance their effectiveness for teaching and learning”, Figure 0,40% Highly
0,10%
3 shows that 71% of the lecturers mostly agreed, 18% 4,40% dissatisfied
definitely agreed and 11% neither agreed nor disagreed.
33,90% Dissatisfied
61,20%
Neutral
11% Definitely agree
18%
Satisfied
Mostly agree
Neither agree nor

71%
disagree Figure 6. Learner’s satisfaction of teaching skills
According to the survey results with 1439 students who are
learning at LHU, Figure 4 shows that 34,80% of the students
Figure 3. Overall effectiveness were satisfied with the materials for teaching and learning, and
59,80% were highly satisfied. Figure 5 shows that 34,50% of
the students were satisfied with professional knowledge of
Before 2018, the use of digital technology in teaching lecturers, and 60,90% were highly satisfied. Figure 6 shows
hasn’t been systematized, so the lecturers often use the that 33,90% of the students were satisfied with the teaching
teaching support and management software such as Google skills of the lecturer, and 61,20% were highly satisfied. On the
Classroom, Kahoot, etc. From 2018, LHU realized that other hand, 33,60% of students were satisfied in terms of
integrating digital technology into teaching and learning is course assessment, and 61,70% were highly satisfied. Some
necessary. From that, LHU has developed an online teaching other aspects related to “Lecturer creates excitement via online
system based on Moodle. Up to 2020, in the context of Covid- interactive activities (group/personal)” with 95,30% of
19 pandemic, LHU has applied Zoom software and created a students agreed; “Lecturer encourages students to enhance
system for online teaching including simultaneous and non- their critical thinking” with 95,60%; “Lecturer’s fairness of
simultaneous interactions. Accordingly, benchmark with the evaluation of students” with 90,08% of students agreed;
those survey results, there is an exact match. But now, there is “Lecturer encourages students to enhance self-study and
not a standard to evaluate the effect of using those research” with 96,10% of students agreed.
tools/systems in classroom at LHU and whether they have Along with learners’ overall satisfaction, there are also
brought positive energy/environment for their students’ some other opinions such as: The academic staff need to
learning or not. So, as mentioned above, in order to have a record during the subject and then post it in E-learning system.
foundation for evaluations objectively, the study integrated Therefore, learner can review it anytime, anywhere; create
more the data related to student survey on lecturer’s teaching more quiz game; create comfortable and friendly environment
activities in the academic years 2021-2022. for teaching and learning; have more interaction with their
students; some lecturers have a limit on using the software;
Internet connection is not stable; some of students are not
87
active because of there are many students in a class and limit learning. So, the academic staff need to be oriented, and
of class management skills; divide the students into the small redesign their lecture with the tight timing, therefore they will
groups to link with members each together; variety of teaching have more positive and various lectures both in direct or
methods to attract learners to join actively in their course; use online teaching. On the other hand, traditional teaching and
tight time, avoid too slow or too quick teaching; etc. learning methods and skills also need to be modified, as an
effort to help their learners believe, understand and learn
Generally, via the detailed evaluations of both lecturers more actively according to the ways which we have built with
and students, we can assess that a number of technologies the lecture, guidance and support them. Therefore, when both
being used at LHU are effective and suitable for teaching and the lecturer and learner understand, change their mindset and
learning as well as student satisfaction. However, to achieve apply the active technologies, the technologies can become
sustainable development education, as mentioned, lecturer not the green technology and help them add meaning value for
only deliver knowledge related to their subject, but also work-life environment and sustainable development in the
convey, share knowledge of green technology and sustainable future.
development to their students. So, they have a chance to
rethink their learning methods as well as adding meaning Finally, redesign the teaching and learning environment
value for work-life environment and sustainable development. with green action and space from school to classroom, library,
Accordingly, there are also some challenges which we face self-study areas, etc. The purpose of this solution aims to
including information technology foundation, Internet quality, spread green positive energy to lecturers and students and
workshop/training quality, the fear of change some lecturers connect people with nature better. Based on this, an old
and support of leaders. educational environment will be replaced by a new green
environment like a green ecosystem where our learner will
V. CONCLUSION participate to study together in a more friendly, intriguing and
Overall, green technology cannot be created the positive way. The most vital thing is to enhance the awareness
sustainable development education by itself. Accordingly, of lecturers and learners in terms of a new educational
green technology needs to incorporate awareness, change and environment via green materials and technology approaches
readiness of stakeholders especially lecturers and students. to achieve results and wishful sustainable development for
themselves, LHU and society.
In order to achieve sustainable development education,
Lac Hong University needs to create the systematical change. In general, for sustainable development education, along
Accordingly, some strategies to be researched and used in with the readiness of most of the academic staff, however, this
green education include: is a long process and besides other factors in terms of
information technology foundation, Internet quality,
Firstly, LHU needs to build a “green education project” workshop/training quality, the fear of change some lecturers
with specific steps. Accordingly, both the manager and the and support of leaders, etc. LHU needs to build a control
lecturer need to rethink the present and the past in terms of system systematically which can evaluate, monitor and
teaching and learning methods, tools, skills, environment, etc. continuously improve the above issues in the coming
So, LHU should build the criteria/standards/system to semesters.
evaluate, monitor and conduct the improvement for that in the
coming time. If the lecturer can meet the criteria/ standards, REFERENCES
they will receive the rewards in accordance with their [1] T.A.Aina, “Beyond reforms: The politics of higher education
competence level in a semester/academic year. transformation in Africa”, African Studies Review, Volume 53, (1),
2010, pp. 21–40.
Secondly, it is necessary to have more advanced training [2] P.Rao & P. S. Aithal, “Green education concepts & strategies in
workshops and coaching for academic staff. These workshops higher education model”, International Journal of Scientific
will be designed to provide a brief overview and basic Research and Modern Education, SSN (Online): 2455 – 5630,
principles of green technology and how to apply green Volume I, Issue I, 2016.
technology in teaching and learning actively and effectively. [3] D.S. Rathod, “Education for Sustainable Development (ESD)”,
After completing the training workshops, participants will be International Journal for Research in Education, ISSN:2320-091X,
2013.
received certificates. But the most important thing is how to
measure training effectiveness for those lecturers [4] F.M.Yusof, A.S.Rosman, S.Mahmood, S.H.M.Sarip and T.U.Noh,
“Green Technology Management in the Muslim World”, Journal
systematically. In general, we can use a mix of types of Technology, ISSN 0127–9696, 2013.
metrics via self-assessment report of lecturer, report for class [5] R.B. Mustapha, “Green and Sustainable Development for TVET in
observation, course completion rate, task performance and Asia”, International Journal of Technical and Vocational
survey on student’s satisfaction, etc. Education, 2015, pp. 133-142.
[6] G.D.Soni, “Advantages of green technology”, International Journal
Thirdly, increase the role of Faculty and academic staffs of Research, Vol.3, 2015, pp. 2394-3629.
in redesigning the green curriculum as well as monitoring and [7] Monu Bhardwaj & Neelam. The Advantages and Disadvantages of
assessing for their academic staff competence in accordance Green Technology, Journal of Basic and Applied Engineering
with the vision, mission of University and program learning Research. Volume 2, Issue 22; October-December, 2015, pp. 1957-
outcomes of Faculty. In this way, the first thing we should 1960.
begin is to change awareness and action of lecturers in their [8] R.Mathar, “The concept of whole school approach – a platform for
teaching activity. Accordingly, besides joining with school development with focus on sustainable development”,
Retrieved from http://esd- expert.net/assets/130314-Concept-
redesigning the green curriculum at program level, lecturers paper- ESD-Whole-school-approach-general introduction.pdf.
need to continuously change their lecture design and [9] V.G.Sadh, “Green Technology in Education: Key to Sustainable
presentation style with higher thinking. The first point of the Development”, International Conference on Recent Advances in
lecture is to not only convey the content of every subject, but Interdisciplinary Trends in Engineering & Applications, 2018.
also be a source of inspiration and motivation for teaching and
88
Unique Competitive Advantages of Vietnam’s

Garment Industry in the Sustainable Development
Trend
Quang-Tri Tran Thanh-Nhan Nguyen Tho Alang
Faculty of Fashion and Tourism Faculty of Fashion and Tourism School of Business
Ho Chi Minh City University of Ho Chi Minh City University of Ho Chi Minh City International
Technology and Education Technology and Education University
tritq@hcmute.edu.vn nhanntt@hcmute.edu.vn alangtho@hcmiu.edu.vn
Tuyet-Anh Truong Kim-Chi Le Nguyen Thi Le

The Social Studio School in Faculty of Fashion and Tourism Faculty of Garment Technology
Collingwood, Victoria Ho Chi Minh City University of and Fashion Design
Melbourne, Australia Technology and Education Hanoi University of Industry
anh@thesocialstudio.org Ho Chi Minh City, Vietnam Hanoi, Vietnam
chilmk@hcmute.edu.vn le.nguyenthi@haui.edu.vn
Abstract: The COVID-19 pandemic has caused disruptive accounting for 31.6 market share regardless of a 7 percent
consequences on the garment industry's demand and supply decrease in 2020 with an export value of 142 billion USD [2].
sides. However, despite the considerable hardship across the Currently, the COVID-19 pandemic has hit almost every
global garment industry, Vietnam’s garment industry is country in the world, devasting the global economy. Vietnam,
taking advantage of its unique competitive advantages to a developing country in South East Asia, is no exception as its
bounce back. This study aims at providing genuine economy is also damaged by this pandemic. In particular, the
information on what unique competitive advantages the garment industry, one of the leading industries in Vietnam, has
sector is leaning on to impressively resilience and growth. been facing with challenges of breaking both the supply chain
This study used a sequential mixed method approach of raw materials and stable human resources when the
including qualitative research with in-depth interviews and government implemented social distancing rules [3]. As a
quantitative research with online surveys. Five result, garment orders slumped dramatically, down 10.91%
generalizations were found and verified in this study compared to 2019. This is because Vietnam’s garment
including possessing a skillful, quick-learning workforce at a industry has been based so much on labour intensive, instead
relatively competitive cost; being blessed with a stable social, of using modern technology, automatic robot, or semi-
economic, and political environment; being born with a automatic robot for their production process. Also, Vietnam’s
convenient geographical location; the product quality of garment industry has been focusing on the lowest production
Vietnam’s garment industry is getting better and better and mode of the value chain, CMT (Cut-Make-Trim, aiming to
accepted by high demanding markets such as USA, EU, and exploit the intensive workforce to solve a big problem in
JAPAN; and establishing interconnections among customers, human resources and welfare of Vietnam [4].
suppliers, and manufacturers. Such a study should give However, in comparison with its rivals in terms of profit
insights into Vietnam’s garment industry, which could help losses across global Vietnam’s garment industry only suffered
take into account and take advantage of developing a 10.5% decrease, achieving $35 billion while its rivals faced
effectively this sector during the period of the new normal a loss of up to 15-20% [5]. This showed that Vietnam’s
state post-COVID-19. garment industry has been possessing competitive advantages
to lean on even though a global pandemic managed to uphold
Keywords: competitive advantage, Vietnam, garment many downfalls. This phenomenon motivates this study to
industry, sustainable development, COVID-19 pandemic explore the reasons behind this result.
I. INTRODUCTION While research on the global garment industry has been

much offered, little is known about the unique competitive
Vietnam’s garment industry is one of the three world’s advantage in the specific context of Vietnam’s garment
largest garment producers and exporters together with China industry. Therefore, the research question applied to guide this
and Bangladesh [1]. In 2021, Vietnam has become the second- study is what unique advantages are Vietnam’s garment
largest garment producer and exporter of ready-made industry leaning on to impressively resile and bound back in
garments after surpassing Bangladesh. Specifically, the difficult circumstances for sustainable development?
Vietnam’s garment export surged by 6.4 percent in 2020 with
an export value of 29 billion USD, while Bangladesh’s The structure of the paper is as follows. Firstly, we present
garment export fell from 6.8 percent to 6.3 percent in 2020. the methodologies applied in this study. Secondly, the key
Meanwhile, China endured the largest producer and exporter

results are outlined before the discussion and the conclusion  defining and operationalizing variables
are presented.  measuring and observing variables using an
instrument to obtain scores [9]
II. METHODOLOGY
In this study, the authors exploited mixed-method Data analysis was carried out via the useful and available
research, utilizing both qualitative research in the form of in- online tool of Qualtrics. It supported analyzing and creating
depth interviews and quantitative research in the form of reports with graphs, charts, and tables exported in forms of
surveys. There were two phases in this study: Word or PDF. Also, to have a deeper research analysis, it
provided a one-step direct download into SPSS or Excel [11].
Phase I was qualitative research through in-depth
interviews with eight participants representing various aspects III. RESULTS
of Vietnam’s garment industry[6][7][8]. Data were collected
through an interview questionnaire developed by the authors. A. The result of Phase I:
To ensure reliable data, a variety of participant groups Theme 1: Possessing a skillful, quick-learning workforce
represented diverse aspects of Vietnam’s garment industry at a relatively competitive cost.
including major government officials, academic professors,
and industrial managers or heads of departments. Although Due to the special characteristics of the garment industry,
the number of participants was limited, further findings in possessing a skillful workforce is considered a competitive
phase I was double-checked in the larger sample size in phase advantage for an industry heavily based on intensive labour.
II to ensure the reliability of the findings. For data analysis, Participant 3: That the Vietnamese workforce is skilled
data collected from participants were converted into text form, and very quick to grasp new technology is a decisive
translated into English, and made ready to develop advantage of the development of Vietnam’s garment
generalizations in categories following Creswell’s steps as industry.
below:
Participant 2: Previously, our company also had a
 gathering general information subsidiary in Cambodia. So, the training of a Cambodian
 asking open-ended questions of participants worker is more difficult than that of a Vietnamese worker.
 analyzing data to form categories This may be due to the labor–traditional nature of diligent
 looking for broad generalizations from categories Vietnamese people.
 relating patterns or generalizations to past experiences In practice, when placing garment orders, customers are
and literature [9] interested in three key elements including good price, good
quality, and fast delivery time. So, any garment enterprise
Table 1. Interview participant profiles possessing a stable workforce and the ability to meet a variety
Participants Title of Type of Organization
of customer needs is holding a big competitive advantage.
Participant Participant 8: Vietnam's workforce is abundant and
Participant 1 General Director Garment Joint Stock Company rapidly adapts to quick changes in technology and science.
As there are over million adult workers each year, the
Vice President Vietnam Garment Association domestic labor workforce possesses a huge potential for
Participant 2 Marketing Garment Joint Stock Company Vietnam’s garment industry.
Manager
Participant 3 Sales Manager Garment Limited Liability However, although Vietnam’s garment industry has the
Company advantage of a skilled and quick-learning workforce, it has
Participant 4 Vice Dean Public University - Faculty of
Garment Industry and Fashion
continuously trained to be proactive in the production process.
Participant 5 Vice Manager Garment Corporation Participant 3: The workforce in Vietnam’s garment
Participant 6 Head of General Garment Joint Stock enterprises has gradually been trained professionally to
Department Company deploy technological processes appropriately and improve
Participant 7 Head of Public University - Department the future competitive advantage.
Department of Technical Textile
Participant 8 Principal Public College Participant 7: Vietnam’s workforce is trained and able to
quickly grasp the technical requirements of the updated
garment device or appliances.
Phase II was quantitative research through online surveys
with eighty-three respondents. Data were collected through a Simultaneously, Vietnam’s garment enterprises need to
survey questionnaire via the useful and available online tool – have a well-organized structure of personnel to maximize their
Qualtrics [10]. Based on findings from phase I, a survey competitive advantage.
questionnaire was constructed to double-check and verify the
authenticity and then to support the interpretation of Participant 4: The exchange of experience between well-
generalizations. Together with a group variety of the trained workforces should be paid attention to both
participants like phase I, the research sample in phase II was enhance the product quality and improve the competitive
selected from a wider representative group about geographies advantage of Vietnam's garment industry.
such as North, Central, and South Vietnam to enhance the Previously, the cheap labor cost used to be a strong
accuracy of the findings. In this phase, authors will: competitive advantage for Vietnam’s garment industry.
However, this is no longer happening as wages for workers
 testing or verifying generalizations in each category
are increasing.
 testing hypotheses or research questions
90
Participant 2: For labor costs, Vietnam is relatively to deliver their products from factories to the ports for
cheaper. At present, the average salary of a worker in export, as an exporting country, Vietnam’s factories
China is around 600 USD / month while it is only around situated near the ports, will not waste time in the delivery
300 USD / month in Vietnam. Therefore, to make higher stages.
profits, customers will find garment producers who are
more competitive and able to offer good quality products However, it should be paid attention to weather problems
but relatively cheap labor costs including Vietnamese in Vietnam because there are rainy and stormy months every
producers. year from March to September due to weather characteristics
of Vietnam, affecting the transportation of Vietnam’s garment
Currently, the relatively cheap labor cost is still a products.
competitive advantage of Vietnam’s garment industry.
However, this advantage cannot last longer and Vietnam’s Theme 4: The product quality of Vietnam’s garment
garment industry is approaching differently to encounter industry is getting better and better and accepted by high-
difficulties in the fierce global competitive trend nowadays. demanding markets such as the USA, EU, and JAPAN.
Participant 8: Vietnam Textile and Apparel Association Regarding the sample development stage and labour
(VITAS) is currently focused on the central of Vietnam productivity, the garment industry of Vietnam certainly
because there are many potential resources on land, labor, cannot compete with that of other strong competitors like
and other preferential policies to attract investments in this China or Bangladesh.
region. Participant 3: In the sample development stage, China
Theme 2: Being blessed with a stable social, economic and has always had abundant and available raw materials. It
political environment. has even had its factories specialized in supplying raw
materials within the corporation.
The stability of the social-political-economic environment
is also a competitive advantage that creates a well competitive However, in terms of quality and price, the garment industry
advantage for Vietnam’s garment industry. of Vietnam is appreciated more than that of its competitors.
Interviewee 2: In Vietnam, it is rare to see strikes or riots Participant 2: Right now, there are a lot of customers
heavily in the textile and garment fields to satisfy the moving from China down to Vietnam to order garment
claims of workers. But, in other countries, the strikes are products.
seriously affecting the process production of enterprises. In recent years, Vietnam’s garment industry has been
This is a really strong advantage of Vietnam’s garment accepted by high demanding markets like the U.S, EU, and
industry. Japan. Also, other countries have also welcomed importing
garment products from Vietnam more and more. This is a very
This advantage is seen as a significant factor affecting
good competitive advantage in this sector.
whether or not customers will choose Vietnam‘s garment
producers as a destination to do business. Participant 8: In terms of garment product quality,
Vietnam enterprises are good because the majority of raw
Participant 3: When our enterprise contacts customers, materials and machinery are imported from the developed
they would love to do business with Vietnamese producers countries in combination with the diligent labor nature of
because of the high appreciation of the security level of Vietnamese people fitting the main characteristic of the
social-political and economic environments in Vietnam, garment industry.
bringing peace of mind to customers.
A good example is that enterprises possessing well-known
Based on the available conditions of a stable social, brands in the world maintained and trusted doing business
economic and political environment, the government needs to with Vietnam to produce garment products until now.
constantly focus on facilitating better and clearer working
Participant 5: Enterprises preserving their reputation with
environments to attract more and more investments in this
customers would like to work with garment producers who
sector
act in compliance with customer requirements. Otherwise,
Participant 4: Vietnam government needs to keep state Vietnam producers would never get orders with poor
undertakings, introduce original terms of the investment quality products and would never achieve an average
policies, and regulate “the subtle form of fees” from growth rate of export of Vietnam’s garment industry at
“harassment members” to create the most convenient 20% per year.
conditions for the enterprises to be stable and keep the However, in response to this trust from the key markets
long-term development in Vietnam. such as the US, EU, and Japan, Vietnam’s garment industry
Theme 3: Being born with a convenient geographical has to keep going to ensure the requirements of product
location. quality, and compliance in the initial commitment to facilitate
Vietnam has a certain competitive advantage in shipping stability and sustainability of this sector.
and transiting time compared with China, especially in the Theme 5: Establishing interconnections among
fashion industry where customers pay much attention to the customers, suppliers, and manufacturers.
"fast market".
In the textile and garment supply chain, Vietnam’s
Participant 4: Thanks to the long coastline, the geography garment enterprises are not only producers for their customers
of Vietnam is somewhat more favorable for the trade than but also buyers for raw material manufacturers. Thus, to
other countries smoothly proceed with orders, it is sure that Vietnam’s
garment enterprises have set up interconnections among
Participant 2: While China will need to add 5 to 7 days suppliers, manufacturers, and customers.
91
Participant 2: Manufacturers, suppliers, and customers IV. DISCUSSION AND CONCLUSION

sat down and worked together to solve promptly problems 1. 71% of respondents indicated that the main advantage
and find correlating solutions that would be highly of Vietnam’s garment industry at the current time is
beneficial overall. possessing a skillful, quick-learning workforce at
Setting up a synchronized supply chain among textile relatively competitive labour cost. Vietnam is still a
companies, garment companies, and customers through developing country with low income. Especially, the
organizations such as AFTEX (ASEAN Federation of Textile workforce in Vietnam's garment industry is located in remote
Industries), SAFSA (Source ASEAN Full Service Alliance), areas or areas with difficult living conditions. They must
CPTPP (Comprehensive and Progressive Agreement for approach a variety of miserable occupations to earn a living.
Trans-Pacific Partnership), and EV FTA (European-Vietnam So, as there are better working environments like industrial
Free Trade Agreement) … is one of the approaches to create parks in general and garment industrial zones in particular,
greater competitive advantages for this sector. they are like opportunities to have a better life. So, they are
Participant 2: Our enterprise has also joined SAFSA. willing and dedicated to learning and learning quickly to make
Although fees paid for SAFSA are more than $ 500 per money. Besides, the difficult living conditions trained the
month, the benefits brought are much greater than the costs workforce to be diligent and willing to approach difficult
that our enterprise would bear to find material suppliers sewing stages to improve product quality as well as labor
and new customers if doing them by ourselves. Also, in productivity to directly bring them higher income and
negotiating with customers, we are more favorable than indirectly bring their enterprises higher profits.
when negotiating with third parties. So, our enterprise However, it should be noted that labor cost is only
would be more active and had more opportunities to get relatively competitive without low labor costs as in previous
orders. years. The reason can be mentioned that firstly, people's living
In the fierce competition nowadays, if two enterprises are standards are getting higher and higher and inflation is always
offering both the same price and the same delivery time, any quite high), secondly, workers understand labor law more and
enterprise can meet the other requirements of customers from more when there is union support. Secondly, the workforce in
its policy. Surely, customers will do business with its the garment industry is increasingly scarce owing to the
company. mushroom of the number of garment enterprises in Vietnam
recently. Meanwhile, the working environment in the garment
Participant 6: With the slogan "Better and better meet
industry is quite stressful, but the salary is not satisfying with
customer needs", our company has actively built strong
what they worked to attract workers for the garment industry
relationships with not only customers but also with
compared to other industries. Therefore, in order not to have a
suppliers to meet the best requirements from customers.
shortage of workforce for the garment industry, there is no
B. The result of Phase II: choice for the employers to raise the salary to retain the
Table 2. unique competitive advantages of Vietnam’s garment workforce to have enough labor force to deploy orders and
industry ensure the delivery progress with customers, otherwise, they
will be compensated and lose key customers.
Average
# Question 0 1 2 3 4
Value The traditional nature of a diligent Vietnamese workforce
Possessing a skillful, at a relatively competitive cost is a decisive advantage in the
quick-learning development of Vietnam’s garment industry. By improving
1 workforce at a 0 1 23 22 37 71.08 the quality of human resources management, attracting stable
relatively competitive workforces, and deploying rigorous technological processes
cost. in the production lines, this sector will gain more favorable
Being blessed with a conditions to enhance competitiveness for Vietnam’s garment
stable social, enterprises.
2 0 3 23 21 36 68.67
economic and
political environment. 2. 69% of respondents showed that Vietnam has the
advantage of being blessed with a stable social, economic
Being born with a
3 convenient 3 16 25 17 22 46.99
and political environment. Vietnam’s garment industry was
geographical location. highly appreciated and selected by the customers due to the
security of social economic and political environments. This
The product quality of is also considered an outstanding strength or an available
Vietnam’s garment competitive advantage of Vietnam. Vietnam is always an ideal
industry is getting
destination for foreign-invested enterprises investing not only
better and better and
4 0 0 21 23 39 74.70 in the garment industry but also in other industries in Vietnam.
accepted by high
demanding markets Specifically, the export turnover of the garment industry has
such as USA, EU and always grown at a steady rate of around 15 % every year. And,
Japan. it is always one of the leading industries of Vietnam in
Establishing contributing to the national export turnover. Although
interconnections negatively impacted by the COVID-19 pandemic, Vietnam’s
5 among customers, 0 5 22 26 30 67.47 garment industry has overcome difficulties and shown
suppliers and
manufacturers.
positive signals through optimistic results in recent years the
second half of 2021 and the beginning of 2022.
(Using a scale of 0 = Not at all important to 4 = Very This has brought customers peace of mind in doing
important) business with Vietnam’s garment industry. Based on those
92
above favorable environments, however, Vietnam’s garment supply among textile suppliers, garment producers, and
enterprises need to constantly enlist and train skilled customers to create greater competition is a critical
manpower and improve their working environment and requirement for any garment producer including Vietnam.
cultural values better to attract a skillful and stable workforce.
Besides, the Vietnamese government needs more Possessing a closed supply chain from input, process, and
transparency in the investment incentive policies to create the output will help the garment industry respond quickly and
most convenient conditions for the enterprises to be stable and effectively to the needs of its customers. Specifically, the
keep sustainable development in Vietnam. industry can quickly solve the three most important criteria of
the garment industry, including price, lead time, and labor
3. Being born in a convenient geographical location is productivity. Especially, when Vietnam has been being a
also one of the prominent competitive advantages of the member of critical FTAs such as the FTA and the CPTPP. It
garment industry in Vietnam although it is not a strong one has been promoting its effectiveness through a commitment to
with only 46% of respondents designating it. The geography abundant material inputs, transparent certificate of origin as
of Vietnam for trade is somewhat more favorable than other well as preferential prices for member countries in the
countries regarding sea freight. Vietnam is a country located agreements. In addition, the agreements will make an
"next to" the East Sea. The East Sea is a particularly important important stepping stone to ensure the output of Vietnamese
trade bridge on the world maritime map. Every year, hundreds garment products. Hence, Vietnam’s garment industry can
and thousands of cargo shipments take place in the East Sea. penetrate the production stage to better meet customers’
Therefore, Vietnam is highly appreciated as a country with an needs. This will be facilitated for all third parties when they
outstanding advantage in the international maritime industry. gather together to solve promptly the problems, bringing
The advantage of 3260 km of coastline allows Vietnam to practical benefits for all taking part in this interconnection.
develop its infrastructure system and logistics service supply
network with dozens of large and small seaports to serve both V. RESEARCH LIMITATIONS
domestic and international trades such as Asia and the world. Additional qualitative interviews with several participants
Thanks to Vietnam's coastline stretching across the whole per institution and region could provide a more concrete
country and industrial zones located near the seaports, it is illustration of producers therefore results could be more
convenient for Vietnam’s garment enterprises to transport confidently generalized.
and distribute garment orders to partners, as well as to save While this research concentrates in much detail on the
delivery time when customer increasingly limits the delivery garment industry, future research could examine influences
time. within other industries such as textile, accessory, or fashion
4. 75% of respondents signified that the quality of industries.
Vietnam’s garment industry is getting better and better This research implies a multidisciplinary approach and
and accepted by high demanding markets such as the requires academic and practical knowledge in the fields of
USA, EU, and JAPAN is the strongest advantage of this management, macroeconomics, and technology. Meanwhile,
sector. In recent years, high-demanding markets like the U.S, the authors haven’t yet presented an in-depth analysis of the
EU, and Japan have really opened to welcome garment export subject of research.
products from Vietnam. After difficulties from the early
stages of formation, Vietnam's garment industry successfully VI. CONTRIBUTION OF THE STUDY
produced difficult and large garment orders with international This research reveals and provides genuine information on
partners. They have gradually become more developed and the transition of Vietnam’s garment industry in an attempt to
confident with the products they have been producing. In fact, contribute to its business practices in the global market and
garment products “made in Vietnam” are considered a encourage a better comprehensive understanding of
criterion for selecting a garment product (besides prerequisite Vietnam’s garment industry.
criteria such as design, price, and brand...). Garment products
in Vietnam are highly appreciated for the care and REFERENCES
meticulousness in the stitches, the durability of the products, [1] T. T. Quang, T. Tran, A. Tho, and J. Burgess, "Chances and
as well as the uniformity of quality in the garment products. A Challenges of Vietnam’s Garment Industry in the new Trend of
good example is that Vietnam's garment products have Sustainable Development," 2020 5th International Conference
on Green Technology and Sustainable Development (GTSD),
gradually convinced large and demanding customers from the 2020, DOI: 10.1109/GTSD50082.2020.9303100.
US, EU, and Japan and continuously expand its market share
[2] N. Okonjo-Iweala, World Trade Statistical Review 2021,
in these key markets. Importantly, even when EV FTA and World Trade Organization, 2021.
CPTPP agreements have been and are in effect, Vietnam's [3] K. Nguyen, and N. Vu, “Assess the Impact of the COVID-19
garment products will increase dramatically as tariffs on these Pandemic and Propose Solutions for Sustainable Development
key markets will favorably be adjusted to the member for Textile Enterprises: An Integrated Data Envelopment
countries of the agreement. However, in response to this trust, Analysis-Binary Logistic Model Approach”, Journal of Risk
Vietnam has to keep going to ensure product quality and and Financial, 2021.
compliance requirements to facilitate sustainability and [4] T. Tran, T. Tran, J. Burgess, S. Müller, and M. Weber, “An
exploration of extraordinary characteristics impacting on the
stability. improvement of competitive advantage of Vietnam’s garment
5. Establishing interconnections among customers, industry”, 2018 4th International Conference on Green
Technology and Sustainable Development (GTSD), 2018,
suppliers and manufacturers is also one of the principal DOI: 10.1109/GTSD.2018.8595545.
advantages of Vietnam’s garment Industry which received the [5] Vietnam News Agency, “Textile - garment industry may
agreement of 67% of respondents. Nowadays, in the global flourish in 2022: insiders”, Accessed online at
competition trend, the establishment of synchronized chain https://en.vietnamplus.vn/textile-garment-industry-may-
93
flourish-in-2022-insiders/216012.vnp. Accessed date [9] J. W. Creswell, Research Design: Qualitative, Quantitative,

02/03/2022. and Mixed Methods Approaches, Third Edition, SAGE
[6] F. Uwe, An Introduction to Qualitative Research, London, Publications Inc 2009.
United Kingdom: Sage Publications Ltd, 2014. [10] Qualtrics, “Qualitative vs Quantitative Research: What is it and
[7] A. Bryman, and E. Bell, Business Research Methods, New when should you use it?”, Accessed online at
York, NY, United States of America: Oxford University Press, https://www.qualtrics.com/blog/qualitative-research/,
2015. Accessed date 06.12.2021.
[8] M. Miles, and M. Huberman, Qualitative data analysis: An [11] Qualtrics, “Data Export Format”, Accessed online at
expanded sourcebook (2nd ed.), Thousand Oaks: Sage https://www.qualtrics.com/support/survey-platform/data-and-
Publications 2014 analysis-module/data/download-data/export-formats/,
Accessed date 06.12.2021.
94
Driving Factors of Green Economy for Smart Cities

in the Context of Developing Countries
Tiep Nguyen Nghia Hoai Nguyen Leonie Hallo Bao Van Pham
Department of Civil Department of Civil Adelaide Business School Department of Civil
Engineering, International Engineering, International The University of Adelaide, Engineering, International
University, Ho Chi Minh City, University, Ho Chi Minh City, South Australia 5005, University, Ho Chi Minh City,
Vietnam Vietnam Australia Vietnam
Vietnam National University, Vietnam National University, Adelaide, South Australia, Vietnam National University,
Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam Australia Ho Chi Minh City, Vietnam
nvtiep@hcmiu.edu.vn nhnghia@hcmiu.edu.vn leonie.hallo@adelaide.edu.au pvbao@hcmiu.edu.vn
Abstract: The “smart city” concept, based on information living, which these authors assert are fundamental
and communication technology (ICT), is widely used to characteristics of a smart city.
describe a new trend in using the advantages of technology to In his analysis of the planned Tehran 2025, Mohseni [4]
deal with complicated issues arising in cities regarding high discussed definitions of smart cities and in particular the place
energy consumption, sustainable infrastructure development, of citizen power within them. He noted three generic broad
and natural environment protection. The green economy is types of governance which influence citizen involvement:
considered as a core component of smart cities with the bottom-up, beneficial, and techno smart. He concluded that
purpose of creating a sustainable economy with respect to about 20% of definitions of smart city governance are based
maintaining natural ecological systems as well as improving upon the bottom-up concept where citizens have the power to
citizens’ quality of life through the creation of intelligent engage fully in policy-making in smart cities. About 70% of
solutions. Many studies have discussed the role of the green definitions are based on the idea that smart cities provide the
economy and its competitive advantages for smart city public with a variety of benefits and opportunities and
development, but few studies have focused on investigating improve their quality of life. This often includes sustainability
the driving factors of the green economy for smart city as a major attribute. However, citizens may not play an
development in the context of developing countries. This important part in influencing decisions. Less than 10% of
definitions do not consider citizens at all and focus instead on
paper aims to identify critical driving factors of the green
technology. Many more recent definitions highlight public
economy through a review of literature during the period
collaboration and participation in decision-making as an
from 2010 until 2021. The initial findings show that 16 important aspect of a city being smart. Ghaffarianhoseini, et
factors, categorized mainly into the technological, social, and al. [6] differentiated between smartness and intelligence
environmental groups contribute to the motivation of green within the design of built environments and concluded that
economy development. These factors should be carefully intelligent buildings have the critical roles of meeting users’
revised to provide investment prioritization for smart city requirements using smart technology and embracing
development, including in the context of developing sustainability through appropriate green and active design.
countries, as well as to suggest blueprints for economic Thus we see the connection with sustainability. The smart city
development policy formulation. concept has developed well beyond the ICT approach to
include sustainability goals such as lower emissions and
Keywords: smart cities, green economies, sustainable energy saving. In fact, smart cities have always been related
development to sustainability and better quality of life, where the economy
and ecology are optimally combined, and sustainability is a
I. BACKGROUND core feature.
A. Smart City and Green Economy Fundamentals In regards to green economy definition, there are many
The term smart city dates back to the early 1990s [1]. definitions of the concept, as noted by OECD (2017), often
There are many definitions of the concept of a smart city. focused on the idea of the green growth. Green economy (GE)
Early definitions, for example that of Katagiri, et al. [2], is a concept which aims at developing sustainable
focused upon IT as a way of solving urban problems. management of resources. The United Nations Conference on
Subsequent efforts to define the concept were more focused the Environment and Development in 1992 adopted the
on greater citizen involvement in decision-making. While following definition of the green economy: “… a development
some authors focus on features such as infrastructure, that meets the needs of the present without compromising the
transportation, energy and healthcare, for example Mohanty, ability of future generations to meet their own needs” [7]. GE
et al. [3], others including Mohseni [4] take a broader view. In has been defined as a concept that “… results in improved
his analysis of the qualities of a smart city he listed human well-being and social equity, while significantly
connectable, accessible, sociable, and shareable. One widely reducing environmental risks and ethological scarcity” [8].
cited definition is that of Giffinger, et al. [5] who refer to the The focus here is on preserving natural capital including
six components of smart economy, smart people, smart natural resources and ecosystem, based upon low carbon,
governance, smart mobility, smart environment and smart efficient use of resources, and social inclusiveness. Barbier [9]
wrote about ‘green growth’ and explored the relationship

between green growth and sustainable economic place to consider how this transition can best play out. Ferrara
development. This relationship is tenuous due to ecological [13] asserted that within Europe the progress towards smart
scarcity through human exploitation. Various failures, cities is highly connected with the green economy and energy
whether they be market policy or institutional, prevent us from efficiency. According to this author, there is often an
recognising the problems of ecological scarcity caused by unwanted hiatus between the two concepts, sometimes caused
environmental degradation arising from human activity. by legal impediments. How possible is it to reconcile
Unsustainable use of natural resources has resulted in environmental protection and new technology development?
degradation of the environment. The green economy is supposed to be about integrating
A report by OECD [10] suggested that green growth is economic growth and environmental values and the smart city
more than growth of the eco industry, but really is focused is a concept which overlaps with the concept of the sustainable
upon growth of the entire economy through catalysing city. According to a report by the European Commission [14]
investment and innovation to produce sustained growth and concerning smart cities, a smart city can use technology to
new opportunities. Thus the green economy is viewed as an address major challenges including climate change, ageing
alternative to economic development and its accompanying populations and pressures on finances: “A smart city…uses
costs. A report by the World Bank [11] focused upon green digital technologies to enhance performance and well-being,
growth as being based upon the efficient use of natural to reduce costs and resource consumption and to engage more
resources, and minimisation of pollution and environmental effectively and actively with the citizens. Key smart sectors
damage. Loiseau, et al. [12] overviewed concepts concerning include transport, energy, health care, water, and waste. A
the green economy and noted that links with sustainability are smart city should be able to respond faster to urban and global
not always apparent. These authors suggest that the term green challenges than one with a simple ‘transactional’ relationship
economy is an umbrella concept that is not clearly understood. with its citizens”.
There is a need for clarification regarding how to implement a A paper by Delitheou, et al. [1] strongly links new
green economy in the transition to sustainability. technologies and green growth to improving the quality-of-
The systematic review conducted by these authors life of people living in smart cities. These authors highlight
identified main keywords for green economy and revealed that that there is now a strong motivation for transition from
over half of keywords out of 877 research outputs with “green traditional cities towards smarter cities in response to
economy” in the title were related to environmental and population density, climate change, polluted air, and
economic dimensions; that is, environmental issues such as protection of the natural environment, et cetera. The smart
climate change, economic issues such as development and green city overcomes these difficulties through recycling,
growth, with a lesser proportion relating to the social bicycle lanes, high water quality and community parks. It’s
dimension, that is employment and impact on people. There not just about protecting the environment; the focus is also
were some keywords concerning how to implement the green strongly on quality of life, which was categorised by Donald
economy in practice, and over 10% referred to governance: [15] as including social cohesion, quality and accessible
how to define and analyse approaches that can support the healthcare, social support, training and development, security,
concept. In addition, the challenge of raising enough finance quality housing, accessible public transport, quality of
to maintain weakening ecosystems cannot be underestimated. environment, and culture and leisure opportunities. These
This problem is particularly bad in some countries including attributes contain elements of both greenness and smartness.
Africa, Latin America, the Middle East, and Asia; that is Gültekin and Erenoğlu [16] investigated the link between the
developing countries. The funding gap occurs due to present green economy and sustainable development, considering the
economic disincentives and addressing this gap requires difficulties for developing countries who wish to increase their
developing adequate mechanisms to provide compensation prosperity but also need to address environmental degradation
and to raise enough funding for these taxation mechanisms to and increase social welfare. As they say: “Concepts such as
work, according to Barbier [9], who also notes:”.. Developing green economy, green growth, environmentally friendly
countries will continue to be the direct and opportunity costs growth contribute to the integration of the idea of environment
of ecosystem conservation for the foreseeable future. Clearly and economy in the sustainable development plan.” [16].
this perpetuates the unsustainability problem, especially given But how to pay for greenness and smartness occurring
rising global ecological scarcity.” [9] together? Taxation is the usual approach. These benefits for
people and the environment can be achieved through
B. The role of Green Economy in Smart City Development environmental tax reform via green taxes used efficiently to
Several authors have considered how the green economy reduce environmental degradation. Green taxes would be
is or should be connected to smart cities. Many publications levied on all harmful economic activities and are used to
present the model of the smart city where economy and promote environmentally friendly activities through
ecology are reconciled, and protecting the environment sits incentivization. The shift to green taxation is gaining
alongside of developing new technologies. Transiting to a acceptance in some constituencies [17]. However, in
green economy is now recognised as an urgent task for big developing countries, environmental tax reform aimed at
urban cities because such cities are increasing in size and creating a green economy sometimes has unexpected
economic growth as well as in consumption and pollution. consequences and may be politically difficult to implement
Cities are where economic growth and environmental [18]. Gramkow and Anger-Kraavi [19] in their study of green
degradation collide. Many cities have already started to innovation policies in Brazil point out that green fiscal policies
transition in response to ecological crises, undertaking actions are generally introduced in stages, through an ad hoc stage in
such as mixed and compact use of land, low carbon transport, which a specific environmental tax is introduced, followed by
green buildings, renewable energy, clean technologies, and expansion into higher levels of environmental taxation and
improved waste management. Innovation and creativity are incentives and putting aside some sections of public revenue
behind these actions. It is clear that the smart city is the ideal for environmental spending, followed by a final stage in which
96
we see shifting the tax burden toward pollution intensive Table 1. Driving Factors Green Economy For Sustainable Smart
activities and technologies; for example, fossil fuel taxation. City Development
The authors conclude:”… Fiscal policies have an important List Critical Factors Factor Description Referred by
role in promoting the transition to a green economy in Brazil Digital Economy The digital Delitheou, et al. [1],
and potentially in other developing countries.” [19]. (TECHNOLOGICAL) economy is not just Ferrara [13], Limna
Nevertheless, it should be apparent that additional taxation to a separate sector, P. [22], Caprotti
protect the environment will be harder to implement in but a system of [23], Savchenko
developing economies where finances are stretched. economic and social and Borodina [24],
1
relations based on Yudono, et al. [25],
Implementing environmental policies will be politically the use of digital Hao, et al. [26],
difficult in these circumstances [18]. technology. Qian, et al. [27],
Thus, this study focuses on such challenges to investigate Tamakhina, et al.
[28]
implementation barriers as well as to seek alternative solutions
Citizen Engagement The engagement Granier and Kudo
for the implementation of green economy within smart cities (SOCIAL) level of citizens in [29], Cardullo and
in the context of developing countries. Drivers for the green decision-making Kitchin [30], Capra
economy bring significant long-term competitive advantages regarding economic [31], Kusumastuti
to sustainable smart city development. strategies and Rouli [32],
2 Przeybilovicz, et al.
II. RESEARCH METHOD [33], Berntzen and
Johannessen [34],
In this study, a scoping review approach was adopted in Tadili and Fasly
the study to conduct a clear and comprehensive overview of [35], Snow, et al.
available research and findings on this topic. The purpose of [36]
this work is to clarify the relationship between green economy Urban green The master Anguluri and
planning planning priority Narayanan [37],
and smart city development. In accordance with Grant, et al. (TECHNOLOGICAL) based on Liu, et al. [38],
[20], the scoping review aims to identify the nature and extent environmental Gavrilidis, et al.
of research evidence (usually including ongoing research), enhancement with [39], Chen [40];
3
and further provides a preliminary assessment of the potential many benefits for Reinwald, et al.
size and scope of available research literature. The other inhabitants, the [41], Narmada and
economy, and the Bhaskaran [42],
motivation for the scoping review is its ability to identify the ecology Gupta, et al. [43],
types of available evidence in a given field, and the ability to Wang, et al. [44]
examine and clarify definitions that are used in the literature Intelligent The level of Reinwald, et al.
[21]. Such an approach provides insights on investigating Governance technology adopted [41], Pereira, et al.
barriers as well as identifying driving factors of green (TECHNOLOGICAL) regarding regional [45] ; Fitsilis and
economy that contribute to the development of smart cities 4 governance that Kokkinaki [46]
improves
The SCOPUS database was selected, and the following communication
strings of TITLE-ABS-KEY were used, including: ‘Green effectiveness
Economy’; AND ‘Smart City’. The initial search retrieved Smart Social The involvement of Mohseni [4];
dialogue technology in Kusumastuti and
200 articles for subsequent refinement. The scope was then (TECHNOLOGICAL- negotiation, Rouli [32]; Fitsilis
narrowed with the following string searched: ‘Green SOCIAL) consultation or and Kokkinaki [46]
Economy’; AND ‘Smart City’; AND Sustainable Economy simply exchange of
within the specific time period of 2010 until 2021. This 5
information
resulted in 112 studies being identified. Through reading the between, or among,
representatives of
abstract, the search was narrowed to those articles with a key governments,
focus on the role of the green economy in smart city employers, and
development; and the context of developing countries. workers, on issues
Finally, 68 articles were selected for analysis and synthesis. of common interest.
From the literature, 16 critical factors of the green economy Renewable energy Use of ICT to Fitsilis and
that may have a significant influence on smart city (TECHNOLOGICAL) manage energy in a Kokkinaki [46];
6
sustainable way Xin, et al. [47];
development were surfaced and examined. Yadav, et al. [48]
Ecosystem balance This refers to the Pandey, et al. [49],
III. FINDINGS AND DISCUSSION (ENVIRONMENTAL) equilibrium Passarelli [50],
From the literature, there is a positive relationship between between living Embarak [51]
green economy and smart city development. Even though the organisms such as
7
human beings,
achievement of the main goals of green economy development plants, and animals,
is challenging for developing countries due to the limitation of as well as their
sources, such goals are always at the top of economic environment
development priorities of local governments in their long-term Sustainability To what extent do Girardi and
planning. The involvement of new technology, particularly in standards smart cities meet Temporelli [52],
(SOCIAL - social or Estevez, et al. [53],
ICT, is a promising approach that enables local governments 8 ENVIRONMENTAL) environmental La Greca and
to re-allocate resources to achieve strategic goals. Table 1 sustainability Martinico [54],
below presents critical driving factors contributing to smart standards? Haupt [55], Martín,
city development with a strong focus on ICT application. et al. [56]
Biodiversity Policies to Colding, et al. [57],
9 (ENVIRONMENTAL) safeguard Wendling, et al.
biodiversity when [58]
97
List Critical Factors Factor Description Referred by particularly for least developed countries.” [967] Attracting
smart technologies foreign investment from developed countries is one way to go.
are unfolded
Another benefit of ICT for developing countries is that it
Smart use of water Sensors and Big Fitsilis and
source Data gathering Kokkinaki [46]; reduces energy use and improves efficiency. Reducing energy
(TECHNOLOGICAL - allow real time Pawliczek, et al. use through, for instance, intelligent operation of
10 ENVIRONMENTAL) monitoring of water [59] transportation systems will be beneficial. Bhattacharya, et al.
quantity and [68] proposed a framework for sustainable smart city
quality. development in developing countries and noted that is
Urban innovation The growth in Haupt [55];
(TECHNOLOGICAL)
necessary to ensure facilities such as an assured supply of
innovation Caragliu and Del
activities, often Bo [60] electricity and water, safety and security, clean air and lack of
11 focused on pollution, reliable transport, suitable housing, sanitation and
technological many other basic factors that will be taken for granted in
solutions for local developed countries. Urban sustainability can only be realised
needs in developing countries if social, economic and environmental
Smart waste Better ways of Fitsilis and
management recycling waste Kokkinaki [46];
sustainable factors are operating well together. Another
12 approach to a conceptual model for smart city development in
(ENVIRONMENTAL - using technology Chen [61]
TECHNOLOGICAL) developing countries is that of Joia and Kuhl [69] who noted
Climate intelligent Using technology to Fitsilis and the special needs that should be considered in this context.
cities optimise energy use Kokkinaki [46];
13 (ENVIRONMENTAL) and maintain Pee and Pan [62] While digital inclusion is increasing rapidly, not everybody is
appropriate climate able to use this to its full advantage and local content and
Efficient resource Removal of Fitsilis and education are necessary, and should be customised for
utilization inefficient practices Kokkinaki [46]; particular communities or target groups. Other barriers might
(ENVIRONMENTAL - to improve the Singh and Sobti
TECHNOLOGICAL)
include poor basic infrastructure, poverty, poor legislation and
14 efficiency of [63]
resource utilisation
regulations, and economic instability [70].
through monitoring IV. CONCLUSION
via ICT
Smart ways of Dealing with Fitsilis and Smart cities offer the opportunity for efficiency and higher
15
dealing with security and privacy Kokkinaki [46]; quality of life based upon ICT. The green economy offers the
privacy issues issues through use Deebak, et al. [64] opportunity to move away from economic development at any
(TECHNOLOGICAL) of technology
cost and reconcile economic considerations and motivations
Healthy and Smart cities attract Fitsilis and
educated workforce knowledge workers Kokkinaki [46]; with environmental protection. The global crisis was a
16 (SOCIAL) significant factor leading to the rise of enthusiasm for the
Agrawal and
Challenges [65] green economy [71]. The green economy concept is based on
the idea that the separation of economic development and
Clearly these critical factors will present some additional environmental policies is an artificial one. Following the
challenges for developing countries, which face many barriers global financial crisis, many global bodies such as the UN, the
in terms of ICT. While the situation is improving, there are OECD, and the World Bank expressed the need for a rethink
still deficits under very constrained financial resource about alternative ways of overcoming the crisis of
situations. Both developed and developing countries are environmental degradation and transitioning toward a green
facing environmental challenges and wish to improve energy economy. The debate is around increasing human well-being
efficiency and waste management. Air pollution, water quality as well as reducing environmental problems. Many countries
and availability, and biodiversity are other factors affecting all are pursuing policies that move away from the traditional
countries, but particularly developing countries. ICT can be economic model towards a greener approach that places
very beneficial in tackling such challenges within developing ecology as central to development. There have been various
countries. Technology can play both a negative and positive attempts, some more successful than others. Transition to a
role in the environment. Digitisation affects many industries green economy can be undertaken at many levels including
and provides efficiencies. The production of ICT equipment regional/subnational/national, and the focus upon smart cities
and so on does require energy consumption and materials as makes sense because this is where populations are
well as producing e-waste. Smart applications can provide concentrated and where economic growth is high. Most
many advantages which will be highly beneficial for people live in cities and this number is rising. Cities are the
developing countries. Detection and early warning of disaster places where most of the environmental problems occur; for
events, efficient energy usage, and smart capability building instance, they account for about 75% of global energy and
such as technology awareness and transfer are key resource consumption with associated CO2 emissions [72].
opportunities for developing countries [66]. ICT is about
information and communication, as well as sustainable This research has reviewed literature concerning factors
development. N’dri, et al. [67] examine differences between that are critical in implementing a green economy within smart
developed and developing countries concerning reaching the cities. A comprehensive literature review revealed critical
sustainable development goals of the UN. Their study factors that from literature during the time period 2010 until
indicates that the stage of development of a country does 2021. These factors fall largely into the three groups of
influence CO2 emissions, and that less-developed countries technology, social and environmental. The implications of
use a lot of nonrenewable energy due to the higher cost of these factors for the implementation of a green economy and
renewable energy, and that: “… ICT promotion can be a smart cities in developing countries are considered and
powerful tool to fight such environmental degradation, suggestions are made for implementation. Future research
could collect data from decision-makers about the perceived
importance of each of these factors in a range of developing
98
countries. Having an understanding of the prioritisation of Brazilian manufacturing sectors," Climate Policy, vol. 18, no.
these factors would assist policymakers who are engaged in 2, pp. 246-257, 2018.
planning smart city developments within the context of a [20] M. J. Grant, A. J. H. i. Booth, and l. journal, "A typology of
developing country. reviews: an analysis of 14 review types and associated
methodologies," vol. 26, no. 2, pp. 91-108, 2009.
[21] Z. Munn, M. D. Peters, C. Stern, C. Tufanaru, A. McArthur,
ACKNOWLEDGEMENT and E. J. B. m. r. m. Aromataris, "Systematic review or
scoping review? Guidance for authors when choosing between
This research is funded by International University, a systematic or scoping review approach," vol. 18, no. 1, pp.
VNUHCM under grant number T2021-01-CE. 1-7, 2018.
[22] Limna, P., Kraiwanit, T. and Siripipatthanakul, S.,. "The
REFERENCES growing trend of digital economy: A review article".
International Journal of Computing Sciences Research,vol.6,
[1] V. Delitheou, V. Meleti, and C. G. Athanassopoulos, "Green pp.1-11, 2022
economy and smart city," Journal of Reliable Intelligent [23] F. Caprotti, "Smart eco-cities in the green economy," The
Environments, vol. 5, no. 4, pp. 235-240, 2019. Routledge Companion to Smart Cities, 2020.
[2] Y. Katagiri, M. Bono, and N. Suzuki, "Capturing [24] A. Savchenko and T. Borodina, "Green and Digital Economy
conversational participation in a ubiquitous sensor as a Means for Sustainable Development of Urban Areas,"
environment," in Pervasive 2004 Workshop on Memory and Izvestiya Rossiiskoi Akademii Nauk. Seriya
Sharing of Experiences, 2004, pp. 101-106. Geograficheskaya, vol. 84, no. 2, pp. 310-320, 2020.
[3] S. P. Mohanty, U. Choppali, and E. Kougianos, "Everything [25] A. Yudono, D. Satria, and A. Erlando, "Toward Inclusive
you wanted to know about smart cities: The internet of things Development Through Smart Economy in Malang Regency,"
is the backbone," IEEE Consumer Electronics Magazine, vol. in IOP Conference Series: Earth and Environmental Science,
5, no. 3, pp. 60-70, 2016. 2019, vol. 328, no. 1, p. 012008: IOP Publishing.
[4] H. Mohseni, "Public engagement and smart city definitions: A [26] L. Hao, X. Lei, Z. Yan, and Y. ChunLi, "The application and
classifying model for the evaluation of citizen power in 2025 implementation research of smart city in China," in 2012
Tehran," GeoJournal, vol. 86, no. 3, pp. 1261-1274, 2021. international conference on system science and engineering
[5] R. Giffinger, C. Fertner, H. Kramar, and E. Meijers, "City- (ICSSE), 2012, pp. 288-292: IEEE.
ranking of European medium-sized cities," Cent. Reg. Sci. [27] Y. Qian, J. Liu, Z. Cheng, and J. Y.-L. Forrest, "Does the smart
Vienna UT, vol. 9, pp. 1-12, 2007. city policy promote the green growth of the urban economy?
[6] A. Ghaffarianhoseini et al., "Intelligent or smart cities and Evidence from China," Environmental Science and Pollution
buildings: a critical exposition and a way forward," Intelligent Research, vol. 28, no. 47, pp. 66709-66723, 2021.
Buildings International, vol. 10, no. 2, pp. 122-129, 2018. [28] A. Y. Tamakhina, S. I. Mezhov, and S. V. Yungblyudt,
[7] G. Brundland, "World commission on environment and "Digital workforce for the “green” economy “smart”
development. Our common future oxford," ed: University enterprises: training and management features," in Institute of
Press.. Oxford, 1987. Scientific Communications Conference, 2019, pp. 345-353:
[8] U. UNEP, "Towards a green economy: Pathways to Springer.
sustainable development and poverty eradication," Nairobi, [29] B. Granier and H. Kudo, "How are citizens involved in smart
Kenya: UNEP, 2011. cities? Analysing citizen participation in Japanese``Smart
[9] E. Barbier, "The policy challenges for green economy and Communities''," Information Polity, vol. 21, no. 1, pp. 61-76,
sustainable economic development," in Natural resources 2016.
forum, 2011, vol. 35, no. 3, pp. 233-245: Wiley Online [30] P. Cardullo and R. Kitchin, "Being a ‘citizen’ in the smart city:
Library. Up and down the scaffold of smart citizen participation in
[10] OECD, "Towards green growth: monitoring progress," ed: Dublin, Ireland," GeoJournal, vol. 84, no. 1, pp. 1-13, 2019.
OECD: Paris, France, 2011. [31] C. F. Capra, "The Smart City and its citizens: Governance and
[11] M. Fay, Inclusive green growth: The pathway to sustainable citizen participation in Amsterdam Smart City," International
development. World Bank Publications, 2012. Journal of E-Planning Research (IJEPR), vol. 5, no. 1, pp. 20-
[12] E. Loiseau et al., "Green economy and related concepts: An 38, 2016.
overview," Journal of cleaner production, vol. 139, pp. 361- [32] R. Kusumastuti and J. Rouli, "Smart City Implementation and
371, 2016. Citizen Engagement in Indonesia," in IOP Conference Series:
[13] R. Ferrara, "The smart city and the green economy in Europe: Earth and Environmental Science, 2021, vol. 940, no. 1, p.
A critical approach," Energies, vol. 8, no. 6, pp. 4724-4734, 012076: IOP Publishing.
2015. [33] E. Przeybilovicz et al., "Citizen participation in the smart city:
[14] European Commission, "Integrated Sustainable Urban findings from an international comparative study," Local
Development, Cohesion Policy 2014–2020," ed: European government studies, vol. 48, no. 1, pp. 23-47, 2022.
Commission Brussels, Belgium, 2012. [34] L. Berntzen and M. R. Johannessen, "The role of citizen
[15] B. Donald, "Economic competitiveness and quality of life in participation in municipal smart city projects: Lessons learned
city regions: compatible concepts?," Canadian Journal of from Norway," in Smarter as the new urban agenda: Springer,
Urban Research, pp. 259-274, 2001. 2016, pp. 299-314.
[16] Ö. F. Gültekin and B. Erenoğlu, "A research on policies for [35] J. Tadili and H. Fasly, "Citizen participation in smart cities: a
green economy in developed and developing countries within survey," in Proceedings of the 4th International Conference on
the scope of sustainable development," Global Economic Smart City Applications, 2019, pp. 1-6.
Observer, vol. 6, no. 1, pp. 33-38, 2018. [36] C. C. Snow, D. D. Håkonsson, and B. Obel, "A smart city is a
[17] H. Ward and X. Cao, "Domestic and international influences collaborative community: Lessons from smart Aarhus,"
on green taxation," Comparative Political Studies, vol. 45, no. California Management Review, vol. 59, no. 1, pp. 92-108,
9, pp. 1075-1103, 2012. 2016.
[18] K. Kuralbayeva, "Environmental taxation, employment and [37] R. Anguluri and P. Narayanan, "Role of green space in urban
public spending in developing countries," Environmental and planning: Outlook towards smart cities," Urban Forestry &
resource economics, vol. 72, no. 4, pp. 877-912, 2019. Urban Greening, vol. 25, pp. 58-65, 2017.
[19] C. Gramkow and A. Anger-Kraavi, "Could fiscal policies [38] Q. Liu, L. Hou, S. Shaukat, U. Tariq, R. Riaz, and S. S. Rizvi,
induce green innovation in developing countries? The case of "Perceptions of spatial patterns of visitors in urban green
spaces for the sustainability of smart city," International
99
Journal of Distributed Sensor Networks, vol. 17, no. 8, p. 155, [56] C. V. Martín, E. Rainer, and H. Schnitzer, "Smart city Graz
2021. from the vision to the action " Places and Technologies 2014
[39] A. A. Gavrilidis, M. R. Niță, D. A. Onose, D. L. Badiu, and I. p. 276, 2014.
I. Năstase, "Methodological framework for urban sprawl [57] J. Colding et al., "Applying a systems perspective on the
control through sustainable planning of urban green notion of the smart city," Smart Cities, vol. 3, no. 2, pp. 420-
infrastructure," Ecological indicators, vol. 96, pp. 67-78, 429, 2020.
2019. [58] L. A. Wendling, A. Huovila, M. zu Castell-Rüdenhausen, M.
[40] Z. Chen, "Application of environmental ecological strategy in Hukkalainen, and M. Airaksinen, "Benchmarking nature-
smart city space architecture planning," Environmental based solution and smart city assessment schemes against the
Technology & Innovation, vol. 23, p. 101684, 2021. sustainable development goal indicator framework," Frontiers
[41] F. Reinwald et al., "Urban green infrastructure planning as a in Environmental Science, vol. 6, p. 69, 2018.
contribution to the smart ‘green’ city," in REAL CORP 2014– [59] A. Pawliczek, J. Chlopecký, O. Oberreiterová, and L.
PLAN IT SMART! Clever Solutions for Smart Cities. Moravec, "Havířov: “The City of Green” and Its Fourth Sector
Proceedings of 19th International Conference on Urban Sustainability Activities. One Case Study from the Czech
Planning, Regional Development and Information Society, Republic," in Social Innovation and Entrepreneurship in the
2014, pp. 757-761: CORP–Competence Center of Urban and Fourth Sector: Springer, 2021, pp. 171-211.
Regional Planning. [60] A. Caragliu and C. F. Del Bo, "Smart innovative cities: The
[42] K. Narmada and G. Bhaskaran, "Evaluation and planning of impact of Smart City policies on urban innovation,"
urban green space network in landscape planning of Ponneri, Technological Forecasting and Social Change, vol. 142, pp.
an emerging smart city in Tamil Nadu," International Journal 373-383, 2019.
of Humanities and Social Sciences vol. 5, pp. 39-46, 2016. [61] X. Chen, "Machine learning approach for a circular economy
[43] K. Gupta, K. Puntambekar, A. Roy, K. Pandey, and P. Kumar, with waste recycling in smart cities," Energy Reports, vol. 8,
"Smart environment through smart tools and technologies for pp. 3127-3140, 2022.
urban green spaces," in Smart Environment for Smart Cities: [62] L. Pee and S. L. Pan, "Climate-intelligent cities and resilient
Springer, 2020, pp. 149-194. urbanisation: Challenges and opportunities for information
[44] K.-L. Wang, S.-Q. Pang, F.-Q. Zhang, Z. Miao, and H.-P. Sun, research," International Journal of Information Management,
"The impact assessment of smart city policy on urban green vol. 63, p. 102446, 2022.
total-factor productivity: Evidence from China," [63] D. K. Singh and R. Sobti, "Long-Range Real-Time
Environmental Impact Assessment Review, vol. 94, p. Monitoring Strategy for Precision Irrigation in Urban and
106756, 2022. Rural Farming in Society 5.0," Computers & Industrial
[45] G. V. Pereira, P. Parycek, E. Falco, and R. Kleinhans, "Smart Engineering, p. 107997, 2022.
governance in the context of smart cities: A literature review," [64] B. Deebak et al., "Seamless privacy-preservation and
Information Polity, vol. 23, no. 2, pp. 143-162, 2018. authentication framework for IoT-enabled smart eHealth
[46] P. Fitsilis and A. Kokkinaki, "Smart Cities Body of systems," Sustainable Cities and Society, vol. 80, p. 103661,
Knowledge," in 25th Pan-Hellenic Conference on Informatics, 2022.
2021, pp. 155-159. [65] A. J. S. C. O. Agrawal and Challenges, "Exploring the
[47] Q. Xin, M. Alazab, V. G. Díaz, C. E. Montenegro-Marin, and Attributes of Smart City from Organisation’s Perspective," pp.
R. G. Crespo, "A deep learning architecture for power 615-631, 2020.
management in smart cities," Energy Reports, vol. 8, pp. [66] J. W. Houghton, "ICT and the Environment in Developing
1568-1577, 2022. Countries: an Overview of Opportunities and Developments,"
[48] S. Yadav, S. Patel, D. J. Killedar, S. Kumar, and R. J. J. o. E. Communications & Strategies, no. 76, p. 39, 2009.
M. Kumar, "Eco-innovations and sustainability in solid waste [67] L. M. N’dri, M. Islam, and M. Kakinaka, "ICT and
management: An indian upfront in technological, environmental sustainability: any differences in developing
organizational, start-ups and financial framework," vol. 302, countries?," Journal of Cleaner Production, vol. 297, p.
p. 113953, 2022. 126642, 2021.
[49] R. U. Pandey, T. Mitra, M. Wadwekar, J. Nigam, and K. [68] T. R. Bhattacharya, A. Bhattacharya, B. Mclellan, and T.
Trivedi, "Green Infrastructure as a Tool for Improving Tezuka, "Sustainable smart city development framework for
Livability of Area Based Development Projects Under Smart developing countries," Urban Research & Practice, vol. 13,
City Mission," in Geospatial Technology and Smart Cities: no. 2, pp. 180-212, 2020.
Springer, 2021, pp. 447-468. [69] L. A. Joia and A. Kuhl, "Smart city for development: A
[50] D. Passarelli, "The Future of City/Cities," in Economic and conceptual model for developing countries," in International
Policy Implications of Artificial Intelligence: Springer, 2020, conference on social implications of computers in developing
pp. 161-169. countries, 2019, pp. 203-214: Springer.
[51] O. Embarak, "Smart City Transition Pillars With Layered [70] D. O. Aghimien et al., "A fuzzy synthetic evaluation of the
Applications Architecture," Procedia Computer Science, vol. challenges of smart city development in developing
191, pp. 57-64, 2021. countries," Smart and Sustainable Built Environment, 2020.
[52] P. Girardi and A. Temporelli, "Smartainability: a [71] L. Georgeson, M. Maslin, and M. Poessinouw, "The global
methodology for assessing the sustainability of the smart city," green economy: a review of concepts, definitions,
Energy Procedia, vol. 111, pp. 810-816, 2017. measurement methodologies and their interactions," Geo:
[53] E. Estevez, K. Cenci, P. Fillottrani, and T. Janowski, "Review Geography and Environment, vol. 4, no. 1, p. e00036, 2017.
of International Standards and Policy Guidelines for Smart [72] R. Simpson, "Introduction: A green economy for green cities,"
Sustainable Cities," Smart Cities and Smart Governance, pp. in The Economy of Green Cities: Springer, 2013, pp. 13-16.
69-99, 2021.
[54] P. La Greca and F. Martinico, "Energy and spatial planning: a
smart integrated approach," in Smart Energy in the Smart
City: Springer, 2016, pp. 43-59.
[55] W. Haupt, "The Sustainable and the Smart City:
Distinguishing Two Contemporary Urban Visions," Springer,
2021.
100
Impacts of Green Training and Green Competencies

on Organizational Green Performance: Empirical
Case of Companies in Dong Nai Parks
Thanh-Lam Nguyen Doan Thi Chuyen
Dong Nai 810000, Vietnam Bien Hoa National Institute of Forensic Psychiatry
lamnt@lhu.edu.vn Dong Nai 810000, Vietnam
doanthichuyen.inv@gmail.com
Nguyen Thi Phuong Thao Doan Van Ly

Lac Hong University Dong Nai University Lac Hong University
Dong Nai 810000, Vietnam Nong Lam University - Ho Chi Minh City, Ninh Thuan
nguyenthiphuongthao24783@gmail.com Branch Ninh Thuan, Vietnam
ly.dovan@hcmuaf.edu.vn
Abstract: Due to the recent climate change and ruined implementation of green business procedures [10]. Chams &
environment, businesses have especially contemplated the Garcia-Blandon [8] also found that employees are more likely
abrupt natural environment in the development of key to accomplish long-term objectives once they are engaged in
competencies among their employees for the sake of their pioneering processes and new practices beneficial to the
sustainable growth and green performance. This study aims at environment. Mircetic et al. [2] stated that an individual is
investigating the statistical impacts of green training and considered pro-environmentally conscious if he/she
green competencies on organizational green performance. demonstrates green competencies and responsible
From a formal survey with self-completed questionnaires willingness to help others with eco-friendly and creative
delivered to key personnel who took part in relevant green methods to improve not only organizational performance in
training supported by the companies in Dong Nai industrial terms of long-term development, innovation, pro-
parks, it is found that green training is crucial for them to environmental and financial performance. Moreover, existing
continuously improve the green competencies of their literatures reveal that there is a positive relationship between
employees and their green performance. Thus, for their the green training provided by organizations and their
sustainable development, companies should organize regular employees’ green competencies [11], [12]. However, such
training courses to enhance employees’ green competencies relationships in the case of companies in industrial parks are
and preferably recruit those with green competencies for their still open for further research. As a consequence, this study
business. aims at investing the possible impacts of green training and
green competencies on their green performance with the
Keywords: Green training, Green competencies, Green empirical case of the companies located in Dong Nai
performance industrial parks so that we can propose some feasible
managerial implications to improve their green performance.
I. INTRODUCTION II. LITERATURE REVIEWS
The recent advances in science and technology as A. Green performance
well as the serious crises have had substantial impacts on
human daily activities and the living environment [1]–[3], Chen et al. [13] defined green performance of an
resulting in several challenges to be dealt with [4]. Thus, organization as the positive impacts of its green initiatives on
organisations are urged to have proper strategies for their the natural environment. Lirn et al. [14] and Yang [15]
resilience and sustainable development. In order to be more proposed several indicators to measure the green
competitive and sustainable, organisations must keep their performance in terms of reduction in energy consumption,
business continuously moving, more innovative, more reduction in superfluous waste, reduction in emission of
comprehensive and more pro-environmental [5], [6]. greenhouse gases, increase in workers’ appreciation of
Sustainability is a long-standing target which should be well environmental protection, regulation establishment and
integrated into their development strategies [7]. Chams & enforcement, overall green cost, and green image of a
Garcia-Blandon [8] pointed out that accomplishing ethical company, etc. Perramon et al. [16] pointed out that when
and environmental objectives should be considered as key environmental protection measures, for example saving
drivers for organisations to survive in the contemporary energy and water, are implemented on daily operation, the
business environment. To do so, it is critical for organisations overall environmental performance of the organization is
to well circulate the green awareness to their employees [9] significantly improved. Similarly, Yang [15] found that
because they are one of the vital constituents in the successful internal green practices in 129 container shipping companies

and agencies in Taiwan positively influence their green and environmental problems in the development of products/
performance. services and operations [40]. Thus, the vivid skills may
include professional skills, vocational skills, and generic
B. Green competencies skills which are all important for both personal development
Green performance can be also defined as the ultimate and organizational sustainability [41]. Kanyimba et al. [42]
output or evident conduct resulting from the amalgamation claimed that green skills support educational initiatives for
of natural, acquired and adapting competencies [17]–[20]. sustainable development and enhance pro-environmental
Rodriguez et al. [21] and Murray [22] described activities, resulting in positive impacts on environmental
competencies as the exposition of personal skills, active management systems, the financial performance and the
knowledge, capabilities, behaviours and other characteristics environmental performance at organizational level [29].
relating to a particular task that improves job performance or Hence, such skills provide a good foundation for data analysis
professional functions. Due to the recent climate change and and knowledge formation to produce feasible solutions and
environmental degradation, companies have taken several actions to deal with real-life challenges relating to the
actions to develop their employees’ competencies towards the environment [43].
current natural environment to ensure their sustainable 3) Green abilities: Goffin & Woycheshin [44] defined
development. Such competencies are called green abilities as “the inborn characters that make the learning
competencies which are defined as “the requisite ecological easier and/or results in better performance”. Gerhart [45]
knowledge, skills and other socio-economic behaviour an found that green competencies of an employee are promoted
individual has to help him/her behave and act rightly and from his/her green abilities. Green abilities allow employees
responsibly towards the overall well-being of his/her to perfect themselves and boost their performance to conquer
immediate environment” [23]. Due to the ambiguity of the environmental protections by promoting their altruism, life-
term “green competencies”, several researchers have value perceptions, and positive biospheres in the organisation
proposed similar terminologies such as pro-environmental with their green behaviour, attitude and knowledge [46].
competency [24]–[26], sustainable capabilities [27], [28],
green abilities [29], and environmental competence, green 4) Green awareness: Green awareness, also called
skills [30], [31], among others. However, green competencies environmental awareness, is depicted as “being aware of the
can be classified into two groups, including natural green natural environment and making choices that benefit rather
competencies and acquired green competencies [23], where than hurt the earth”. It has been well regarded as one of the
natural green competencies comprise of inborn characters vital factors in the successful implementation of
such as green concern, while acquired ones include green environmental management systems. Green awareness in an
knowledge, green skills, green behaviour and the expertise organisation enables its employees to pay more attention to
acquired through personal learning and experiences. avoid possible adverse effects of their actions on the
Basically, acquired green competencies are usually environment and take immediate remedy for harmful impacts
considered more important than natural ones because green [47]–[51].
attitude towards conservation and engagement in green
behaviour are believed more practical and valuable than 5) Green attitude: Ajzen [52] defined attitude as “the
meager green concern. Green competencies are a multi-facet degree to which a person has a favourable or unfavourable
scale of six different dimensions, including green knowledge, evaluation or appraisal of the behaviour in question”. And
green skills, green abilities, green awareness, green attitude green attitude has been defined as the attitude of an individual
and green behaviour [29], [32]–[34]. The following in assessing the value of environmental protection and the
subsections provide more details about these concepts. engagement in environmental performance [53]. It is a key
driver that urges constituents to commit green behaviour [54].
1) Green knowledge: Fryxell & Lo [35] defined green Hence, Major et al. [55] concluded that providing
knowledge as “the general knowledge about the facts, environmental education is vital to develop the green attitude
concepts, and relationship regarding the natural environment among employees targeted for the organization sustainability
and the entire ecosystem”. Kollmuss & Agyeman [36] and environmental awareness to promote the green traits
emphasized the linkage between the personal knowledge and among the employees.
the existing environmental problems as well as feasible
solutions to deal with such problems in a green manner. The 6) Green behavior: Ones & Dilchert [63] defined green
green knowledge is classified into two categories, including behaviour as the scalable action and behaviour of an
objective green knowledge (factual knowledge about individual aiming at providing positive contributions to
environmental problems) and subjective green knowledge environmental sustainability. The “green” mentioned in green
(knowledgeable actions against the identified environmental behaviour is interchangeably used with other terminologies
problems) [37]. Thus, green knowledge is found valuable to such as environmental responsible, environmentally
employees as it helps employees to have the ecosystem sustainable, and ecofriendly [64]. Typical examples of green
mindset and the capability to think and behave responsibly behaviour include the purchasing and using green products
towards their initiatives in environment protection [38]. that are efficient in energy consumption, reusable and
recyclable as well as emit least pollutants [65], developing
2) Green skills: Green skills are defined as “the green production and regular monitoring environmental costs
professional and vocational skills, as well as the generic skills which lead to constructive impacts on environmental
(such as sustainable approaches, innovation and problem- management and significant increase in overall financial
solving) required for new green jobs and the greening of performance [66]. Moreover, Casalo & Escario [67] claimed
existing jobs across all industry sectors as a response to that there is a heterogeneous relationship between green
climate change and sustainability imperatives” [39]. They attitude and green behaviour; specifically, employees with
are also defined as “vivid skills” that consider climate change stronger green attitude towards the environment protection
102
tend to have stronger green behaviour. rationality of the word usage and the meaning of each
statement listed. Their responses were carefully considered in
C. Green training the second stage to incorporate into a refined questionnaire.
Teixeira et al. [56] defined green training as “a type of The questionnaire consists of 20 observed variables, and each
training related to relevant environmental topics, which of them was evaluated on a 5-Likert scale where 1 indicates
enables all staff to integrate the firm’s performance with the lowest level and 5 indicates the highest level. From a pilot
environmental issues”. Green training is actually a continuing study with 35 valid observations, by using reliability analysis
education which is designed to provide employees with with Cronbach’s Alpha coefficient, the items in the
relevant updates on green knowledge and skills to achieve questionnaire are all validated as their corrected item-total
organization sustainability [56]–[58]. Ren et al. [11] pointed correlations are in [0.43,068]; thus, the questionnaire is used
out that providing green training is beneficial to its in the third stage which is actually a formal survey and data
organization as it helps to develop not only the employees’ analysis. In this study, an average value of the observed items
attitude and behaviour but also their green knowledge, green of each factor is assigned as the value of the factor. The
skills and green abilities to enhance their job performance official survey was conducted from February 15, 2022 to
with green awareness. Green training results in significant March 15, 2022. We totally delivered 200 hard copies of the
improvement of not only the awareness and accountability for final questionnaire to 100 companies with 200 key personnel
environmental problems [59], [60], but also the technical (including managers/vice managers of personnel office and
knowledge and relevant skills of employees in dealing with production office in each company) who took part in relevant
environmental problems [12], [58]. Consequently, green green training supported by the companies in Dong Nai
training has been proved to associate with certain expected industrial parks. There are 149 out of 200 pieces of completed
organizational outcomes, such as environmental practice of questionnaires collected. Among them, there were 28 pieces
firms [61]–[67], green procurement [57], green supply chain invalid; so, data from 121 valid observations were finally
management [56], [62], successful implementation of analyzed in this study. This study employs traditional linear
environmental management [73], etc. regression models to test the proposed hypotheses.
Specifically, Table 1 briefly presents relevant models to be
Literally, as a continuing education process, green investigated in this study (GT: Green training; GC: Green
training usually has four key phases, including: (1) identify competencies; GP: Green performance).
practical needs of green training, (2) design appropriate green
training plans, (3) implement green training sessions, and (4) Table 1. Linear regression models investigated in this study
evaluate the training effectiveness [74]. Unnikrishnan & Model No. Dependent Var. Independent Var. Hypotheses
Hegde [75], Brenner et al. [76], and Renwick et al. [77] LRM1 GC GT H1
claimed that there are two content areas of green training,
LRM2 GP GT, GC H2, H3
including: (1) awareness and responsibility, and (2) technical
knowledge and skills. Thus, it is critical to have all members Table 2. Analysis results of regression models
in an organization to fully capture the great benefits of green
training and responsibly engage in the training process for the Model Dependenta Independentb S.Co.c Sig.d Hypotheses
sake of the sustainable development goals of the LRM1 GC GT 0.384 *** H1 supported
organization [78]. LRM2 GP GT 0.258 *** H2 supported
GC 0.427 *** H3 supported
D. Research hypotheses
From the above-mentioned concepts, this study will Notes: a Dependent variable; b Independent variable; c Standardized
investigate the following hypotheses and the proposed Coefficients; d Significance level; *** ≤ 0.001
research model is presented in Figure 1:H1: Green training
positively affects personal green competencies. IV. EMPIRICAL RESULT
• H2: Green training positively affects organizational Table 2 shows that all of the three hypotheses are well
green performance. supported. This means that there is a significant impact of
green training on employees’ green competencies, especially
• H3: Personal green competencies positively affect the acquired ones, which further agrees with the findings of
organizational green performance. Ren et al. [11] and Pellegrini et al. [78]; and the
organizational green performance is significantly affected by
both green training and employees’ green competencies.
These figures indicate that companies providing green
training to their employees will have good gains in their green
performance in terms of energy consumption, unnecessary
waste, emission of greenhouse gases, awareness of
Figure 1. Proposed research model
environmental protection, overall green cost, and
organization’s green image, etc. as the claims by Lirn et al.
III. RESEARCH METHOD [14] and Yang [15], i.e., the green training and employees’
This study is conducted in three stages. Specifically, the green competencies lead to the attainment of the sustainable
first stage aims at creating a list of possible statements for a development goals of the organization [78]. Compared to
self- complete questionnaire to be used in an official survey. green training, personal green competencies are found to
In this stage, we conducted a qualitative research with five have stronger impacts on the organizational green
experts working as human resource managers at five performance while the green training helps to develop the
companies in Dong Nai industrial parks to gauge the acquired competencies of employees. Hence, it is
103
recommended that companies should organize regular management research in emergence: A review and future directions,
training courses to enhance not only employees’ Asia Pacific Journal of Management, 1-35, 2017.
consciousness and accountability for environmental problems [12] M. Pinzone, M. Guerci, E. Lettieri, E. Huisingh, 2019. Effects
[59], [60], but also their green knowledge and skills in dealing of ‘green’ training on pro-environmental behaviors and job
with environmental problems [12], [58]. In addition, they satisfaction: evidence from the Italian healthcare sector. J. Clean.
should also integrate a preference of candidates with green Prod. 226, 221- 232, 2019.
competencies (either natural or acquired or both) during [13] F. Chen, T. Nagistedema, S. Li, A cross-country comparison
their recruitment. of green initiatives, green performance and financial performance.
Manag. Decis. 56 (5), 1008-1032, 2017.
V. CONCLUSION [14] T.C. Lirn, Y.C.J. Wu, Y.J. Chen, Green performance criteria
for sustainable ports in Asia. Int. J. Phys. Distrib. Logist. Manag. 43
Due to the recent climate change and ruined environment, (5/6), 427-415, 2013.
businesses have especially contemplated the abrupt natural [15] C.S. Yang, An analysis of institutional pressures, green
environment in the development of key competencies among supply chain management, and green performance in the container
their employees for the sake of their sustainable growth and shipping context. Transport. Res. Transport Environ. 61, 246-260.
green performance. Providing green training is crucial for Part B, 2018.
organizations to continuously improve the green [16] J. Perramon, M.D.M. Alonso-Almeida, J. Llach, L.B. Bagur-
competencies of their employees and their green Femenias, Green practices in restaurants: impact on firm
performance. The statement is further validated in the case of performance. Operations Management Research 7 (1e2), 2-12,
companies located in Dong Nai industrial parks. As such, for 2014.
their sustainable development, companies should organize [17] S.B. Follows, D. Jobber, D. Environmentally responsible
regular training courses to enhance employees’ green purchase behaviour: A test of a consumer model. Eur. J. Mark. 34,
competencies and preferably recruit those with green 723-746, 2000.
competencies. [18] B.C. Tan, The role of perceived consumer effectiveness on
value-attitude-behavior model in green buying behavior context.
REFERENCES Aust. J. Basic Appl. Sci. 5, 1766–1771. 2011.
[1] M. Popovic, An MCDM approach for personnel selection [19] T. Mainieri, E.G. Barnett, T.R. Valdero, J.B. Unipan, S.
using the CoCoSo method. J. Process Manag. New Technol. 9, 78- Oskamp, Green buying: The influence of environmental concern on
88, 2021. consumer behavior. J. Soc. Psychol. 137, 189-204, 1997.
[2] B. Dordevic, A process for assessing organization. J. Process [20] H.R. Kaufmann, M.F.A.K. Panni, Y. Orphanidou, Factors
Manag. New Technol. 9, 30-44, 2021. affecting consumers’ green purchasing behavior: An integrated
[3] V. Mircetic, S. Vukotic, D. Cvijanovic, The concept of conceptual framework. Amfiteatru Econ. 14, 50-69, 2012.
business clusters and its impact on tourism business improvement. [21] D. Rodriguez, R. Patel, A. Bright, D. Gregory, M.K. Gowing,
Econ. Agric. 66, 851- 868, 2019. Developing competency models to promote integrated human
[4] T.V. Pham, Y.F. Huang, M.H. Do, The Effect of Critical Work resource practices, Human Resource Management, 41(3), 309-324,
Factors on Job Motivation and Satisfaction: An Empirical Study 2002.
from Vietnam. J. Asian Financ. Econ. Bus. 9, 399-408, 2022. [22] P. Murray, Organisational learning, competencies, and firm
[5] T.C. Ambos, K. Tatarinov, Building Responsible Innovation performance: empirical observations, The Learning Organization,
in International Organizations through Intrapreneurship. J. Manag. 10(5), 305-316, 2003.
Stud. 59, 92-125, 2022. [23] N. Subramanian, M.D. Abdulrahman, L. Wu, P. Nath, Green
[6] V.B. Arsic, M. Latinovic, N. Petrovic, Sustainable switching competence framework: evidence from China, The International
option in green hotels construction. In Proceedings of the INBAM Journal of Human Resource Management, 27(2), 151-172, 2016.
2013 - 3rd Conference of the International Network of Business and [24] V. Corral-Verdugo, A structural model of pro-environmental
Management Journals, Lisbon, Portugal, 13-17 June 2013. competency, Environment and Behavior, 34(4), 531-549, 2002.
[7] O.R. Dania, E.O. Okobia, Staff awareness of the role of [25] B. Fraijo-Sing, V. Corral-Verdugo, C. Tapia-Fonllem, D.
education in sustainable development goals, Evidence from the Gonzalez- Lomeli, Promoting pro-environmental competency, in
University of Benin. In Implementing the Sustainable Development Psychological Approaches to Sustainability, Nova, 225-246, 2010.
Goals in Nigeria Barriers, Prospects and Strategies, 1st ed.; Ekhator, [26] B. Fraijo-Sing, C. Tapia-Fonllem, V. Corral-Verdugo,
E.O., Miller, S., Igbinosa, E., Eds.; Routledge: London, UK, 2022. Designing environmental education: modeling pro-environmental
[8] N. Chams, J. Garcia-Blandon, On the importance of competency, in Appleton,
sustainable human resource management for the adoption of J. (Ed.), Values in Sustainable Development, Routledge, 244-254,
sustainable development goals. Resour. Conserv. Recycl. 141, 109- 2014.
122, 2019. [27] I. Thomas, T. Day, Sustainability capabilities, graduate
[9] V. Tornjanski, M. Cudanov, Towards Society 5.0 Era: capabilities, and Australian universities, International Journal of
Organisational Empowerment of the Sustainable Future. In Sustainability in Higher Education, 15(2), 208-227, 2014.
Proceedings of the 3rd Virtual International Conference Path to a [28] I. Thomas, J. Depasquale, Connecting curriculum,
Knowledge Society-Managing Risks and Innovation, Complex capabilities and careers, International Journal of Sustainability in
System Research Centre, Nis and Mathematical Institute of the Higher Education, 17(6), 738-755, 2016.
Serbian Academy of Sciences and Arts, Nis, Serbia, 15–16 [29] M.H. Wu, W. Thongma, W. Leelapattana, M.L Huang, Impact
November 2021; 413–421. of hotel employee’s green awareness, knowledge, and skill on
[10] D.W.S. Renwick, Contemporary Developments in Green hotel’s overall performance, Advances in Hospitality and Leisure,
Human Re- source Management Research, Towards Sustainability 12, 68-81, 2016.
in Action? Rout- ledge: London, UK, 2020. [30] K. Cleverley, What type of green skills will you need to Be
[11] S. Ren, G. Tang, E.S. Jackson, Green human resource successful in the business sector? Sustainability, Journal Record,
104
7(2), 92-93, 2014. air pollution effect in Chinese stock market. J. Clean. Prod. 185,
[31] G. Dlimbetova, Z. Zhylbaev, L. Syrymbetova, A. Aliyeva, 446-454, 2018.
Green skills for green economy: case of the environmental education [52] I. Ajzen, The theory of planned behavior. Organ. Behav.
role in Kazakhstan’s economy, International Journal of Hum. Decis. Process. 50, 179-211, 1991.
Environmental and Science Education, 11(8), 1735-1742, 2016. [53] K. Lee, Opportunities for green marketing: young consumers.
[32] B. Zareie, N.J. Navimipour, The impact of electronic Market. Intell. Plan. 26, 573-586, 2008.
environmental knowledge on the environmental behaviors of [54] S. Bergin-Seers, J. Mair, Emerging green tourists in
people, Computers in Human Behavior, 59, 1-8, 2016. Australia: Their behaviours and attitudes. Tourism Hospit. Res. 9,
[33] C. Cabral, C.J. Chiappetta Jabbour, Understanding the human 109-119, 2009.
side of green hospitality management, International Journal of [55] L. Major, Z. Namestovski, R. Horak, A. Bagany, V.P. Krekic,
Hospitality Management, 102389, 2019. Teach it to sustain it! Environmental attitudes of Hungarian teacher
[34] C. Cabral, R.L. Dhar, Green competencies: construct training students in Serbia. J. Clean. Prod. 154, 255-268. 2017.
development and measurement validation”, Journal of Cleaner [56] A.A. Teixeira, C.J.C. Jabbour, H. Latan, Green training and
Production, 235, 887-900, 2019. green supply chain management: Evidence from Brazilian firms. J.
[35] G.E. Fryxell, C.W.H. Lo, The influence of environmental Clean. Prod. 116 (1), 170e176, 2016.
knowledge and values on managerial behaviours on behalf of the [57] Z. Liu, X. Li, X. Peng, S. Lee, Green or non-green
Environment: An empirical examination of managers in China. J. innovation? Different strategic preferences among subsidized
Bus. Ethics 46, 45-69, 2003. enterprises with different ownership types. J. Clean. Prod. 245,
[36] A. Kollmuss, J. Agyeman, Mind the gap: Why do people 2020.
behave environmentally and what are the barriers to pro- [58] M. Pinzone, M. Guerci, E. Lettieri, T. Redman, Progressing
environmental behaviour. Environ. Educ. Res. 8, 239-260, 2002. in the change journey towards sustainability in healthcare: The role
[37] A. Pagiaslis, A.K. Krontalis, Green consumption behavior of ‘green’ HRM. J. Clean. Prod. 122, 201-211, 2016.
antecedents: environmental concern, knowledge, and beliefs. [59] G.M. Perron, R.P. Cote, J.F. Duffy, Improving environmental
Psychol. Market. 31, 335-348. 2014. awareness training in business. J. Clean. Prod. 14 (6-7), 551-562,
[38] M. Laroche, Targeting consumers who are willing to pay 2006.
more for environmentally friendly products. J. Consum. Mark. 18 [60] K. Sammalisto, T. Brorson, Training and communication in
(6), 503-520. J. Consum. Mark. 18, 503-520, 2001. the implementation of environmental management systems (ISO
[39] M. Brown, The development of green skills through the local 14001): A case study at the University of Gavle, Sweden. J. Clean.
TAFE Institute as a potential pathway to regional development. Int. Prod. 16 (3), 299- 309, 2008.
J. Train. Res. 11, 27-43. 2013. [61] J. Sarkis, P. Gonzalez-Torre, B. Adenso-Diaz, Stakeholder
[40] DEBIS (Department for Business Innovation and Skills), pressure and the adoption of environmental practices: the mediating
2011. D. Of E. and C., Change and Department of Environment, F. effect of training. J. Oper. Manag. 28 (2), 163-176, 2010.
and R.A. (England) (Skills for a green economy). [62] C.J.C. Jabbour, A.B.L. de Sousa Jabbour, Green human
[41] CEDEFOP (European Centre for the Development of resource management and green supply chain management: Linking
Vocational Training), 2012. Green skills and environmental two emerging agendas. J. Clean. Prod. 112, 1824-1833, 2016.
awareness in vocational education and training. [63] C.J.C. Jabbour, A.B.L. de Sousa Jabbour, Low-carbon
[42] A.T. Kanyimba, B.W. Richter, S.P. Raath, The effectiveness operations and production: Putting training in perspective. Ind.
of an environmental management system in selected South African Commer. Train. 46 (6), 327-331, 2014.
primary schools. J. Clean. Prod. 66, 479-488. 2014 [64] C.J.C. Jabbour, D. Jugend, A.B.L. de Sousa Jabbour, A.
[43] H. Karimzadegan, H. Meiboudia, Exploration of Gunasekaran, H. Latan, Green product development and
environmental literacy in science education curriculum in primary performance of Brazilian firms: measuring the role of human and
schools in Iran. Procedia - Soc. Behav. Sci. 46, 404-409, 2012. technical aspects. J. Clean. Prod. 87, 442-451, 2015.
[44] R.D. Goffin, D.E. Woycheshin, An empirical method of [65] C.J.C. Jabbour, A.L. Mauricio, A.B.L. de Sousa Jabbour,
determining employee competencies/KSAOs from task-based Job Critical success factors and green supply chain management
analysis. Mil. Psychol. 18, 121-130. 2006. proactivity: shedding light on the human aspects of this relationship
[45] B. Gerhart, Human resources and business Performance: based on cases from the Brazilian industry. Prod. Plan. Control 28
findings, unanswered questions, and an alternative approach. Manag. (6-8), 671-683, 2017.
Rev. 16, 174-185, 2005. [66] S.E. Jackson, D.W.S. Renwick, C.J.C. Jabbour, M. Muller-
[46] I. Rajiani, H. Musa, b. Hardjono, Ability, motivation and Camen, State- of-the-art and future directions for green human
opportunity as determinants of green human resources management resource management: Introduction to the special issue. Ger. J. Hum.
innovation. Res. J. Bus. Manag. 10, 51-57, 2016. Resour. Manag. 25 (2), 99-116, 2011.
[47] D.L. Gadenne, J. Kennedy, C. McKeiver, An empirical study [67] S.K. Singh, J. Chen, M. Del Giudice, A.N. El-Kassar,
of environmental awareness and practices in SMEs. J. Bus. Ethics Environmental ethics, environmental performance, and competitive
84, 45-63, 2009. advantage: role of environmental training. Technol. Forecast. Soc.
[48] S. Garcia, A. Cordeiro, I. Naas, P.L. Costa Neto, The Chang. 146, 203-211, 2019.
sustainability awareness of Brazilian consumers of cotton clothing. [68] D.S. Ones, S. Dilchert, Employee green behaviors. In:
J. Clean. Prod. 215, 1490-1502, 2019. Jackson, E., Ones, D.S., Dilchert, S. (Eds.), Managing Human
[49] X. Peng, Y. Liu, Behind eco-innovation: Managerial Resources for Environmental Sustainability. Jossey - Bass, San
environmental awareness and external resource acquisition. J. Clean. Francisco, CA, 85-116, 2012.
Prod. 139, 347- 360, 2016. [69] Y.F. Wang, Modeling predictors of restaurant employees’
[50] F. Shrouf, B. Gong, J. Ordieres-Mere, Multi-level awareness green behavior: Comparison of six attitude-behavior models. Int. J.
of energy used in production processes. J. Clean. Prod. 142, 2570- Hosp. Manag. 58, 66e81, 2016.
2585, 2017. [70] A. Paço, C. Shiel, H. Alves, A new model for testing green
[51] X. He, Y. Liu, The public environmental awareness and the consumer behaviour. J. Clean. Prod. 207, 998-1006, 2019.
105
[71] H. Cheng, X. Hu, R. Zhou, How firms select environmental [75] S. Unnikrishnan, D.S. Hegde, Environmental training and
behaviours in China: The framework of environmental motivations cleaner production in Indian industry: A micro-level study. Resour.
and performance. J. Clean. Prod. 208, 132-141, 2019. Conserv. Recycl. 50 (4), 427-441, 2007.
[72] L.V. Casalo, J.J. Escario, Heterogeneity in the association [76] J.C. Brenner, J.G. Hamilton, T. Drake, J. Jordan, Building
between environmental attitudes and pro-environmental behavior: A local environmental knowledge in undergraduates with experiential
multilevel regression approach. J. Clean. Prod. 175, 155-163, 2018. wilderness skills and awareness training: the case of environmental
[73] C.J.C. Jabbour, Environmental training in organisations: sentinels. J. Environ. Soc. Sci. 3 (4), 404-415, 2013.
From a literature review to a framework for future research. Resour. [77] D.W.S. Renwick, T. Redman, S. Maguire, Green human
Conserv. Recycl. 74, 144-155, 2013. resource management: a review and research agenda. Int. J. Manag.
[74] A.A. Teixeira, C.J.C. Jabbour, A.B.L. de Sousa Jabbour, Rev. 15 (1), 1-14, 2013.
Relationship between green management and environmental [78] C. Pellegrini, F. Rizzi, M. Frey, The role of sustainable human
training in companies located in Brazil: A theoretical framework and resource practices in influencing employee behavior for corporate
case studies. Int. J. Prod. Econ. 140 (1), 318-329, 2012. sustainability. Bus. Strateg. Environ. 27 (8), 1221-1232, 2018.
106
Mathematical modelling of combined infrared and

heat pump drying of squid
Pham Van Toan Phan Nhu Quan Nguyen Hay Le Anh Duc
Lac Hong University Lac Hong University Nong Lam University Nong Lam University
Buu Long Ward, Bien Hoa Buu Long Ward, Bien Hoa Linh Trung Ward, Thu Duc Linh Trung Ward, Thu Duc
City, Viet Nam City, Viet Nam District, Ho Chi Minh City, District, Ho Chi Minh City,
toanlhu@gmail.com phannhuquan@lhu.edu.vn Viet Nam Viet Nam
ng.hay@hcmuaf.edu.vn leanhduc@hcmuaf.edu.vns
Abstract: This study proposed a mathematical model for pre- microwave drying in terms of a lower product shrinkage, a
dicting the temperature and moisture content of squid during brighter white color, and a lower hardness. Deng et al. [9]
a combined infrared and heat pump drying process. The input conducted an experimental investigation into the effects of
infrared energy during the drying process was determined by infrared-heat pump drying on the protein quality of squid.
Lambert's law under the assumption of two infrared sources The results showed no significant difference between the
placed above and below the squid, respectively. The simulta- amino acid and protein contents of squid samples dried using
neous heat and mass transfer equations were solved by the the heat pump method alone or with the combined infrared-
finite difference method. The results showed that the use of heat pump method. However, the use of infrared heating in
infrared heating increased the temperature of the squid be- the drying process significantly reduced the aerobic bacteria
yond that of the hot drying air. Moreover, the drying time de- index in the dried product compared to that of the samples
creased as the drying air temperature and infrared power in- dried using the heat pump method alone.
creased. The calculated drying curves and temperature
Many studies have shown that when infrared radiation is
change during the drying process were found to be consistent
projected directly onto a material, the water molecules in the
with the experimental results presented in the literature. In
material absorb the heat radiation. As a result, the tempera-
particular, the calculated values of the drying time and drying
ture is rapidly and evenly increased over the entire volume of
rate deviated from the experimental results by just 4.20% and
the material and the diffusion process is therefore enhanced.
0.15%, respectively, at a drying air temperature of 50°C, and
The faster diffusion process increases the drying rate and
13.80% and 11.10%, respectively, at 60°C.
hence minimizes the loss of flavor, color and nutritional con-
tent in the dried product ([10], [11]). The faster drying rate
Keywords: Heat pump; Infrared irradiation; Squid; Mathe-
additionally shortens the drying time. Consequently, the in-
matical modelling
frared radiation method not only improves the quality of the
I. INTRODUCTION dried product, but also reduces the drying cost ([12], [13]).
Dried squid is not only delicious, but also has a high nutri- Although the literature contains several studies on the use
tional content ([1], [2]). As a result, it is one of the most com- of infrared-heat pump techniques to dry squid products, most
mon snacks in Vietnam and throughout the rest of South East of these studies are experimental studies aimed at evaluating
Asia. In preparing dried squid, it is necessary to heat the the effects of the infrared irradiation on the drying time and
squid to reduce the initial moisture content from 84% [3] to the quality of the finished product. By contrast, the present
around 25% [4]. One of the most common methods for drying study adopts a mathematical approach to evaluate the temper-
squid is simply to lay it under the sun. However, while this ature and moisture content within squid products processed
method has the advantages of simplicity and a low investment using the combined infrared and heat pump method. The va-
cost, it runs the risk of damaging the product through the ac- lidity of the mathematical results is confirmed by way of
cumulation of dust and other airborne contaminants and de- comparison with the experimental findings reported in the lit-
pends on the weather [5]. Accordingly, various more techno- erature.
logically advanced methods for drying squid have been pro-
posed in recent years. Among these methods, hot air drying II. MATHEMATICAL MODEL FORMULATION
is one of the most common due to its simplicity and low cost. A. Mathematical Model of Heat and Mass Transfer
However, the drying time is relatively long [6].
Figure 1 illustrates the basic setup of the infrared-heat
As a result, the potential for using microwave and infrared pump drying method considered in the present study. As
drying methods to dry squid has attracted growing attention shown, the squids are placed on a tray mesh and heated from
in the literature. Chen et al. [7] studied the effects of hot air above and below by two infrared sources. The radiation en-
drying, combined infrared-convective drying, and low-fre- ergy of the infrared waves is propagated to the surface of the
quency microwave drying on the quality of squid cubes. The dried material and is absorbed till the penetration depth of in-
results showed that the optimal drying performance was ob- frared radiation. The penetration depth of infrared energy is
tained using the infrared-convective technique at a tempera- relatively low due to the short wavelengths involved (e.g., 0.1
ture of 60ºC. Wang et al. [8] showed that the combined infra- – 0.5 mm) [14]. Moreover, the penetration depth depends on
red-convective drying method outperforms hot air drying and

the physical properties of the particular material concerned Based on the classical laws of heat and mass conservation,
(e.g., a depth of 2.6 – 18 mm for potatoes) [15]. However, in the following differential equations can be derived from the
the present study, the squid is assumed to have a thickness of heat and mass transfer processes in the squid:
just 6 – 7 mm. Therefore, it is reasonable to assume that the T  2T M
infrared energy emitted by the two heaters is absorbed  pcp = kp 2 + h fg k + Qia (1)
t x t
throughout the whole volume of the heated product.
M 2 M
The heat exchange process at the surface of the squid con- = Dm (2)
t x 2
sists of four components, namely (1) heat conduction ex-
The corresponding initial conditions are given as
change with the region of the squid just below the surface; (2)
At t = 0: T ( x, 0 ) = T0 ; M ( x, 0 ) = M 0 (3)
convective heat exchange between the surface of the squid
and the drying air supplied by the pump; (3) radiant heat While the boundary conditions ( t > 0 ) have the form
transfer from the squid to the infrared sources; and (4) evap- At x = 0
oration of the moisture content from the squid’s interior into T M
the environment. As shown in Fig. 1, the squids are placed on = 0, =0 (4)
x x
a mesh grid during the drying process. Consequently, the area
of heat exchange with the external environment on the lower At x = 
side of the squid is less than that on the upper side. However, hc A (Ta − Ts ) − h fg  k hm A ( M s − M e )
the reduction in the heat exchange area is very small com-  T  x T (5)
pared to the overall surface area of the squid. Thus, the heat + AQax −  −k p A  = cp  p A
2  x  2 t
exchange processes at the two surfaces of the squid were as-
M M x
sumed to be the same in the present analysis. hm A ( M e − M s ) + Dm A = A (6)
Infrared heater x t 2
B. Energy Absorbance of Dried Squid
In accordance with [16], the intensity of the infrared en-
ergy incident on the squid surface can be determined as
Tray
 (TFIR
4
− Ts4 )
Qr = -1
1-ε H  
Squids
Dryer Dryer 1 1-ε SQ
+  A H FH-SQ +  +
chamber chamber
AH ε H 
 (1/AH FH-C ) + (1/A SQ FSQ-C )  A SQ ε SQ
Mesh grid
(7)
where  H and  SQ are the emissivity of the infrared radia-
tion source and squid, respectively, and have values of 0.9
Infrared heater [17] and 0.93 [18]. The heat energy absorbed by a dried prod-
Figure 1. Heat exchange process within squid drying chamber un- uct reduces with an increasing thickness. Thus, in formulat-
der infrared-heat pump drying method. ing the mathematical model, the absorption process was as-
sumed to conform to the exponential Lambert-Beer law [19],
In addition to the assumptions given above, the mathemat-
i.e.,
ical model of the drying process assumed that the squid prop-
erties (e.g., moisture content, temperature) were uniform and Qia = (1 − R)Qr e− Kx (8)
isotropic and no chemical reactions nor internal heat genera- The radiation heat transfer process at the squid surface de-
tion took place. In addition, the heat and mass transfer pro- pends on the size and direction of the heaters relative to the
cesses within the squid were considered to be one-way; the squid surface and can be formulated as follows [20]:
initial temperature and moisture distribution in the squid were 1  X  Y  Y  X 
both uniform, and the squid and drying chamber walls were FASQ − AH 1 =  tan −1  + tan −1  
2  1 + X 2
 1+ X 
2
1+ Y 2
 1+ Y
2
 
both gray materials. Finally, thermal losses to the external en-
vironment were assumed to be negligible and the dried mate- (9)
rial remained flat (i.e., no warpage occurred) throughout the W1 W2
entire drying process. where X = ,Y = ;
D D
x C. Heat and Mass Transfer Coefficients
The heat transfer coefficient hc and mass transfer coeffi-
cient hm were derived in accordance with a simple flat plate
 model (see Fig. 2) [21], i.e.,
2 ka
0 y hc = 0.664 Re1/L 2 Pr1/3 (10)
− L
D
Figure 2. Flat plate model for drying squid hm = 0.664 Re1/L 2 Sc1/3 a (11)
L
108
x
Note that for simplicity, the radiation heat transfer coeffi-
cient was ignored s
D. Thermo-physical Properties and Equilibrium Moisture xi-5
i-5
Content
xi-4
i-4
The specific heat and density of the squid product were de-
termined respectively as [22]. xi-3 
i-3
c p = 3.108 + 0.006M (12) i-2

xi-2
M   M  xi-1
 p = 2059 − 71
i-1
 − 736exp  0.247  (13)
 0
M  M 0  i
xi
0
The thermal conductivity of the squid was taken from Rah-
man [23] as kp = 0.49. In addition, the equilibrium moisture Figure 3. The thickness of the dried half-squid  separated by Ni
content within the dried squid was evaluated as [24]. layers
Moreover, the average temperature of the dried squid at
M e = [exp(−1.383 − 0.029T )]1/1.267 (− ln RH )−1/1.267 (14)
time t was calculated as
E. Effective Diffusivity nx
The drying behavior of the squid was modeled using Fick’s V T i

t
i
t
second law given the assumptions of the slab geometry, mois- T =

t
av
i
nx
(21)
ture migration under diffusion, a constant diffusion coeffi- V i
t
cient, and negligible shrinkage. In particular, the moisture ra- i
tio within the squid was computed as [25].

B. Evaluation of Predictive and Experimental Results
8  1  −(2i + 1) DM  2t 
MR = 2  exp   (15)
 i =0 (2i + 1) 2
 4 L2 
The fit of the theoretical model to the published experi-
mental results was evaluated using the root mean square error
For long drying periods, Eq. (15) can be rewritten as (RMSE) and mean absolute error P metrics, defined respec-
M − Me 8  − D  2t  tively as
MR = = 2 exp  m 2  (16)
M0 − Me   4L  0.5
1 N 
From Eq. (16), the graph of ln ( MR ) versus the drying time RMSE =   ( M exp − M pre )2  (22)
 N I =1 
has the form of a straight line with a slope of
1 N  M exp − M pre 
D
Slope =  2 m2 (17) P=
N
  M pre


(23)
4L i =1
 
The temperature dependence of the effective moisture dif-
fusivity, Dm can be modeled using the following Arrhenius IV. RESULTS AND DISCUSSION
type equation:
A. Comparison of Calculated Drying Curves with Experi-
 E 
Dmt = D0 exp  − a  (18) mental Results
 R T
 ug  The calculation results obtained for the moisture content
Based on the experimental data presented in [26], the mois- of the dried squid were compared with the experimental re-
ture diffusion coefficient was calculated to be: sults reported by Chen et al. [7] for the infrared convection
 42810.909  drying method and Vega-Gálvez [28] for the convection dry-
Dmt ,i = 2.82110−3 exp  −  (19) ing method. Note that to ensure a fair comparison between
 8.314(Ti + 27315)  the various results, the input data used in the calculations (i.e.,
III. METHODOLOGY the initial moisture content, the drying temperature, the air
velocity, and the moisture diffusion coefficient) were as-
A. Mathematical Modeling Approach
signed the same values as those reported in the experimental
The temperature distribution and moisture content of the studies.
squid were calculated by solving the heat and mass transfer
equations using a self-written finite difference solution pro- Figure 4 compares the predicted values of the moisture
cedure with the thickness of the dried half-squid  separated content within the squid at drying temperatures of 50°C and
by Ni layers (see Fig. 3) by implemented in Matlab 2016a [27]. 60°C, respectively, with the experimental results reported in
The calculation process was terminated when the difference [7]. It is observed that a good qualitative agreement exists be-
in the average moisture content obtained in consecutive time tween the two sets of results at both drying temperatures. For
steps reduced to less than 10-5. The average moisture content a drying temperature of 50°C, the maximum deviation be-
of the dried squid at time t was calculated as tween the two sets of results is found to be 6.6%. Moreover,
nx
the RMSE and absolute error metrics have values of just
V i
t
M it RMSE = 0.15 and P = 8.9%, respectively, when evaluated
over the entire drying process. The predicted and experi-
M avt = i
nx
(20)
mental results for the maximum drying rate are also found to
Vi t
i
be in good agreement, i.e., 0.1343 (kg/kg dry matter.min) and
109
0.1492 (kg/kg dry matter.min), respectively. (Note that both moisture diffusion coefficient varies slightly as the dried
values are evaluated over the first 25 minutes of the drying product undergoes warpage. However, in the present mathe-
process.) The average calculated drying rate over the full dry- matical model, the warpage phenomenon is ignored and the
ing process is equal to 0.0289 (kg/kg dry matter.min), and product is assumed to remain flat throughout the entire drying
deviates by just 0.15% from the experimental value (0.0293 process. Second, the experimental thickness of the squid var-
(kg/kg dry matter.min). Finally, the calculation results indi- ied between 6 and 8 mm, whereas in the mathematical model,
cate that the drying time (i.e., the time to reach an average the squid was assumed to have a constant thickness of 7 mm.
moisture content of 20%) is 313 minutes, while the experi- Third, slight differences may exist between the emissivity
mental results show that the drying time is 300 minutes. In factors and heating distances considered in the experimental
other words, the calculation results are overstated by just studies and mathematical model, respectively. Finally, the as-
4.2%. For a drying temperature of 60°C, the maximum devi- sumption in the present model of full heat transfer symmetry
ation between the predicted and experimental values of the between the upper and lower surfaces of the squid is not
moisture content is 9.1%, while the RMSE and absolute error strictly valid due to the effects of the mesh grid in partially
metrics have values of RMSE = 0.24 and P = 10.9%, respec- masking the lower surface. Nonetheless, the close agreement
tively, over the full drying process. The calculated and exper- between the predicted drying curves and the experimental
imental drying times are 209 minutes and 180 minutes, re- drying curves demonstrates the basic validity of the mathe-
spectively, corresponding to a deviation of 13.8%. Finally, matical model and confirms its suitability for use in further
the drying rates are found to deviate by 11.1%. Overall, the investigating the moisture change process of the squid mate-
results presented in Fig. 3 confirm that the proposed mathe- rial during drying.
matical model accurately reproduces the experimental results
reported in the literature for the infrared convection drying
5.5 Predicted 50˚C
method. In other words, the basic validity of the proposed
model is confirmed. Moisture content, kg/kg dry matter 5 Experimental 50˚C of Vega-Gálvez
4.5 Predicted 60˚C
10 Predicted 50˚C Experimental 60˚C of Vega-Gálvez
4
9 Experimental 50˚C of Chen
3.5
Moisture content, kg/kg dy matter
Predicted 60˚C
8
Experimental 60˚C of Chen 3
7 2.5
6 2
5 1.5
4 1
0.5
3
0
2 0 100 200 300 400 500 600 700
1 Drying time, min
Figure 5. Comparison between calculation results for moisture
0
0 50 100 150 200 250 300 350
content and experimental results reported by Vega-Gálvez et al.
[28]
Drying time, min
Figure 4. Comparison between calculation results for moisture B. Temperature and Moisture Distribution
content and experimental results reported by Chen et al. [7]
The validated mathematical model was used to evaluate
Figure 5 compares the calculation results for the drying the average temperature change within the squid given three
curve of the squid with the experimental results reported by different drying temperatures (40C, 45C and 50C) for a
Vega-Gálvez et al. [28] using the convection drying method. squid thickness of 6 mm, an air velocity of 1.5 m/s, a humid-
From inspection, the mean deviation between the two sets of ity of 35% by the combined infrared and heat pump (HP+IR)
results is equal to 17.9% and 15.6% at drying temperatures of (with an infrared power of 500W) and the heat pump alone
50°C and 60°C, respectively. The calculation results indicate (HP). The corresponding results are shown in Figure 6. As
drying times of 676 and 459 minutes at drying temperatures shown, for the combined infrared and heat pump method, for
of 50°C and 60°C, respectively. By contrast, the experimental each drying temperature, the temperature curve comprises
drying times are 600 and 370 minutes, respectively. The dif- two phases, namely a rapid heating phase and a constant heat-
ference between the two sets of results for the drying time and ing phase. During the rapid heating phase, the average tem-
drying rate, respectively, is thus equal to 11.2% and 25.1% at perature of the squid reaches the drying air temperature of
a temperature of 50°C and 19.3% and 18.5 % at a temperature 40C, 45C and 50C after 35, 40 and 45 minutes, respec-
of 60°C. tively. From inspection, the squid temperature increases at a
rate of approximately 0.42, 0.50 and 0.56 (C.min) during the
The difference between the mathematical and experi- rapid heating phase at drying temperatures of 40C, 45C and
mental results presented in Figs. 4 and 5 can be attributed to
50C, respectively. In the latter phase, the temperature of the
several factors. First, in practical drying applications, the
110
squid remains approximately constant as the drying process those of the heat pump method alone at the same drying tem-
proceeds, but is slightly higher than the corresponding drying perature. The average drying rates of the combined infrared
air temperature due to the additional heat energy provided by and the heat pump method were found to be 0.0096, 0.0075
the infrared radiation sources (i.e., 42.6C, 47.5C, and and 0.0053 (kg/kg dry matter.min), respectively, while the
52.5C, respectively). Due to the support of the infrared radi- heat pump method alone was found to be 0.0082, 0.0061 and
ation and the convective heat transfer from the drying air into 0.0042 (kg/kg dry matter.min), respectively. In other words,
the dried material, the temperature of the dried product was the drying rates of the combined infrared and the heat pump
higher than the temperature of the drying air, which was dif- were 20%, 18% and 15% faster than those of the heat pump
ferent from the process of heat pump method alone. In normal, alone.
the temperature of the dried material has reached only the
maximum temperature of the drying air. The present results V. CONCLUSION
are similar to those reported by Wang et al. [8] for the infrared This study proposes a mathematical model for predicting
convection heating method and confirm that the addition of the heat and mass transfer process during the drying of squid
infrared heating is beneficial in enhancing the heating rate. products using a combined infrared and heat pump method.
The results have shown that the temperature and moisture
54
content values obtained using the model are in good qualita-
52 tive agreement with those reported in the literature. Moreover,
50 the calculated drying curves are also consistent with the ex-
48 perimental results. Thus, the basic validity of the mathemati-
46 cal model as a tool for examining the heat and mass transfer
44 phenomena during the drying of squid products has been con-
Temperature, ˚C
42 firmed. Notably, the proposed method serves as a useful

40
means of optimizing the combined infrared and heat pump
method in such a way as to save energy (i.e., reduce the dry-
38 HP + IR 50˚C
ing time).
36 HP + IR 45˚C
HP + IR 40˚C
34
HP 50˚C
Nomenclature
32 HP 45˚C AH Infrared heater surface area m2
HP 40˚C
30 AH1 25% of infrared heater surface area m2
28 ASQ Area of incidence of infrared heat energy on
26 squid surface m2
0 50 100 150 200 250 300 cp Heat capacity of squid J/kg K
Drying time, min
Dm Effective moisture diffusivity m2/s
Figure 6. Calculated mean temperature in squid product given dif- t
Dm,i Effective moisture diffusivity at time t and posi-
ferent drying temperatures
tion i of squid m2/s
5.5 HP + IR 50˚C Da Effective moisture diffusivity of air m2/s
5 HP + IR 45˚C D Distance between squid and infrared heater m
HP + IR 40˚C D0 Integration constant m2/s
Moisture content, kg/kg dy matter
4.5
HP 50˚C Ea Activation energy J/mol
4 HP 45˚C F View factor
3.5 HP 40˚C FASQ – AH1 Fraction of energy leaving surface ASQ which
3 reaches surface AH1
2.5 FASQ - AH Fraction of energy leaving surface ASQ which
reaches surface AH
2
FH-SQ Fraction of energy leaving surface H which
1.5 reaches surface SQ
1 FH-C Fraction of energy leaving surface H which
0.5 reaches surface C (chamber)
FSQ-C Fraction of energy leaving surface SQ which
0
0 150 300 450 600 750 900 1050 1200
reaches surface C
hc Heat transfer coefficient W/m2
Drying time, min
K
Figure 7. Calculated drying curves for squid product given differ- hfg Latent heat of vaporization J/kg
ent drying temperatures (40°C, 45°C and 50°C) hm Mass transfer coefficient m/s
kp Thermal conductivity of squid W/m K
Figure 7 shows the calculated drying curves for the drying ka Thermal conductivity of air W/m K
air temperatures of 40°C, 45°C and 50°C. The required dry- K Attenuation factor
ing times of the combined infrared and the heat pump drying M Moisture content kg/kg dry mat-
method to reach 0.33 (kg/kg dry matter) moisture content ter
were 914 minutes, 623 minutes, and 477 minutes correspond-
ing to 225 minutes,148 minutes, and 103 minutes faster than
111
Me Equilibrium moisture content kg/kg dry mat- REFERENCES

ter [1] Zhang, Y., Dai, B., Deng, Y., & Zhao, Y. “In vitro anti-inflammatory
M0 Initial moisture content kg/kg dry mat- and antioxidant activities and protein quality of high hydrostatic pres-
ter sure treated squids (Todarodes pacificus)”. Food chemistry, vol. 203,
pp. 258–266, 2016.
Ms Moisture content at squid surface kg/kg dry mat- [2] Deng, Y., Luo, Y., Wang, Y., & Zhao, Y. “Effect of different drying
ter methods on the myosin structure, amino acid composition, protein di-
Mav Average moisture kg/kg dry mat- gestibility and volatile profile of squid fillets”. Food Chemistry, vol.
ter 171, pp. 168–176, 2015.
[3] Sukkhown, P., Jangchud, K., Lorjaroenphon, Y., & Pirak, T. “Fla-
MR Moisture ratio vored-functional protein hydrolysates from enzymatic hydrolysis of
M it Moisture content at time t and position i of squid dried squid by-products: Effect of drying method”. Food Hydrocol-
loids, vol. 76, pp. 103-112, 2018.
kg/kg dry mat- [4] TCVN 5651-1992, T., Dried squids’ specifications for export. Minis-
ter try of science and technology of the socialist republic of Vietnam, the
Mexp Experimental values Direction of Standards, Metrology and Quality published.
Mpre Predicted values [5] Jain, D. and P.B. Pathare, “Study the drying kinetics of open sun dry-
ing of fish”. Journal of Food Engineering, vol. 78, p. 1315–1319,
N Number of experimental data points 2007.
L Length m [6] Wang Y., Zhang M. and Mujumdar A.S., “Trends in processing tech-
Pr Prandtl number % nologies for dried aquatic products”. Drying Technology, vol. 29, pp.
P Mean relative deviation 382–394, 2011.
[7] Chen, H., Zhang, M., Fang, Z., & Wang, Y. “Effects of different dry-
Qr Incident power of infrared source on squid sur- ing methods on the quality of squid cubes”. Drying Technology, vol.
face W 31, pp. 1911–191, 2013.
Qia Absorbed infrared energy in squid over x W [8] Wang, Y., Zhang, M., Mujumdar, A. S., & Chen, H. “Drying and qual-
ity characteristics of shredded squid in an infrared-assisted convective
Rug Universal gas constant J/kmol dryer”. Drying technology, vol. 32, pp. 1828–1839, 2014.
K [9] Deng, Y., Wang, Y., Yue, J., Liu, Z., Zheng, Y., Qian, B., ... & Zhao,
R Reflection coefficient Y. “Thermal behavior, microstructure and protein quality of squid fil-
lets dried by far-infrared assisted heat pump drying”. Food con-
Re Reynolds number trol, vol. 36, pp. 102–110, 2014.
RMSE Mean square error [10] Ning, X., J. Lee, and C. Han, “Drying characteristics and quality of
RH Relative humidity % red ginsengusing far-infrared rays”. Journal of ginseng research,
Sc Schmidt number 39(4): pp. 371–375, 2015.
[11] Riadh, M. H., Ahmad, S. A. B., Marhaban, M. H., & Soh, A. C. “In-
T Temperature of squid °C frared heating in food drying: An overview”. Drying Technology, vol.
Tav Average temperature °C 33, pp. 322–335, 2015.
Ta Air temperature °C [12] Xu, M., Tian, G., Zhao, C., Ahmad, A., Zhang, H., Bi, J., ... & Zheng,
t J. “Infrared Drying as a Quick Preparation Method for Dried Tange-
Ti Temperature at time t and position i of squid °C rine Peel”. International journal of analytical chemistry, 2017.
TFIR Temperature of infrared heater °C [13] Özdemir, M. B., Aktaş, M., Şevik, S., & Khanlari, A. “Modeling of a
T0 Initial temperature of squid °C convective-infrared kiwifruit drying process”. International Journal
of Hydrogen Energy, vol. 42, pp. 18005–18013, 2017.
t Time s [14] Delgado, J. M., & de Lima, A. G. B. (Eds.). Drying and energy tech-
x Distance m nologies. Springer International Publishing, 2016.
Vi t Volume at time t and position i of squid m3 [15] Pan, Z., & Atungulu, G. G. (Eds.). Infrared heating for food and ag-
ricultural processing. CRC Press, 2010.
W1 Width of infrared heater m [16] Holman, J.P., Heat Transfer - McGraw-Hill, Science Engineering
W2 Length of infrared heater m Math, 2009.
x Depth of layer m [17] Donghai Wenyi Quartz Products Co., L., Infrared heating tube, 2016.
[18] Hui, Y.H. and F. Sherkat, Handbook of food science, technology, and
Greek letters engineering-4 volume Set. 2005: CRC press.
 SQ Emissivity of squid [19] Erdogdu, S. B., Eliasson, L., Erdogdu, F., Isaksson, S., & Ahrné, L.
“Experimental determination of penetration depths of various spice
H Emissivity of infrared heater commodities (black pepper seeds, paprika powder and oregano leaves)
p Density of squid kg/m3
under infrared radiation”. Journal of Food Engineering, vol. 161, pp.
75–81, 2015.
k Absolute density of dry matter kg/m3 [20] Onwude, D. I., Hashim, N., Abdan, K., Janius, R., Chen, G., & Kumar,
C. “Modelling of coupled heat and mass transfer for combined infra-
a Air density kg/m3 red and hot-air drying of sweet potato”. Journal of Food Engineer-
 Stefan-Boltzmann constant 5.67 x 10 W/m2
-8 ing, vol. 228, pp. 12–24, 2018.
[21] Jaturonglumlert, S. and T. “Kiatsiriroat, Heat and mass transfer in com-
K4 bined convective and far-infrared drying of fruit leather”. Journal of
Food Engineering, vol. 100, pp. 254–260, 2010.
ACKNOWLEDGMENT [22] Pham Van Toan, Le Anh Duc and Nguyen Hay. “Determination of
thermal physical properties of squid”. Science and Technology jour-
The authors would like to express their sincere thanks to nal of Agriculture and Rural Development, 2018.
all of the lecturers in the Faculty of Mechatronics and Elec- [23] Rahman, M.S. and P.L. Potluri, “Thermal conductivity of fresh and
tronics at Lac Hong University for their detailed insights and dried squid meat by line source thermal conductivity probe”. Journal
of food science, 56(2): pp. 582–583, 1991.
discussions throughout the course of this study. The authors [24] Pham Van Toan and Le Anh Duc, “Moisture Desorption Isotherms
are also grateful for the assistance and support of all the mem- of squids”. Asia Pacific Journal of Sustainable Agriculture, Food and
bers of the Postgraduate Office of Nong Lam University, Energy, vol. 6, pp. 7–12, 2018.
HCM City, Vietnam. [25] Crank, J. The mathematic of diffusion (2nd ed.). Oxford, UK:
Oxford University Press, 1975.
112
[26] Pham Van Toan and Le Anh Duc, “Determination of Drying Method [28] Vega-Gálvez, A., Miranda, M., Clavería, R., Quispe, I., Vergara, J.,
for Squid”. Science and Technology journal of Agriculture and Rural Uribe, E., ... & Di Scala, K. “Effect of air temperature on drying ki-
Development, vol. 16, pp. 63–68, 2016. netics and quality characteristics of osmo-treated jumbo squid (Dosid-
[27] Anh Duc, L., “Study on Circulating Concurrent Flow Dryer for Dry- icus gigas)”. LWT-Food Science and Technology, vol. 44, pp. 16–23,
ing Rapeseed”. International Journal on Advanced Science, Engi- 2011.
neering and Information Technology, vol. 5, pp. 367–371, 2015.
113
Improving the Tensile and Compressive Strength of

Cement-Based Materials by Hybrid Electrospun
Nanofibers
Tri N. M. Nguyen Xuan Tung Nguyen Thanh Toan Dao
Faculty of Civil Engineering, Faculty of Civil Engineering, Faculty of Electrical-Electronic
Campus in Ho Chi Minh City Campus in Ho Chi Minh City Engineering
University of Transport and University of Transport and University of Transport and
Communications Communications Communications
Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam Hanoi, Vietnam
Corres: trinnm_ph@utc.edu.vn tungnx_ph@utc.edu.vn daotoan@utc.edu.vn
Huy Q. Nguyen Jung J. Kim

Department of Civil Engineering Department of Civil Engineering
Kyungnam University Kyungnam University
Gyeongsangnam, Republic of Korea Gyeongsangnam, Republic of Korea
nguyenquochuy@muce.edu.vn jungkim@kyungnam.ac.kr
Abstract: In this study, electrospun nylon 66 nanofibers response to develop the self-sensing material system [4],
containing nano silica fume were introduced into the cement improving the radioactive resistance to apply in radiation
paste. The observations from mechanical strength tests at 28 industry or storage the harmful wastes [5], enhancing the
days showed significant improvements in tensile and mechanical properties of the hardened cement pastes [6], and
compressive strength, i.e., the increases were observed of 41 % so on, have been reported.
and 33 % when adding the hybrid electrospun nanofibers into It is worth mentioning that, with a wide range of
the paste, respectively. Besides, field emission scanning applications of electrospun in numerous fields of the
electron microscopy and thermal gravimetric analysis were industry, strengthening cement paste by electrospun
conducted to estimate the microstructure of the cement matrix. nanofibers has rarely been reported. Electrospinning is
Based on these microscopic analysis results, the bead-on-string known as a flexible and economic technique to fabricate
shape of the electrospun fiber is found. Besides, the bridging nanosize fiber. When the surface tension of the viscoelastic
and filling effect of the fiber inside the cement matrix, together solution is sufficiently charged by the high voltage power
with the consumption of the calcium hydroxide phase by the supply, in combination with the electrostatic repulsion
added nano silica fume, have clarified the improvement of the between two electrodes, the fiber in nanosize is spun out and
cement structure. Finally, these microscopic results are good observed on the collector [7]. Besides, based on the
ex-planations for the improvement of selected mechanical observations from previous work [8-11], the influence of
strengths of cementitious materials. nanosize fibers fabricated by the electrospinning system on
Keywords: electrospinning, nylon 66, nano silica fume, improving the tensile strength of cement paste, and the
cement, tensile strength, compressive strength feasibility of using this additional fiber on cement have been
clarified. Continuing this strategy, the work presents the
I. INTRODUCTION results when modified cement pastes by the hybrid version of
electrospun nylon 66 nanofiber with nano silica fume (NSF-
Recently, modifying the microstructure of the cement
N66 NFs). In this present work, the mechanical strength test
matrix to improve its performance has become a priority
as tension and compression test had been carried out to
topic for material researchers. From the literature reviews,
observe the change in the mechanical properties of hardened
silica fume has shown its role in strengthening cement
cement paste. Besides, results from microscopic analyses
properties. Especially, nano silica, with the advantage of
such as field emission scanning electron microscopy (FE-
nanosize, can fill up the porous structure of the cement
SEM) and thermal gravimetric analysis (TGA) had been
matrix, as a result, a more compacted structure formed,
presented to investigate the behaviors of these electrospun
improving the compressive behaviors of the cement paste
nanofibers in the cement matrix.
[1]. Besides, the extra pozzolanic reaction formed by the
added silica and the generated calcium hydroxide (CH) from II. MATERIALS AND SAMPLES PREPARATION
the hydration process. As a result, calcium silicate hydrate A. Materials
(CSH) is produced, promoting the strength forming of the
cement matrix at the early age [2]. The work [3] has A type I Ordinary Portland Cement (OPC) compliance
presented a comprehensive study on the positive effects in with ASTM C150 from Ssangyong Co, Korea; Nano silica
terms of cost, time, as well as environmental issues, when fume grade 940-U from Elkem Microsilica (Norway) were
adding silica fume into cement materials. From another utilized in this study. Besides, nylon 66 pellets and solvent
perspective, the effectiveness of nanosize filaments on chemical as chloroform and formic acid were ultilized
cement characteristics such as investigating the electrical similiarty to the work [9]. Cement and NSF properties can be
seen in Table 1 and Table 2.

Table 1. Cement properties

CaO Al2O3 SiO2 SO3 MgO Fe2O3 Ig. loss Specific surface area Compressive strength,
(cm2/g) 28-day (MPa)
61.33 6.40 21.01 2.30 3.02 3.12 1.40 2800 36
Table 2. Nano silica fume properties

Property Unit Value
SiO2 % 90.0
H2O % 1.0
Ig. Loss % 3.0
Coarse particles > 45 μm % 1.5
3
Bulk density Kg/m 200
Table 3. Component proportion of the viscoelastic solution

(weight %)
B. Samples preparation
In this study, the NSF-N66 NFs were observed by Figure 1. Schematic of the electrospinning process
spinning the viscoelastic solution under the electrospinning
system. The procedure can be found in [8, 9]. The
viscoelastic solution was prepared by melting nylon 66
pellets in a solvent, in which, the solvent contains formic
acid and chloroform with ratio of 4:1 by volume. Then, a
proper NSF amount was added into the as-solution and
conducted the ultrasonic vibration for 2 hours. It is worth
noting that, in terms of polymer chain relaxation, the final
viscoelastic solution should be kept in the laboratory
condition for 24 hours. The component proportion of the
viscoelastic solution is presented in Table 3.
After spinning out of the spinneret, an improved collector
was utilized to blend the nano fiber with cement powder, the Figure 2. Results from tension and compression tests
weight proportion of viscoelastic solution and cement was In general, there is an increasing trend for both tensile
set of 5:95 [8, 9] (see Fig.1 for the electrospinning process and compressive properties of cement when adding the
schematic). The final composite containing cement and proposed electrospun nanofibers. For instance, the increase
nanofibers then conducted preparing the samples for tension of 41% and 33% in tensile and compressive strength were
and compression tests, three samples for each test. The observed. Previous work concerned with the modification of
water/binder was 0.5. The cement samples without cement paste by electrospun nylon 66 nanofibers showed the
electrospun nanofibers were prepared for control samples. increase in tensile strength of 28% and slightly increase in
All samples were demold after 24 hours and cured in the compressive strength of 8% compared to those of the control
water under the laboratory condition for 28 days (23 ± 2°C , paste [9]. The increase in tensile strength could be
RH of 50%). For microscopic analyses, the broken samples understood thanks to the bridging effect of electrospun
from mechanical strength test were collected and prepared nanofibers inside the cement matrix as observing in previous
for the TGA and SEM. Besides, the raw nanofibers were also work. For the increase in compresive strength, an assumption
collected on a carbon tape for SEM analysis. of the occurrent of a pozzolanic reaction between the
III. RESULTS AND DISCUSSION attached silica fume and the generated CH from the cement
hydration process is proposed, and the results of this process
C. Mechanical tests are promoting the CSH generation, leading to an increase in
Fig. 2 shows the mechanical strength results of the hardened the strength of the paste. Besides, the adhesion between the
cement paste containing electrospun nanofiber compare to hydration products and the fibers would be increased thanks
those of the control paste. The compressive strength test was to this reaction. Hence, the more compacted structure
conducted by means of the 1000 kN-UTM compliance with formed, the more the compressive properties of the cement
ASTM C109, the tensile strength test was conducted by paste were improved. Above all, from the mechanical results,
means of the mortar tensile strength test apparatus with the 5 the effectiveness of the NSF-N66 NFs compared to the raw
kN-capacity, compliance with ASTM C307. Three briquette electrospun nylon 66 nanofibers are clear. The below section
samples and three cubic samples for the modified cement will present the microscopic analyses of the proposed
paste and the control paste were prepared for the tests at 28 nanofibers and the cement matrix. Then, the behavior of the
days. NSF-N66 NFs inside the cement matrix, as well as the above
assumption, could be clarified. Therefore, the explanation for
the increase of the mechanical strength could be derived.
115
effects of the fibers inside the cement pastes. Above all,

together with the adhesive effect between the fiber and
cement hydration products and the bridging and filling effect
of the fiber, a more compacted structure formed, leading to
improve mechanical properties of the cement matrix.
F. TGA results of the cement matrix
Fig. 5 shows the TGA-DTG (thermalgravimetric analysis
- derivative thermogravimetry) of the hardened control paste
and the paste containning NSF-N66 NFs. The thermal
analysis was conducted by means of the TA instrument SDT-
Q600, with the input parameters as the heating range from
room temperature to 1000°C, under the Nitrogen
Figure 3. SEM image of electrospun nanofibers
atmosphere, the flow rate of 100 ml/min, and the heating
D. SEM image of electrospun nanofiber velocity of 10°C/min.
Fig. 3 presents the SEM image of the electrospun NSF- As shown in the literature [15, 16], the thermal analysis
N66 nanofibers. The SEM analysis was conducted by means of the control paste (Fig. 5a) can be divided into three
of the Zeiss Merlin Compact system, using the input segments, i.e., the decomposition of water from CSH and CH
parameters as the accelerating voltage of 3 to 5 kV, the phases, and the decarbonation of the calcite. The temperature
working distance of 7.1 to 7.9 mm, and the samples were range is presented as follows:
coated with a 5Å-platinum layer.
Before 200°C: The evaporation of arbitrary water and
As can be seen from Fig. 3, the bead-on-string shape of water from CSH phase
the electrospun nano-fiber containing nylon 66 and nano
silica fume has been observed, in which, the nano silica fume From 400°C to 500°C: The evaporation of water from
particles were attached alongside the nylon 66 fiber. Based CH phase
on the result of SEM, the mean diameter of the fiber is 150 From 550°C to 900°C: The evaporation of carbon
nm and the mean diameter of the silica particle is 290 nm. As dioxide (CO2) from calcite phase.
a comparative result, the observation from previous work Comparing the thermal analysis results between the
shows the mean diameter of electrospun nylon 66 nanofiber control paste and the paste containing the electrospun
is 180 nm [14], bigger than the present observatuon. It can be nanofiber, overall, the shape of TGA, as well as DTG curves,
explained due to the reduction of the nylon 66 content in the are similar between the two results. Three main phases of the
present work compared to that in the previous one, i.e., 7 cement matrix are clearly shown in the graph. Focusing on
wt% and 10 wt%, respectively. the weight loss of CSH and CH phases, the percent weight
E. Microscopic analysis of the cement matrix loss of at 200°C from the result of the control paste show
Fig. 4 presents the cement matrix microscopic analysis lower than that of the modified paste, i.e., 11.1% and 19.33
from SEM. It is worth noting the difficulty of finding the %, respectively. Since the water/binder was kept constant in
existence of the NSF-N66 NFs inside the cement matrix. The this work, therefore, it can be concluded the CSH generated
fibers are hidden within the cement hydration products. In from the modified matrix is higher than that from the control
contrast with what can be found from previous work [9], that matrix. In addition, observing the range from 400°C to
show the clear existence of the fibers inside the cement 500°C, i.e., the decomposition of CH phase, the contrast
matrix. As can be found from the literature, an extra results are observed. The percent weight loss of CH phase in
pozolanic reaction between the added silica fume and the CH the control paste is higher than that in the modified paste,
product from the cement hydration process, which is i.e., 3.74% and 2.78%, respectively. The above observations
promoting the generation of CSH, the main component of clarify the consumption of generated CH from the hydration
forming the strength of the cement paste [2, 12-14]. The process by the added silica fume and also meet with the
main reactions of the hydration process can be described as initial assumptions. Finally, the more CSH phase generated,
follows: the higher strength of the cement matrix observed.
IV. CONCLUSION
C3A + 6H → C3AH6 (1)
This study presents the effectiveness of the hybrid
2C3S + 6H → C3S2H3 + 3CH (2) electrospun nylon 66 nanofiber containing nano silica fume
2C2S + 4H → C3S2H3 + CH (3) on the mechanical strength and microstructure of the cement
matrix. Based on the presented results, the following
Hence, compared with the observation from SEM, the conclusions can be drawn:
attached silica fume has shown its role in forming the extra
pozzolanic reaction, as a result, the morphology of the fibers • The bead-on-string shape of the electrospun
is transformed completely. The fibers are covered by the nanofiber is observed, in which, the silica fume particles are
hydration products (see Fig. 4b), which confirmed the good attached alongside the nylon 66 nanofiber.
adhesion between the fiber and the cement matrix. Besides, • The added nano silica fume has proven its role in
from Fig. 4a, the bridging effect and filling of the fibers can creating the extra pozzolanic reaction. Besides, the bridging
be observed. The beginning and ending of fibers are
observed from other parts of the cement hydration products.
This phenomenon is confirmed by the bridging and filling
116
Figure 4. SEM image of cement matrix: a) electrospun nanofiber inside the cement matrix, b) deformed nanofibers morphology
(a) (b)
Figure 5. TGA-DTG results of the hardened cement paste: a) control paste, b) NSF-N66 NFs cement paste
and filling effects of the proposed fiber inside the cement iron lag and titanate nanofibers to produce advanced containment for
matrix are observed. As a result, the adhesion between fiber radioactive waste, Construction and Building Materials, 200, (2019),
135-145.
and cement hydration products is improved, more CSH [6] M. J. Chinchillas-Chinchillas, A. Gaxiola, C. G. Alvarado-Beltrán, V.
phase is generated, more compacted cement structure is M. Orozco-Carmona, M. J. Pellegrini-Cervantes, M. Rodríguez-
formed, and finally, the strength of the cement matrix is Rodríguez, and A. Castro-Beltrán, A new application of recycled-
enhanced. PET/PAN composite nanofibers to cement–based materials, Journal
of Cleaner Production, 252, (2020), 119827.
Above all, the positive observations from this present [7] S. Thenmozhi, N. Dharmaraj, K. Kadirvelu, and H. Y. Kim,
study show the feasibility of the NSF-N66 NFs in improving Electrospun nanofibers: New generation materials for advanced
applications, Materials Science and Engineering: B, 217, (2017), 36-
the cement properties from both mechanical and structural 48.
perspectives. However, other extended studies should be [8] T. N. M. Nguyen, D.-Y. Yoo, and J. J. Kim, Cementitious material
conducted in the future to find more detail on the effect of reinforced by carbon nanotube-Nylon 66 hybrid nanofibers:
this proposed fiber on cement properties, as well as the Mechanical strength and microstructure analysis, Materials Today
Communications, 23, (2020), 100845.
applications of this reinforced agent in the concrete industry. [9] T. N. M. Nguyen, J. Moon, and J. J. Kim, Microstructure and
ACKNOWLEDGMENTS mechanical properties of hardened cement paste including Nylon 66
nanofibers, Construction and Building Materials, 232, (2020),
The authors would like to acknowledge the National 117134.
Research Foundation (NRF) grant funded by the Korean [10] T. N. M. Nguyen, D. H. Lee, and J. J. Kim, Effect of Electrospun
Nanofiber Additive on Selected Mechanical Properties of Hardened
government (2020R1A2B5B01001821). Cement Paste, Applied Sciences, 10, 21, (2020).
REFERENCES [11] J. J. Kim, T. N. M. Nguyen, X. T. Nguyen, and D. H. Ta, Reinforcing
Cementitious Material Using Single-walled Carbon Nanotube - Nylon
[1] S. Zheng, T. Liu, B. Qu, C. Fang, L. Li, Y. Feng, G. Jiang, and Y. Yu, 66 Nanofibers, Transp. Commun. Sci., 71, 1, (2020), 46-55.
Experimental investigation on the effect of nano silica fume on [12] P. K. Mehta and P. J. M. Monteiro, Concrete: Microstructure,
physical properties and microstructural characteristics of lightweight Properties, and Materials - 4th edition, McGraw-Hill Education, USA,
cement slurry, Construction and Building Materials, 329, (2022), 2006.
127172. [13] A. M. Neville, Properties of Concrete, Prentice Hall; 5 edition
[2] X. Ma, T. He, Y. Xu, R. Yang, and Y. Sun, Hydration reaction and (September 2, 2012), 2005.
compressive strength of small amount of silica fume on cement-fly [14] H. F. W. Taylor, Cement chemistry - 2nd edition, Thomas Telford
ash matrix, Case Studies in Construction Materials, 16, (2022), Publishing, USA, 1997.
e00989. [15] J. J. Kim, E. M. Foley, and M. M. Reda Taha, Nano-mechanical
[3] H. Eker and A. Bascetin, Influence of silica fume on mechanical characterization of synthetic calcium–silicate–hydrate (C–S–H) with
property of cemented paste backfill, Construction and Building varying CaO/SiO2 mixture ratios, Cement and Concrete Composites,
Materials, 317, (2022), 126089. 36, (2013), 65-70.
[4] H. Wang, J. Shen, J. Liu, S. Lu, and G. He, Influence of carbon [16] E. M. Foley, J. J. Kim, and M. M. Reda Taha, Synthesis and nano-
nanofiber content and sodium chloride solution on the stability of mechanical characterization of calcium-silicate-hydrate (C-S-H) made
resistance and the following self-sensing performance of carbon with 1.5 CaO/SiO2 mixture, Cement and Concrete Research, 42, 9,
nanofiber cement paste, Case Studies in Construction Materials, 11, (2012), 1225-1232.
(2019), e00247.
[5] H. M. Saleh, S. M. El-Sheikh, E. E. Elshereafy, and A. K. Essa,
Mechanical and physical characterization of cement reinforced by
117
Controlling Crystal Morphology via Crystallization

Processes, Case Studies of Zinc Lactate and KDP
Tam Le-Minh Cuong Nguyen Van Venkata Subbarayudu Sistla
Department of Chemical Department of Basic Subject Department of Chemical
Engineering Industrial University of Ho Chi Minh Engineering
HCMC University of Technology City Rajiv Gandhi Institute of Petroleum
and Education Ho Chi Minh City, Vietnam Technology
Ho Chi Minh City, Vietnam nvc@iuh.edu.vn Jais, India
tamlm@hcmute.edu.vn vsistla@rgipt.ac.in
Abstract: In chemical process engineering, ripening, etc. Among them, using additives and seeds are
crystallization is usually considered a high-performance unit highly interested in this work. According to Zhang et al. [6],
for separation and purification of many valuable additives can be adsorbed on different active sites of crystals
compounds, especially in pharmaceuticals, food industry, and prohibit building materials on the related surfaces,
agriculture, etc. In many cases, monitoring and controlling therefore, resulting selectively growth effects. Contrarily,
crystal morphology are the major tasks which affect another report by Dowling et al. [7] proved that additives
downstream process designs. In fact, crystal morphology is can accelerate crystal growth. Thus, the effects of additives
an important property besides purity of the final products are complicated and will be clarified in this work. Besides
since different crystal morphologies show a strong influence additive, effects of nuclei on selective crystal growth are
on ability of, for example, drug formulation as well as also considered. Existing nuclei offer specific surfaces
solubility of tablets, capsules, etc. In this work, some simple according to different crystal lattices that create driving
experiments were performed to figure out the effects of forces to selectively build up crystals in different directions.
additives, seed loads and supersaturated degrees on In short, the effects of seeds and additives will be studied in
preferential crystal growth according to different crystal this work for the cases of zinc lactate while effects of
lattices. The processes will be carried out under isothermal different cooling profiles will be investigated for a case of
crystallization conditions. In fact, this task is often potassium dihydrogen phosphate (KDP).
challenging because of its complexity relating to primary
and/or secondary nucleation, crystal growth, aggregation, II. MATERIALS AND METHODS
attrition and crystal breakage, etc. In this research, instead of A. Experimental part
growing a single crystal, our approach is based on analyses
Chemicals used in this study including Zinc lactate were
of population density functions for statistical solutions. Two
bought from Cosphatec, Germany (98% purity); KDP was
systems will be experimentally investigated including zinc
used from Xilong (99.5% purity); Acid lactic was purchased
lactate and potassium dihydrogen phosphate (KDP).
from Musashino (90% purity); Distilled water was used as
Keywords: crystal morphology, KDP, zinc lactate the solvent.
Laboratory apparatus: A jacketed two-liter crystallizer
I. INTRODUCTION
coupled with a thermostat (Stuart SRC5); Mettler Toledo
Controlling crystal morphology is one of the major tasks ME203 (electronic balance); IKA CMAG HS7 (magnetic
which decide many post-processes of particulate systems stirrer); SEM JM400 Plus; BH1705 laser module coupled
[1,2]. Different morphologies of the same materials could with a microcontroller.
even change significantly properties and, therefore, the
applicability of these materials such as dissolution rates of B. Methodology
pharmaceuticals in various media [1-3]. Indeed, different
crystal lattices can grow and/or dissolve preferentially
depending on specific systems [1-4]. To obtain the right
morphology, crystallization processes are frequently applied
in which conditions such as additive, super-saturation
degree, seed, ripening, etc. must be well understood. In
crystallization, crystal growth is considered the most
important process to manipulate the size and shape of
crystals [5]. Herein, material transferred from liquid phases
to the interfaces of solids and build up crystals according to
different crystal lattices. Consequently, the final crystals
will be obtained with different morphologies such as cubic,
elongated, needle-like shapes, etc. [1,2,5]. Figure 1. Simplification of crystal dimensions [8]
Practically, crystal morphology could be controlled by In general, a 3D crystal projected onto a 2D plane can be
various techniques such as temperature profiles, Ostwald considered a rectangular with appropriate dimensions as

seen in Fig. 1. Regarding mathematical approximation, D. Seed preparation

cylinders/rectangular prisms were converted into objects An amount of 100 g zinc lactate was completely
with characteristic dimensions Llength and Lwidth (i.e. dissolved into 1000 g water at 80oC for 60 mins in a glass
Llength x Lwidth = a x b). vessel. Then, the vessel was immersed in ice-cold water for
Herein, the volumetric shape factor kv is introduced as 30 mins to recrystallize zinc lactate while strong agitation
the following (1). This equation presents a volume ratio of was applied to accelerate primary nucleation. After that,
the cylinder with respect to an equivalent cubic. Physically, suspension was separated by filtration and seeds were dried
the variation of kv indicates the development of crystals at 80oC in 12 hours. A small quantity of seeds was randomly
along different directions. When kv increases, the crystals collected and presented in Fig. 3. Obviously, besides
tend to form elongated shapes. Otherwise, rectangular plate- relatively big crystals (50–100 µm), plenty of fragments and
like crystals will be preferred in cases of kv decreased. tiny imperfect crystals (1–20 µm) were also included in this
𝜋𝑏 2 𝑎 population.
𝑘𝑣 = (1)
4𝑎3
Theoretically, there are two approaches in study of
crystal growth including (a) a single crystal growth and (b) a
population growth. Each of them is suitable for specific
systems with different intrinsic advantages and
disadvantages. In fact, (a) single crystal growth experiments
need a very well controlled apparatus, and the selected
crystal must be large enough to be hanged in the fluid inside
the fluidic grow-chamber in front of a microscope. In this
work, the second approach (b) was applied in which crystal
size distribution (CSD) of the population will be measured
instead of just only one single crystal as the previous Figure 3. Seed crystals of zinc lactate obtained from a
method (a). For evaluating CSD of every population, at least recrystallization step
500 crystals were measured to obtain statistical results. E. Effects of seed loads
Nowadays several in-situ image analytical tools are
available, for instance PVM (particle vision & Effects of adding 1% seed. The same experiment setup as
measurement) of Mettler Toledo. This tool is powerful for the previous part was prepared. Additionally, 1.0±0.1 g of
measurements of relatively large particles. However, it is dried seed was also added into the crystallizer when the
not suitable for such small particles as zinc lactate, system was cooled down from 80 to 40 oC. Based on the
especially in cases of high concentration relating to the seed, secondary nucleation and crystal growth occurred.
subtract background issue. Therefore, offline image analysis Statistical analysis for variation of crystal length and width
was conducted in this work in which solid products at at different process times was shown in Fig. 4a. The
certain intervals were analyzed via a SEM equipment. For individual dimensions of these particles were recorded as
each of measurements, the suspensions were collected using histograms and known as CSD (crystal size distribution). In
syringes, filtered and dried at 80oC for 5h. After that, more this process, tiny crystals were continuously newly formed
than 500 crystals in several images of a population were (phase 1st) while existing crystals were also continuously
randomly measured. growing (phase 2nd). Depending on competition rates of
these phases, the average size of whole population can be
increased or decreased.
C. Crystal growth behavior of zinc lactate As presented in Fig. 4a, the average length tended to
In the preliminary experiments, different morphologies decrease while the average width was seen as plateaus. This
of zinc lactate were obtained under various conditions as observation was related to the developing of crystals with
presented in Fig. 2. Obviously, depending on competition different directions. Indeed, the average length decreased
rates of the growing faces, crystals will be selectively because of abundantly forming of new crystals. However,
developed along length or width axes which result in these crystals are not developing along the length direction.
rectangular plate-like or elongated crystal shapes. In the Otherwise, these crystals seemed to grow preferably along
next section, important effects will be investigated to control width direction. Consequently, kv-variation was slightly
these morphologies. Two factors will be studied including increased that indicated the final population tended to
the effects of seeds and additives. broader their width dimension more than crystal length as
. seen in Figure 4b. Thus, seed load showed influence on
preferentially growth of zinc lactate crystals. Increasing load
of seeds will be performed in the next section to evaluate
further seed effects.
Figure 2. Preferential development of crystals leads to different

morphologies
119
Figure 4. a) Effects of 1% seed load, every point was calculated Figure 6. Effects of 3% acid lactic
from average of about 500 crystals; b) SEM of a randomly
collected population representing the final-time product
Effects of 6% acid lactic. In contradiction to the previous
Effects of adding 3% seed. In this part, the same case, selective growing effects were clearly showed up in
experiment conditions were conducted but the seed load was the case of added 6% additive as seen in Fig. 7. Obviously,
increased to 3%. As seen in Fig. 5a, the average length tends both length and width were increasing but with different
to growth faster than in the previous case while the average rates. In fact, the length grew faster than the width did when
width is almost remaining unchanged. As a result, kv was comparing the slopes of the corresponding curves in Figure
found as gradually decreased values which indicate that 7a. Thus, acid lactic was selectively adsorbed at different
crystals build up materials along the length direction. These crystal lattices which prohibited growing certain faces.
phenomena were clearly observed in the final-time product Consequently, at the end of this process, the final crystals
which was shown in Fig. 5b. were obtained as elongated shape.
Figure 5. Effects of 3% seed Figure 7. Effects of 6% additive
Thus, as seen in the above experiments, exposed surfaces In short, morphology of zinc lactate can be varied under
of seed showed effects on selective growth of zinc lactate influences of seed and additive as seen in this part. For
crystals. Interestingly, various amounts of seeds resulted in quantifying these effects, more details experiments are
different effects. Using 1% seed load supported developing needed but will not be discussed herein. In general, another
width crystals while 3% used seeds makes crystals grow possibility to control crystal morphology relates to different
faster in length dimension. This can be explained due to the cooling profiles, which are equivalent to the matter of
fact that exposed seed surfaces can promote secondary different super-saturated degrees. In the next section, KDP
nucleation happening with different rates. These newly will be investigated to clarify the possible effects from
formed crystals possess different specific surface areas cooling strategies.
which in turn decide crystal growth rates as observed from
the above two experiments. G. Controlling KDP morphology via crystallization
Compared to the case study of zinc lactate in the
F. Effects of additive
previous part, another strategy of controlling morphology
Using 3% acid lactic as additive. In the literature [9], acid will be applied herein. In this section, special cooling
malic and succinic acid were reported that they can prohibit profiles will be used to change crystal habits with a case
growing for some certain faces of zinc lactate crystals. In study of KDP. To perform crystallization, saturated
this part, acid lactic was added into the system to study this solutions of KDP in water at 45oC were cooled down to 35
behavior. With 3% of acid lactic as additive, both length and and 30oC that corresponds to two different super-saturated
width were seen as growing with the same rates (see Fig. degrees. A special setup was used in which a laser beam
6a). For that reason, rectangular plate-like crystals appeared (680 nm, 5 mW) shone through the crystallizer and focuses
in this case as seen in Fig. 6b. Thus, such small amount of on a laser module. The laser transmittances were recorded
3% additive did not show preferential growing effects for and presented in Fig. 8. Obviously, crystallization trajectory
zinc lactate system. It could be explained that the adsorption in case of larger supersaturated degree (i.e. tend at 30oC)
of additive was not effective at different active sites of the prolonged better than that in case of tend at 35oC. At the end
crystals. Increasing amount of additive will be implemented of the two processes, crystals were harvested, and their
in the next part. morphologies were analyzed.
120
first case study with zinc lactate/water system, an amount of

1% seed preferred to form rectangular plate-like crystals
while 3% of seed tended to prolong the length resulting
elongated crystals. In another scenario, when using acid
lactic as additive, relative larger amount of this acid can
differ selective growth for different crystal lattices. In fact,
3% of this additive did not show the interested effects as the
case of 6% did. For the second studied system with
KDP/water, crystal habits were changed with cooling
profiles. High supersaturated degree helped developing the
crystal length, but it also resulted in more tiny crystals. In
short, effects of selective growth are interesting phenomena
and partly seen in this simple work. However, more
experimental efforts are needed to quantify these effects.
Furthermore, for future works, molecular simulation is
suggested to have a deeper understanding about mechanism
Figure 8. Crystallization profiles
of crystal growth which allows explaining in detail the
As seen in Fig. 9a, the crystals which were obtained in phenomena observed in this work.
case of the moderate supersaturated degree at 35oC
crystallized as larger sizes and more uniform in habits. REFERENCES
Hereby, both length and width of these crystals were [1] Mullin, J.W.: Crystallization, 4th ed. Butterworth-Heinemann
relatively good developed to form perfect prisms. Different (2001).
behavior was seen in case of higher super-saturated degree [2] Lewis, A.E., Seckler, M., Kramer, H., van Rosmalen, G.:
when the system was cooled down to 30oC. Larger driving Industrial Crystallization: Fundamentals and Applications.
forces promoted secondary nucleation which led to form Cambridge University Press (2015).
tiny crystals as seen in Fig. 9b. Furthermore, these crystals [3] Myerson, A.S.: Handbook of Industrial Crystallization,
appeared as elongated shapes. Thus, cooling profiles Butterworth-Heinemann (2001).
showed selective growing effects in the studied case, i.e. [4] Leubner, I.H.: Precision Crystallization: Theory and Practice
KDP. However, these observations were just first evidences of Controlling Crystal Size, 1st ed. CRC Press, Boca Raton,
to prove selective growing effects. For quantitative FL, USA (2009).
assessments, more experimental and simulation efforts are [5] Randolph, A.D.: Theory of Particulate Processes, 2nd ed.
needed and, hereby, suggested for future works. Academic Press, San Diego, CA, USA (1988).
[6] Zhang, J., Zhang, M., Wan, J., Li, W.: Theoretical study of
the prohibited mechanism for ethylene/vinyl acetate co-
polymers to the wax crystal growth. J Phys Chem B. 112(1),
36 – 43 (2008).
[7] Dowling, R., Davey, R.J., Curtis, R.A., Han, G, Poomachary,
S.K., Chow, P.S and Reginald, B.H.T. Acceleration of crystal
growth rates: An unexpected effect of tailor-made additives.
Chemical Communications. 46(32):5924-6 (2010).
Figure 9. KDP crystals crystallized at (a) 35oC and (b) 30oC [8] Tam, L.M.; Thanh, T.P.; Hong, N.N.T.; Minh, V.P. A Simple
Population Balance Model for Crystallization of L-Lactide in
IV. CONCLUSION & OUTLOOKS a Mixture of n-Hexane and Tetrahydrofuran. Crystals 12, 221
(2022).
This paper presented some first results of an
[9] François, F., Gilles, F. A method of characteristics for solving
investigation about effects of selective growth which are
Population Balance Equations (PBE) describing the
interesting for controlling morphology of crystals. adsorption of impurities during crystallization processes.
Isothermal crystallization processes were carried out in Chemical Engineering Science 65, 10, 3191 – 3198 (2010).
consideration of influences of seed, additive and super-
saturated degree on morphology of various crystals. In the
121
A Green Solution for Kitchen Waste Treatment

Using Earthworm, Experimental and Mathematical
Approaches
Tam Le-Minh Phuong Pham Thi Hong Nhu Vo Thi Thu
Department of Chemical Department of Chemical Department of Chemical
HCMC University of Technology Industrial University of Ho Chi Minh HCMC University of Technology
and Education City and Education
tamlm@hcmute.edu.vn phamthihongphuong@iuh.edu.vn nhuvtt@hcmute.edu.vn
Abstract: Huge amounts of kitchen waste are daily needs of sustainable treatments.
produced, particularly in urban areas. These organic
The main method of solid waste treatment in Vietnam is
pollutants are highly harmful to environment’s protection. In burial since it requires simple techniques, but consuming
principle, these sources can be converted into valuable large landfills is the main disadvantage of this method [1,2].
materials in suitable ways, but most of them have been In fact, the imported waste treatment methods from foreign
frequently either burned or buried in landfills for decades. countries applied to Vietnam are not effective because the
For civilization, more sustainable solutions should be waste has not been classified at the source [2]. Currently, the
considered urgent needs. In this work, kitchen waste will be amount of garbage in urban areas sent to allowed landfills
treated to produce organic fertilizers with assistant of accounts for only 60 – 65%, the rest is thrown in ponds,
earthworms. Obviously, dual benefits are herein obtained lakes, canals, and roadsides [1,2]. Recently, the incineration
based on this approach. On the one hand, kitchen waste will method is also applied in treating domestic solid waste, but
be solved instead of releasing pollutants into the this source often contains high humidity, so the burning
environment. On the other hand, green fertilizers for crops method is not effective [1-3]. Additionally, the investment
will be produced from these wasted materials. Purposely, in an industrial incinerator often requires quite large capital.
raising earthworms herein will be carried out in limited If incineration technology is operated incorrectly, it will
spaces such as balconies or small gardens that make this produce air pollutants, so it is also necessary to have an
approach applicable even to families in both urban and rural exhaust gas treatment system [3]. Recently, regarding the
areas. In this work, besides the experimental approach, a environmental protection law, green solutions are urgently
mathematical description will be applied to model the birth required in Vietnam. A potential method is based on
biotechnology in which eco-friendly species such as
and growth rates of an initial population of earthworms over
bacteria, insects, algae, etc. are preferred [1,2]. Among
time under controlled conditions of nutrients, temperature
them, earthworms are especially interesting due to many
and humidity, etc. The applied model, based on population advantages. In the last two decades, there have also been
analysis, is quite sufficient for describing the raising many studies on the ability to use earthworms to decompose
earthworm processes. organic matter, and certain achievements have been
achieved [3,4]. The concept is since the worm's intestines
Keywords: earthworm, kitchen waste, population contain millions of aerobic bacteria that play the role of
balance equation decomposing organic biomass and chemicals [3-5]. Raising
a population of 15,000 worms produces billions of bacteria
I. INTRODUCTION
in a short time. This population of bacteria plays a role in
According to the national environmental report [1,2], the breaking down the organic matter of the waste. At the same
amount of domestic solid waste generated in the period time, the intestine of worms also contains many enzymes
2011-2015 increased rapidly compared to the previous time. such as protease, lipase, amylase, cellulose, chitinase, etc.
The statistic in 2011 showed that the total amount of which are also the agents that degrade protein- and fiber-
domestic solid waste nationwide was 44,400 tons/day in rich materials in organic waste [5-6]. In the process of
2007 [1]. In 2019, this number increased to 64,658 tons/day decomposing waste, the worms excrete fluid from the
in which the volume of urban domestic solid waste was intestines and this fluid has the effect of killing pathogenic
about 35,624 tons/day [1]. Particularly in big cities, e.g., in bacteria in the compost [6]. Thus, the waste when treated by
Hanoi and HCM Cities, the volume of domestic solid waste earthworms will reduce the toxicity to the environment.
approximated 6,500 tons/day and 9,400 tons/day, Moreover, these substances obtained throughout earthworm
respectively [1]. According to calculations, comparing life’s circle can make good fertilizers for plants [1-3,5,6].
numbers in 2010 and 2019, the growth rate increased by
about 46% [1]. Thus, with a population of more than 97 In this work, using earthworm for kitchen waste
million people, amounts of waste generated in Vietnam are treatments will be reported. This work is particularly
very large. These statistical numbers partly show up a meaningful for urban areas where green solutions are highly
serious situation of pollution in Vietnam and the urgent needed. On the one hand, this solution helps to reduce
pollution in big cities. On the other hand, organic fertilizers

are also produced that provide clean vegetables and creates data will be used for parameterization of the PBE model as
green fields for urban spaces. Firstly, experiments are being described in the next section. Then, this population
carried out for raising earthworms under controlled was further cultivated under the same conditions as the
conditions about feeding, temperature, humidity. Secondly, previous part until 40 days. The obtained data at this time,
a mathematical description based on the well-known method which serve as an independent cross-check, will be used for
(population balance equation - PBE) is also implemented to model validation.
model birth and growth of an initiating earthworm Finally, compost was blended with commercial soil at a
population. In fact, PBE is a complex mathematical ratio 1:1. Seeds of various vegetable were warmed up in
approach for multiple dispersion systems. Historically, it water at 60 oC for 1 h and remained for 10 h before placed
was successfully used to predict the exploration of the in growing boxes. These boxes were watered twice per day.
number of kangaroos in Australia [7]. Nowadays, PBE is After 7 days, compost was added to these boxes to
widely applied in chemical and bioprocesses such as compensate nutrition for growing of vegetables.
crystallization, catalysis reaction, fermentation, etc. [8-12].
B. Applying Population Balance Equation (PBE) model
II. MATERIALS AND METHODS In model formulation, two major processes are
A. Experimental part considered including birth and growth [13-15] of
earthworms as described in Equations (4 and 5). These ODE
Earthworms (Perionyx Excavatus) were purchased from
equations were actually transformed from the PDE (1)
TQCC Company, Cu Chi District, Vietnam. During
which describes variation of number and length of the
transportation, a minor part of worms could die due to
population as a function of time. The initial condition was
mechanical impact and changed environment, thus, a
used as the first population added into the system as seen in
preparation step is needed. Both earthworm and biomass are
(2). Three assumptions were made: 1. New born earthworms
rest in 3 days at mild temperature, humidity and avoid direct
have very small sizes which can be neglected as assumed in
sunlight. After that, about 100 earthworms were randomly
(3); 2. These species are growing with the same rates as seen
collected as the initiate population and transferred into the
in Equation (5); 3. To simplify complexity of the PBE
investigated environment. This environment was prepared
model, kinetics of only two major processes (i.e. birth and
as the following: 20 kg clean soil (commercial product for
growth) will be taken into account while the other
flower and vegetable growing) was placed in a 40x60x40
parameters were assumed to be constants.
cm EPS box with a special netted cover which allows
aspiration, but worms cannot escape. Several small spots at
the bottom are properly made to allow leaking only liquid Initial population
out of the box. Temperature and humidity are frequently Initial parameters
checked using a Beurer MH16 (Germany). Indeed,
temperature and humidity were kept as close to natural
conditions as possible. The temperature varied in the range
of 28 - 32oC while the humidity was found in a range of 58-
75%. Due to the habit of earthworms, the top of the raising
box was covered with canvas to ensure the dark conditions
to help the growth and development of the earthworm
populations. Three boxes were investigated under the same
conditions to perform parallel experiments. Besides, a
control box was also prepared in which the same conditions
as the raising boxes were applied but without presence of
earthworms. Figure 1 described some healthy earthworm
during cultivation and feeding.
End Parameter
Population estimation
Model validation
Figure 1. Raising earthworm in EPS boxes
Feeding was repeated every day with about 5 kg of
kitchen waste containing about 70% vegetable and 30%
End
bonds of fish, chicken, pig, shell, and shrimp, egg, etc.
These materials were directly poured on the surface of the
box. After that, it was covered with a thin layer of sold and a
net on top before being stored in dark place. Figure 2. Systematic approach for modeling of birth and growth of
After 20 days, the population inside the box will be earthworm. Where n is number density function; (kb, b) and (kg, g)
are kinetic parameters of birth (B) and growth (G) terms,
collected and carefully detected number of earthworms and respectively; θ is driving force; L is the characteristic length; and
measured their length with an M&G linear. The obtained µ3 is the third moment of the population.
123
To solve the above ODE system, the ode45 suit in recently predominated by birth term. The shape of the
Matlab was used. Indeed, lsqnonlin algorithm was applied histogram in Fig. 3c was seen as a positively screwed
for estimation of two parameters including b and g by distribution which supposes birth phase dominated in
minimizing the objective function F which was defined as comparison to the growth term. To capture this tendency,
(6). This equation computes the differences between the Fig. 4 presented the increment of earthworms in a period of
simulated and experimental results of the number and size 40 days.
of the worms.
(6)

C. Raising earthworm
Firstly, the cultivation was carried out in 20 days. The
temperature and humidity were regularly controlled at
values of 30±5oC and 65±10%, respectively. The number
and size of initial population were measured and presented
in Fig. 3a. Indeed, this population was randomly collected
from the healthy commercial worms. Herein, 102 individual
worms with various sizes from 0.50 cm to 18.00 cm. The
average diameter of this population was found at a value of
7.49 cm (standard deviation SD of 4.65). These worms
consumed disposing kitchen waste for birth and growth
phases. As seen in Fig. 3a and 3b, after 20 days, the number Figure 4. Number of earthworms as a function of time
increased from 102 to 143 earthworms, equivalent to an
increment of 40.2% in quantity. However, the average size D. Characterization of PBE and model validation
of the new population was decreased to a value of 7.07 cm In general, the characteristic length of earthworm was
(SD 3.27). This evidence proved that little worms were born found in a range from 0.5 to 25 cm. In this part, a linear
with high rate. Therefore, the number of species in the discretization was performed with a step size was selected at
population increased but their averaged size decreased. a value of 0.1 cm. Data in these histograms (Fig. 3a and 3b)
relating to initial and end (after 20 days) populations were
Secondly, the above population was then sent back to the used as input and output for characterization of the PBE
previous environment for further cultivation. After 40 days, model of which the programming procedure was already
this population was again measured with the same procedure described in Fig. 2. In the calculation, shape factor k v
as the previous one. The number was found at 426 (0.056), specific growth and birth factor (k g and kb) were
earthworms with a mean size of 6.98 cm as seen in Fig. 3c. assumed to be unchanged during the cultivation period.
This observation again proved high order of power law These parameters were listed in Table 1.
relating to birth rate of earthworms under cultivation
conditions. Table 1. Parameter estimation
Estimated
Parameter Pre-parameters Fixed value
Value
b, [-] 2.3 kb, [[#.m− 3.day− 1] 1.25102
g, [-] 0.7 kg, [m.day− 1] 7.110-3
Regarding to parameter estimation, firstly, the local

optimization was performed. Then, the global solution was
searched using multi-start algorithm. Consequently, a high
order of birth term was found at 2.3 while the lower value
for growth term at 0.7 was proven. These values reflected
relatively well the observation in the previous section.
In the next section, model validation was carried out.
The obtained parameters were used to check agreement
between the experimental and modeling approaches. A
comparison was performed at a process time of 40 days. Fig.
Figure 3. Development of a population of earthworms in (a) initial 5 showed the experimental and model distributions. From
population, (b) & (c) populations after 20 and 40 days, respectively the similarity of two histograms, an agreement between the
Besides, growth rate of these worms were also two approaches in both quantity and size can be derived.
significant. As seen in this case, the growth phase can However, due to simplified hypotheses, the deviation
compensate the average length of the population which was between these approaches was significant at values of 2.7%
and 4.3% for number and size, respectively. After that,
124
biomass obtained from the above experiments will be decompose rate was also significantly slower than the
evaluated in the next section to check whether they can be cultivation boxes due to activities of earthworm.
used as organic fertilizers. ● Flavor and color of vegetable growing from the obtained
biomass were better than commercial products.
● The yield of crop was significantly increased compared to
the case without biomass.
● Good feasibility for application at urban areas can be
concluded. With a small space about 0.5 m3, kitchen
waste of a medium-size family (e.g.5 members) was
completely solved without any discharged into the
environment. Treatment capacity was estimated about 300
– 400 kg.month-1.m-3.
IV. CONCLUSION & OUTLOOKS
● This paper presented a method to solve kitchen waste
using earthworms. The method worked properly and
showed potential applicability due to several reasons. At
first, it helps to reduce serious pollution situation
happening in many urban areas such as Hanoi and HCM
cities. Earthworms are well known as unharmful species,
Figure 5. (a) Experimental and (b) model correlation of the
earthworm population after 40 days easy for raising, short life-circle and strong activities in
decomposing organic waste. At second, biomass from
E. Organic fertilizers produced from kitchen waste using decomposition can be used as an organic fertilizer with
earthworms excellent quality. This supplies a clean vegetable source
The obtained compost was herein fed for plants on a for daily meals, and therefore it helps to avoid several
balcony with a small area of about 10 m2. In this part, first diseases such as cancers relating to unsafe products in the
experiments using these products as fertilizers were market. Finally, it provides green fields for urban areas,
interesting. Some photos were presented in Fig. 6 for particularly in small spaces such as balconies, roofs, etc.
growing of different plants. These plants were fertilized ● Besides experiments, a mathematical description was
with only the above biomass but did not use any extra applied for this bioprocess and obtained certain
inorganic fertilizers such as NPK. After 1.5 months, the crop achievements. Kinetics of birth and growth of earthworm
was harvested as seen in photos in which vegetable quality were evaluated with a conclusion that high order of power
were remarkably good. Thus, these fertilizers are suitable law dominated birth rate but low order-controlled growth
for various kinds of vegetables such as salad, tomato, bitter rate. Even though a good agreement between experiment
melon, water spinach, cabbage, melon fruit, menthol, etc.
and model correlation was observed, there was a deviation
between two approaches which could relate to
assumptions used in simplifying mathematical
description.
● For future works, improving measurements with finer
measured intervals is suggested. This could help to
capture growing population in more detail. Furthermore,
combining more sophisticated events in the PBE model
Cabbage Tomato Bitter melon will be also considered. Additional events could be
considered relating to death phase, changing conditions of
feeding, temperature and humidity.
REFERENCES
[1] Tran Hong Ha, Vo Tuan Nhan, Nguyen Van Tai, Hoang Van
Thuc: National environmental status report in 2019. Ministry
of Resouces and Environment. Dan Tri Publisher (2020).
Cheri Tomato melon fruit Chili [2] Ministry of natural resources and environment: Vietnam
environmental performance assessment (EPA) report, Greater
Figure 6. Plants growing with organic fertilizers produced from Mekong Subregion – Core environmental program (2012).
kitchen waste
[3] Nicholas P. C. Handbook of Solid Waste Management and
The following benefits were proved during the experiment: Waste Minimization Technologies. Butterworth-Heinemann
● There was no bad smell relating to decomposing of (2003).
kitchen waste thanks to activities of earthworm. In a [4] L Peng, P Zheng and I Siswanto: Good earthworm-kitchen
different manner, several problems were observed in the waste decomposition device. J. Phys.: Conf. Ser. 1456 012029
control box, for instance bad smell of decomposition was (2020).
clearly recorded. Moreover, besides leakage, the [5] Sreejata Biswas, Pulak Lahiri and Satadal Das: Isolation of
predominant bacterium from gut of earthworm Lampito
125
mauritii for effective use in soil fertility. Current science, vol [11] Ehsan Vafa Mohammad Shahrokhi Hossein Abedini: Solution
107, 10, 105-109 (2014). of Population Balance Equations in Emulsion Polymerization
[6] Katheem Kiyasudeen S, Mahamad Hakimi Ibrahim, Shlrene Using Method of Moments. Chemical Engineering
Quaik, Sultan Ahmed Ismail: Prospects of Organic Waste Communications, 200, 1, (2013).
Management and the Significance of Earthworms. Springer [12] Madras, G.; McCoy, B.J.: Reversible crystal growth–
International Publishing (2016). dissolution and aggregation–breakage: Numerical and
[7] P. Bayliss: The Population Dynamics of Red and Western moment solutions for population balance equations. Powder
Grey Kangaroos in Arid New South Wales, Australia. II. The Technol, 143–144, 297–307 (2004).
Numerical Response Function. Journal of Animal Ecology. [13] Basanta Kumara Behera & Ajit Varma: Microbial Biomass
Vol. 54, No. 1, pp. 127-135 (1985). Process Technologies and Management, 257–298, Springer
[8] Miller, S.M.: Modelling and Quality Control Strategies for Link (2017).
Batch Cooling Crystallizers. Ph.D Thesis, The University of [14] Czapla, F.; Haida, H.; Elsner, M.P.; Lorenz, H.; Seidel-
Texas at Austin, Austin, TX, USA (1993). Morgenstern, A.: Parameterization of population balance
[9] Saleeby, E.G.; Lee, H.W.: On the solution of the PBE with models for polythermal auto seeded preferential
agglomeration and random growth rate dispersion. Chem. crystallization of enantiomers.Chem. Eng. Sci. 64, 753–763
Eng. Sci. 50, 1971–1981 (1995). (2009).
[10] Ouchiyama, N.; Rough, S.L.; Bridgwater, J.: A population [15] Tam, L.M.; Thao, P.T.; No, N.T.H.; Vuong, P.M.: A Simple
balance approach to describing bulk attrition. Chem. Eng. Sci. Population Balance Model for Crystallization of L-Lactide in
60, 1429–1440 (2005). a Mixture of n-Hexane and Tetrahydrofuran. Crystals, 12 (2),
221 (2022).
126
Effect of Reinforcement Corrosion on Crack

Development in Concrete Under Load
Vo Van Nam Tran The Truyen
Ho Chi Minh City University of Transport University of Transport and Communications
Ho Chi Minh City, Viet Nam Ha Noi City, Viet Nam
nam.vo@ut.edu.vn truyenestc@gmail.com
Abstract: Corrosion of reinforcement in reinforced and simulated in the experimental environment most closely
concrete structures is a common phenomenon for coastal compared to the actual working conditions of the beam
traffic infrastructure works, reinforced concrete beam members.
structures are affected by aggressive environment, leading to Cracking in reinforced concrete beams is a very common
damaged reinforcement corrosive; Under the simultaneous phenomenon and is studied by many scientists. In Vietnam,
effect of load and corrosion of reinforcement, cracks in author T. T. Pham and his colleagues published research
concrete form early and develop faster than normal, then results on the formation and widening of perpendicular
promote corrosion of reinforcement. This paper conducts cracks in reinforced concrete beams when subjected to
studies to evaluate the influence of loads on the degree of short-term loads according to Russian standards - SP
corrosion of reinforcement in concrete beams; analyzing the 63.13330. 2012 [4]. Author T. M. Nguyen publishes a study
simultaneous influence of reinforcement corrosion and on the comparative analysis of the limit crack width
simultaneous loading on crack development in concrete. regulations and the influence of these regulations on the
Research results allow us to assess more accurately the design of reinforced concrete works with the requirements
degree of corrosion of reinforced concrete constructions in for anti-corrosion works coastal program [5].
coastal areas. A research team from Cairo University, Egypt has
published research results comparing the corrosion of
Keywords: cracking, corrosion of reinforcement, load,
reinforcement on prestressed reinforced concrete beams and
coastal areas, concrete normal reinforced concrete beams, the research results have
I. THE PROBLEM shown that the influence of corrosion caused by the
aggressive environment has a strong effect and rapidly
The effect of reinforcement corrosion in reinforced reduces the bending resistance for normal reinforced
concrete girder members on crack development and vice concrete beam structures. This has less impact on the
versa is always an issue that needs to be carefully prestressed reinforced concrete beam structure with the
considered during the operation of transportation same experimental environmental conditions, the group also
infrastructure project [1]. Coastal project using reinforced pointed out that the cause of the prestressed reinforced
concrete structures, under the effect of loads, cracks appear concrete is less affected because the cracks in the concrete
and develop in girder structures, strongly aggressive develop less. In addition, high-strength steel is surrounded
environment and the corrosive agents penetrate and and protected by concrete mortar and gene pipe, so it is
penetrate the core through the cracks, causing corrosion to difficult for corrosive agents to contact to promote corrosion
the reinforcement in the reinforced concrete structure. of reinforcement [6]. In 2018, a group of graduate students
Corrosion process of reinforcement increases the volume Hussain, H., & Miteva, D. of Chalmers University of
of reinforcement and reduces cross sectional area of the Technology, Sweden completed a research thesis on the
reinforcement thereby reducing the flexural resistance of behavior of corroded reinforced concrete structures using
reinforced concrete members, under the effect of concrete the 3D FE model, The authors have evaluated the
structural loads in the tensile zone, many cracks appear performance of the corroded reinforced concrete beams by
because the reinforcement in this area is corroded and the experiments and finite element (FE) model, the research
bending resistance is reduced. In order to protect the results have concluded that the corrosion process reduces
reinforcement from corrosion during the manufacture of the diameter of the reinforcement, reducing the adhesion
concrete structures, people usually arrange a protective force between concrete and reinforcement thereby reducing
concrete layer with a thickness of 2cm to 4cm. However, the load capacity of reinforced concrete beams [7].
concrete is not a good waterproof material, so the
penetration of corrosive agents still takes place silently and II. EXPERIMENT PROGRAM
the larger the thickness of the protective concrete layer, the A. Material and sample creation
more obvious the crack development process [2], [3].
C30 concrete has compressive strength at the age of 28
In order to be able to evaluate seriously and accurately days is expected to be fc' = 30 (Mpa). The aggregate
the influence of reinforcement corrosion on crack composition of concrete is designed according to ACI
development of reinforced concrete beams under load, it is 211.1-91 standards [3].
necessary to conduct experiments in conditions of fast
corrosive environment with the natural conditions designed Steel used to make beams: D10, smooth round, fy = 240
(Mpa).

Table 1. Material for 1 cubic meter of concrete

Composition Weight
Concrete 395 (kg)
Sand 848 (kg)
Soil 1060 (kg)
Water 170 (lit)
Samples of beams manufactured for experiment have

dimensions of 500x100x100 (mm) [8].
Figure 3. Bending sample to get PMax result
Quantity: 10 beams.
Table 4. Number of beam samples for loading level
4 points bending test Amount of Load
girder
Girder group 01 – No corrosion soaking 02 0
Girder group 02 – No corrosion soaking 02 0,8. Pmax
Girder group 03 – with corrosion soaking 04 0,8. Pmax
Figure 1. Dimension of experiment samples The beam samples are shackles loaded in pairs, ensuring
that the beam can withstand 4-point bending; 02 girder head
bearings are 25 mm from the beam end, 2 inside girder
bearings are 150 mm apart from the girder head support,
divide the beam into 3 segments 150 mm each, 0.8.Pmax, by
tightening the set bolts of reinforcement fixtures
Figure 2. The experiment samples

B. Experimental process
The objective of the research direction is to analyze the
effect of the reinforcement corrosion process on the crack Figure 4. The girder sample is loaded without soaking
development of reinforced concrete beams under the effect
of loads.
The study was conducted on 10 samples of reinforced
concrete beams with dimensions of 500x100x100 (mm), of
which 02 samples were used for bending to determine
Pmax, 02 samples without loading, not soaking in corrosion;
02 samples were loaded with 0.8.Pmax, no corrosion
soaking process and 04 samples of girders were added level
load. 0.8.Pmax, immersed in a rapidly corrosive
environment.
Sample group 01: 02 samples without loading, not
soaking in corrosion; Figure 5. Girder specimen loaded in corrosive environment
Sample group 02: 02 samples level loading 0,8.Pmax,
not soaking in corrosion;
Sample group 03: 04 level loading sample 0,8.Pmax,
immersed in a rapidly corrosive environment
[9,10,11,13,14].
Table 2. The number of sample to get Pmax
Experiment Amount of girder Symbol Load

4 point bending test Pmax 02 M01 , M02 Pmax
Table 3. PMAX value statistics

Figure 6. Girder specimen unloaded in corrosive environment
Sample Pmax (KN) σ (Mpa)
M01 Pmax1 = 28 σ1 = 150
The team determines the shackle force through the
loadcell, and controls it through strain and stress, by electric
M02 Pmax2 = 27,66 σ2 = 148,2 resistor plates on the shackle equipment and girder devices.
128
The loading levels will be determined through the bending

stress on the reinforcement, when the concrete stress is
equal to the cracking modulus of the concrete. This control
is implemented during the girder loading process [3], [4],
[9], [12].
Figure 11. Experiment sample
Figure 7. Description of beam loading device
Figure 8. Description of electrode corrosion equipment
Figure 12. Destructive bending of samples after corrosion test
Figure 9. Experimental practice of electrode corrosion

After the girde samples are shackled into pairs according
to the load class as shown in Table 4, the girders group is
soaked for 30 days in a rapidly evolving corrosive
environment; It is then taken out and bent to destruction.
Figure 13. Destructive bending of samples without corrosive

Figure 10. Experiment sample after corrosive process immersion
129
III. EVALUATION AND ANALYSIS Initial steel volume: is the volume of 02 smooth round
D10 steel bars that are weighed after cutting according to
C. Evaluation of destructive force and crack drawing dimensions, before manufacturing girders.
characterization on beams
Volume of steel after soaking: is the weight of 02 steel
Table 4. Destructive statistics and crack R of samples bars in the girders after the sample has been soaked for 30
Destructive
days. The steel bar is removed and cleaned of rust, concrete
Destructive Characteristic clinging to the bar and then weighed.
Initial bending
bending force of crack
Sample load force after
No after locations on
name force immersion
immersion beams
(KN) average
(KN)
(KN)
1 Sample 0 28.01 Location of
1.1 tensile zone,
perpendicular
2 Sample 29.03 28.7 to normal
1.2 stress, crack
direction of
beam
perpendicular
to beam axis,
wide below,
small upper. Figure 15. Corrosion of steel reinforcement after the test
3 Sample 0 22.10 Location of
2.1 tensile zone,
perpendicular
4 Sample 19.96 to normal
2.2 21.03
stress, crack
direction of
beam
perpendicular
to beam axis,
wide below,
small upper.
5 Sample 0,8.Pm 14.40 Tensile zone
3.1 ax location Figure 16. Weigh the reinforcement sample to determine the level
Appearance of of corrosion
6 Sample 0,8.Pm 12.20 12.41
widening,
3.2 ax Table 5. Statistics of lasting steel steel groups of beam
dissecting
7 Sample 0,8.Pm 10.10 cracks along
the steel Corrosion Average
3.3 ax Sample Initial load
No. ration corrosion ratio
8 Sample 0,8.Pm 12.95 name force (KN) (%)
(%)
3.4 ax
1 Sample 1.1 0 0
Relationship of destructive load and degree of corrosion 2 Sample 1.2 0 0 0
3 Sample 2.1 0 26.7
50 29,41
21,0325,95 4 Sample 2.2 25.2 25.95
12,41
0 0 5 Sample 3.1 0,8.Pmax 29.8
0 6 Sample 3.2 28.2
Group 1 Group 2 Group 3 7 Sample 3.3 30.2 29.41
8 Sample 3.4 29.5
Destructive bending force % Corrosion
Figure 14. Relationship of destructive load and degree of Environmental relations, loading force and degree of
corrosion corrosion
D. Evaluate the amount of steel corroded after the test 40

29,41
30 25,95
Controlling the amount of steel corroded during the test
for girders assemblies. 20
The girder samples belonging to 3 groups of samples, 8
10
after bending to destructive, will proceed to destroy the 0 0 0
0
protective concrete layer, the steel bars will be removed and
Group 1 Group 2 Group 3
cleaned of the surrounding steel rust, the concrete clinging
to the bar, and then weighed accurately, determine the Force % corrosion
volume. The amount of steel wear is calculated by the
difference in mass between the steel bars before and after
soaking. Figure 17. Environmental relations, loading force and degree of
corrosion
130
IV. CONCLUSION AND SUGGESTION

REFERENCES
On the basis of evaluation and analysis of experimental [1] M. Young, The Technical Writer’s Handbook. Mill Valley,
results of girder sample groups, assessment of the degree of CA: University Science, 1989.
corrosion and crack development on experimental beam T. V. Nguyen, T. T. Tran, T. H. Xuan, “Permeation and
groups; The research team made some conclusions and corrosion of reinforced concrete structures”. Construction
recommendations: Publisher, 2014.
[2] H. K. Dong, H. T. T. Duong. “Corrosion of reinforced
Reinforcing steel in girder group 01 - girder group is not concrete and the method to avoid corrosion for reinforced
loaded, and at the same time is not immersed in a rapidly concrete structures in Vietnam’s sea”. VNCOLD, 2014.
evolving corrosive environment; zero corrosion mass; Thus, [3] Concrete mix design. Standard ACI211.1. 1991.
when in normal environment, few corrosive agents plus no [4] T. T. Pham, D. Q. Pham, T. V. Dinh, & Q. V. Nguyen, “A
load, during immersion, cracks do not appear and develop, study on the appearance and development of the vertical crack
reinforcement is always protected by outer layers of of the reinforced concrete beam under short-term
concrete, so they do not corrode; loading”. Journal of Construction Science and Technology.
2018. No. 12, pp. 3-10.
In girders group 02 - group of unloaded girders, [5] T. M. Vo, “Effect of limited crack width in design of
immersed in a rapidly corrosive environment, the average reinforced concrete structures in coastal area”. Journal of
corrosion volume of 2 girder samples is: 25.95%; girders Construction Science and Technology. 2021. No. 3.
appear only longitudinal cracks, dissection along the beam [6] M. Mohamed, H. El-Karmoty, A. El Zanaty, “Behavior of
corroded bonded fully prestressed and conventional concrete
body; can be seen when under the influence of the beams”. HBRC Journal, 2016; Issue 14, pp. 137-149.
environment many corrosive agents, with the process of [7] Hussain. H., Miteva. D. “Structural behavior of corroded
permeating these agents through the concrete layer and reinforced concrete structures - A study based on detailed 3D
corroding the reinforcement; Corrosion level is much larger FE analyses”. Master dissertation. Chalmers University of
than that of girders group 01. Technology, 2018.
[8] Y. Xu, J. Shen, Y. Zheng, J. Mao, P. Wu. “Corrosion
In beam group 03 - girder samples were loaded to the Characteristics of Reinforced Concrete Under the Coupled
level of 0.8.Pmax and immersed experimentally in a rapidly Effects of Chloride Ingress and Static Loading: Laboratory
evolving corrosive environment; the amount of corroded Tests and Finite Element Analysis”. Materials Science. 2018.
reinforcement is significantly larger than this process in No. 2. Vol. 24.
girder groups 01 and 02 mentioned above. [9] H. T. Nguyen, C. C. Van, H. D. Nguyen, K. D. Dao. “Study
on the reduction of flexural strength of reinforced concrete
Thus, it can be concluded that the degree of beams by corrosion level”. Vietnam Journal of Construction.
reinforcement corrosion depends on the working 2020; No. 624, pp. 128-132.
environment conditions and the load acting on the girder; in [10] T. H. Nguyen, A. T. Le, D. D. Nguyen. Bending strength
diagnosis for corroded reinforced concrete beams with
the environment containing many corrosive agents, the attendance of deterministic, random and fuzzy parameters.
reinforcement will be exposed to the corrosive environment Journal of SI & M. 2020; Vol. 5, Issue 3, pp. 183-189.
through the process of seepage or directly through the [11] A. N. Vu. Corrosion of reinforcement and its influence on the
cracks caused by the action of the load. behavior of reinforced concrete structures.Vietnam National
Committee on Large Dams and Water Resources
Corrosion reinforcement often swells in volume, Development – VNCOLD, 2014.
cracking the protective concrete structure, through the [12] T. D. Cao, K. V. Pham, H. Q. Le. “Summary report of KT-KT
cracks the ability to penetrate the corrosive agent progresses project Anti-corrosion and protection of concrete and
faster and more complicatedly. After that, the level of reinforced concrete works in coastal areas”. Vietnam Journal
corrosion became more and more serious. of Construction, 2003.
[13] Zhou. Y., Gencturk. B., Willam. K., & Attar. A. Carbonation-
Along with the process of increasing the corrosion of the induced and chloride-induced corrosion in reinforced concrete
reinforcement, the bending resistance of the beam is reduced structures. Journal of Materials in Civil Engineering, 2015;
due to the reduction of the bearing cross section of the No. 27, Issue 9.
reinforcement. [14] Coronelli. D, Gambarova. P. Structural assessment of
corroded reinforced concrete beams: Modeling guidelines.
On the surface of the beam experimentally corroded, Journal of Structural Engineering, 2004, Vol. 130, pp. 1214-
widening and destructive cracks appeared along the girder 1224.
body, dissection along the position of the reinforcement bar [15] A.El-Hefnawy. “A New Statistical Approach for Predicting
the Residual Capacity of Reinforced Concrete Beams Having
in the corroded reinforced concrete beam structure. The Corroded Main Steel”. Ph.D. Thesis. Cairo University, 2000.
cause may be due to corroded reinforcement, causing
expansion of reinforcement volume, causing local
longitudinal cracking in the concrete area around the
corroded reinforcement.
It is recommended to further study the influence of
reinforcement corrosion on the adhesion between concrete
and reinforcement in reinforced concrete beam structures.
131
Study on Synthesizing and Size Controlling of

Silver Nanoparticles by Using a System of Two
Protectants Trisodium Citrate and
Polyvinylpyrrolidone
Hien Chuc Mai Quynh Nguyen Thi Nhu Thuan Hoang Duc
Faculty of Pharmacy Faculty of Pharmacy Faculty of Pharmacy
7B Military Medical Hospital Lac Hong University Lac Hong University
Bien Hoa, Viet Nam Bien Hoa, Viet Nam Bien Hoa, Viet Nam
chucmaihien@gmail.com quynhnguyen@lhu.edu.vn thuanhoang@lhu.edu.vn
Du Cao Van Cuong Ngo Van Dung Duong Thi Ngoc

Lac Hong University Lac Hong University Lac Hong University
Bien Hoa, Viet Nam Bien Hoa, Viet Nam Bien Hoa, Viet Nam
ducao@lhu.edu.vn vancuong@lhu.edu.vn dungduong@lhu.edu.vn
Abstract: In this research, we successfully synthesized catalytic reduction, microwave-assisted synthesis, irradiation
small-sized silver nanoparticles by the chemical reduction reduction, the microemulsion method, biochemical method,
method. Silver nanoparticles are formed from the reduction etc., but these chemical methods have been reported along
of silver nitrate in an aqueous solvent, using two protectants with various drawbacks, including the use of toxic solvents,
including a space protectant (PVP - polyvinylpyrrolidone) generation of hazardous products, high energy consumption,
combined with a charge protectant that also acts as a which pose potential risks to human health and the
reducing agent (TSC - trisodium citrate). The size of silver environment [2, 4-7]. Up to now, several methods for
nanoparticles was controlled through survey parameters synthesizing silver nanoparticles have been proposed to limit
these problems such as green synthesis and biological
including the concentration of reducing agent TSC,
synthesis [2, 7-9].
temperature and time of reaction, and the ratio of Ag+/PVP.
At the best parameters, the silver nanoparticles formed had In this research, we present the process of synthesizing
surface plasmon resonance at 399 nm through UV-Vis silver nanoparticles by chemical reduction method with a
measurement, the average size was 6 ± 4 nm determined by system of 2 protectants PVP and TSC. In which TSC acts as
TEM image, and Fcc structure was determined by X-ray both a reducing agent and a protectant, this synthesis has the
diffraction. advantages of a fast reaction, a simple reaction system, and
the size of silver nanoparticles is easily controlled through
Keywords: silver nanoparticles, polyvinylpyrrolidone, experimental parameters. The results of the research are the
trisodium citrate basis to apply silver nanoparticles to cosmetics, medical
supplies, and functional foods.
I. INTRODUCTION
Silver nanoparticles (AgNPs) have broad-spectrum II. MATERIALS AND METHODS
antimicrobial activity against various bacteria, fungi, and A. Materials
viruses. Because of their versatility, they have been used in
biomedical applications, water, and air purification, foods, Polyvinylpyrrolidone (PVP), Mw = 40000 g/mol was
cosmetics, clothing, numerous household products, wound purchased from India. Trisodium citrate (TSC) and silver
dressings, medical devices, deodorant sprays, and fabrics. nitrate were purchased from Merck – Germany.
Silver nanoparticles might exhibit additional antimicrobial B. Synthesis of silver nanoparticles
capabilities which are not exerted by ionic silver. Their small
A quantity of 0.2 M AgNO3 solution and 0.2 g PVP (20
size and large surface-to-volume ratios lead to both chemical
ml of solution 1 g/100 mL) are added to the beaker on a
and physical differences in comparison with their bulk
magnetic stirrer at the reaction temperature. Then trisodium
counterparts. AgNPs can be produced in various sizes and
citrate is added slowly to the mixture. The final product has a
shapes depending on the fabrication methods, among which
characteristic yellow color signaling silver nanoparticles are
the most widely used is the chemical reduction method [1-3].
formed. The total volume after synthesis is 30 mL. Survey
Several chemical methods have been developed for the parameters include the concentration of TSC reducing agent,
synthesis of silver nanoparticles including chemical reaction temperature, reaction time, and silver nitrate
reduction, aqueous chemical reduction, nonaqueous chemical content.
reduction, the template method, electrochemical reduction,
C. Identification of synthesized silver nanoparticles
ultrasonic-assisted reduction, photo-induced or photo-
UV-vis spectra were determined on UV-2450-Shimadzu

Figure 1. XRD diffraction pattern of the silver nanoparticles

machine (Japan) at Vietnam Academy of Science and different concentrations of TSC reducing agents. The results
Technology - Ho Chi Minh City. The structure of silver showed that, with the concentration of TSC reducing agent
nanoparticles was determined by an XRD diffraction pattern 0.2 M, the produced silver nanoparticles are mostly in quasi-
made on D8 Advance-Bruker (Germany) at the Institute of spherical form and evenly distributed with the size of 25 ±
Applied Materials Science - Ho Chi Minh city. The size and 10 nm (Figure 3). When increasing the concentration of the
distribution of silver nanoparticles were determined from TSC reducing agent to 0.4 M and 0.6 M, the silver
TEM images performed on the JEM-1400 (Japan) combined nanoparticles have a smaller, more uniform size in the size
with IT3.Tool particle counting software at University of range of 13 ± 5 nm (Figure 4) and 15 nm ± 5 nm (Figure 5).
Science and Technology - Ho Chi Minh City. The TEM image results are consistent with the previous
results of predicting the size change of silver nanoparticles
III. RESULTS AND DISCUSSION by UV-Vis analysis in Figure 2.
A. XRD analysis results
The results from the X-ray diffraction pattern in Figure 1
show that the 5 peaks with the highest intensity completely
coincide with the standard spectrum of metal Ag at the
position of the angle values 2θ = 38o (d = 2.35837Å); 44.2o
(d = 2.05520Å); 64.4o (d = 1.44606Å); 77.6o (d = 1.23137Å)
and 81.6o (d = 1.17804Å) correspond to the faces {111},
{200}, {220}, {311} and {222} belonging to the Bravais
lattice in the structure Fcc of Ag metal [6,8].
B. Effect of TSC concentration Figure 2. UV-Vis spectra of AgNPs at different TSC
To investigate the effect of TSC reducing agent concentrations
concentration, the process was conducted with the following
parameters: reaction temperature 50 oC; reaction time 5
minutes; 0.2 g PVP (20 ml of solution 1g/100 mL); the
amount of AgNO3 was determined from 0.2 M AgNO3
solution to ensure a mass ratio of Ag+/PVP of 3% (w/w).
TSC reducing agent concentration varies from 0.1 M ÷ 0.6
M.
UV-Vis analysis results of the synthesized silver
nanoparticles are shown in Figure 2. The results show that Figure 3. TEM image and size distribution of AgNPs at 0.2 M TSC
there is a shift in the position of the absorption peaks when
increasing the concentration of the TSC reducing agent.
Specifically, the positions of the absorption peaks are 412
nm (0.1 M), 410 nm (0.2 M), 400 - 401 nm (from 0.3 to 0.6
M). According to the surface plasmon resonance
phenomenon, this result allows predicting the silver
nanoparticles synthesized with a large concentration of TSC
reducing agent (from 0.3 M to 0.6 M) have smaller size and
are more stable than the silver nanoparticles synthesized with
Figure 4. TEM image and size distribution of AgNPs at 0.4 M TSC
a low concentration of TSC reducing agent (0.1 M and 0.2
M) [5]. Figures 3, 4, and 5 are TEM images and size
distribution diagrams of silver nanoparticles synthesized with
133
Figure 5. TEM image and size distribution of AgNPs at 0.6 M TSC

Thus, the concentration of the TSC reducing agent has a Figure 7. UV-Vis spectra of the AgNPs at different temperature
strong influence on the size and distribution of the formed
silver nanoparticles. When nanoparticles are synthesized Figures 8, 9, and 10 are TEM images and size
with low concentrations of reducing agents (0.1 and 0.2 M), distribution diagrams of silver nanoparticles synthesized at
the amount of nuclei produced is small, and the nucleation different temperatures. The results show that, at the
process takes place quickly. In contrast, the growth and temperature of 40oC and 60oC, the silver nanoparticles are
formation process of nanoparticles are prolonged, the mainly in quasi-spherical form, uniform distribution, and
particles tend to agglomerate during growth to form larger respective sizes of 9 ± 4 nm (Figure 8) and 9±5 nm (Figure
particles. When increasing the concentration of TSC 9). When the temperature is increased to 80oC, the formed
reducing agents (0.3 to 0.6 M), the nucleation and growth silver nanoparticles still have uniform distribution but are
rates are compatible, so that the silver nanoparticles formed larger (15 ± 5 nm, Figure 10). This result is consistent with
are smaller and more uniform in size. In addition, this result the results of UV-Vis spectra in Figure 7.
can also be explained by the role of TSC, the excess amount
of TSC also acts as a charge protectant to help stabilize the
formed silver nanoparticles. Mechanism of PVP and TSC in
protecting silver nanoparticles is proposed as shown in
Figure 6.
Figure 8. TEM image and size distribution diagram of AgNPs at

40oC
Figure 6. Protective synergism of TSC and PVP : (a) before the

reaction, (b) after the reaction to form silver nanoparticles
C. Effect of reaction temperature
The effect of temperature was investigated on the
following parameters: concentration of reducing agent TSC Figure 9. TEM image and size distribution diagram of AgNPs at
0.3 M; reaction time 5 minutes; 0.2 g PVP and the amount of 60oC
AgNO3 were determined from 0.2 M solution to ensure the
ratio of AgNO3/PVP = 3%; The reaction temperature varies
from 40 to 90oC.
The obtained silver nanoparticles give UV-Vis analysis
results as shown in Figure 7. The results show that when the
temperature increases from 40 to 70oC, the position of
absorption peaks is almost unchanged in the wavelength
range from 400 to 401 nm. At higher (80 and 90oC), the
position of absorption peaks shifts towards the larger
wavelength with values of 407 and 425 nm, respectively.
Figure 10. TEM image and size distribution diagram of AgNPs at
This result allows predicting that silver nanoparticles are 80oC
small in size and stable when synthesized in the temperature
range from 40 to 70oC. However, when the temperature is D. Effect of reaction time
higher (80 and 90oC), collisions and agglomeration will The reaction time was determined from the time TSC
occur during particle formation, so the obtained silver reducing agent was completely injected into the reaction
nanoparticles will have a larger size. This result will be system and the temperature and stirring speed are stable.
verified through analysis of TEM images. Time is a factor affecting the size of the formed silver
134
nanoparticles. When the reaction time is suitable, the residual

TSC will play an important role, silver nanoparticles will be
protected by both PVP and TSC at the same time.
To investigate the effect of reaction time, the process was
carried out with the following specific parameters: 0.2 g
PVP; 0.3 M TSC reducing agent; the amount of AgNO3 was
determined from 0.2 M AgNO3 solution to ensure enough
Ag+/PVP ratio of 3%; reaction temperature 60oC, reaction
time varies from 2 to 7 minutes.
The obtained silver nanoparticles give UV-Vis analysis Figure 13. TEM image and size distribution of AgNPs at 7 minutes
results as shown in Figure 11. The results show that when the
reaction time is suitable (2 to 5 minutes), the position of the E. Effect of Ag+/PVP ratio
absorption peaks is almost unchanged with wavelengths To investigate the influence of the Ag+/PVP ratio, the
between 400 and 401 nm. When the reaction time is longer process was carried out with the following specific
(7 min) the absorption peak position has a larger wavelength parameters: PVP 0.2 g; 0.3 M TSC reducing agent; reaction
shift with a wavelength value of 408 nm. This result can temperature 60 oC; reaction time 3 minutes. The amount of
show that the generated silver nanoparticles have a stable, AgNO3 was determined from 0.2 M AgNO3 solution to
small size when reaction time from 2 to 5 minutes. When the ensure that the ratios of Ag+/PVP = 1%, 2%, 3%, 4%, 5%,
reaction time is longer (7 minutes), collision, and 6% and 7%.
agglomeration of the particles will form silver nanoparticles
UV-Vis spectra of silver nanoparticles are presented in
of larger size.
Figure 14. The results show that when the Ag+/PVP ratio
The results of UV-Vis analysis in Figure 11 were verified increases (from 1 to 5%), the peak absorption peak intensity
by TEM images in Figures 12 and 13. The results showed gradually increases, however, the change in position of the
that the reaction time of 3 minutes creates the silver absorption peaks is not much, in the range of 399 ÷ 401 nm.
nanoparticles in quasi-spherical form with uniform With higher Ag+/PVP ratios (Ag+/PVP = 6 or 7%), the peak
distribution and a size of 8 ± 4 nm (Figure 12). When the positions are 407 and 412 nm respectively. Thus, it can be
time was increased (7 min), collision and agglomeration led predicted that with the Ag+/PVP ratios varying from 1 to 5%,
to the formation of larger silver nanoparticles with the size of the silver nanoparticles formed will be small and stable.
15 ± 5 nm (Figure 13). Those results are also consistent with However, when increasing the Ag+/PVP ratio to 6% and 7%,
the results of UV-Vis spectra in Figure 11. the size of the silver nanoparticles will increase. This result
was verified through the analysis of TEM images.
Figure 11. UV-Vis spectra of synthesized AgNPs at different

reaction time
Figure 14. UV-Vis spectra of the AgNPs at different ratios of

Ag+/PVP
Figures 15, 16, and 17 are the results of TEM image
analysis of silver nanoparticles with the Ag+/PVP ratios of 1,
3, and 5%. The results show that the formed silver
nanoparticles have a uniform distribution in the form of a
sphere with a size range of 7 ± 4 nm (Figure 15), 9 ± 4 nm
(Figure 16), and 6 ± 4 nm (Figure 17) respectively. Those
results are consistent with the absorption peaks in the UV-
Figure 12. TEM image and size distribution of AgNPs at 3 minutes
Vis spectra (Figure 14).
135
REFERENCES
[1] S. S. Jeremiah, K. Miyakawa, T. Morita, Y. Yamaoka, and
A. Ryo, “Potent antiviral effect of silver nanoparticles on
SARS-CoV-2”, Biochemical and Biophysical Research
Communications, 533(1), 2020, pp. 195–200.
[2] Mariana Guilger-Casagrande and Renata de Lima,
“Synthesis of silver nanoparticles mediated by fungi: a
review” Front. Bioeng. Biotechnol, 7, 2019, pp. 287. DOI:
10.3389/fbioe.2019.00287.
Figure 15. TEM image and size distribution pattern of AgNPs at [3] Gloria E. Campillo, Ederley Vélez, Gladis Morales, César
the ratio Ag+/PVP = 1% Hincapié, Jaime Osorio, Oscar Arnache, José Ignacio Uribe,
Franklin Jaramillo (2017), “Synthesis of Silver nanoparticles
(AgNPs) with Antibacterial Activity”, IOP Conf. Series:
Journal of Physics: Conf. Series 850 (012023), 2017.
[4] Zongtao Zhang, Bin Zhao and Liming Hu, “PVP Protective
Mechanism of Ultrafine Silver Powder Synthesized by
Chemical Reduction Processes”, Journal of Solid-State
chemistry 121, article no.0015, 1996, pp. 105-110.
[5] Vincenzo Amendola, Osman M. Bakr & Francesco Stellacci
(2010), “A Study of the Surface Plasmon Resonance of
Silver Nanoparticles by the Discrete Dipole Approximation
Figure 16. TEM image and size distribution pattern of AgNPs at Method: Effect of Shape, Size, Structure, and Assembly”,
the ratio Ag+/PVP = 3% Plasmonics 5, 2010, pp. 85–97, DOI 10.1007/s11468-009-
9120-4.
[6] Peter Logeswari, Sivagnanam Silambarasan and Jayanthi
Abraham, “Synthesis of silver nanoparticles using plants
extract and analysis of their antimicrobial property”, Journal
of Saudi Chemical Society, 2012
http://dx.doi.org/10.1016/j.jscs.2012.04.007.
[7] Hinal Gandhi and Shabib Khan, “Biological Synthesis of
Silver Nanoparticles and Its Antibacterial Activity”, J
Nanomed Nanotechnol, 7:2, 2016, DOI: 10.4172/2157-
Figure 17. TEM image and size distribution pattern of AgNPs at 7439.1000366.
the ratio Ag+/PVP = 5% [8] Hongshui Wang, Xueliang Qiao, Jianguo Chen, Shiyuan
Ding, “Preparation of silver nanoparticles by chemical
IV. CONCLUSION
reduction method”, Colloids and Surfaces A: Physicochem.
Silver nanoparticles have been successfully synthesized Eng. Aspects 256, 2005, pp. 111–115.
by a simple chemical reduction method, in a short time, [9] Chhangte Vanlalveni, Samuel Lallianrawna, Ayushi Biswas,
silver nanoparticles are formed with small size, stability, and Manickam Selvaraj, Bishwajit Changmai and Samuel
little variation in a wide range of synthesis parameters such
Lalthazuala Rokhum, “Green synthesis of silver
as temperature (from 40 to 80oC), TSC reducing agent
nanoparticles using plant extracts and their antimicrobial
concentration (≥ 0.3 M), reaction time (2 to 5 minutes),
Ag+/PVP ratio (from 1 to 5%). At the best synthesis activities: a review of recent literature”, RSC Adv, 11, 2021,
parameters, the formed silver nanoparticles are stable, and pp. 2804-2837 DOI: 10.1039/d0ra09941d.
uniformly distributed with the smallest size of 6 ± 4 nm. This [10] Jing Xiong, Ye Wang, Qunji Xue, and Xuedong Wu,
result was achieved due to the protective synergism of PVP “Synthesis of highly stable dispersions of nanosized copper
and trisodium citrate, in which TSC plays an important role particles using L-ascorbic acid”, Green Chem., 13, 2011, pp.
as both a reducing agent and a protectant. This result is the 900-904.
basis for our further studies on silver nanoparticle application
as well as contributing to the development of reaction
systems for the synthesis of small, stable metal nanoparticles
in general.
136
Study on Chemical Composition and In-vitro

Biological Activities of Salvia officinalis L. in
Lamdong, Vietnam
Thao Tran Thach Cuong Ngo Van Xuan Nguyen Bang
Bien Hoa, Dong Nai Hoa, Dong Nai Hoa, Dong Nai
thachthao@lhu.edu.vn vancuong@lhu.edu.vn xuanhcm2444789@gmail.com
Abstract: The aim of this research is to evaluate the for further studies of S. officinalis in Vietnam and the
chemical composition as well as antioxidant, antimicrobial, application of sage to medicinal products.
and anti-inflammatory activities of essential oil (SoEO) and
water extract (SoWE) from the leaves of Salvia officinalis. II. MATERIALS AND METHODS
Gas chromatography analysis revealed that camphor A. Plant material
(32.63%) was the major chemical constituent found in Leaves of S. officinalis were collected in Lamdong,
SoEO. The other main components are α-thujone (17.11%), Vietnam, from July to December 2021. Specimens were
viridiflorol (8.64%), β-thujone (6.88%), epimanool (6.57%), identified species by amplifying DNA through PCR reaction
and eucalyptol (6.46%). By using chemical reactions, and comparing the similar levels of sample in NCBI gene
triterpenoids, alkaloids, flavonoids, tannins, and saponins bank. The identification was conducted by Professor Huynh
were found in SoWE. The SoEO and SoWE were also Ky, Faculty of Agriculture, Can Tho University.
screened for their antimicrobial activity. While SoEO
recorded moderate inhibitions for all tested microorganisms, B. Bacterial and fungal strains
the SoWE only shows an effect on Staphylococcus aureus Five bacteria and one fungus were used in the in-vitro
and Candida albicans. The antioxidant activity determined test including Staphylococcus aureus ATTC 23235,
by 2,2-diphenyl-1-1picrylhydrazyl radical-scavenging, Staphylococcus epidermidis ATTC 14990, Shigella sp.
identify IC50 are 62.74 and 6653.36 µg/mL on SoWE and ATTC 23354, Escherichia coli ATTC 25922, Pseudomonas
SoEO, respectively. In the egg albumin denaturation aeruginosa ATTC 15442, Candida albicans ATTC 10231.
method, the value IC50 of SoWE was 0.058 µg/mL whereas C. Extraction
standard diclofenac was 0.038 µg/mL. In conclusion, SoEO The essential oil (SoEO) has been extracted from dried
has strong antimicrobial effects, while SoWE has leaves by hydrodistillation for 3 h, using a Clevenger-type
antibacterial, antioxidant, and anti-inflammatory effects. apparatus. The remaining water extract (WE) in the flask was
filtered and evaporated in the water bath to get the SoWE
Keywords: salvia officinalis, antioxidant, antimicrobial, fraction. Both SoEO and SoWE were stored at 0oC until use.
anti-inflammatory, essential oil, water extract
D. Gas chromatography-mass spectrometry (GC-MS)
I. INTRODUCTION Detecting the components in SoEO was conducted at the
Salvia is the largest genus of the Lamiaceae family, Center of Analytical Services Experimentation of Ho Chi
which is native to the Mediterranean area and includes about Minh City (CASE) by using GC Agilent 6890N machine,
900 species. The plants of Salvia are used in folk medicine 5973 INERT Mass Selective Detector, and HP5-MS column
for their antibacterial, antitumoral, antidiabetic, and (0.25 µm, 0.25 mm, 30 m). The oven temperature was kept
antioxidant activities. They are used traditionally in food at 50oC for 2 min, gradually raised to 80oC at 2oC /min, then
preparation, herbal tea, flavoring agents in perfumery, and raised to 150oC at 5oC/min, raised to 200oC at 10oC/min, and
cosmetics [1]. finally raised to 300oC at 20oC/min, keep at 300oC in 5
minutes. Helium was the carrier gas. The sample was diluted
In Vietnam, numerous Salvia species were investigated. in hexane and 1 μL was injected into the system with a split
Among them, Salvia officinalis (sage) is encountered in cool ratio of 1:50. The determination of components in the
weather places such as Lamdong and is considered a essential oil is done by using the pre-installed software on
medicinal herb with a significant essential oil (EO) potential. the machine to compare the MS and GC-MS spectra of each
Because the phytochemistry components and biological component separated on the chromatogram with the standard
activities of plants could be affected by several factors such spectra available in the spectrum library. Quantitative data of
as geographical origin, climate, and soil... this study aims to the individual components were obtained from peak area
determine the chemical compositions of S. officinalis in percent data.
Vietnam as well as to assess its antimicrobial, antioxidant,
and anti-inflammatory activities to compare with plants of
this species in other places. The research results are the basis

E. Identification of organic compounds value of DPPH. The tests were repeated 1 more time and the
The organic compounds were identified by using average value of the 2 trials was taken. The IC50 values were
characteristic chemical reactions or reagents in Table 1. determined graphically by linear regression.
H. Anti-inflammatory activity
Table 1. The characteristic chemical reactions or reagents
Using the egg albumin denaturation method of Dey P. et
Organic al. with modifications (2011) [4]. Substances with protein
No. Chemical reactions or reagents
compounds protection ability make the protein density measured higher
Precipitation with Mayer, Dragendorff, than the protein concentration without the protectant at 660
1 Alkaloid
and Bouchardat reagents nm. 2 mL of different concentrations of samples prepared in
distilled water was added to 0.2 mL egg albumin and 2.8 ml
2 Anthranoid Borntrager reaction (shake with NaOH) of phosphate buffer solution (pH 6.3). The samples were
Increasing fluorescence in alkali at UV
mixed, incubated at 37oC for 20 min, then increased to 60oC
3 Coumarin
365 nm
and kept for 5 min. It is then cooled at about 20°C for 15 min.
Measure the absorbance of the samples at 660 nm.
Flavonoid Cyanidin reaction Absorbance values of these solutions were recorded on an
4
(-pyrol) (turn to red with Mg + HCl) ultraviolet and visible (UV–Vis) spectrometer (Thermo
Scientific Evolution 300). Diclofenac is a positive control.
5 Organic acids Bubble with Na2CO3
Anti-inflammatory activity (AIA) is calculated as the
following equation:
6 Polyphenol Form blue complex with FeCl3 1%
AIA (%) = [(Abstest - Abscontrol) / (Abstest)] x 100
7 Reducing agent Fehling reagent (red precipitate)
where AIA is the percent of inhibition (%), Abstest is the
absorbance value of the samples, Absblank is the absorbance
8 Saponin Foam when shaken in the water value of sample containing distilled water instead of extracts.
The tests were repeated 1 more time and the average of the 2
9 Tanin Precipitation with gelatin solution trials was taken. The IC50 values were determined
graphically by linear regression.
10 Triterpenoid Lierbermann-Buchard reaction
III. RESULT AND DISCUSSION
F. Agar well diffusion assay A. Chemical composition of S. officinalis essential oil
The antibacterial activities of the SoEO and SoWE were 24 constituents have been identified from the essential oil
determined by Agar paper diffusion assay techniques. In this extracted from S. officinalis. The volatile compounds and their
method, 100 μL of standardized inoculum of each test relative percentages are shown in Table 2. Camphor (32.63%),
bacterium was spread onto sterile Muller-Hinton Agar plates. α-thujone (17.11%), viridiflorol (8.64%), β-thujone (6.88%),
After that, 6 mm diameter papers filled with essential oil, or epimanool (6.57%), eucalyptol (6.46%), … are the main
extract were put onto the surface of Muller-Hinton Agar compounds.
plates. The plates were kept at room temperature for 1 h to
allow proper diffusion of the samples into the agar and then The composition of SoEO of sage in Lamdong, Vietnam
incubated at 37oC for 24 h. The samples having antimicrobial slightly differs from that reported by studies on this species in
activity inhibit the microbial growth and the clear zones were other countries. In Italia (South Europe), camphor (16.16–
formed. The zone of inhibition was measured in millimeters. 18.92%), 1,8-cineole (8.80–9.86%), β-pinene (3.08–9.14%),
The standard antibiotic paper (ceftriaxone) is positive control camphene (6.27–8.08%), and α-thujone (1.17–9.26%) are
and sterile water-filled paper is negative control [2]. identified as the most abundant constituents [5] Another
research about sage in Brazil (South America) showed that
G. DPPH free radical scavenging activity the major constituents of the oil were α-thujone (40.90%),
The ability to scavenge the 2,2- camphor (26.12%), α-pinene (5.85%) and β-thujone (5.62%)
diphenyl-1-picrylhydrazyl (DPPH) radical was measured by [6]. In Tunisia (North Africa), 49 components were found in
the method of Sahin et al. (2004) with modifications [3]. 1 which camphor (25.14%), α-thujone (18.83%), 1,8-cineole
mL of different concentrations of samples prepared in (14.14%), viridiflorol (7.98%), β-thujone (4.46%) and
methanol was added to 1 mL of a methanolic DPPH solution β-caryophyllene (3.30%) were the major components [1]. It is
(0.5 mM) and 2 mL of methanol. The measurement of clear that the different geography makes differences in the
absorbance was made against a blank prepared for each chemical components of plants.
concentration at 517 nm after 30 min of incubation in the
B. Organic compounds in water extract
dark at room temperature. Absorbance values of these
solutions were recorded on an ultraviolet and visible (UV– Based on chemical reactions, triterpenoids, alkaloids,
Vis) spectrometer (Thermo Scientific Evolution 300). Acid flavonoids, tannins, saponins, and reducing agents were
ascorbic is a positive control. The results were expressed as identified in the water extract of Salvia officinalis. Those
percent inhibition (AOI %) using the following equation: organic compounds bring biological activities for SoWE.
AOI % = [(Absblank - Abstest) / Absblank)] × 100 C. Antimicrobial activity

The results of antimicrobial and antifungal rings were
where AOI is the percent of inhibition (%), Abstest is the
shown in Figure 1 and Table 3.
absorbance value of the samples, Absblank is the absorbance
138
Table 2. Chemical composition of SoEO
Molecule Molecule Yield

No. Retention time Name Yield (%) No. Retention time Name
weigh weigh (%)
1 7.16 α-Pinene 136 1.38 13 19.56 Isothujol 154 0.43
2 7.73 Camphene 136 2.92 14 20.11 Terpinene-4-ol 154 0.44
3 8.86 Sabinene 136 0.14 15 20.80 α-Terpineol 154 0.24
4 8.95 β-Pinene 136 1.55 16 21.02 Myrtenol 152 0.21
5 9.80 β-Myrcene 136 0.26 17 24.61 Bornyl acetate 196 2.47
6 11.61 D-Limonene 136 1.06 18 28.90 Caryophyllene 204 2.03
7 11.72 1,8-Cineole 154 6.46 19 29.89 Humulene 204 3.74

(Eucalyptol)
8 16.09 α-Thujone 152 17.11 20 33.12 Caryophyllene oxide 220 0.28
9 16.76 β-Thujone 152 6.88 21 33.30 Viridiflorol 222 8.64
10 18.02 Thujanol 154 0.23 22 33.64 Humulene oxide II 220 0.46
11 18.30 Camphor 152 32.63 23 39.01 Epimanool 290 6.57
12 19.48 Borneol 154 3.67 24 39.24 Unknown - 0.20
A B A B A B
D C D
D
C C
E. coli Shigella sp S. epidermidis
B A B B
A A
D D D
C C
C
S. aureus P. aeruginosa C. albicans
Figure 1. Determination of antibacterial and antifungal activities

A. SoWE B. SoEO C. Distilled water D. Ceftriaxone
139
Table 3. Diameter of the sterile ring Essential oils did not show anti-inflammatory ability
because the linear equation was not reliable, the cause may be
Diameter of the sterile ring (mm)
Strains that the essential oil itself has no anti-inflammatory effect or
SoEO SoWE the essential oils cannot be dissolved in solvents as well as
Ceftriaxone
(100%) (100 mg/mL) evaporated by heat.
E. coli 12 - 31
IV. CONCLUSION
Shigella sp. 12 - 37
Water extract of S. officinalis owns triterpenoids,
S. epidermidis 11 - 21
alkaloids, flavonoids, tannins, saponins, and reducing agents
S. aureus 16 11 20 and has antimicrobial, antioxidant, and anti-inflammatory
P. aeruginosa 11 - 30 activities. The essential oil of S. officinalis has major
C. albican 11 11 27 components including camphor, α-thujone, viridiflorol,
β-thujone, epimanool, and eucalyptol. Essential oil shows
SoEO has stronger antibacterial and fungal activity than strong antibacterial activity but shows no antioxidant (DPPH
SoWE at the tested concentrations. SoEO is active against all test) and anti-inflammatory (egg albumin denaturation test)
tested bacteria and fungi, however, the SoWE is only effects.
resistant to S. aureus and C. albicans. The antibacterial
diameter of the SoEO is also wider than SoWE, of which the REFERENCES
strongest is on the strain S. aureus (16 mm). [1] M. R. B. Khedher, S. B. Khedher, I. Chaieb, S. Tounsi, and M.
Hammami, “Chemical composition and biological activities of
There are many articles that announced the antibacterial Salvia officinalis essential oil from Tunisia”, EXCLI Journal, 16,
activities of SoEO from S. officinalis [1], [5] all of them 2017, pp. 160-173.
showed that SoEO has strong activity against many bacterial [2] B. Mahesh, and S. Satish, "Antimicrobial Activity of Some
strains. But on the other hand, studies of water extracts from Important Medicinal Plant Against Plant and Human Pathogens",
S. officinalis are fewer. Research by Stanojević et al. (2010) World Journal of Agricultural Sciences, 4, 2008, pp. 839-843.
showed that water extract has the ability to fight against [3] Sahin et al., “Biological activities of the essential oils and methanol
many spoilage bacteria [7]. extract of Origanum vulgare in the Eastern Anatolia region of
Turkey”, Food Control, 15(7), 2004, pp.549-557
D. Antioxidant activity [4] P. Dey, P. Chatterjee, S. Chandra, and S. Bhattacharya,
"Comparative in-vitro evaluation of anti-inflammatory effects of
Table 4. Results of anti-oxidant activity aerial parts and roots from Mikania scandens", J Adv Pharm Educ
Res. 1, 2011, pp. 271-277.
Linear equation R2 IC50 (µg/mL)
[5] R. Tundis, M. Leporini, M. Bonesi, S. Rovito, and N. G. Passalacqua,
“Salvia officinalis L. from Italy: A comparative chemical and
SoWE y = 647.45x + 9.374 0.9971 62.74
biological study of its essential oil in the Mediterranean
context”. Molecules, 25(24), 2020, pp. 5826.
SoEO y = 6.0906x + 9.477 0.9948 6653.36
[6] A. Porte, R. L. O. Godoy, and L. H. Maia-Porte, “Chemical
Acid ascorbic y = 8.8698x + 2.9915 0.9938 5.30 composition of sage (Salvia officinalis L.) essential oil from the Rio
de Janeiro State (Brazil)”, Revista Brasileira de Plantas
Medicinais, 15, 2013, pp. 438-441.
The results in Table 4 show that SoWE has an
antioxidant activity with IC50 that is about 11.8 times higher [7] D. Stanojević, L. Čomić, O. Stefanović, and S. Solujić-Sukdolak,
“In-vitro synergistic antibacterial activity of Salvia officinalis L. and
than acid ascorbic. IC50 value of SoEO is up to 1330 times some preservatives”, Archives of biological sciences, 62(1), 2010, pp.
higher than acid ascorbic, showing that essential oil has no 167-174.
antioxidant effect. Some researchers showed that water [8] I. Alayan, M. Mroueh, and C. Daher, “Salvia officinalis water extract:
extract has antioxidant activity such as Alayan et al. (2010) a potential hypolipidemic, hypoglycemic and anti-ulcerogenic
[8], Ollanketo et al. (2002) [9], and Fattah et al. (2013) [10]. remedy”, Planta Medica, 76(12), 2010, pp. 625.
[9] M. Ollanketo, A. Peltoketo, K. Hartonen, R. Hiltunen, and M. L.
E. Anti-inflammation Riekkola, “Extraction of sage (Salvia officinalis L.) by pressurized hot
The determination of anti-inflammatory activity is shown in water and conventional methods: antioxidant activity of the
Table 5. extracts”, European Food Research and Technology, 215(2), 2002,
Table 5. Results of anti-inflammatory activity pp. 158-163.
[10] S. M. Abd El Fattah, N. M. El-Fatih, and T. M. Fahim, “Effect of
Linear equation R2 IC50 (µg/mL) Using Aqueous Extract of Salvia officinalis L. Leaves on Some
Antioxidants Status in Irradiated Rats”, Egyptian Journal of
SoWE y = 961.16x – 5.432 0.9973 0.058 Radiation Sciences and Applications, 26(1-2), 2013, pp. 27-39.
[11] S. F. Vieira, H. Ferreira, and N. M. Neves, “Antioxidant and
SoWE y = 1.8406 + 151.89 0.2255 -55.36 anti-inflammatory activities of cytocompatible Salvia officinalis
extracts: A comparison between traditional and Soxhlet
Diclofenac y = 1365.3x – 1.5113 0.9965 0.038 extraction”. Antioxidants, 9(11), 2020, pp. 1157.
[12] E. Y. Qnais, M. Abu-Dieyeh, F. A. Abdulla, and S. S. Abdalla, “The
SoWE showed its ability to protect the egg albumin against antinociceptive and anti-inflammatory effects of Salvia officinalis leaf
temperature denaturation. IC50 of SoWE and diclofenac were aqueous and butanol extracts”. Pharmaceutical Biology, 48(10),
2010, pp. 1149-1156.
0.058 µg/mL and 0.038 µg/mL, respectively. The
anti-inflammatory effect of SoWE is about 1.5 times lower than
that of the control sample diclofenac. The previous researches
of other authors by other different tested models also show that
the water extract from S. officinalis has anti-inflammatory
activity [11], [12]
140
Dynamic Analysis of Plates Under Moving

Discontinuous Impulsive Load on Viscoelastic
Foundation
Trong Phuoc Nguyen
The Tuan Nguyen
Faculty of Civil Engineering
phuoc.nguyen@ou.edu.vn
tuannt.21ac@ou.edu.vn
Abstract: This paper presents the dynamic analysis of discontinuous impulsive load, taking into account the loss of
plates on viscoelastic foundation subjected to moving contact between the load and the plate, and the collision
discontinuous impulsive load. Numerical investigation using when return.
the finite element method is applied to analyze the dynamic
II. FORMULATION
equation of the plate under the impulse of the load. The
obtained numerical results describe typically the actual From the classic plate theory, the partial differential
wheel load when traveling not always in contact with the equation for the deflection of a plate on a viscoelastic
road surface. foundation subjected to a moving load as
Keywords: dynamic analysis, plate, viscoelastic 𝜕4𝑤 𝜕4𝑤 𝜕4𝑤

𝐷( + 2 + ) + 𝑘𝑓 𝑤+
foundation, moving discontinuous impulsive load, finite 𝜕𝑥 4 𝜕𝑥 2 𝜕𝑦 2 𝜕𝑦 4
(1)
element method 𝜕𝑤 𝜕2𝑤
(𝑐𝑝 + 𝑐𝑓 ) + 𝜌ℎ 2 = 𝐹(𝑥, 𝑦, 𝑡)
I. INTRODUCTION 𝜕𝑡 𝜕𝑡
where w is the plate deflection in the direction of the
The model of the plate on the foundation is a basic normal which is satisfied the above (1); D is the plate
problem, which has been widely studied and applied, stiffness; E is Young’s modulus; μ is Poisson’s ratio; kf is the
especially in transportation construction. There are many foundation stiffness; cp and cf are damping constants of the
studies for the load acting on this model, from the simplest plate and the foundation; ρ and h are the density and the
static problems with concentrated and distributed loads to the thickness of the plate; t is the time variable; x and y are
dynamic problems such as harmonic, impulsive loads at rectangular Cartesian coordinates in the plane of the plate;
different fixed positions. In order to model the load more F(x,y,t) is the loads and forces with (x,y) coordinates with
closely to the behavior of vehicles acting on the road, the respect to the time variable t acting on the plate resting on
moving load with constant value was studied, from moving the viscoelastic foundation.
load problems with constant velocity [1]-[4] to variable
velocity during movement [5]-[6].
However, during the movement of vehicles, the load
acting on the road surface always changes in magnitude.
Many studies have modeled vehicle loads into harmonic
ones, typically a dynamic foundation model for the analysis Figure 1. The model of the plate on the viscoelastic
of plates on foundation to a moving oscillator [7], critical foundation
velocities of a beam on a nonlinear elastic foundation under
harmonic moving load [8], dynamics of a beam on a bilinear The moving discontinuous impulsive load is shown in
elastic foundation under harmonic moving load [9], analysis Fig. 2. If Pz is the amplitude and 𝜔 is the frequency of the
of a uniform Euler–Bernoulli beam on Winkler foundation load P, the moving discontinuous impulsive load P is
subjected to harmonic moving load [10], analytical solution described as follows
for a finite Euler–Bernoulli beam with single discontinuity in 𝐼𝑓𝑃𝑧 𝑠𝑖𝑛(𝜛𝑡) < 0then𝑃 = 𝑃𝑧 𝑠𝑖𝑛(𝜛𝑡)
the section under arbitrary dynamic loads [11], dynamic (2)
analysis of railway track on variable foundation under 𝐸𝑙𝑠𝑒𝑃 = 0
harmonic moving load [12]. In addition, during high-speed
movement, the load is not always in contact with the road
surface, and when in contact again, the collision between the
wheel and the road surface also causes a significant
interaction force. There are currently no studies that clearly
describe this process.
This paper is going to fully describe the process of
Figure 2. The moving discontinuous impulsive load
moving load with variable magnitude by a moving

The load P described is close to the load that appeared by

the vehicle moving on the road, and it is modeled [𝑀] = ∫[𝑁]𝑇 𝜌ℎ[𝑁]𝑑𝑥𝑑𝑦 (9)
approximately as a sinusoid. 𝑒
When the load is not in contact with the road surface for [𝐾𝑝 ] = ∫[𝐵]𝑇 [𝐷][𝐵]𝑑𝑥𝑑𝑦 (10)
some reason, and then returns, it is described as the process 𝑒
of turning off the load and then impacting with a completely The plate damping matrix is used Rayleigh damping as
inelastic collision between the load (the wheel) and the road
surface. The forces of wheel friction, braking, and the wind [𝐶𝑝 ] = 𝛼[𝑀] + 𝛽([𝐾𝑝 ] + [𝐾𝑓 ]) (11)
are considered to be completely perpendicular to the 0z axis, with 𝛼 and 𝛽 are determined by the first two modes
the force f due to the collision is calculated based on the through modal analysis. And the load vector is
balance of momentum. Considering the interactions between
the load, the elements of the plate, and the foundation, let
mv, vv be the mass and the pre-collision velocity of the load [𝐹] = ∫[𝑁]𝑇 𝐹(𝑥, 𝑦, 𝑡)𝑑𝑥𝑑𝑦 (12)
𝑒
in the z-direction, mp, vp be the mass and the pre-collision
velocity of the element of the plate in the z-direction, mf, vf From the above analysis steps, equation (6) can be solved
be the mass and the pre-collision velocity of the element of by Newmark’s method. The problem can be considered with
the foundation in the z-direction, v is the post-collision different boundary conditions.
velocity of the elements of the progress of the completely III. NUMERICAL INVESTIGATION
inelastic collision, Δt is the time of the acting momentum
force f. Assuming the plate and the foundation are not In order to implement numerical investigation and
separated during oscillation (vp = vf), expressed as analysis, a computer program written in Python
programming language [13] with open source libraries
𝑚𝑣 ⃗⃗⃗
𝑣𝑣 + 𝑚𝑝 ⃗⃗⃗⃗
𝑣𝑝 + 𝑚𝑓 ⃗⃗⃗⃗
𝑣𝑓 = (𝑚𝑣 + 𝑚𝑝 + 𝑚𝑓 )𝑣 = 𝑓𝛥𝑡 (3) including numpy [14], matplotlib [15], and openseespy [16]
is used.
(𝑚𝑣 + 𝑚𝑝 + 𝑚𝑓 )𝑣
𝑓= (4) A. The verification of the written computer program
𝛥𝑡
The numerical example from the program was compared
with the study of M.-H. Huang and D. P. Thambiratnam [1]
in Fig. 4 and Fig. 5
Figure 3. The completely inelastic collision

Then, the force F acts with respect to time variable t as
𝐹 = 𝑃⃗ + 𝑓 (5)
In order to solve (1), the finite element method is
adopted. Based on the results of the study on the dynamic
response of plates on elastic foundation to moving loads [1],
the dynamic equation of the plate-foundation system is
[𝑀]𝑑¨ + ([𝐶𝑝 ] + [𝐶𝑓 ])𝑑˙ + ([𝐾𝑝 ] + [𝐾𝑓 ])𝑑 = 𝐹 (6) Figure 4. The results of M.-H. Huang [1]
Where [M] is the mass matrix; ([Cp]+[Cf]) are the
damping matrices of the plate and the foundation; [Kp], [Kf]
are the stiffness matrices of the plate and the foundation; F is
the load vector; d is the complete displacement vector. From
the theory of the finite element method for the plate bending
element, let [N] and [B] be the shape function matrix and the
gradient matrix, respectively, e is the plate bending element
under consideration of minimum potential energy. The
stiffness and damping matrices of the foundation are
obtained
[𝐾𝑓 ] = ∫[𝑁]𝑇 𝑘𝑓 (𝑥)[𝑁]𝑑𝑥𝑑𝑦 (7)

𝑒
[𝐶𝑓 ] = ∫[𝑁]𝑇 𝑐𝑓 [𝑁]𝑑𝑥𝑑𝑦 (8)

𝑒
The mass matrix [M], the plate stiffness matrix [Kp], the Figure 5. The written computer program
plate damping matrix [Cp] are taken according to the above
study [1], specifically the mass matrix and the plate stiffness This is shown that the computer program using the finite
matrix are given by element method and dynamic analysis step by step is
reliable.
142
B. Numerical analysis
A simply supported LxB=100(m)x10(m) thin plate with
thickness thk=0.3(m) resting on the viscoelastic foundation.
The data for the plate and the load amplitude are given by:
E=3.1x1010(N/m2), μ=0.25, ρ=2,440(kg/m3), damping ratio
ξ=5(%), Kf=107(N/m3), Cf=100(Ns/m3), Pz=1,000(N).
The load moves along the center line and parallel to the
long edge of the plate. It moves into the plate with an initial
velocity of v=20(m/s) and this velocity remains constant.
The numerical analysis with the following cases is given:
• Case 1: the displacement spectrum by the load with
frequencies of 8.5(Hz), 13.8(Hz), 1(Hz), 17(Hz)
respectively, and compare with constant-moving-
load Pz;
Figure 7. The momentum force
• Case 2: the contour of the displacement spectral by
the load with a set of frequency range from 1(Hz) to The momentum force in this case reached values of 97(N)
20(Hz), step 0.2(Hz), and compare with constant- and -75(N), equivalent to ~10(%) of the amplitude of load P.
moving-load Pz;
The plate is divided with 100 elements in the longitudinal
direction, and 10 elements in the horizontal direction
(N=1,111 nodes) with a ratio of two-element sides of 1:1 to
achieve the best results. Then the stiffness of each spring is
k=KxLxB/N=9,000,900.1(N/m), and the damping coefficient
is c=CxLxB/N=90.01(Ns/m) . The dynamic responses of the
plate with various load case 1 and case 2 are as follows.
1) Case 1
From the recorded statistics [17] with the speed range
from 50(km/h) to 110(km/h) (from 13.89(m/s) to 30.56(m/s),
the analyzed load velocity is 20(m/s)=72(km/h)), the
frequency of the wheel load acting on the road surface is
from 8.5(Hz) to 18(Hz), popular is 13.8(Hz).
a) In the case of 8.5(Hz) load frequency, the load P Figure 8. The displacement spectrum
and the momentum force f between the wheel and the road
surface: The displacement spectrum of the plate caused by the
8.5(Hz) load frequency (maximum value is |-3.574.10-6|(m))
is smaller than the displacement spectrum caused by the
constant-moving-load Pz (maximum value is |-5.454.10-
6
|(m)). This occurs when the dynamic magnification factor
dissipates the load, because the frequency of the load and the
natural frequency of the element are relatively different,
causing displacement to decrease by about 35(%) compared
to the Pz.
It should also be noted that when the constant-moving-
load Pz enters the plate at an initial speed of v = 20(m/s), that
is the load that has just transitioned from the extremely high
stiffness foundation to the finite stiffness foundation (from
the support to the plate), it will provide a significant initial
velocity to the plate element at time t = 0(s), resulting in a
relatively large oscillation at this position in comparison to
Figure 6. The load with 8.5(Hz) frequency other positions, including the other end of the plate.
b) In the case of the popular frequency that usually
appeared during the recorded statistic, 13.8(Hz), the load P
and the momentum force f between the wheel and the road
surface:
143
speed range. The load P and the momentum force f between

the wheel and the road surface:
Figure 9. The load with 13.8(Hz) frequency
Figure 12. The load with 1(Hz) frequency

The momentum force in this case reached values of
263(N) and -263(N), equivalent to ~25(%) of the amplitude
of load P.
The momentum force in this case reached small values,

4(N) and -3(N), equivalent to ~5(%) of the amplitude of load
P.
Figure 11. The displacement spectrum
The displacement spectrum of the plate caused by the

13.8(Hz) load frequency (maximum value is |-5.697.10-6|(m))
is larger than the displacement spectrum caused by the
constant-moving-load Pz. This occurs when the dynamic
magnification factor gains the load plus the momentum force
is relatively large, equivalent to 25(%) of the Pz, causing a
displacement increase of about 5(%) compared to the Pz. Figure 14. The displacement spectrum
c) In the case of low frequency, 1(Hz), this is also a
suitable frequency for driver or passengers [18], but this It is clear that at low frequency the load causes plate
frequency does not appear in the speed range from 50(km/h) displacement equivalent to the constant-moving-load Pz. In
to 110(km/h) in reality. This case occurs only in the lower this case, the dynamic magnification factor is close to 1.
144
d) In the frequency range from 1(Hz) to 20(Hz), step frequency and the natural frequency of the system approach
0.2(Hz), the load applied at a frequency of 17(Hz) causes approximately equal values, plus a momentum force
the largest displacement of the plate. The load P and the equivalent to 25(%) of the Pz, causing a displacement
momentum force f between the wheel and the road surface: increase of about 30% compared to the Pz.
2) Case 2
Considering all the frequencies of the load from 1(Hz) to
20(Hz), step 0.2(Hz), then plotting the contour of the
displacement spectrum of these frequencies. It will give a
relatively general view of the dynamics of the plate resting
on the viscoelastic foundation under moving discontinuous
impulsive load:
Displacement (m)
Figure 15. The load with 17(Hz) frequency
Time (s)
Figure 18. The contour of the displacement spectrum

The moving load model with a constant magnitude and
always in contact with the road surface will result in
displacements about 30(%) smaller than the current analysis.
IV. CONCLUSION
The influence of the vehicle's oscillation and the loss of
contact between the load and the road surface then return
Figure 16. The momentum force which is modeled by the moving discontinuous impulsive
load, causing a relatively large additional load, and creating
The momentum force in this case reached values of plate displacement greater than 30(%) in comparison with
236(N) and -147(N), equivalent to ~25(%) of the amplitude the always-in-contact-moving load.
of load P. The design process using the static load multiplied by the
impact coefficient of 1.25, or the study process of the
moving load assuming a condition that the load is always in
contact still lacks the collision force between the wheel and
the road surface. This momentum force is equivalent to about
25% of the amplitude of the moving load.
Not all discontinuous moving loads with collision cause
larger displacements than in the usual problem. However, the
stochastic of the vehicle moving, in reality, is always likely
to cause cases like the ones presented in this article. The
problem has much room to extend as consider the separation
between the plate and the foundation; vehicle load can be
modeled as triangular, rectangular impulsive loads, or actual
recorded data for better results.
REFERENCES
Figure 17. The displacement spectrum [1] M.-H. Huang and D. P. Thambiratnam, “Dynamic Response
of Plates on Elastic Foundation to Moving Loads,” J. Eng.
The displacement spectrum of the plate caused by the 17(Hz) Mech., vol. 128, no. 9, pp. 1016–1022, Sep. 2002, doi:
10.1061/(ASCE)0733-9399(2002)128:9(1016).
load frequency (maximum value is |-7.141.10-6|(m)) is larger
[2] R. K. Praharaj and N. Datta, “Dynamic response of plates
than the displacement spectrum caused by the constant- resting on a fractional viscoelastic foundation and subjected to
moving-load Pz. This situation occurs when the load a moving load,” Mechanics Based Design of Structures and
145
Machines, pp. 1–16, Jun. 2020, doi: [11] H. Yu, Y. Yang, and Y. Yuan, “Analytical solution for a finite
10.1080/15397734.2020.1776621. Euler–Bernoulli beam with single discontinuity in section
[3] N. D. Beskou and D. D. Theodorakopoulos, “Dynamic effects under arbitrary dynamic loads,” Applied Mathematical
of moving loads on road pavements: A review,” Soil Modelling, vol. 60, pp. 571–580, Aug. 2018, doi:
Dynamics and Earthquake Engineering, vol. 31, no. 4, pp. 10.1016/j.apm.2018.03.046.
547–567, Apr. 2011, doi: 10.1016/j.soildyn.2010.11.002. [12] H. Phadke and O. Jaiswal, “Dynamic analysis of railway track
[4] N. T. Phuoc and P. D. Trung, “The influence of mass of two- on variable foundation under harmonic moving load,”
parameter elastic foundation on dynamic responses of beams Proceedings of the Institution of Mechanical Engineers, Part
subjected to a moving mass,” KSCE J. Civ. Eng., vol. 20, no. F: Journal of Rail and Rapid Transit, vol. 236, no. 3, pp.
7, pp. 2842–2848, 2016, doi: 10.1007/s12205-016-0167-4. 302–316, Mar. 2022, doi: 10.1177/09544097211020838.
[5] M.-H. Huang and D. P. Thambiratnam, “Deflection response [13] G. van Rossum and F. L. Drake, The Python language
of plate on Winkler foundation to moving accelerated loads,” reference, Release 3.0.1 [Repr.]. Hampton, NH: Python
Engineering Structures, vol. 23, no. 9, pp. 1134–1141, Sep. Software Foundation, 2010.
2001, doi: 10.1016/S0141-0296(01)00004-9. [14] C. R. Harris et al., “Array programming with NumPy,”
[6] N. D. Beskou and E. V. Muho, “Dynamic response of a finite Nature, vol. 585, no. 7825, pp. 357–362, Sep. 2020, doi:
beam resting on a Winkler foundation to a load moving on its 10.1038/s41586-020-2649-2.
surface with variable speed,” Soil Dynamics and Earthquake [15] J. D. Hunter, “Matplotlib: A 2D Graphics Environment,”
Engineering, vol. 109, pp. 222–226, Jun. 2018, doi: Comput. Sci. Eng., vol. 9, no. 3, pp. 90–95, 2007, doi:
10.1016/j.soildyn.2018.02.033. 10.1109/MCSE.2007.55.
[7] P. T. Nguyen, T. D. Pham, and H. P. Hoang, “A dynamic [16] F. McKenna, M. H. Scott, and G. L. Fenves, “Nonlinear
foundation model for the analysis of plates on foundation to a Finite-Element Analysis Software Architecture Using Object
moving oscillator,” Struct. Eng. Mech., vol. 59, no. 6, pp. Composition,” J. Comput. Civ. Eng., vol. 24, no. 1, pp. 95–
1019–1035, 2016, doi: 10.12989/sem.2016.59.6.1019. 107, Jan. 2010, doi: 10.1061/(ASCE)CP.1943-5487.0000002.
[8] D. Froio, E. Rizzi, F. M. F. Simões, and A. Pinto da Costa, [17] Prażnowski, K., Mamala, “Classification of the road surface
“Critical velocities of a beam on nonlinear elastic foundation condition on the basis of vibrations of the sprung mass in a
under harmonic moving load,” Procedia Engineering, vol. passenger car,” IOP Conf. Ser.: Mater. Sci. Eng., vol. 148, p.
199, pp. 2585–2590, 2017, doi: 012022, Sep. 2016, doi: 10.1088/1757-899X/148/1/012022.
10.1016/j.proeng.2017.09.348. [18] T. D. Gillespie, “Fundamentals of Vehicle Dynamics,” SAE
[9] D. Froio, E. Rizzi, F. M. F. Simões, and A. Pinto Da Costa, International, Warrendale, PA, SAE Technical Paper R-114,
“Dynamics of a beam on a bilinear elastic foundation under Feb. 1992. Accessed: Mar. 25, 2022. [Online]. Available:
harmonic moving load,” Acta Mech, vol. 229, no. 10, pp. https://www.sae.org/publications/technical-papers/content/R-
4141–4165, Oct. 2018, doi: 10.1007/s00707-018-2213-4. 114/
[10] M. Jiya and A. Shaba, “Analysis of A Uniform Bernoulli –
Euler Beam on Winkler Foundation Subjected to Harmonic
moving load,” Journal of Applied Sciences and
Environmental Management, vol. 22, no. 3, p. 368, Apr. 2018,
doi: 10.4314/jasem.v22i3.13.
146
Isolation and Quantitative Determination of

Geniposide from Gardenia jasminoides Ellis. Using
High-Performance Liquid Chromatography
Vo Thi Nga Truong Thi Khanh Van Bui Trung Huu*
Faculty of Chemical and Food Faculty of Chemical and Food Faculty of Chemical and Food
Technology Technology Technology
Ho Chi Minh City Univerisity of Ho Chi Minh City Univerisity of Ho Chi Minh City Univerisity of
ngavt@hcmute.edu.vn 17128085@student.hcmute.edu.vn bhtrung@hcmute.edu.vn
*Corresponding author
Abstract: Geniposide, a representative bio-active iridoid studies have confirmed that G. jasminoides possesses various
glycoside found in Gardenia jasminoides Ellis. fructus has activities in vitro and in vivo, such as anti-inflammatory,
been proved by in vitro assays to own diverse neuroprotective activity, anti-depressants, and positive
pharmacological properties and is used as a marker effects on the cardiovascular and digestive systems.[1]
compound in the quality control of gardenia fructus. The Geniposide is a well-known iridoid glycoside that is a
growing therapeutical evidence has led to the increasing main bioactive component of gardenia fructus (3.3–8.56 wt%
need of isolating geniposide as well as its monitoring from in dry weight). It is also found in nearly 40 herbs with lower
potential plants/vegetables. Here, we report a procedure for geniposide contents, such as Rehmannia glutinosa Libosch.
the isolation of geniposide from gardenia fructus using some roots (0.2035–0.4381%), Scrophularia ningpoensis Hemsl.
basic and conventional chemicals and means in the roots (0.0699–0.1135%), and Eucommia ulmoides Oliv.
laboratory. The preparative separation and purification of barks (0.0173–0.5811%) [2]. Recently, pharmacological
geniposide were performed through fat removal, silica gel research has demonstrated geniposide to be responsible for
chromatographic column, gel filtration chromatography, and the traditional medical properties due to its beneficial effects
crystallization process with a yield of 5.53%. The structure on anti-diabetes, anti-inflammatory, antiarthritis,
and purity of geniposide were evaluated using 1H-NMR, antithrombotic, antiangiogenic, antidepressant,
13
C-NMR spectra with a purity of 95.3%. The resultant neuroprotective, hepatoprotective, cardioprotective,
geniposide was used as a standard to build an analytical immunomodulatory, and antitumor activities [3,4].
procedure for determining geniposide from plant extracts Therefore, geniposide could be proposed as an effective
using a high performance liquid chromatography (HPLC) component for the prevention and treatment of diabetes and
its complications, Alzheimer's disease, Parkinson's disease,
method. The HPLC method was validated with a suitable
and liver disorders [5]. However, potential hepatotoxicity of
linear range of 50–300 g/mL (R2 = 0.9992), and RSD < geniposide has also been reported [6,7]. In the oral chronic
2% for the method specificity, precision and accuracy. The toxicity study of geniposide in rats, the results showed that it
geniposide content in the methanol extract was determined affected serum biochemistry, urinalysis, hematological
to be 35.2% corresponding to 7.6% in the gardenia fructus, parameters, relative organ weights, and pathological
with a recovery rate of 101.7±0.13%. This study provided abnormalities in liver and kidney tissues. The treatment with
an effective procedure for isolating geniposide as well as a a high dose of geniposide (100 mg/kg) for 26 weeks could
suitable quantitative method for testing the analysis of this induce obvious liver and kidney damage [8].
compound in plant samples.
The clinical application of G. jasminoides medicinal herb
Keywords: geniposide, isolation, HPLC, quantitative is mainly in the form of traditional Chinese preparations
method, gardenia fructus instead of purified geniposide[4]. Gardenia yellow dyestuff
always includes an undetermined amount of geniposide
I. INTRODUCTION extracted simultaneously from G. jasminoides[1]. The
consumption of geniposide in high contents and for a long
Gardenia jasminoides Ellis (G. jasminoides) is a shrub time poses a health risk; therefore, it is necessary to quantify
belonging to Rubiaceae, which is widely distributed in geniposide in herbal materials or to separate and purify
tropical and subtropical countries. In Vietnam, the tree grows geniposide from gardenia materials. Generally, geniposide
throughout the delta and midland provinces. Since ancient from G. jasminoides has been extracted, separated, and
times, gardenia fruit has been used as a natural dyestuff for purified by solvent extraction methods with or without
colorant food additives or folklore medicine. G. jasminoides microwave or ultrasound-assistance[3], classic column
has been traditionally characterized as antipyretic, clearing chromatography (CC), high-speed countercurrent
heat and toxic materials, resolving phlegm, and promoting chromatography [9,10], matrix solid-phase dispersion
blood circulation to remove blood stasis. Pharmacological extraction [11], aqueous two-phase extraction [12]. In this

report, we presented a simple procedure for the isolation of and 5% sulfuric acid in 96% ethanol-water solution,
geniposide from gardenia fructus using silica gel CC, gel followed by heating at 110oC.
filtration chromatography (GFC), and crystallization process;
and an analytical procedure for determining geniposide from The MeOH extract (3 g) was applied to an open CC (150
plant extracts using high-performance liquid chromatography g silica gel, diameter 4.5 cm, length 30 cm) with varying
coupled with a photodiode array detector (HPLC-PDA). CHCl3:MeOH ratio from 100:0, 99:1, 98:2, 95:5, 90:10,
80:20, 60:40 to 100% MeOH (v/v). The CC process was
II. MATERIAL AND METHOD monitored by TLC experiments with the proper solvent
system as CHCl3:MeOH (7:3, v/v) or EtOAc:IPA:H2O
A. Material (6:3:1, v/v). Nine fractions M1-M9 were collected with the
Gardenia jasminoides Ellis fructus was purchased from mass of 0.21 g, 1.01 g, 0.23 g, 0.18 g, 0.15 g, 0.11 g, 0.37 g,
TanPhat Co Ltd. and authenticated. A voucher specimen (No 0.14 g, 0.19 g, respectively. Fraction M2 (1.01 g) was loaded
UTE–A003) was deposited at the Department of Chemical on the GFC column (100 g Sephadex LH-20, diameter 2.5
Technology, Ho Chi Minh City University of Technology cm, length 60 cm) eluted with the solvent system
and Education, Vietnam. CHCl3:MeOH (1:4, v/v). The subfractions having similar
TLC performances using TLC eluent of EtOAc:IPA:H2O
B. Instrument and chemicals (6:3:1, v/v) were collected separately and coded as M2.1-3.
Nuclear magnetic resonance (NMR) spectra were The subfraction M2.2 (0.93 g) was recrystallized from
recorded on a Brüker Avance III spectrometer, 600 MHz for acetone (approximate 5 mL) along with decolored by
1
H NMR and 150 MHz for 13C NMR (Brüker BioSpin, activated charcoal, then dried in the air to yield a pure
Switzerland). Chemical shifts are referenced to the residual compound, coded G (0.825 g).
solvent signal (CD3OD) at H 3.34 ppm and C 49.9 ppm.
E. Purity determination
The analytical procedure for geniposide was conducted The purity of isolated geniposide (G) was evaluated using
using an HPLC system (Jasco, Japan) consisting of a the relative (100%) qHNMR (quantitative proton nuclear
CrestPak C18T-5 reversed-phase column (4.6 mm × 250 magnetic resonance) method developed by Pauli et al.[13]
mm, 5 𝜇m), PU-2089 Plus pump, LC-NetII/ADC, CO–2065 The purity calculation, P, is conducted following the Eq. (1)
thermostatic chamber, PDA MD-2018 Plus detector.
nInt t . MWt
CC experiments were carried out on silica gel (0.040– P[%] = × 100 (1)
nInt t . MWt + ∑u1(nInt u . MWu )
0.063 mm) from HiMedia (Mumbai, India). Thin-layer
chromatography (TLC) analyses were performed on pre- where nInt is the integrals for one proton (nInt [per 1H]),
coated silica gel 60 F254 purchased from Merck (Darmstadt, MW is the molecular weight (MW), u is the number of
Germany). Sephadex LH-20 (25–100 m) was applied for impurities, and t is the target analyte.
GFC from GE Healthcare Bio-Sciences AB (Sweden).
F. HPLC procedure
Methanol (MeOH) used for HPLC analysis was of Analysis of geniposide in the prepared MeOH extract
chromatographic grade from Fisher (USA). Deionized water was performed using an HPLC instrument (Jasco, Japan),
was obtained from a Yamato WGH201 purification unit equipped with a PDA detector (MD-2018 Plus). The
(Yamato, Japan) for all solutions and dilutions. Vanillin was
chromatographic data were recorded and analyzed using
of analytical grade from Rhovanil (France) for spot-
visualizing reagent in TLC analyses. The solvents for CC, ChromNAV software. Geniposide was separated using a
TLC, and GFC were of analytical grade from Chemsol Co. reverse-phase C18 column (4.6 mm × 250 mm, 5 𝜇 m) at
Ltd. (Vietnam); or Xilong Chemical Engineering Co. Ltd. 30oC, flow rate of 1 mL/min. The mobile phase containing 1
(China). % CH3COOH in deionized water (A) and MeOH (B) (v/v)
was degassed before running. The absorbance of geniposide
C. Preparation of crude extract was collected at a maximum wavelength of 238 nm. Method
Dried gardenia fructus were removed from the peel, then validation parameters including the linear range, precision,
finely ground and sieved through a 40 mesh sieve. The dried method specificity, and recovery rate were performed
gardenia powder (100 g) was defatted using the maceration accordingly.
extraction method at ambient temperature with chloroform Geniposide stock solution of 1000 µg/mL was prepared
(CHCl3) (500 mL, triplicate). After decanting the CHCl3 in MeOH, which was used to prepare working standard
extraction solution, the defatted gardenia crude was extracted solutions by dilution with a mixed solution of MeOH:H2O
with MeOH (500 mL, triplicate) at ambient temperature with (1:1, v/v).
magnetic stirring. Removal of the solvent from the filtrate An exact mass of the MeOH extract from the gardenia
under low pressure to obtain the MeOH extract (20.1 g). powder (25 mg) was weighed and dissolved in a mixed
D. Separation and Purification solution of MeOH:H2O (1:1, v/v) assisted by an ultrasonic
bath (5 min) using a 10 mL volumetric flask. An aliquot of
For TLC method, different solvent systems
the prepared solution was then diluted 10 times using the
CHCl3:MeOH (9:1 to 5:5, v/v) or ethyl acetate (EtOAc):
same solvent solution and filtered through a 0.22 μm filter
isopropyl alcohol (IPA):H2O (6:3:1, v/v/v) were applied
before being injected into the HPLC system. The geniopside
properly as mobile eluents for TLC separation depending on
content in the MeOH extract, Y (%), and in gardenia fructus,
the polarity of the samples (MeOH extract, fractions, or pure
Z (%), can be calculated in Eq. (2) and (3).
substance). The TLC plates were visualized under white
light, next under ultraviolet (UV) light at 254 nm, and then CT × V
by dipping in the visualizing reagent containing 1% vanillin Y (%) = × 100% (2)
mC
148
Y X 100 experiments with applying an increasing CHCl3:MeOH ratio

Z (%) = × × × 100% (3) to be suitable for the polar property of the MeOH extract.
100 100 100 − a
Among of nine fractions M1-9 collected, the fraction M2
where, mC (mg) is the mass of the MeOH extract, CT is the (1.01g) was yellow powder, predicted as the major
concentration of geniposide in the prepared solution, V geniposide containing fraction. The fraction M2 could be
(mL) is the volume of the prepared solution, X (%) is the stained by crocins having higher MW (977-1138 g/mol) than
extraction yield of MeOH extract from the gardenia powder, geniposide (388.4 g/mol). Therefore, the GFC method was
and a (%) is the moisture content of the gardenia powder. further applied to separate crocins from geniposide in the M2
III. RESULTS AND DISCUSSION fraction (1.01 g), using a Sephadex LH-20 chromatographic
column and CHCl3:MeOH (1:4, v/v) eluent. The GFC
The combination of silica gel CC, GFC, and resulted in three subfractions, coded M2.1-3. The subfraction
recrystallization on the defatted dry gardenia fructus powder M2.2 obtained as a light yellow powder (0.93g) was
led to the isolation of pure geniposide. The structure and recrystallized from acetone to yield a pure compound, coded
purity of geniposide were evaluated using NMR G (0.825g) with a yield of 5.53% from dried gardenia fructus
spectroscopic method. The resultant geniposide was used as powder. The performance of the MeOH extract, fraction M2,
a standard chemical to build an analytical procedure for subfraction M2.2, and isolated compound G on TLC plates
determining geniposide from plant extracts using the HPLC- developed in the solvent system EtOAc:IPA:H2O (6:3:1,
PDA method. v/v/v), were exhibited in Fig. 2.
G. Separation and Purification
The separation and purification of geniposide from
Gardenia jasminoides Ellis fructus started with the defatting
process using chloroform. The TLC experiments were
performed for preliminary analysis of the chemical
composition of the MeOH extract, to determine the
appropriate solvent for CC separation, or to monitor the CC
progress. Investigating the MeOH extract with different
eluents led to the selection of the solvent system
EtOAc:IPA:H2O (6:3:1, v/v/v) for the highest resolution
separation (Fig. 1). Under white light, only yellow spots
appeared on the TLC plate indicating crocins of gardenia
yellow pigment. Under UV light at 254 nm, a main dark spot
at Rf = 0.56 appeared obviously along with five other blur Figure 2. TLC performance of MeOH extract, M2, M2.2 and G
developed in EtOAc:IPA:H2O (6:3:1, v/v/v) and visualized using
dark spots on the fluoresce green background. As treated
chemical stain vanilin/sulfuric acid, heating
with vanillin/sulfuric acid reagent and heating, the TLC plate
exhibited six spots, including two green ones appearing at Analysis of the 1H NMR spectrum (600 MHz, CD3OD,
the same position of yellow crocin spots under white light Table 1) revealed G to be an iridoid glucoside with the major
and four purple spots predicted to belong to iridoid
features of an acetal proton at  5.17 (1H; d; 7.2 Hz; H-1)
glycosides or their derivatives, which are the main
and a characteristic anomeric proton at  4.71 (1H; d; 7.8
components of gardenia fructus. The main spot at Rf =0.56
Hz; H-1’). A doublet at δH 7.51 (1H; d; 1.2 Hz; H-3) and a
was predicted to a representative of geniposide, the main
broad singlet at δH 5.80 (1H; brs; H-7) indicated the presence
component of gardenia fructus, therefore, it became the
of two olefinic protons. Resonances ascribed to two protons
target of isolation.
of a coupled oxymethylene group were present at 4.31 (1H;
d; 14.4 Hz; H-10a) and 4.19 (1H; dd; 14.4; 1.8 Hz; H-10b).
Besides, the presence of a methoxy was determined by a
singlet at δH 3.71 (3H; s; H-12). Furthermore, two resonances
at δH 3.86 (1H; brd; 10.8 Hz; H-6’a) and 3.64 (1H; ddd; 12.0;
4.2; 1.2 Hz; H-6’b) were assigned for an oxymethylene along
with the signals in the zone of 3.21-3.40 ppm for
oximethines of a glucopyranosyl unit.
The 13C-NMR (150 MHz, CD3OD, Table 1) presented
that compound G contained 17 carbons, including a methyl,
three methylenes, ten methines, and three quarternary carbon
Figure 1. TLC performance of MeOH extract developed in groups (including carbonyl carbon). The characteristic acetal
EtOAc:IPA:H2O (6:3:1, v/v/v). Visualization: (1) under white carbons of the aglycon and the glucose unit were found at δC
light, (2) under UV light at 254 nm, (3) chemical stain 98.3 (C-1) and δC 100.3 (C-1’), respectively. The presence of
vanilin/sulfuric acid, heating an ester was demonstrated by signals of a carbonyl at δC
169.5 (C-11) and a methoxyl at δC 36.6 (C-12) along with δH
The CC experiment was conducted using an open glass 3.71 (3H; s; H-12). The signal at δC 61.4 (C-10) confirmed
column with a column length of approximately 8 times its the presence of the hydroxymethylene. The signals at δC
internal diameter. Because of the sample complexity (Fig. 1), 153.3 (C-3) and 112.5 (C-4) together with the doublet signal
the CC experiment was packed with silica gel at a silica gel: at δH 7.51 (H-3) demonstrated the occurrence of a double
extract ratio of 50:1 (3 g of MeOH extract and 150 g of silica bond. Moreover, the signals at δC 128.3 (C-7) and 144.8 (C-
gel). Monitoring the CC process was operated by TLC 8) together with the broad singlet signal at δH 5.80 (H-7)
149
were evidence for the second double bond. The signal at δC According to Pauli, in cases the unknown impurities are
62.6 (C-6’) was characteristic for the hydroxymethylene of present, the generally accepted approach is to assume that the
glucose unit, and the four other signals, including δC 74,9 (C- unknown impurity is an isomer of the target analyte and they
2’); 78,4 (C-3’); 71,5 (C- 4’); 77,8 (C-5’) confirmed this have the same MW[13]. Therefore, the last impurity was
information. assigned for the geniposide’s isomer with molecular formula
of C17H24O10 (MW=388.14). The 1H-NMR signals of
The 1H and 13C-NMR data of compound G showed good geniposide were summarized in Table 1, while those of
compatibility with those of geniposide [10,14]. The chemical crocin-2 were obtained from a previous study [15]. The 1H
structure of geniposide was depicted in Fig. 3. NMR signals of grease were identified following Fulmer et
al. [16], showing multiplet signals in the zone of H 0.85-
0.96 for the terminal methyl groups and a broad singlet at H
1.29 for methylene groups. Then, the integral values of
geniposide and impurities at all observed 1H resonances were
carried out and applied to Eq. (1). As result, the content of
geniposide, crocin-2, grease, and the geniposide’s isomer in
the sample G was found to be 95.3%, 1.2%, 1.0%, and 2.5%
(w/w), respectively.
H. Analysis of geniposide using HPLC-PDA method
The HPLC-PDA was employed for building a
quantitative determination of geniposide in the prepared
MeOH extract of gardenia fructus. The HPLC conditions
Figure 3. Chemical structure of compound G elucidated as
were optimized using a C18 column (4.6 mm × 250 mm, 5
geniposide 𝜇m) at 30 oC and an isocratic elution of the mobile phase
A:B of 70:30 (v/v). The chromatograms show a peak of
Table 1. NMR spectroscopic data of the purified geniposide G geniposide standard and MeOH extract at the same retention
time of 6.3 min (Fig. 4 (a) and (b)). The peak of geniposide
Type of Compound G (CD3OD) had a high symmetry with a coefficient of 1.08, be suitable
Position carbon 𝛅𝐇 (J)a 𝛅𝐂 b for further validation experiments.
Aglycone Linearity of geniposide was determined using five levels
1 -O-CH-O- 5.17 (1H; d; 7.2) 98.3 of concentration between 50 and 300 μg/mL, showing the
3 -CH= 7.51 (1H; d; 1.2) 153.3 linear regression equation of peak area against with
4 >C= - 112.5 concentration and its correlation coefficient (R2) of y =
5 >CH- 3.19 (1H; ddd; 16.2; 7.8; 1.2) 47.0 28017x + 323803 and 0.9992, respectively. This range was
6a 2.82 (1H; dd; 16.2; 8.4) optimized to be suitable for the geneposide content in the
-CH2- 39.7
6b 2.09 (1H; ddt; 16.2; 8.4; 2.4) prepared extract.
7 -CH= 5.80 (1H; brs) 128.3
8 >C= - 144.8 The precision test was done with six repeated runs of the
9 >CH- 2.73 (1H; dd; 7.8; 7.2) 51.7 same concentration geniposide. The RSD was obtained to be
10a 4.31 (1H; d; 14.4) 1.08% and 1.48% for the variances of retention time and
-CH2-OH 61.4 peak area, respectively. The symmetric coefficient among the
10b 4.19 (1H; dd; 14.4; 1.8)
11 -COO- - 169.5 peaks was in the range of 1.047 and 1.095. The results
12 -O-CH3 3.71 (3H; s) 36.6 indicate the good precision of the proposed method.
Glucose
Method specificity was tested by comparing the
1’ -O-CH-O- 4.71 (1H; d; 7.8) 100.3
chromatogram of a blank sample, geniposide standard and
2’ >CH-OH 74.9
3’ >CH-OH 78.4
extract sample at the optimal conditions. Fig. 4(c) shows the
3.21-3.40 (4H)c appearance of geniposide peak at the exact time of 6.33 min
4’ >CH-OH 71.5
5’ >CH-OH 77.8 for both cases of the standard and extract chromatograms,
6’a 3.86 (1H; brd; 10.8) while no peak was observed in that of the blank. The peak of
-CH2-OH 62.6 geniposide in the extract was well separated from others.
6’b 3.64 (1H; ddd; 12.0; 4.2; 1.2)
a) b) c)
600 Mhz 150 MHz overlapping signals
A standard spiked method at 100% level of geniposide
The purity of compound G (P) was determined using the
was employed to validate the recovery test of the extract
relative (100%) quantitative 1H-NMR (qHNMR) method[13]
solution. The spiked geniposide was found to be 101.7%
following the Eq. (1). Based on the structure elucidation
with RSD of 0.13% (n = 3). In general, these results show
analysis, the signals of isolated geniposide (G) were assigned
that the proposed method can achieve high precision,
and integrated into a four-decimal place. Also, its molecular
specificity and recovery rate all confirming the requirement
formula was determined as C17H24O10 with a MW of 388.14.
test for Pharmaceuticals for Human Use (RSD <2%) [17].
The identity of impurities was established based on the
chemical composition of gardenia material and the residual The average content of geniposide in the MeOH extract
solvent for NMR experiments. Firstly, crocin-2 (C38H54O19 was 35.16±0.53%, corresponding to about 7.64±0.12% in
and MW=814.32), the main gardenia colorant that would be dry gardenia fructus. This result was consistent with the
co-eluted with geniposide in the separation process, maybe literature reported on the geniposide content in gardenia
an impurity. Secondly, grease (C10H22 and MW=142.17), a fructus[18][19].
common residual solvent, could be a foreign substance.
150
IV. CONCLUSION [3] W. Xiao, S. Li, S. Wang, and C.-T. Ho, “Chemistry and bioactivity
of Gardenia jasminoides,” Journal of Food and Drug Analysis, vol.
From Gardenia jasminoides fructus, the preparative 25, no. 1, pp. 43–61, Jan. 2017, doi: 10.1016/j.jfda.2016.11.005.
separation and purification of geniposide were performed [4] Y.-X. Zhou, R.-Q. Zhang, K. Rahman, Z.-X. Cao, H. Zhang, and C.
through fat removal, silica gel chromatographic column, gel Peng, “Diverse pharmacological activities and potential medicinal
filtration chromatography, and crystallization process with a benefits of geniposide,” Evidence-Based Complementary and
Alternative Medicine, vol. 2019, pp. 1–15, Apr. 2019, doi:
yield of 5.53%. The structure and purity of geniposide were 10.1155/2019/4925682.
evaluated using 1H-NMR and 13C-NMR spectra with a purity [5] W. Zhang et al., “The emerging possibility of the use of geniposide
of 95.3%. The HPLC-PDA was adopted and validated for in the treatment of cerebral diseases: a review,” Chin Med, vol. 16,
determining the content of geniposide in MeOH extract from no. 1, p. 86, Dec. 2021, doi: 10.1186/s13020-021-00486-3.
gardenia fructus, showing a good linear regression (R2 = [6] Y. Ding, T. Zhang, J. Tao, L. Zhang, J. Shi, and G. Ji, “Potential
hepatotoxicity of geniposide, the major iridoid glycoside in dried
0.9992), excellent recovery rate (101.7%), and high precision ripe fruits of Gardenia jasminoides (Zhi-zi),” Natural Product
and specificity. Therefore, this method is feasible and cost- Research, vol. 27, no. 10, pp. 929–933, May 2013, doi:
effective in the quality control of plant samples. 10.1080/14786419.2012.673604.
[7] J. Wei et al., “Proteomic investigation of signatures for geniposide-
induced hepatotoxicity,” J. Proteome Res., vol. 13, no. 12, pp.
5724–5733, Dec. 2014, doi: 10.1021/pr5007119.
[8] J. Tian et al., “Oral chronic toxicity study of geniposide in rats,”
Journal of Ethnopharmacology, vol. 213, pp. 166–175, Mar. 2018,
doi: 10.1016/j.jep.2017.11.008.
[9] T. Zhou, G. Fan, Z. Hong, Y. Chai, and Y. Wu, “Large-scale
isolation and purification of geniposide from the fruit of Gardenia
jasminoides Ellis by high-speed counter-current chromatography,”
Journal of Chromatography A, vol. 1100, no. 1, pp. 76–80, Dec.
2005, doi: 10.1016/j.chroma.2005.09.026.
[10] Z. Liang et al., “Isolation and purification of geniposide, crocin-1,
and geniposidic acid from the fruit of Gardenia jasminoides Ellis
by high-speed counter-current chromatography,” Separation
Science and Technology, vol. 49, no. 9, pp. 1427–1433, Jun. 2014,
doi: 10.1080/01496395.2013.879179.
[11] Y. Gao et al., “A practical and rapid method for the simultaneous
isolation, purification and quantification of geniposide from the
fruit of Gardenia jasminoides Ellis by MSPD extraction and UFLC
analysis,” Anal. Methods, vol. 5, no. 16, p. 4112, 2013, doi:
10.1039/c3ay40638e.
[12] H.-P. Zhang, J.-Y. Guo, and T. Li, “Selective separation of
geniposide and gardenia yellow from gardenia fruit by
isopropanol/salt aqueous two-phase system,” Separation Science
and Technology, vol. 51, no. 7, pp. 1124–1132, May 2016, doi:
10.1080/01496395.2015.1109660.
[13] G. F. Pauli et al., “Importance of purity evaluation and the potential
of quantitative 1H NMR as a purity assay: miniperspective,” J.
Med. Chem., vol. 57, no. 22, pp. 9220–9231, Nov. 2014, doi:
10.1021/jm500734a.
[14] Z. Li et al., “A quantitative 1H nuclear magnetic resonance
(qHNMR) method for assessing the purity of iridoids and
secoiridoids,” Fitoterapia, vol. 100, pp. 187–194, Jan. 2015, doi:
10.1016/j.fitote.2014.12.001.
[15] H. Zhang et al., “Semi-preparative isolation of crocins from saffron
Figure 4. HPLC-PDA chromatograms: (a) Geniposide standard solution (Crocus sativus L.),” Chromatographia, vol. 59, no. 11–12, Jun.
(200 µg/mL), (b) MeOH extract solution, and (c) comparison among blank, 2004, doi: 10.1365/s10337-004-0311-z.
geniposide standard and MeOH extract solutions. [16] G. R. Fulmer et al., “NMR chemical shifts of trace impurities:
common laboratory solvents, organics, and gases in deuterated
ACKNOWLEDGMENT solvents relevant to the organometallic chemist,” Organometallics,
Our thanks are presented to Ho Chi Minh City University vol. 29, no. 9, pp. 2176–2179, May 2010, doi: 10.1021/om100106e.
[17] C. Tietje and A. Brouder, Eds., “International conference on
of Technology and Education for providing the facilities harmonisation of technical requirements for registration of
necessary to complete this project. pharmaceuticals for human use,” in Handbook of Transnational
Economic Governance Regimes, Brill | Nijhoff, 2010, pp. 1041–
REFERENCES 1053. doi: 10.1163/ej.9789004163300.i-1081.897.
[1] L. Chen et al., “Gardenia jasminoides Ellis: Ethnopharmacology, [18] E. J. Lee, J. K. Hong, and W. K. Whang, “Simultaneous
phytochemistry, and pharmacological and industrial applications of determination of bioactive marker compounds from Gardeniae
an important traditional Chinese medicine,” Journal of fructus by high-performance liquid chromatography,” Arch. Pharm.
Ethnopharmacology, vol. 257, p. 112829, Jul. 2020, doi: Res., vol. 37, no. 8, pp. 992–1000, Aug. 2014, doi: 10.1007/s12272-
10.1016/j.jep.2020.112829. 013-0293-1.
[2] M. Shan et al., “A Review on the Phytochemistry, Pharmacology, [19] F. Yin et al., “Quality control of gardeniae fructus by
Pharmacokinetics and Toxicology of Geniposide, a Natural HPLC-PDA fingerprint coupled with chemometric
Product,” Molecules, vol. 22, no. 10, p. 1689, Oct. 2017, doi: methods,” Journal of Chromatographic Science, vol. 53, no.
10.3390/molecules22101689. 10, pp. 1685–1694, Nov. 2015, doi:
10.1093/chromsci/bmv072.
151
Oxygen-LPG Torch for Thermal Spraying

Ngo Thanh Binh Le Van Canh Pham Huy Dong
Southern branch Faculty of Mechanical Engineering An Binh Technological Mechanical
Vietnam-Russia Tropical Center Tran Dai Nghia University Company Limited
Ho Chi Minh, Vietnam Ho Chi Minh, Vietnam Ho Chi Minh, Vietnam
binhbauman85@gmail.com canhle195@gmail.com phunphu.anbinh@gmail.com
Abstract: The aim of this study is to design and produce kinetic energy into thermal energy, which further enhances in
an experimental high velocity oxygen fuel (HVOF) torch - creating dense coatings[7], [8].
an axial fuel gas powder injection spray torch which was In comparison to conventional spray guns, HVOF torches
built based on the patented ideas vortex technology of have the advantage of their typical extended nozzle, which is
author James A. Browning. It has the ability to work well considerably longer. The achievable gas velocity depends on
with LPG fuel gas, with a working pressure of the the type and design of the individual torch, the geometry of
combustion chamber not exceeding 5 bar. Fuel gas and the nozzle, and the type of fuel [5]. Classified by combustion
oxidized gas are injected with high vortex flow. The high chamber design, HVOF systems have two main design types,
rotational components of fuel and oxygen create a shroud which are the system with an open combustion chamber and
effect, limiting the heat transfer between the flame stream the system with a closed combustion chamber. While closed
and the torch component walls due to the outermost layer of system torches are equipped with an additional expansion
separated oxidizer in the shield barrel. Hence, this may nozzle, the open system ones have no separate combustion
reduce or partly eliminate the need for additional cooling. In chamber, and they are cooled by air or water. In the closed
addition, the microstructure of some coatings fabricated by systems, particle velocity is higher, but this also brings along
this torch has been introduced. a risk that the molten particles will stick to the extended
nozzle and may even completely clog it, compared to the
Keywords: high velocity oxygen fuel, HVOF, vortex opened systems [2]. The design presented below in this paper
technology, vortex flow, thermal spray is the first type of design – open combustion chamber
design..
I. INTRODUCTION
II. THE CHARACTERISTICS OF AN OXYGEN-LPG
The HVOF process was invented in 1958 by Praxair TORCH FOR THERMAL SPRAYING
Surface Technologies, Inc. but it did not become
commercially significant until 1983 when Kennametal Vortex technology is a new HVOF technology patented
Stellite™ introduced the HVOF Jet Kote system. James A. by James A. Browning [9]–[11]. The commercial product
Browning, an American engineer, is one of the pioneers in Diamond Jet™ Vortex ID-125 by Oerlikon Metco is based
this field, renowned for the invention of the HVOF process on a patent [9]. This thermal spray ID (inner diameter)
[1], [2]. HVOF spray torches use a combustion process to HVOF torch is capable of spraying IDs as small as 125mm,
heat the gas flow and the coating material then accelerate the well suited for use in confined spaces. According to [12], the
gas/particle two-phase flow to high velocities. The fuels can oxidizing gas is injected with a high rotational flow
be gasses (hydrogen, methane, propane, propylene, component, and hydrogen fuel gas is injected on a straight
acetylene, natural gas, etc.) or liquids (kerosene (aviation trajectory. The fuel flow is disturbed by the rotational
grade), ethanol, etc.). The gas and particle velocities in the component of the oxidizing flow in the zone of accelerated
HVOF process are commonly much higher than those combustion. This design allows for rapid mixing and
achieved using plasma spraying [3]. Metallic, composite, and combustion, reducing the burner size and complexity of the
ceramic coatings are being applied to substrates to improve torch.
wear resistance, abrasion resistance; corrosion protection, Designs based on vortex technology, developed by James
thermal, and electrical insulation. Applying coatings using A. Browning, allow the production of HVOF devices that are
HVOF spraying has been proven to be higher in density inexpensive, easy to operate and maintain. However, the use
(with less than 1-2 % porosity), good bond strength of different fuel gases requires appropriate internal designs to
(generally exceeds 60 MPa between coatings and substrates), provide the highest stability and performance. Gas LPG
and low oxide inclusion [4]–[6]. In comparison to plasma (mixture of butane and propane) is rarely used in HVOF
spray, it brings about reducing the degree of particle melting technology because the supplier's gas mixing ratio is not
and oxidation, and lower average particle temperatures. stable. The use of LPG gas requires a vaporizer to increase
However, high coating densities are still achieved through the supply pressure for the equipment to operate. In the
HVOF high particle impact velocities, which deform paper, a torch design is introduced using oxygen-LPG
particles that may not be well melted. Furthermore, some working gas. LPG gas was selected for the study because it is
particles are heated by the impact of the conversion of widely used in Vietnam, as well as cheaper than other gases,
for example, about 10 times cheaper than propane.

152
Figure 1. Cross section of the powder high velocity oxygen-LPG torch and experimental torch
Fig. 1 illustrates an oxygen-LPG high velocity flame jet oxygen pressure 10 bar, LPG pressure 6 bar, carrier gas
torch which has a body section and a barrel attached thereto. pressure 5 bar.
The barrel is covered inside a water-cooling jacket which
composes an annular water passage around the barrel. The Table 1. HVOF spraying parameters for coatings
water-cooling jacket has a water inlet and a water outlet
which help to supply and to lead cooling water out. Parameters Value
The high-velocity stream of combustion products is LPG flow rate, NLPM 50

provided by the ignition of fuel and oxidizer combined in the Oxygen flow rate, NLPM 210
stratified stream. The outermost layer of unmixed oxidizers
in the barrel works as a shield limiting the heat transfer Powder carrier gas (nitrogen), NLPM 9.4
between the flame stream and the torch component walls. At
Barrel length, mm 150-200
a result, it reduces the need for additional cooling; therefore,
most of the heat generated by the burning is maintained in Powder feed rate, g/min 40-90
the combustion products when it passes down the barrel.
Spraying distance, mm 180-220
The oxygen and LPG gas are introduced into the
passages first and the second section via tangential passages, III. COATINGS CHARACTERISTICS AND
which causes the flows of oxygen and LPG to swirl within APPLICATIONS
the sections of the passage about the central axis of the torch.
The swirling fuel is introduced into the swirling oxygen via Fig. 2-6 illustrate the microstructure of coatings sprayed
multiple angled LPG passages. The use of multiple angled by an experimental torch. There are no cracks in the bonding
LPG passages provides more even distribution of fuel. The surface area. Coatings produced by the experimental torch
high rotational components of oxygen and LPG allow for the have a characteristic microstructure of lamellae, pores, and
shroud effect, limiting the heat transfer between the flame oxide clusters. Porosity is not high. The bonding strength of
stream and the torch component walls due to the outer layer coatings was carried out by pulling off “the pin method”
of unmixed oxidizer in the vortex barrel shields and reducing [13], all coatings surpass 60 MPa. Compositions, powder
or eliminating the need for additional cooling. size of powders, microhardness, and porosity of coatings
show in Table 2.
Table 1 gives information about the main specifications
of applied coatings, and the working conditions of the torch:
153
Figure 2. Cross-sectional structure of SS420 coatings sprayed at low and high magnification
Figure 3. Cross-sectional structure of SS430 coating
Figure 4. Cross-sectional structure of SS630 coating
154
Figure 5. Cross-sectional structure of Cr3C2-NiCr coating
Figure 6. Cross-sectional structure of NiCrBSi coating
Figure 7. Roller for the paper manufacturing industry, Woka 3625 (WC/10Co4Cr) coating
155
Figure 8. The torch in use applying coating Woka 36259 to parts for the petrochemical industry
Figure 9. Coating NiCrBSi on automotive crankshaft journals (4 crankpin journals + 5 main journals restored)
Table 2. HVOF spraying parameters for coatings
Powder type Powder size Hardness HV1 Hardness HV0.3 Porosity, %

SS420 -45/+15µm 392 405 1.8
SS430 368 376 1.9
SS630 311 334 1.5
Cr3C2-NiCr 1050 1120 2.0
NiCrBSi 497 531 1.3
Fig. 7-9 show some applications of a designed torch for temperature can be controlled by the ratio of oxygen and fuel
thermal spraying. It can be seen that the flame of the torch is LPG, so they can be controlled to form higher dense coatings
obviously centralized and powder distributes uniformly in with high adhesive strength, low porosity, and oxide [14].
the flame. IV. CONCLUSION
The quality of the coating is mainly related to the Based on the principles of vortex technology of
condition of the particles before deposition, especially James A. Browning as well as changing the design of
particle velocity and temperature. The flame velocity and some details of the torch, a new oxygen-LPG torch was
156
successfully developed. The experimental torch possesses 51, vol. 5, no. 1, pp. 53–61, 1996, doi:
some typical characteristics: 10.1007/BF02647518.
- The oxygen-LPG torch has most of the advantages of [5] S. Grainger and J. Blunt, Engineering Coatings: Design
and Application, 2nd ed. Woodhead Publishing, 1998.
the HVOF.
[6] “Oerlikon Metco Provides Materials and Surface
- Coating quality is high. Solutions.” https://www.oerlikon.com/metco/en/ (accessed
- Low capacity of a cooling system. Apr. 12, 2022).
- The equipment is easy to make, and the spare parts [7] L. Thakur and H. Vasudev, Thermal Spray Coatings. CRC
are cheap. Press, 2021.
[8] P. L. Fauchais, J. V. R. Heberlein, and M. I. Boulos,
- The cost of equipment after completion and
“Thermal spray fundamentals: From powder to part”,
optimization will be many times lower than that of Therm. Spray Fundam. From Powder to Part, pp. 1–1566,
industrial systems. Jan. 2014, doi: 10.1007/978-0-387-68991-3.
ACKNOWLEDGMENT [9] James A. Browning, “Method and apparatus for
combusting fuel employing vortex stabilization”,
This study is granted by Vietnam-Russia Tropical US7628606B1, 2009.
Center, Hanoi, Vietnam, under grant number 2029/QD-
[10] James A. Browning, “Rapidly-mixing high velocity flame
TTNDVN. Its support is gratefully acknowledged by the torch and method”, US8455056B1, 2013.
authors.
[11] James A. Browning, “HVOF torch with fuel surrounding
REFERENCES oxidizer”, US8827176B2, 2014.
[1] J. R. Davis, Handbook of Thermal Spray Technology. [12] J. Gutleber and J. Tewes, “Novel HVOF Technology for
ASM International, 2004. Small ID Spraying”, in HVOF Colloquium 2018, 2018,
pp. 123–132.
[2] F.-W. Bach, A. Laarmann, T. Wenz, and C. B.
Nakhosteen, Modern Surface Technology. Wiley-VCH [13] L. K. Baldaev, Thermal Spray (in Russian). Moscow:
Verlag GmbH, 2006. Market DC, 2007.
[3] S. Joshi, “Special issue: Advances in thermal spray [14] W. Tillmann, E. Vogli, I. Baumann, G. Matthaeus, and T.
technology”, Materials, vol. 13, no. 16. 2020, doi: Ostrowski, “Influence of the HVOF gas composition on
10.3390/MA13163521. the thermal spraying of WC-Co submicron powders (-8 +
1μm) to produce superfine structured cermet coatings”, J.
[4] W. L. Oberkampf and M. Talpallikar, “Analysis of a high- Therm. Spray Technol., vol. 17, no. 5–6, pp. 924–932,
velocity oxygen-fuel (HVOF) thermal spray torch part 1: 2008, doi: 10.1007/S11666-008-9234-9.
Numerical formulation”, J. Therm. Spray Technol. 1996
157
Predicting Land Use Change in Buon Ma Thuot

City, Dak Lak Province by Integrating GIS and
Markov Chain
Nguyen Thi Ngoc Quyen Nguyen Thi Tinh Au
Faculty of Agriculture and Forestry Faculty of Chemical and Food Technology
Tay Nguyen University HCMC University of Technology and Education
Dak Lak province, Viet Nam Ho Chi Minh City, Viet Nam
ntnquyen@ttn.edu.vn tinhau@hcmute.edu.vn
Nguyen Cong Tai Anh Tran Thi Xuan Phan

Faculty of Agriculture and Forestry Faculty of Agriculture and Forestry
Tay Nguyen University Tay Nguyen University
Dak Lak province, Viet Nam Dak Lak province, Viet Nam
nctaianh@ttn.edu.vn ttxphan@ttn.edu.vn
Abstract: Evaluating and predicting land use changes is a using Landsat satellite imagery was done by Yu et al.,
critical task in different parts of the world. This study aims to (2011)[3]. Therefore, GIS is an essential tool for analyzing
predict land use change to 2030 in Buon Ma Thuot city, Dak multi-layered databases capable of handling multiple data in
Lak province, based on land use change evaluation in space for facilitating land use planning and management. Dao
previous years using GIS and Markov chain. The predicted and Le (2003), Nguyen et al., (2010), also carried out some
results show that agricultural land and forestry land will research on land use change assessment through GIS and
decrease to promote and develop non-agricultural land. The remote sensing combination in the period 2000-2010 in
achieved outcomes also reveal that the transformation matrix Vietnam[4][5]. Regarding the application of the Markov
of land use types consistently varied as Buon Ma Thuot city chain, Weng (2001) implemented land use change analysis in
the Delta of China integrating remote sensing satellite images,
is still in progress, causing land policies to change
GIS, and Markov model[6]. The study shows that the
accordingly. Overall, the Markov chain was suitable for integration of remote sensing satellite imagery and GIS is an
stable development sites. However, this study’s results effective method for the oriented analysis, speed and spatial
provide a general framework for authorities in land use pattern of land use change. Combination of these two
planning in Dak Lak province. technologies with the Markov model is beneficial in
describing and analyzing land use change processes; or a
Keywords: GIS, Markov, land use change, predict, Buon study employed the Markov chain model to identify land use
Ma Thuot. changes in the Twin Cities [7]; in another study conducted in
Nha Trang city, Khanh Hoa province and Dong Hoa District,
I. INTRODUCTION Phu Yen Province Huynh et al., (2017) or Nguyen et al.,
In the current context of the development of human (2021) indicated an integration of GIS and the Markov chains
society, the population growth, the fast industrialization - in analyzing and predicting land use changes in the
modernization at the global level and the growing demand for future[8][9]. Accordingly, many studies have suggested that
land, land use change has increased gradually. Although the automation in volatility assessment provides a greater
land use change takes place in particular areas, it has impacts advantage than us a greater benefit than the traditional method
on all aspects of the society, region, and the globe. Therefore, of volatility evaluation.
it is essential to monitor, evaluate and forecast the changes in
land use regarding effective land management. These Buon Ma Thuot covers an area of 37,718 hectares. It is not
activities provide a comprehensive and general view of the only the central area of Dak Lak province but also of the
land use situation in an area as well as the ability to assess the Central Highlands. The city has a fast-growing economy, and
production potential, and, the orientation of development can people's life improves significantly. Therefore, there has been
be addressed. causing an increase in demand for land, making it scarce and
valuable (Fig. 1). Buon Ma Thuot city is facing general
The community of scientists has recognized the enormous pressures on population and socio-economic development.
impacts of land use change on the entire globe and called for The situation of changing land use is increasingly complicated
research. After that, a series of studies on the current situation and difficult to control, the increase in demand for land use for
and land use change prediction, since then, have been various purposes, especially agricultural and non-agricultural
conducted in many countries. For example, Mas (1999) lands for development activities have created great variation
employed remote sensing technology to detect land use from unused land group to cultivation land and non-
changes in the studied area[1]; Muller and Munro (2007) used agricultural land. Therefore, the study of land use changes and
statistical method and considered it as a powerful tool to study evaluation of land use efficiency to manage and use land
land use change[2]; and the land use changes analyzed in resources effectively, economically ensuring the beneficial
Daqing City, Heilongjiang Province, China from 1997 to 2007 harmonization between economic growth and environmental

protection, social stability and food security are tasks set out d) Markov Chain application method
for managers, scientists, and all land users.
The Markov Chain was applied to determine the
possibility of changing land use patterns based on the
evolution of land use types and the land use maps to identify
the proportion of factors influencing the change (Fig. 2)[10].
Land use type at the first time (to) Land use type at the second time (ti)
(1). Forest land Y11 (1). Forest land
Y21 Y12
(2). Agricultural land (2). Agricultural land

Y13
Y14
(3). Water (3). Water
Y15
Y25
(4). Non-agricultural land Y44 (4). Non-agricultural land
Y54 Y45
(5). Barren land (5). Barren land
(Source: http://gadm.org/data.html)
Figure 1. The location of Buon Ma Thuot city Figure 2. The matrix LUC in Markov chain
(Note: Yij is the probability of change determined from overlapping land
From that fact, this study aims to analyze the current use maps at two different times)
context of land use changes in the study area over the stages:
2010-2015, 2015-2020 and forecast land use changes in the The Markov chain can be applied to predict the
period 2020-2030 as the basis for policymakers to manage and distribution of different land use types in the next periods (see
use land rationally, economically and effectively. Fig. 3).
II. MATERIALS AND METHODS

The change ratio of Transitional The change ratio of
A. Data and materials land use types in the probabilities of land land use types in the
first time use types second time
Statistics and land inventory data were achieved from
Buon Ma Thuot city in 2010, 2015, and 2020; Buon Ma
Thuot land use maps in 2010, 2015 and 2020 were collected Figure 3. The calculation formula for the change ratio of LUTs
at Dak Lak Department of Natural Resources and using the Markov chain
Environment. The area of land use types is calculated as follows [11].
B. Methods Vt2 = M * Vt1 (1)
a) Approach Methods
Where: M represents the rate of change of land use types
The study used the land use maps collected, normalized throughout data collection and Vt1 represents the area of land
and intersected for 2010, 2015 and 2020 to develop land use use types at the first time; Vt2 describes the area of land use
change maps and analyze land use transitional probabilities types at the second time.
in Buon Ma Thuot city according to two periods: 2010-2015 Based on the assessment results of land use changes in the
and 2015-2020. Based on the land use change analysis, the study area from 2010-2015 and 2015-2020, equation (2) is
Markov Chain model is applied to forecast the future trend of used to determine the estimated period [12].
land use change with the period corresponding to the previous
years (period 2020-2025 and 2025-2030). TDB = TCT + (TCT - TCD) (2)
b) Methods of collecting and inheriting secondary data Where: TDB: prediction time; TCT: Sup timeline of the
assessment process; TCD: Inf timeline of the assessment
The existing datasets of Buon Ma Thuot were collected at process.
the city and the communes People’s Committees, particularly
data on natural conditions, topography, meteorology, III. RESULTS AND DISCUSSION
hydrology, and socio-economic information related to the C. Land use change in Buon Ma Thuot city from 2010 to
development and exploitation of natural resources in Buon 2015
Ma Thuot city, Dak Lak province. Data on the current state
of land use, management mechanism on land; statistical Based on the current land use map for 2010 and 2015 in
yearbook, study results were carried out in the study area. Buon Ma Thuot city, there are 51 types of land. However,
some of them have a small proportion compared to the natural
c) GIS application method area. This will have a significant impact on the forecast of
QGIS software was applied to produce land use maps in land use changes. Therefore, the study edited and grouped the
2010, 2015, and 2020. The study then overlayed the two maps same land use patterns at two time periods. As a result, there
at two different periods to determine the change rate of land are 10 types of land use classified in this study: non-
use types. agricultural land (CDG); unused land (CSD), forest land
(LNP), agricultural land (NNP), land used for cemeteries,
159
graveyards, funeral service centers and cremation centers The land use change matrix from 2010 to 2015 in Buon
(NTD); aquaculture land (NTS), residential land (OTC), Ma Thuot city was produced from overlaying current status
Land with rivers, streams, canals, springs and special-use layers displayed in Table 2 and Fig. 5. The result of area
water surface (SON); Land used by belief institutions (TIN), statistics is shown in Table 3. According to Tables 2 and 3,
land used by religious institutions (TON). After that, a matrix the area of agricultural land, residential land, aquaculture
of land use change in the period 2010-2015 was made and land, land with rivers, streams, canals, springs, special-use
created as a change map. The final output is the current land water surface, and religious land increased between 2010 and
use maps of 2010 and 2015 (Fig 4a and 4b) and statistics on 2015. In particular, aquaculture land has the highest increase
the land use types extracted from the map are shown in Table by 85.01% (149.41ha) since 2010, followed by agricultural
1. land, which has increased by 7.06% (1,851.84ha). The results
Table 1. The area of land use types in Buon Ma Thuot city in 2010 indicate the economic development orientation of the Dak
and in 2015 Lak province from 2010 to 2015, prioritizing the
development of agricultural production while paying
2010 2015
CODE attention to people's lives (The people's Committee of Dak
Area (ha) Percentage (%) Area (ha) Percentage (%) Lak province, 2010). The unused land area has completely
CDG 5.644,75 14,97 4.959,44 13,15 decreased to 96.99% (1,280.61ha) to achieve the local
CSD 1.320,38 3,50 39,77 0,11 development orientation in comparison to 2010, the non-
LNP 1.000,58 2,65 961,42 2,55 agricultural land has declined by 12.14%, and land used for
cemeteries, graveyards, funeral services showed a significant
NTD 214,12 0,57 170,63 0,45
reduction by20.31%. It is appealing to note that land use for
NTS 175,77 0,47 325,18 0,86
belief purposes has reduced by 91.07% but primarily been
OTC 2.307,05 6,12 2.338,87 6,20 converted to religious land. It is clear that this study
SON 776,17 2,06 792,72 2,10 demonstrated the validity of the Markov chain for describing
SXN 26.222,14 69,52 28.073,98 74,43 and projecting land use change in the site. According to the
TIN 6,13 0,02 0,55 0,00 changes in the transition mechanism recorded over the 5-year
TON 50,90 0,13 55,43 0,15 period from 2010 to 2015, the 2010-2015 land use
transitional probabilities matrix based on the Markov chain
Total 37.718,0 100 37.718,0 100
can be used to predict the distribution of land use categories
in the next period (Table 4). To predict changes, there is a
need to obtain a land use transitional probabilities matrix
(Markov matrix). To get this matrix, the area of land use
categories in 2010 which changed to the land use types in
2015 is divided by the total area of the land use types in 2010.
(a)
Figure 5. Land use change map in Buon Ma Thuot city, 2010-

(b) 2015
Figure 4. The land use maps of Buon Ma Thuot city (a. 2010 and
b. 2015)
160
D. Land use change in Buon Ma Thuot city from 2015 to of changes in land use types (Table 6) and a map of land use
2020 changes in Buon Ma Thuot city in the period 2015 -2020 (Fig.
7). Table 5 shows the land use change matrix from 2015 to
Similar to the period 2010-2015, a matrix of land use
2020. From the table, it is clear that there has been
changes over the period 2015-2020 was produced by using
considerable change in the region during the 5 years.
the current database maps of 2015 (Figure 4b) and 2020 (Fig.
Moreover, the changes of land use area were also calculated
6) with 10 land use patterns. As a result, we have a matrix of
over time (Table 6).
land use changes for the period 2015-2020 (Table 5), the area
Table 2. Land use change matrix in Buon Ma Thuot city (2010-2015)
(Unit: ha)
CODE CDG CSD LNP NTD NTS OTC SON SXN TIN TON 2010
CDG .235,44 1,84 26,49 4,83 0,84 683,86 9,43 1.580,88 0,52 40,62 5.644,75
CSD 45,38 24,51 348,69 - 6,35 2,78 16,12 876,40 - 0,16 1.320,38
LNP 157,32 1,08 426,89 0,05 1,17 5,76 6,91 401,40 - - 1.000,58
NTD 13,97 - 1,31 125,90 0,06 3,38 0,23 69,28 - - 214,12
NTS 30,56 0,28 4,79 - 46,98 2,85 12,77 77,53 - - 175,77
OTC 125,44 0,15 0,08 3,16 11,75 961,41 3,13 1.200,41 0,03 1,50 2.307,05
SON 92,88 2,58 15,41 0,07 22,86 1,28 520,16 120,94 - - 776,17
SXN 1.253,28 9,34 137,77 36,57 195,15 672,38 203,93 23.703,10 - 10,61 26.222,14
TIN 0,85 - - - - 1,10 0,00 4,17 - 0,01 6,13
TON 4,32 - - 0,05 0,03 4,07 0,03 39,87 - 2,52 50,90
2015 4.959,44 39,77 961,42 170,63 325,18 2.338,87 792,72 28.073,98 0,55 55,43 37.718,00
Table 3. The change of land use types in Buon Ma Thuot city (2010-2015)
Change
CODE 2010 (ha) 2015 (ha)
Area (ha) Percentage (%)
CDG 5.644,75 4.959,44 -685,31 -12,14
CSD 1.320,38 39,77 -1280,61 -96,99
LNP 1.000,58 961,42 -39,16 -3,91
NTD 214,12 170,63 -43,49 -20,31
NTS 175,77 325,18 149,41 85,01
OTC 2.307,05 2.338,87 31,82 1,38
SON 776,17 792,72 16,54 2,13
SXN 26.222,14 28.073,98 1851,84 7,06
TIN 6,13 0,55 -5,59 -91,07
TON 50,90 55,43 4,53 8,91
Total 37.718,00 37.718,00 0,00 0,00
Table 4. Land use transitional probabilities in Buon Ma Thuot city (2010-2015)

Code CDG CSD LNP NTD NTS OTC SON SXN TIN TON
CDG 0,57 0,00 0,00 0,00 0,01 0,12 0,01 0,28 0,00 0,01
CSD 0,03 0,02 0,26 - 0,00 0,00 0,01 0,66 - 0,00
LNP 0,16 0,00 0,43 0,00 0,00 0,01 0,01 0,40 - -

NTD 0,07 - 0,01 0,59 0,00 0,02 0,00 0,32 - -
NTS 0,17 0,00 0,03 - 0,27 0,02 0,07 0,44 - 0,00
OTC 0,05 0,00 0,00 0,00 0,01 0,42 0,00 0,52 0,00 0,00
SON 0,12 0,00 0,02 0,00 0,03 0,00 0,67 0,16 - 0,00
SXN 0,05 0,00 0,01 0,00 0,01 0,03 0,01 0,90 - 0,00
TIN 0,14 - - - - 0,18 0,00 0,68 - 0,00
TON 0,08 - - 0,00 0,00 0,08 0,00 0,78 - 0,05
161
Table 5. Land use change matrix in Buon Ma Thuot city, 2015-2020

(Unit: ha)
CODE CDG CSD LNP NTD NTS OTC SON SXN TIN TON Năm 2015
CDG 4.561,51 29,38 48,97 1,77 7,81 42,87 51,69 215,03 0,00 0,41 4.959,44
CSD 6,00 21,95 0,40 0,14 3,96 0,84 6,38 0,11 39,77
LNP 9,57 1,58 441,84 1,26 2,60 1,16 4,26 499,15 0,00 961,42
NTD 2,41 0,42 0,00 165,78 0,21 0,00 1,58 0,22 170,63
NTS 4,03 0,01 0,00 280,98 2,99 1,41 35,75 0,00 325,18
OTC 61,63 4,44 2,83 0,46 5,48 1.809,40 0,60 452,78 0,11 1,16 2.338,87
SON 17,05 0,00 0,51 0,00 4,53 0,88 751,07 18,68 0,00 792,72
SXN 557,26 25,71 123,42 18,26 133,34 680,48 24,92 26.506,55 0,04 4,00 28.073,98
TIN 0,00 0,00 0,00 0,00 0,55 0,55
TON 2,78 0,00 0,00 0,00 0,00 0,69 0,00 0,78 0,00 51,18 55,43
Năm 2020 5.222,23 83,48 617,98 187,53 434,88 2.542,64 834,79 27.736,68 0,80 56,97 37.718,00
Table 6. Changes in land use patterns in Buon Ma Thuot city in the period of 2015-2020
2015 2020 Change
CODE
(ha) (ha) Area (ha) Percentage (%)
CDG 4.959,44 5.222,23 262,79 5,30
CSD 39,77 83,48 43,71 109,90
LNP 961,42 617,98 -343,44 -35,72
NTD 170,63 187,53 16,90 9,91
NTS 325,18 434,88 109,70 33,74
OTC 2.338,87 2.542,64 203,77 8,71
SON 792,72 834,79 42,08 5,31
SXN 28.073,98 27.736,68 -337,31 -1,20
TIN 0,55 0,80 0,26 46,63
TON 55,43 56,97 1,54 2,78
Total 37.718,00 37.718,00 0,00 0,00
Table 7. Land use transitional probabilities in Buon Ma Thuot city, 2015-2020

CODE CDG CSD LNP NTD NTS OTC SON SXN TIN TON
CDG 0,92 0,01 0,01 0,00 0,00 0,01 0,01 0,04 0,00 0,00
CSD 0,15 0,55 0,01 0,00 0,00 0,10 0,02 0,16 0,00 0,00
LNP 0,01 0,00 0,46 0,00 0,00 0,00 0,00 0,52 0,00 0,00
NTD 0,01 0,00 0,00 0,97 0,00 0,00 0,00 0,01 0,00 0,00
NTS 0,01 0,00 0,00 0,00 0,86 0,01 0,00 0,11 0,00 0,00
OTC 0,03 0,00 0,00 0,00 0,00 0,77 0,00 0,19 0,00 0,00
SON 0,02 0,00 0,00 0,00 0,01 0,00 0,95 0,02 0,00 0,00
SXN 0,02 0,00 0,00 0,00 0,00 0,02 0,00 0,94 0,00 0,00
TIN 0,00 0,00 0,00 0,00 0,00 0,00 0,00 0,00 1,00 0,00
TON 0,05 0,00 0,00 0,00 0,00 0,01 0,00 0,01 0,00 0,92
Table 8. Area of land use types recorded in 2010, 2015, 2020, 2025 and 2030
(Unit: ha)
Year CDG CSD LNP NTD NTS OTC SON SXN TIN TON Total
2010 5644,75 1320,38 1000,58 214,12 175,77 2307,05 776,17 26222,14 6,13 50,90 37.718
2015 4959,44 39,77 961,42 170,63 325,18 2338,87 792,72 28073,98 0,55 55,43 37.718
2020 5222,23 83,48 617,98 187,53 434,88 2542,64 834,79 27736,68 0,80 56,97 37.718
2025 5479,93 98,14 337,33 181,90 382,34 2578,06 846,94 27166,15 0,80 52,41 37.718
2030 5705,50 108,78 210,95 176,44 337,28 2597,39 861,36 26497,66 0,80 48,22 37.718
162
Table 9. Percentage of land use types at five time periods from 2010, 2015, 2020, 2025 and 2030
(Unit: ha)
2010 14,97 3,50 2,65 0,57 0,47 6,12 2,06 69,52 0,02 0,13 100
2015 13,15 0,11 2,55 0,45 0,86 6,20 2,10 74,43 0,00 0,15 100
2020 13,85 0,22 1,64 0,50 1,15 6,74 2,21 73,54 0,00 0,15 100
2025 14,76 0,26 0,91 0,49 1,03 6,94 2,28 73,18 0,00 0,14 100
2030 15,61 0,30 0,58 0,48 0,92 7,11 2,36 72,51 0,00 0,13 100
Table 10. Area changes in land use types over the 15-year period in Buon Ma Thuot city
(Unit: ha)
2015 -685,31 -1280,61 -39,16 -43,49 149,41 31,82 16,54 1851,84 -5,59 4,53 0,00
2020 262,79 43,71 -343,44 16,90 109,70 203,77 42,08 -337,31 0,26 1,54 0,00
2025 257,70 14,66 -280,65 -5,63 -52,54 35,42 12,15 -570,53 0,00 -4,56 0,00
2030 225,57 10,64 -126,38 -5,46 -45,06 19,33 14,42 -668,49 0,00 -4,19 0,00
337.31 ha, and most of them converted to residential areas.

This is consistent with the overall trend of land use change
across the country. In contrast, some land areas are being
expanded, such as aquaculture land (109.70ha); non-
agricultural land (262.79ha); residential land (203.77ha). In
particular, the unused land area is increasing by 43.71ha as
an inefficient part of the area was converted to agricultural
land in the period 2010-2015. Therefore, it is necessary to
find solutions to make the most of this unused land area.
E. Land use change prediction based on the Markov chain
In performing land use change prediction for 2025 under
the Markov chain, the study established the computation
based on the actual number of observations in land use
change during the same period regardless of the way that the
change process occurred. The expected transitional
probabilities between 2015 and 2020 under the Markov
Figure 6. Land use map of Buon Ma Thuot city in 2020
hypothesis are calculated and displayed in Table 7. They
were obtained by multiplying the area of land use types in
2020 and the period 2015-2020 land use transitional
probabilities used in the following formula (1) where: Vt1: is
current land use area in 2020; M: is the number of transitional
probabilities from land use patterns over data collection
period; Vt2: land use predicted until 2025 (Fig. 8).
Vt1 M
Vt2
Figure 8. Forecast of land use change in 2025 in Buon Ma Thuot city
Similarly, predicting the trend of land use change for
Figure 7. Land use change map in Buon Ma Thuot city
2030 is conducted by multiplying the area of land use patterns
(2015-2020) in 2025 and transitional probabilities of land use types over
the period from 2015 to 2020. The expected results for 2025
As a result, the forest area decreased considerably by
and 2030 are represented in Table 8.
35.72%, (343.44 ha) in the period 2015-2020 and changed
mainly to agriculture; the agricultural land was reduced by
163
Based on the results of land use forecasting at the given REFERENCES

time of the study, it is noticeable that specialized land, [1] Mas, J. F., et al., (1999). Monitoring land-cover changes: A
residential land, land with rivers, streams, canals, springs and comparison of change detection technique, Journal of Remote
special-use water surface, and unused land tend to increase. Sensing, 20 (1), pp. 139 - 152.
Unused land showed an upward trend from 2015 to 2030, [2] Muller, D., and Munro, D., (2007). Issues in spatially explicit
compared to 2010. In contrast, the area in the following years statical land use/cover change (LUCC) models: Examples from
decreased significantly from 1320.38 ha in 2010 to 98.14 ha Western Honduras and the Central Highland in Vietnam, Land
in 2025 and 108.78 in 2030. This trend is a consequence of a use Policy, No 24, pp. 521 – 530.
significant conversion of unused land to agricultural land [3] Yu, W., et al., (2011). Analyzing and modeling land use land
between 2010 to 2015. Conversely, ineffectiveness in the cover change (LUCC) in Daqing city, Heilongjiang Province,
following years led to an increase in this area. China, Applied Geography, No 31, pp. 600 - 608.
[4] Nguyen, K.T., Nguyen, T.T.H., and Pham, V.T., (2010). A case
As shown in Tables 9 and 10, the forest land and study on using satellite image data to develop a map of
agricultural land tend to decrease. By 2025, the area of forest agricultural land use changes in Gia Lam-Long Bien area from
land will be only 337.33 ha, which has reduced by 66.29% 1999-2005. Journal of Soil Science, 33.
compared to 2010. This trend will continue to decrease until [5] Dao, C.T., and Le, T.G., (2003). Understanding of the changes
2030 with the remaining area of 210.95 ha. In addition, the in agricultural and forestry land use in Yen Chau city, Son La
area of agricultural land increased in 2015 before gradually province through remote sensing imagery interpretation
decreasing in the following years. It is likely to see techniques, Journal of Soil Science, 17:169-174.
agricultural land reaching an area of 27,166.15 ha in 2025, [6] Weng, Q., (2001). Land use change analysis in the Zhujiang
and 2,6497.66ha in 2030; The remaining types of land such Delta of China using satellite remote sensing, GIS and
as religious land and belief land tend to increase or decrease stochastic modelling. Department of Geography, Geology, and
Anthropology, Indiana State University, Terre Haute.
irregularly or with little variation.
[7] Michael, I., (2012). A Markov Chain Model of Land Use Change
II. CONCLUSION in the Twin Cities, 1958-2005.
[8] Huynh, V.C., Chau, V.T.T., and Huynh, C.H., (2017).
In the period 2010-2015, land for aquaculture, land with Integrating GIS and Markov chain in studying and predicting
rivers, streams, canals, springs and special-use water surface, land use changes in Nha Trang city, Khanh Hoa province.
residential land, agricultural land, and religious land tends to [9] Nguyen, T.T.A., and Nguyen, T.N.Q., Tran, T.X.P., Nguyen,
increase. The next period (2015-2020) marked a sharp H.A., (2021). Integrating GIS and Markov chain in land use
reduction of forestry land and agricultural land to convert to change in Dong Hoa district, Phu Yen province. Sci. Tech. Dev.
non-agricultural land and residential land. Forecast results for J. - Sci. Earth Environ.;5(SI1):SI95-SI106.
2025 and 2030 show that agricultural land and forestry land [10] Nguyen, K.L., (2005). Application of GIS and Markov chain in
will continue to decrease to focus on promoting the land use change assessment, derived from the 2013 National
development of non-agricultural land. The research process GIS Application Conference, Agriculture Publisher, Ho Chi
shows that the transformation matrix of land use types in Minh City.
different periods is always different because Buon Ma Thuot [11] Mubea K.W., Ngigi T.G., and Mundia C.N., (2010). Assessing
city is still on the way of development, so the change in land application of Markov chain analysis in predicting land cover
change: A case study of Nakuru municipality, Department of
depends on the socio-economic development. It can be said
Geomatic and Geospatial Information Systems, Jomo Kenyatta
that the Markov chain will be suitable for areas with stable University of Agriculture and Technology, Nairobi.
development. However, the research results are expected to
[12] Tran, A.T., and Hoang, T.A., (2010). Applying the Markov-
be a reference for the authorities in the planning of land use Cellular Automata model to predict land cover change in Hanoi
in the area. City, Hanoi, The 10th International Conference of Southeast
Asian Geography Association.
164
Dissimilar Friction Stir Welded Lap-joint of

Aluminum Alloy 6061 and 316 Stainless Steel
Huy Huu Ho Hao Dinh Duong Nam Hoai Quach
Department of Mechanical Department of Engineering Department of Engineering
Engineering Mechanics Mechanics
Nha Trang University, Nha Trang University, Nha Trang University,
Nha Trang city, Vietnam Nha Trang city, Vietnam Nha Trang city, Vietnam
dinhhao@ntu.edu.vn
Thuyen Van Phi Tra Hung Tran

Center of Experiments and Practices Department of Engineering Mechanics
Nha Trang University, Nha Trang University,
Nha Trang city, Vietnam Nha Trang city, Vietnam
tra@ntu.edu.vn
Abstract: The dissimilar friction stir welded lap-joint experimental results indicated that the IMCs layer was
between aluminum alloy 6061 to 316 stainless steel is improved significantly in both thickness and microstructure
fabricated. The attention is focused on the impact of welding compared to that produced by fusion welding method.
speeds on the interface, microstructure, and mechanical However, rotary friction welding method is only suitable for
properties of the joints. The test results show that an connecting circular structure.
increase in the welding rate leads to a finer grain size and Friction stir welding (FSW) seems to be a suitable
thicker oxide layer along the interface. The joint tensile process for joining plate structures. The dissimilar joint of
strength is reduced by increasing welding rate. The joint is aluminum to steel produced by this method was attended by
fractured along the welding interface. The discussion is put researchers in both butt-joint and lap-joint. They tried to
on the welded interface and fracture characteristics. focus on the improvement of joint strength by changing
welding parameters [9,10,11], tool offset [10,12], and
Keywords: friction stir welding, dissimilar joint, tensile welding tool [9,12]. The main aim of these works is to
strength, lap-joint, AA6061, 316 stainless steel evaluate the influence of them on the IMCs layers. They
concluded that the joints attained the high strength
I. INTRODUCTION corresponding to thinner IMCs layer. This implies that
Reduction in weight structure is a crucial issue in many controlling the thickness and composition of the IMCs layer
industries such as aerospace, automotive, shipbuilding, etc. has a big contribution to enhancing joint strength. However,
[1]. One of the solutions to be attained for this purpose is the the characteristic of dissimilar FSWed lap-joint of aluminum
application of light metal and dissimilar joints. Dissimilar alloys 6061 to 316 stainless steel has been limited in
joints of aluminum to steel in composite structures have literature.
many potential employments in high vacuum chambers and The aim of this work is to investigate the impact of
cryogenic pressure vessels [2]. However, this welding has welding speed on the mechanical properties of the dissimilar
been facing a lot of significant challenges because of FSWed lap-joint of AA6061/316SS. The microstructure,
difference in terms of melting temperature, mechanical interface, and mechanical properties of the joints were
properties, and the coefficient of thermal expansion. So, the detailly revealed.
high stress concentration might be easily formed in Fe-Al
intermetallic compounds (IMCs), resulting in degrading the II. EXPERIMENTAL PROCEDURES
joint strength.
A. Fabrication of lap-joint
Weldability of dissimilar joint of steel to aluminum by
fusion welding was revealed in literature [3-5]. The The dissimilar lap-joint of aluminum alloys 6061
dissimilar joint of 5A06AA and ST04Z steel was produced (denoted as AA6061) to 316 stainless steel (denoted as 316
by laser welding [3]. Yan et al. [4] and Dong et al. [5] joined SS) was produced by friction stir welding (FSW) method.
aluminum alloys to low carbon steel by using cold metal AA6061 plate has the thickness of 3.0 mm while 316 SS has
transfer and fusion brazing. However, the joint strength the thickness of 1.0 mm. The welding tool has the shoulder
produced by the welding methods is relatively low owing to with a diameter of 18.0 mm and a cylindrical pin with the
the appearance of high residual stress and thicker IMCs. diameter of 6.0 mm. The pin length is 2.9 mm. The welding
speed was set up from 300 mm/min to 600 mm/min while
Friction welding has been considered as potential the rotational speed was the constant of 800 rpm. The
technique to weld aluminum to steel owing to low welding shoulder penetration is 0.2 mm during the FSW process. The
temperature. So, the thickness of IMCs layer and residual pin axis was obliquely aligned in a vertical direction with an
stress can be significantly improved [6]. Rotary friction angle of 2.0 degrees. The AA6061 plate was placed at the
welding method was utilized to fabricate the dissimilar joint advancing side (AS) while 316 SS plate was at the retreating
of AA6063/304SS [7] and AA5052/304SS [8]. The side (RS), as seen in Fig. 1.

165
B. Analysis process morphology was insignificantly changed after the FSW

The specimen for microstructure, hardness, and tensile process.
test was perpendicularly extracted to welding line. The Due to significant differences in the mechanical
microstructure specimen was etched by Kerller’s reagent properties of materials, four welded zones appeared on the
after it was polished with abrasive paper and alumina. The cross-section of lap-joint after etching. That is stirred zone
microstructure of the joints was observed by a microscope. (SZ), thermo-mechanically affected zone (TMAZ), and heat
The hardness distribution in the stirred zone from the top affected zone (HAZ). The microstructure of them is shown in
surface of AA6061 to the bottom surface of 316 SS was Fig. 4(b-e). Because of the impact of welding temperature,
measured by the means of a micro Vickers. The distance of the microstructure in the HAZ zone is coarse with big grain
measured position is 0.2 mm. The dimension of tensile size (Fig. 4(d)), even though it is coarser than that of bae
specimen is shown in Fig. 2. Testing process was carried out metal (BM) AA6061 (Fig. 4(e)). In contrast, the SZ zone is
by means of universal testing machines Instron under a finer microstructure by the most severe plastic-deformation
movement of 1.0 mm/min. (Fig. 4(b)). The grain size in this zone is 20-30 m
approximately. The elongated grain size can be observed in
the TMAZ zone, as seen in Fig. 4(c). The similar result was
also found in [13].
The impact of welding rate on the microstructure of the
SZ zone is indicated in Fig. 5. It seems that increasing
welding rate decreased grain size. In other words, the
microstructure of the joints is finer corresponding to high
Figure 1. Configuration of FSWed lap-joint between AA6061 and welding speed (Fig. 5(d)). This relationship can be useful for
316 SS enhancing the joint strength [14,15]. Decreasing heat input
corresponding to increasing welding rate is reason for this
result [15]. In other words, the low value in heat-input
produced a fine-grained structure in the joints [14].
However, increasing welding speed seems to be
unsuitable for interface formation in this work, as seen in
Fig. 6. Notably, the dark line along interface was observed in
all welding regimes. The thickness of dark line trends to
increase with growing welding rate. As reported in [16], the
dark line is oxide layer that might be uneliminated after
Figure 2. Geometry of tensile specimen of the lap-joint (dimension welding process. This result degraded the bonding strength
in mm) of interface. The role of interface in the joint strength will be
investigated in the next section.
C. Inspection of dissimilar lap joints
The representative geometry of dissimilar lap-joints
fabricated by 300 mm/min are displayed in Fig. 3. It can be
realized that some defects such as distortion and
solidification cracking that often occur in conventional
fusion welding, is not found on welding surface of the joints.
Furthermore, the welding surfaces are smooth without
groove defects or rough.
Figure 3. Representative of the welding surface of AA6061/316

SS joint produced by 300 mm/min
D. Microstructure of dissimilar lap-joint
The macrostructure of the dissimilar lap-joint of AA6061
to 316 SS produced by 300 mm/min is displayed in Fig. 4(a).
It has clear that the joint is free of defects. After etching, we
easily realized AA6061 sheet, 316 SS sheet, and welding Figure 4. (a) Macrostructure of cross-section and microstructure at
interface. Here, the macrostructure of interface morphology (b) SZ, (c) TMAZ, (d) HAZ of joint produced by 300 mm/min, and
seems to be a strong line. This means that the interface (e) BM AA6061
166
the distribution of welding temperature. The welding

temperature at near interface has higher than that at far away.
As a result, the microstructure and the hardness of joint is
inhomogeneous [18]. Similar to the trend of AA6061, the
high hardness value is reached corresponding to increasing
welding speed.
Figure 5. Microstructure at SZ produced by (a) 300 mm/min, (b)

400 mm/min, (c) 500 mm/min, and (d) 600 mm/min
Figure 7. Effect of welding regimes on distribution of micro-
hardness
F. Tensile properties
The influence of welding rate on the maximum joint
tensile strength is performed in Fig. 8. The diagram indicated
that the joint strength was deeply reduced with growing
welding rate. It emphasizes that a high welding rate is not
suitable for this work. The highest joint strength is about 70
MPa and it is produced by the welding rate of 300-400
mm/min. Then it quickly reduced down to 35 MPa
corresponding to the tool speed of 600 mm/min.
As reported by Wan and Huang [6], the formation of the
thickness of the IMC layer has a big contribution in the
mechanical properties of the joints. It is believed that the
joint strength might be enhanced if the IMCs thickness is
smaller than the critical value of 1-2 m [19,20].
Additionally, the thinner IMCs is only reached at higher
welding speed. In other words, the increase in welding rates
Figure 6. Microstructure at interface produced by (a) 300 mm/min, reduced the thickness of the IMC due to attaining a lower
(b) 400 mm/min, (c) 500 mm/min, and (d) 600 mm/min, as marked heat-input during FSW process. In this work, however, the
in Fig. 4(a) welding defect in terms of oxide layer had been eliminated
out the joint, as seen in Fig. 6. So, degrading joint strength
E. Hardness distribution with increasing welding rate can be impacted by this oxide
The effect of welding rate on the distribution of hardness layer.
value of the dissimilar joints is displayed in Fig. 7. Here, the
average hardness value was measured at three various
positions in stir zone. The hardness distribution of AA6061
is left with smaller value than that of 316 SS is right. In all
welding regimes, the value hardness in the SZ was deeply
reduced in comparison with base metal alloys. The average
value is about 75 HV compared to 110 HV of BM AA6061.
Additionally, the hardness trends to increase with increasing
welding rate. This result has close relationship to change in
microstructure in Fig. 5. Here, the small grain size
corresponding to high welding speed can bring about a high
value of the strength [17].
Dissimilar to the AA6061, the hardness distribution of
the 316 SS has fluctuation. The lower value is obtained at
near welding interface in all cases, then it gradually
increased to BM region. This result can relate to change in Figure 8. Effect of welding rate on untimate joint tensile strength
167
The fracture location of the joints produced by 300 [2] H. Dong, J. Yang, Y. Li, Y. Xia, X. Hao, P. Li, M. Lei. Evolution of
mm/min is showed in Fig. 9(a). Here, the crack location took interface and tensile properties in 5052 aluminum alloy/304 stainless
steel rotary friction welded joint after post-weld heat treatment. J.
place at the welding interface. As presented in previous Manuf. Process. vol. 51, pp. 142–150, 2020.
section, the presence of oxide layer prevented the diffusion [3] J. Huang, J. He, X. Yu, C. Li, D. Fan. The study of mechanical
of two alloys, resulting in reducing bonding strength. The strength for fusion-brazed butt joint between aluminum alloy and
fracture surface is displayed in Fig. 9(b). It is easy to realize galvanized steel by arc-assisted laser welding. J. Manuf. Process. vol.
the bonding area that showed more bright color on the 25, pp. 126–33, 2017.
surface of two metal alloys. The width of bonding area is 6.0 [4] J. Yang, A. Hu, Y. Li, P. Zhang, D. C. Chandra Saha, Z. Yu. Heat
mm that is equal to the diameter of pin. input, intermetallic compounds and mechanical properties of Al/steel
cold metal transfer joints. J. Mater. Process. Technol. vol. 272, pp.
40–46, 2019.
[5] H. G. Dong, W. J. Hu, X. C. Zhang. Detachment of interfacial layers
during arc-brazing of aluminum alloy to carbon steel with filler wire.
Trans. Nonferrous Met. Soc. China. vol. 23, pp. 1583–1588, 2013.
[6] L. Wan, Y. Huang. Friction stir welding of dissimilar aluminum
alloys and steels: a review. Int. J. Adv. Manuf. Technol. vol. 99, pp.
1781–1811, 2018.
[7] M. Kimura, K. Suzuki, M. Kusaka, K. Kaizu. Effect of friction
welding condition on joining phenomena and mechanical properties
of friction welded joint between 6063 aluminium alloy and AISI 304
stainless steel. J. Manuf. Process. vol. 26, pp. 178–187, 2017.
[8] S. Fukumoto, H. Tsubakino, K. Okita, M. Aritoshi, T. Tomita.
Friction welding process of 5052 aluminium alloy to 304 stainless
steel. Met. Sci. Technol. vol. 15, pp. 1080–1086, 1999.
[9] M. Dehghani, A. Amadeh, S. A. A. Akbari Mousavi. Investigations
on the effects of friction stir welding parameters on intermetallic and
defect formation in joining aluminum alloy to mild steel. Materials &
Design. vol. 49, pp. 433–441, 2013.
[10] H. A. Derazkola, H. J. Aval, M. Elyasi. Analysis of process
parameters effects on dissimilar friction stir welding of AA1100 and
A441 AISI steel. Sci. Technol. Weld. and Join. vol. 20(7), pp. 553–
562, 2015.
[11] R. P. Mahto, R. Bhoje, S. K. Pal, H. S. Joshi, S. Das. A study on
mechanical properties in friction stir lap welding of AA 6061-T6 and
AISI 304. Mater. Sci. Eng. A. vol. 652, pp. 136–144, 2016.
[12] K. K. Ramachandran, N. Murugan, S. Shashi Kumar. Effect of tool
axis offset and geometry of tool pin profile on the characteristics of
friction stir welded dissimilar joints of aluminum alloy AA5052 and
HSLA steel. Mater. Sci. Eng.A. vol. 639, pp. 219–233, 2015.
[13] L. Zhang, H. Zhong, S. Li, H. Zhao, J. Chen, L. Qi. Microstructure,
Figure 9. (a) Fracture location and (b) fracture surface of the joint mechanical properties and fatigue crack growth behavior of friction
produced by 300 mm/min stir welded joint of 6061-T6 aluminum alloy. Int. J. Fatigue. vol. 135,
105556, 2020.
IV. CONCLUSION [14] A. Kalinenko, I. Vysotskii, S. Malopheyev, S. Mironov, R.
The dissimilar lap-joint of the AA6061 to 316 SS was Kaibyshev. Relationship between welding conditions, abnormal grain
growth and mechanical performance in friction-stir welded 6061-T6
produced by friction stir welding technology. The aluminum alloy. Mater. Sci. Eng. A. vol. 817, 141409, 2021.
microstructure, interface, and mechanical properties of the [15] A. Kalinenko, K. Kim, I. Vysotskiy, I. Zuiko, S. Malopheyev, S.
joints were revealed. Some conclusions were reached as Mironov, R. Kaibyshev. Microstructure-strength relationship in
followings: friction stir welded 6061-T6 aluminum alloy. Mater. Sci. Eng. A. vol.
793, 139858, 2020.
 The cross-section of the lap-joint is free tunnel defect
[16] D. D. Hao, M. Okazaki, T. H. Tra. Influence of probe length on the
in all cases. formation of an interface in friction stir welded T-lap joints. Mater.
 Increasing welding rate would lead to fine-grained Manuf. Process. vol. 36, pp. 693–701, 2021.
microstructure and thicker oxide layer at the interface. [17] J. F. Guo, H. C. Chen, C. N. Sun, G. Bi, Z. Sun, J. Wei. Friction stir
welding of dissimilar materials between AA6061 and AA7075 Al
 The joint strength was reduced corresponding to alloys effects of process parameters. Mater. Des. vol. 56, pp. 185-192,
increasing welding speed. The highest tensile strength 2014.
was reached 70 MPa at the welding speed of 300 [18] L. Zhou, K. Nakata, T. Tsumura, H. Fujii, K. Ikeuchi, Y. Michishita,
mm/min. M. Morimoto. Microstructure and mechanical properties of 316l
stainless steel filling friction stir-welded joints. J. Mater. Eng.
 The fracture location took place at the welding Perform. vol. 23(10), pp. 3718–3726, 2014.
interface in all cases. The improvement of interface [19] X. Liu, S. H. Lan, J. Ni. Analysis of process parameters effects on
kept a vital role in enhancing dissimilar lap-joint friction stir welding of dissimilar aluminum alloy to advanced
strength. high strength steel. Mater. Des. vol. 59, pp. 50–62, 2014.
[20] M. Movahedi, A. H. Kokabi, S. M. S. Reihani, H. Najafi. Effect of
REFERENCES tool travel and rotation speeds on weld zone defects and joint
[1] P. Gallo, M. Guglielmo, J. Romanoff, H. Remes. Influence of crack strength of aluminum-steel lap joints made by friction stir welding.
tip plasticity on fatigue behaviour of laser stake-welded T-joints made Sci. Technol. Weld .Join. vol. 17, pp. 162–167, 2012.
of thin plates. Int. J. Mech. Sci. vol. 136, pp. 112–123, 2018.
168
A Stochastic Half-Car Model for Vibration Analysis

with Uncertain Parameters
Nguyen Van Thuan
Department of Transportation Engineering
Nha Trang University
Nha Trang, Vietnam
thuannv@ntu.edu.vn
Abstract: This study deals with the response variability body as the principal source of control while the latter is
in the vehicle’s natural frequencies by a perturbation acquired by using fuzzy-logic control as the complementary
method. A four degree of freedom half-car model was used control [5].
to evaluate the vibration response of vehicles with uncertain Some of these reports have shown that the dynamic
parameters. The governing equations for the natural vehicle’s behavior and its vibration control were
frequency of vehicles are derived from Lagrange’s investigated. The disturbances from the road are assumed to
equations. The randomness in suspension stiffness type be sinusoid and the time delay between the disturbances is
stiffness is considered a random variable. Mass matrix and investigated [6]. The optimally controlled active suspension
stiffness matrix can be perturbed concerning the mean of the was used as a model for the active suspension of vehicles
system random variables. The variance of eigenvalue and [7].
mean of eigenvalue are obtained by zeroth-order and first-
order of approximated perturbation equation. The mean The majority of research in vehicle dynamics was
value, standard deviation, and variation coefficient of the conducted on the assumption that all vehicle system
vehicle’s natural frequencies are obtained by using Monte characteristics are deterministic, despite the enormous
Carlo simulation and perturbation method. The results increase of mathematical modeling tools for analysis and
obtained by the perturbation techniques were consistent with computation. The stochastic analysis was a useful analytical
the ones achieved by the Monte Carlo simulation approach. tool to predict the response of vehicle dynamic with random
material properties.
Keywords: perturbation method, vehicle dynamic, For the vehicle dynamics with uncertainties, the Monte
natural frequencies, Monte Carlo simulation Carlo simulation method and perturbation method are
widely used. A two-degree-of-freedom vehicle model with
I. INTRODUCTION
non-linear passive suspension components is taken into
The unevenness of the road surface that the vehicle account, together with active control of the time-varying
travels on is what mostly stimulates an on-road vehicle's response to a stationary random stimulation [8]. A two-
vibration. Due to its crucial role in ride comfort, vehicle degrees-of-freedom linear model is used to describe the
safety, and general vehicle performance, vehicle dynamic dynamic behavior of vehicles running on randomly profiled
dynamics has been a popular research area for many years. roads [9]. Polynomial chaos approaches that explicitly
Numerous studies of the theoretical and experimental consider parametric uncertainty during the modeling of
investigation on the quarter-car, half-car and full-car model vehicle dynamics are presented [10]. A four degree of
of passively and actively suspended road vehicles in order to freedom car model was used to study the vibration response
improve comfort and handling. of vehicles with uncertain parameters [11].
The mathematical model of the quarter-car model is However, the research works concerning stochastic
derived in terms of the state form and the dynamics vehicle dynamic analysis are still limited. In this paper, the
evaluation in the sprung mass displacement and acceleration first-order perturbation technique based on the stochastic
[1]. The geometric parameters of suspension system are Lagrange’s equation was developed for the solution of
optimized using Matlab as an optimization tool [2]. The random eigenvalue analysis. Comparison results have been
quarter-car model was selected to study the vehicle’s carried out with Monte Carlo simulation through the random
response in order to profile imposed excitation with variables [12].
randomly varying traverse velocity and variable
The paper is outlined as follows: In Section 2, a
vehicle forward velocity [3].
formulation for the solution of the natural frequency of the
The spectral decomposition method and fuzzy-logic half-car model is proposed in the context of Lagrange’s
control are widely used to assess the dynamic of half-car equation. Section 3 describes a procedure of using the
model. The spectral decomposition method and fuzzy-logic random vibration analysis with spring of stiffness of half-car
control are applied to compute accurately the RMS values model. The perturbation and Monte Carlo simulation
for the control forces, suspension strokes, and tire deflection methods are addressed for the half-car model with uncertain
at front and rear in a half-car model with preview [4]. The parameters.
former is obtained by vertical acceleration of the vehicle

II. VEHICLE MODEL III. RANDOM VIBRATION ANALYSIS

Figure 1 depicts a half-car model with four degrees of The Lagrange’s equations of the half vehicle model with
freedom. In this model, the denoted sprung and unsprung the spring of stiffness are considered to be random. The
masses correspond to one corner of the vehicle, by m1 and equation for natural frequency analysis can be expressed as
m2, respectively. The suspension system was represented by [K  i M ]X i  0 (5)
a linear spring of stiffness (k1, k2) and the tire was modeled
by a linear spring of stiffness (kt1, kt2). Since tire damping In this situation, the spring stiffness of random variables
was normally very minimal, it was not taken into (r) can be introduced in this context by doing the following:
consideration in this investigation. This model consisted of k1  k1 1  r1 
the body bounce (x), body pitch (), wheel hop (x1, x2) and
independent road excitations (y1, y2). We used the k2  k2 1  r2 
(6)
Lagrange’s equation to determine the equation of motion for
vehicle vibrating model. The kinetic and potential energies
kt1  kt1 1  r3 
of the system are kt 2  kt 2 1  r4 
1
2
 1  m2 x2  I y
 K  mx&2  m1 x&
2
&2 
&2  Following that, all of the natural frequency equation's
components can be affected in the following ways with
1 respect to the mean of the random variables:
V  kt1 ( x1  y1 ) 2  kt 2 ( x2  y2 ) 2 
2
4
  K  r1 , r2 ,...r4   K 0   K i ri 
1
  k1 ( x  x1  a1  ) 2  k2 ( x  x2  a2 ) 2  i 1
2 4  j
Applying Lagrange’s equations can be rearranged in a  j  r1 , r1 ,...r4    j 0   ri  (7)
matrix form i 1 ri
 [M ] X&
&  [K ] X   0  4 X j
X j  r1 , r1 ,...r4   X j 0   ri 
where K and M are the stiffness and mass matrices. i 1 ri
m 0 0 0 Applying the first-order equation, the variances of the
0 I 0 0  natural frequencies can be expressed as
 M    0 0y m1 0   i , j  X iT  K, j  X i (8)
 
0 0 0 m2  Introducing Eq. (8) into the expression of Taylor
expansion on the eigen value , the eigenvalue for i-th can be
 k1  k2 a2 k2  a1k1 k1 k2  obtained as.
a k  a k k a 2  k a 2 a1k1 a2 k2 
 K    2 2k 1 1 1 1a k 2 2 k1  kt1 0 
Var i  r1 , r2 ,...r4  
(9)
 
1 1 1
  k 2  kt 2   Var  i  r1 , r2 ,...r4  
 k 2 a 2 k2 0  
2
x 4
1  i  2
    j
  j 1 4i 0  rj 
   
X 
The coefficient of variation (COV) of natural frequency,
x1
  which is defined as a ratio of the standard deviation (SD) of
 x2  response to the absolute of the mean response.
For Eq. (3) to be valid or to have nontrivial solution, the Var ( )
following needs to be satisfied: COV  (10)

 det  K  -   M  = 0 
IV. NUMERICAL EXAMPLES
xx Natural frequencies of a half-vehicle model, consider a
a2 a1
vehicle with the following characteristics. Table 1 shows
 mean values of the vehicle’s parameter for this study [13].
m, Iy
C Table 1. Uncertain parameters in half vehicle model
k2 k1 Parameters Unit Mean values

x2 x1 m kg 420
m2 m1
m1 kg 53
kt2 kt1 m2 kg 53
y1 y1
k1, k2 N/m 10,000
kt1, kt2 N/m 200,000
Figure 1. The haft car model of the vehicle
170
Table 2. First four natural frequencies

Mode number Natural frequency [Hz]
1 1.11
2 1.46
3 8.43
4 10.02
Table 2 provides the first four natural frequencies of a

haft car model. The results obtained in this study are exactly
the same as given by Jazar [13], showing the adequacy of the
proposed formulation.
Table 3. Error depending on mode number ( f  0.1)
Mode number PM MCS Error [%]

1 4.90 4.92 0.4 Figure 3. COV of natural frequencies of different mode number
for perturbation method and Monte Carlo simulation
2 6.76 6.86 1.5
3 62.95 64.58 2.5 III. CONCLUSION

The uncertainty in eigenvalue of half vehicle model is
The discrepancies between the perturbation method and
analyzed the Lagrange’s equations by the perturbation
Monte Carlo simulation in the mean of the natural
method and Monte Carlo simulation. For all the cases, the
frequencies for the first three modes are shown in Table 3. In
response variability, given in terms of the standard deviation
all cases, the differences between those two results in
and coefficient of variation of natural frequency is revealed
percentile is from 0.4% to 2.5% at the standard deviation of
to be smaller than 50% of the input standard deviation of the
stochastic assumed ( f  0.1). This is can easily be random variable in the case of the random spring of stiffness.
understood the error increases as mode number increases
from 1 to 3. ACKNOWLEDGMENT
The authors would like to acknowledge the support from
Nha Trang University (NTU) under grant number TR2021-
13-12.
REFERENCES
[1] G. A. Hassaan, “Car dynamic using Quarter model and
passive suspension, part I: effect of suspension damping and
car speed”, International Journal of Computer Techniques,
vol. 1, 2014.
[2] K. K. Jagtap and D. R. Dolas, “Simulation of quarter car
model using Matlab”, International Journal of Engineering
Research and General Science, vol. 3, 2015.
[3] S. Turkay and H. Akcay, “A study of random vibration
characteristics of the quarter-car model”, Journal of Sound
and Vibration, pp. 111-124, 2004.
[4] A. G. Thompson and B. R. Davis, “Computation of the rms
state variables and control forces in a half-car model with
preview active suspension using spectral decomposition
method”, Journal of Sound and Vibration, vol. 3, pp. 571-583,
Figure 2. SD of natural frequencies of different mode number for 2005.
perturbation method and Monte Carlo simulation [5] T. Yoshimura, N. Nakaminami, M. Kurimoto and J. Hino,
“Active suspension of passenger cars using linear and fuzzy-
The effect of mode number of the random on the logic control”, Control Engineering Practice, vol. 7, pp. 41-47,
variability of natural frequency is shown in Fig. 3 and Fig 4, 1999.
where the results of MCS and PM are compared. As seen in [6] Q. Zhu and M. Ishitobi, “Chaotic vibration of a nonlinear full-
Figures 3 and 4, these values were found to be vehicle model”, International Journal of Solids and Structures,
approximately 38% of the assumed standard deviation in the vol. 43, pp. 747-759, 2006.
perturbation method and the Monte Carlo simulation at the [7] E. Esmailzadeh and F. Fahimi, “Optimal adaptive active
standard deviation of stochastic assumed  f  0.05, and suspension for a full car model”, Vehicle System Dynamic,
vol. 27, pp. 89-107, 1997.
 f  0.15 in SD and COV of natural frequency cases. Also, [8] S. Narayanan and S. Senthil, “Stochastic optimal active
a good agreement between results of the scheme proposed control of a 2-dof quarter car model with non-linear passive
and MCS was found to exist. suspension elements”, vol. 211, pp. 495-506, 1998.
171
[9] M. Gobbi, F. Levi and G. Mastinu, “Multi-objective Transport and Communications Science Journal, vol. 72, pp.
stochastic optimisation of the suspension system of road 215–226, 2021.
vehicles”, Journal of Sound and Vibration, vol. 298, pp. 1055- [12] J. E. Hurtado and A. H. Barbat, “Monte Carlo Techniques in
1072, 2006. Computational Stochastic Mechanics”, Archives of
[10] G. Kewlani, J. Crawford and K. Iagnemma, “A polynomial Computational Methods in Engineering, vol. 5, pp. 3-30,
chaos approach to the analysis of vehicle dynamics under 1998.
uncertainty”, Vehicle System Dynamic, vol. 50, pp. 749-774, [13] R. N. Jazar, “Vehicle dynamic: Theory and application”,
2012. Springer, New York, 2014.
[11] N. V. Thuan, and T. D. Hien, “Variability in frequency of
vehicle vibration analysis with multiple random variables”,
172
Sources of Payment Risks to Contractors in the

Vietnam Construction Industry
Duong Vuong Thao Huynh Phu Tran
Department of Economics and Department of Transport Economics Department of Transport Economics
Construction Management Ho Chi Minh City University of Ho Chi Minh City University of
Mientrung University of Civil Transport Transport
Engineering Ho Chi Minh City, Viet Nam Ho Chi Minh City, Viet Nam
Tuy Hoa city, Viet Nam thao.huynh@ut.edu.vn phu.tran@ut.edu.vn
vuongthithuyduong@muce.edu.vn
Abstract: The construction contract payment risks are Vietnam disputes with the total value, is up to 10,000 billion
common issues that may significantly influence the VND [6]. Construction project disputes have usually
contractors' resource allocation ability during project stages. occurred because of untimely contractual payments [7] and
Many studies attempt to identify construction project risk financial issues [8]. Payment, which seems to be contractors’
factors but few studies deeply focus on contract payment main income, is an amount of money, that owners should
issues between parties. This study investigates risks and pay to contractors due to contractual agreement [9]. If the
associated risk sources regarding contract payments to owners cannot execute their payment obligations in
contractors. By applying for a large literature review and accordance, contractors get payment risks, which negatively
impacts their corporation’s financial statement. Both owners
holding in-depth interviews of high-experienced experts, the
and contractors are opposite payment risks and bear their
officially critical payment risk factors have been released
consequences at different levels. Therefore, all contractual
and to examine these factors’ importance, the descriptive parties must have a good awareness of critical payment risk
statistical survey data analysis has been conducted. As a resources being a risk kind confronted during the
result, the lack of thorough negotiation preparation, the construction contract implementation.
weakness of the legal department or even the common short
of the legal department in contractors’ corporate structure, The key study aim is to discover the root payment risk
and contractors’ and owners’ acts are found as the most resources to contractors participating in the Vietnam
serious factors leading contractors to payment risks. The construction industry. To obtain this, qualitative and
initial study findings would help contractors be examined quantitative methods, including the literature survey, deeply
carefully during negotiation over the contractual terms and interview, questionnaire survey, and descriptive statistics are
employed to underline the critical resources. This study
conditions and contractual management to mitigate risk
contributes to the literature twofold; firstly, expanding the
consequences in a construction project. current literature in which terms of payment risks are quite
limited, regarding contract payment; secondly, identifying
Keywords: payment risk, risk resources, contractors,
and ranking the importance of the critical payment risk
owners resources to contractors by using the mean score approach.
I. INTRODUCTION This study organizes as follows. Section II presents a
The construction project is inherently complex, dynamic, brief survey of the literature. The research methodology is
and involved multiple feedback processes [1, 2]. Uncertain shown in section III. The empirical result and discussion are
construction operations are significantly influenced by the outlined in section IV.
productivity of labor, equipment, materials, budget, and
implementation methods; and their benefits are capably II. LITERATURE REVIEW
affected by the result of the construction’s industrial The increased payment dispute shows that payment risks
heterogeneous nature. Risk management was formally become common in the construction industry. A large
adopted in the 1990s [3]. The categorized risk sources based number of empirical studies investigated the main payment
on controllable and uncontrollable factors lead to cost risk resources to contractors in construction projects. Latham
overrun, time delay, and poor quality in a project [4]. A [10], for instance, had already examined disputes and
considerable amount of projects do not continuously meet litigations being the problems of procurement and
their due dates, exceed budget; untimely deliver the contractual arrangements in the United Kingdom
specifications, miss quality, underestimate risk; or do not construction industry. Semple, Hartman, et al. [11] pointed
apply to customer satisfaction despite the recognized out that the most common causes of disputes were changes
criticality of project success for organizations [5]. in the work scope, weather, and limited access to the works.
Additionally, communication quality and effort coordination
From 2015 to July 2021, according to the Vietnam among key stakeholders including owners, contractors,
Meditation center’s statistics, there are 143 construction consultants, suppliers were dispute sources. Moreover,
disputes related to late payment, reduced payment, or even Kumaraswamy [12], Henriod and Lantran [13] investigated
no payment, which accounted for 12% of the total amount of that all construction conflicts, claims, and disputes involved

payment issues. Cheung and Suen [14] found that the late roles, project management, project implementation issues.
payment from the owner is the critical factor leading to a
dispute in the construction industry. Accordingly, Pettigrew [18] found that the industry has a
culture of ‘work first, gets paid later’, which adds another
Che Munaaim, Mohd Danuri, et al. [15] investigated the dimension to the payment problems to contractors.
main identified factors for late and non-payment in the Furthermore, Ramachandra and Rotimi [19] found that the
Malaysian construction industry including delays in financial strength of industry players was considered central
certification, paymaster's poor financial management, local to payment problems among contractual issues, the financial
culture/attitude, paymaster's failure to implement good strength of the industry players, disputes, shortcomings of
governance in business, underpayment of certified amounts payment processes, and ‘domino effects.
by the paymaster, and the use of 'pay when paid' clauses in
contracts. This leads contractors face to cash-flow problems, Lam [20] identified 16 risk factors classified into five
stress, and financial hardship. Assaf and Al-Hejji [16] related groups, namely, capability risks, contractual and legal
attempted to diagnose and prioritize payment risks risks, economic risks, physical-factor risks, political risks,
concerning the acts or omissions of the stakeholders but only and societal risks. Hansen [21] examined untimely
in connection with delay and/or cost issues, which is only a contractual payments being the cause of the disputes and
fraction of the holistic payment risk sources. It is worth litigations in the construction industry. A dispute and/or
focusing on the payment risks to contractors. Mbachu [17] litigation could occur when a claim for the rights of one
showed that owners, contractors, consultants, and suppliers party is denied by another. In the Vietnam construction
were respectively evaluated to contribute 24%, 36%, 37%, industry, Vo, Nguyen, et al. [7] discovered the owner’s
and 3% of contractors’ payment risks and cash-flow delayed payment was one of the economic items causing the
problems attributable to stakeholders in a project through dispute between parties.
their acts or omissions were associated with their contractual Disappointingly, there is little research on the contract
Figure 1. Research framework
174
payment issues between parties contributing to contractors’ In step 2, the questionnaire survey has been designed to
payment risks in the construction industry. The payment collect data from Vietnam construction projects from now
issues among contracting parties are rarely analyzed and on. The questionnaire survey has been adapted to the study
systematically identified in the previous research and environment before being conducted. The 5-point scale has
references. Based on the critical payment risk factors, a been applied for assessment of the probability (1: rare; 2:
questionnaire survey is conducted not only to determine unlikely; 3: possible; 4: likely; 5: very likely), and the impact
which factors do occur to contractors but also to evaluate (1: very low; 2: low; 3: medium; 4: high; 5: very high) of
their important levels associated with the probability and officially identified payment risk factors to contractors. The
impact assessment. high-experienced experts are asked to rank the probability
and the impact on contractors of critical payment factors.
III. RESEARCH METHODOLOGY This questionnaire survey covers officially critical resources,
The qualitative and quantitive methods are used to which may occur during the project stages, divided into 6
identify the critical resources leading to the payment risks to groups consisting of key stakeholders’ acts; project
contractors. This study was carried out in 2 steps as shown in characteristics and environment constraints; contractual
Fig. 1. provisions; regulatory, force majeure; and economic
conditions that have been taken into account 21 factors (see
In step 1, the primary resources of contractors’ payment Table 1). The questionnaire survey was handed out using a
risks have been investigated by conducting a systematic convenience sampling approach.
review of literature, and payment terms and conditions of
construction contracts; and clarifying the Vietnam legal To provide reliable results, a sample is randomly
corridor for protection of payment to contractors. These collected from multiple construction projects in Vietnam
primary identified factors then further have been revised and megacities. Additionally, questionnaire respondents have
complemented based on a large number of in-depth been chosen on working experience, educational
interviews of high-experienced experts in both payments background, and their company designation as shown in
under a contract and payment disputes, and litigations in Table 2. More specially, the proportion of respondents
construction. As a result, the officially critical resources of having more than 10 years of working experience in the
payment risks to contractors are presented in Table 1. construction field is approximately 59%.
Table 1. The key payments risk resources to contractors
No Variables Critical factors Critical factor explanation

1 HV Acts of key stakeholders
- Owners do not obey the payment terms and conditions in the
construction contract because of their insolvency;
1.1 HV1 Acts of owners
- Owners deliberately prolong payment timing for the performed
works.
Acts of contractors - Inadequate contractors' experience and capacity;
1.2 HV2 (general contractors and
subcontractors) - Problems with the subcontractors.
- Inadequate consultants' experience and capacity;

Acts of consultant - Poor preparing the contractual document for a project;
1.3 HV3
organizations - Delays either in approval of claims by the clients or in response
to all the queries of contractors lead to any disruption to the works.
- Shortage in material and equipment;
- Delayed delivery to the construction site;
1.4 HV4 Acts of suppliers - Non-conforming material and equipment;
- Financial failure of the key suppliers;
- Supplier relationships are fraught and frustrating.
2 DR Project characteristics and environment constraints
Multi-parties having their own goals in different constraints
Disputes and conflicts
2.1 DR1 participate in a construction project and take complications in
between the parties
project management.
- The project complexity is the most important characteristic,
The project complexity which has a negative influence on the project performance;
2.2 DR2 regarding the technique - A construction project operated by new and more complex
and technology technology will have a high-complexity rating and a
correspondingly high risk.
Environmental
2.3 DR3 Precipitation/flood, unpredicted weather conditions, and pollution.
conditions
175
No Variables Critical factors Critical factor explanation

Seismic activity, the sub-surface unexpected, the archeological
2.4 DR4 Sub-surface conditions
survey was done, and the geotechnical investigation.
- Construction area, access conditions, on-site congestion, delay in
permits and licenses, security requirements, safety regulation, and
2.5 DR5 Site locations differing site conditions;
- Contractor’s poor site management.
3 HD Contractual provisions
Lack of thorough
3.1 HD1
negotiation preparation
The weakness of the
legal department or - Weak negotiation preparation;
even the common short - Contractors have not paid attention to establishing and
3.2 HD2
of the legal department developing their legal department;
in contractors’ corporate - Contractors' apathy with the possibility of payment risk before
structure signing the contract and during its performance.
Contractors often get
3.3 HD3 complacent and tend to
sign without reading
4 PL Regulatory
The risk of - The continuous and rapid changes in laws and regulations can
inconsistency and make most project stakeholders untimely understand;
4.1 PL1
continuous change in - The differences between old regulations and newly amended
regulations and laws ones without detailed implementation instruction documents while
publishing regulations are unable to keep up with requirements in
practical, which causes many difficulties when applying them;
Unreasonableness and
- There are no Vietnam legal control mechanisms promulgated for
4.2 PL2 lack of strictness of
checking the existence of fund resources, listed in the approved
regulations and laws
capital allocation plan;
- Taxation regulations are unclear and insufficient.
According to Clause 1, Article 156 of the Civil Code

91/2015/QH13 [22] and Clause 2 Article 51 of Decree
5 BKK Force majeure
37/2015/ND-CP, the force majeure event must be objective,
unforeseeable, and irremediable
5.1 BKK1 Act of God Sub-surface conditions, site conditions, weather conditions
Unforeseeable or uncontrollable changes to law and legislation,
Epidemics and
5.2 BBK2 such as construction sites were instructed by the government to
pandemics
close at the start of the coronavirus.
Social, political, or
Riots, rebellions, wars, invasions, industrial disputes, terrorism,
5.3 BBK3 other individual or
strikes, and so on.
group actions
6 DK Economic conditions
Monetary and fiscal

6.1 DK1 - Monetary and fiscal policy could lead to changing interest rates,
policy
government spending, and tax rates, which can raise materials,
labor, equipment, etc;
6.2 DK2 Exchange rates
- Because of long project duration, exchange rates, inflation, and
the negative state of the global economy can raise the product cost.
6.3 DK3 Inflation
- Much time is consumed for approval procedures of costs arising
from procurement of all resources that will be charged to the
The state of the global project.
6.4 DK4
economy
176
Table 2. Participant demographics conditions under a construction contract. A fine ranging from
80 million VND to 100 million VND for violations of the
Characteristics Categorization Percentage contractual terms and conditions promulgated in Point c of
5 – 10 years 41.94% Clause 2 of Article 19 of Decree No. 16/2022/ND-CP [24] is
Working not enough threat to prevent all parties from their breakdown
10 – 15 years 25.81% of contractual payment agreement. The insolvency makes
experience
More than 15 years 32.26% owners fail their payment obligation under the contractual
agreement and causes contractors to get financial problems
Graduate 54.84%
in business. Despite the owner’s assurance of adequate funds
Education
Postgraduate 41.94% for payment, which has been mentioned in the approved
background
project appraisal before signing the construction contract, it
Others 3.23%
does not guarantee that contractors will be paid on time and
Owner/ Project in full value. Seriously, the owners deliberately prolong
29.03%
management board payment timing for the performed works to contractors or
Consultant even intentionally push contractors into difficulties in
Designation of 22.58% receiving prompt payment even though he has strong
participants' organizations
solvency.
company Contractors 41.94%
Furthermore, with the accounted mean scores of 15.45
Government 3.23% and 14.81, respectively, HD1 and HD2 are the first and
Others 3.23% second resources, in both contractual provision risk group
and total 21-factor ranking, appropriately. Lack of thorough
To obtain a purpose on identifying the critical payment negotiation preparation might not produce a quality contract
resources to contractors, the collected data is used to rank the and not set the foundation for timely and effective contract
importance based on their mean score of risk index. execution. Contractors lack knowledge, or misunderstand to
apply the laws to their business or do their business by
IV. THE RESULTS AND DISCUSSION ignoring the law, due to the common shortage of the legal
The ranking of 6-compared groups and the critical department. Until any dispute has arisen, contractors, even
resources have been illustrated in Table 3 and Table 4 by though, never read the contractual terms and conditions,
using their mean scores. Out of 6 compared groups to which could have unreasonable and unclear provisions
contractors, key stakeholders’ acts (owner, contractors, deliberately set by either owners or general contractors, such
consultants, and suppliers) having a mean score of 14.97 is as the use of the ‘pay when paid’ clause. Che Munaaim,
the most significant group; the contractual provisions (the Mohd Danuri, et al. [15] found the use of the ‘pay when
weakness of the legal department or even the common short paid’ clause in a contract is also the main factor for late and
of the legal department in contractors’ corporate structure, non-payment in the Malaysian construction industry.
contractors often get complacent and tend to sign without According to Point 8, Clause 8, Article 146 of Construction
reading, and contractors often get complacent and tend to law No. 50/2014/QH13 and Article 45 of Decree No.
sign without reading) is 14.65, the large group leading 37/2015/ND-CP, parties are able to negotiate for disputes
contractors to the payment risks. By contrast, force majeure, settlement by themselves, and their disputes are going to be
and economic conditions, which accounted for 11 and 10.13, settled in order through Dispute Adjudication Board,
respectively, have taken the lowest positions in the ranking Arbitration, or Court under legal regulations in case of own
outcome at mean scores. unsuccessful negotiation [25]. Any dispute costs all parties a
lot of money which bears litigation expenses and wastes their
The findings in this study differ from the previous other own resources.
research. Mbachu [17] and Mbachu and Nkado [23] found
the owner to be the greatest contributor to payment risks Finally, the pandemics and ongoing armed conflict are
faced by contractors and contractors to be the second one. not significant resources that cause contractors to the
On the contrary, in the riskiest group, this study has found payment risks despite their severe impact. The unforeseeable
that contractors are the riskiest sources compared to owners, characteristic may no longer exist because the Covid-19
the second factor evaluating high risky, resources to pandemic began in early 2020 and has lasted for nearly two
contractors’ payment (see Table 4). As for the total critical years; contractual parties have taken actions and remedial
resource ranking, in which these 2 factors get mean scores measures to overcome. As a result, the Covid-19 pandemic is
14.65 and 14.06, appropriately, take the third and fourth. not considered as a critical force majeure because it has been
Almost all contractors, performing themselves may cause predicted and remedied. Data, which was accumulated as of
payment risk due to their unacceptable performance relating February 2022 by the Ministry of Planning and Investment
to their project management such as violation of time frame, of the Socialist Republic of Vietnam, revealed that the
specified quality, and other requirements for performed Russia-Ukraine crisis has little impact on the Vietnam
works under a contract. Contractors’ inadequate experience construction investment because these two markets’ direct
and capacity have been known as the main resources that investment only accounts for 0.23% of total foreign direct
make contractors face this consequence. In addition, investment into Vietnam [26]. Moreover, the questionnaire
problems with subcontractors, and the invalid and survey has been conducted in early March 2022 when the
unsupported payment claims with well-documented evidence Russia-Ukraine conflict has just started.
are also the payment risk resources. The Vietnam legal
regulations for the contractors' payment protection have not
been followed by contractual parties due to a deficiency of
realistic sanctions to make them obey the payment terms and
177
Table 3. Ranking the compared groups of the payment risk to V. CONCLUSION

contractors
Based on both qualitative and quantitative approaches,
this study investigates critical resources causing the payment
Code Variables Mean Ranking
risks to contractors participating Vietnam construction
HV Acts of key stakeholders 14.97 1 industry. The empirical result shows that 6 groups, including
acts of key stakeholders, project characteristics and
Project characteristics and environment constraints, contractual provisions, regulatory,
DR 12.39 3
environment constraints force majeures, and economic conditions, are identified as
HD Contractual provisions 14.65 2 the critical resources causing the payment risk to contractors.
Comparing the 6 groups, this study finds that key
PL Regulatory 11.58 4 stakeholders’ acts and contractual provisions as the first and
BKK Force majeure 11 5 second most common groups cause contractors to face
payment risks. In which, lack of thorough negotiation
DK Economic conditions 10.13 6 preparation, the weakness of the legal department or even the
Table 4. Ranking the critical resources of the payment risk to common short of the legal department in contractors’
contractors corporate structure, and contractors’ and owners’ acts have
emerged as the most serious factors leading contractors to
Ranking payment risks. To ensure contractors could avoid the
Ranking
Code Variables Mean (in each payment risk, several recommendations are produced. In the
(in total)
group)
case of using the contractual standard form, contractors
HV1 Acts of owners 14.06 2 4 should thoroughly read its provisions, fix any identified
Acts of contractors critical errors by writing in more detail, clearly, and oriented.
HV2 (general contractors and 14.65 1 3 Furthermore, contractors should thoroughly prepare for
subcontractors)
negotiation to remove negative terms and conditions in a
Acts of consultant construction contract. To avoid the uncertainties and delays
HV3 13.68 3 5
organizations
involved in relying on the applicable law, contractual parties
HV4 Acts of suppliers 12.23 4 8 could provide a specific regime for force majeure, along with
DR1
Disputes and conflicts
12.87 1 7 a definition of which events qualified for special treatment.
between the parties
The project complexity
The identified sources causing contractors’ payment risks
DR2 regarding the technique 11.06 2 10 on the construction project in Vietnam support contractors in
and technology composing the initial stage of contract negotiation.
DR3 Environmental conditions 10.45 3 15 Moreover, the study findings are significant in practice and
DR4 Sub-surface conditions 9.81 4 18
supply a background for further research on securing
contractors’ payments. Additionally, these findings would
DR5 Site locations 9.65 5 19
help contractors to create the talent strategies to achieve
Lack of thorough desired business results in line with available abilities.
HD1 15.45 1 1
negotiation preparation
Contractors could miss their expected profit on a contract or
The weakness of the legal even go bankrupt if they cannot identify payment risks and
department or even the
HD2 common short of the legal 14.81 2 2
control them without payment risk management. Thus,
department in contractors’ contractors should take the most likely effective strategy or
corporate structure strategy mixture for their payment protection in the event of
Contractors often get the payment terms and conditions violations by owners or
HD3 complacent and tend to 13.61 3 6 general contractors. Having an effective plan can help
sign without reading contractors proactive rather than reactive on payment risk.
The risk of inconsistency
PL1 and continuous change in 10.84 1 12 REFERENCES
regulations and laws [1] T. E. Uher and M. Loosemore, "Essentials of construction
Unreasonableness and project management", Sidney: University of New South
PL2 lack of strictness of 10.42 2 16 Wales Press, 2004.
regulations and law [2] J. D. Sterman, System dynamics modelling for project
BKK1 Act of God 10.48 3 14 management: School of Management, Massachusetts
BBK2 Epidemics and pandemics 11.45 1 9 Institute of Technology, 2012.
Social, political, or other [3] S. M. Renuka, C. Umarani, and S. Kamal, "A Review on
BBK3 individual or group 10.81 2 13 Critical Risk Factors in the Life Cycle of Construction
actions Projects", Journal of Civil Engineering Research, vol. 4, pp.
Monetary and fiscal 331-36, 2014.
DK1
policy
9.1 3 20 [4] B. Akincl and M. Fischer, "Factors affecting contractors’ risk
of cost overburden", ASCE Journal of Management in
DK2 Exchange rates 8.77 4 21
Engineering, vol. 4, 1998.
DK3 Inflation 11 1 11 [5] I. Jose, "Project Risk Management Tools and Their
The state of the global Effectiveness", presented at the 4th DBA & DMC Annual
DK4 10.26 2 17
economy Doctoral Colloquium, Berlin, Berlin, 2014.
[6] C. V. Bac. (2019). Dispute resolution for EPC contract by
commercial arbitration and some recommendations on use of
mediation. Efficient dispute resolution for EPC contract
178
disputes (Unclog the stagnation in projects of significance in [16] S. A. Assaf and S. Al-Hejji, "Causes of Delay in Large
Vietnam). Available: https://www.adr.com.vn/vi/tin-tuc/giai- Construction Projects", International Journal of Project
quyet-hieu-qua-tranh-chap-trong-cac-hop-dong-tong-thau- Management, vol. 24, pp. 349-357, 2006.
epc-khoi-thong-tac-nghen-tai-cac-du-an-trong-diem-cua- [17] J. Mbachu, "Sources of contractor’s payment risks and cash
viet-nam-tp-hcm-1942019. flow problems in the New Zealand construction industry:
[7] K. D. Vo, P. T. Nguyen, and L. H. T. T. Nguyen, "Disputes project team’s perceptions of the risks and mitigation
in Managing Projects: A Case Study of Construction measures", Construction Management and Economics, vol.
Industry in Vietnam", The Journal of Asian Finance, 29, pp. 1027-1041, 2011.
Economics and Business, vol. 7, pp. 635-644, 2020. [18] R. Pettigrew, The legal and contractual background. In:
[8] H. Wenfa, C. Jianguo, R. Zhaomin, and W. George, Payment under construction contracts legislation: Heron
"Identification and control of payment risks in international Quay Thomas Telford, 2005.
construction projects", presented at the International [19] T. Ramachandra and J. O. B. Rotimi, "Causes of Payment
Conference on Multi-national Construction Project, Problems in the New Zealand Construction Industry",
Shanghai, China, 2008. Construction Economics and Building, vol. 15, pp. 43-55,
[9] V. T. T. Duong and T. T. Thiem, "Vietnam legal corridor for 2015.
payment to contractor", Journal of transportation science [20] K. Lam, "Modelling Risk Allocation Decision in
and technology, vol. 11, issue 1, pp.56 - 62, 2022. Construction Contracts", International Journal of Project
[10] M. Latham, "Constructing the Team, Final Report of the Management, vol. 25, pp. 485-493, 2007.
Government/Industry Review of Procurement and [21] S. Hansen, "Challenging arbitral awards in the construction
Contractual Arrangements in the UK Construction Industry", industry: Case study of infrastructure disputes", Journal of
ed. HMSO, London, 1994. Legal Affairs and Dispute Resolution in Engineering and
[11] C. Semple, F. T. Hartman, and G. Jergeas, "Construction Construction, vol. 11, 2019.
claims and disputes: Causes and cost/time overruns", Journal [22] T. N. A. o. t. S. R. o. Vietnam, "The civil code No.
of Construction Engineering and Management, vol. 120, pp. 91/2015/QH13", ed. Ha Noi: 13th National Assembly, 2015.
785-795, 1994. [23] J. Mbachu and R. N. Nkado, "Conceptual framework for
[12] M. M. Kumaraswamy, "Conflicts, claims and disputes in assessment of client needs and satisfaction in the building
construction", Engineering Construction and Architectural development process", Construction Management and
Management, vol. 4, pp. 95-111, 1997. Economics, vol. 24, pp. 31-44, 2006.
[13] E. E. Henriod and J. M. Lantran. (2000, 20 July). Trends in [24] T. V. Government, "Decree No. 16/2022/ND-CP detailed
contracting practice for civil works. Available: penalties for administrative violations in construction", ed.
http://siteresources.worldbank.org/INTROADSHIGHWAYS Ha Noi: The Vietnamese Government, 2022.
/Resources/3389931122496826968/cn_ntk2b.pdf [25] T. N. A. o. t. S. R. o. Vietnam, "Construction law No.
[14] S. O. Cheung and H. C. Suen, "A multi-attribute utility 50/2014/QH13", ed. Ha Noi: The National Assembly of the
model for dispute resolution strategy selection", Socialist Republic of Vietnam, 2014.
Construction Management & Economics, vol. 20, pp. 557- [26] The Ministry of Planning and Investment of the Socialist
568, 2002. Republic of Vietnam, "Report on foreign direct investment in
[15] M. E. Che Munaaim, M. S. Mohd Danuri, and H. Abdul- the first 2 months of 2022", ed: The Ministry of Planning and
Rahman, "Is late or non-payment a significant problem to Investment of the Socialist Republic of Vietnam, 2022.
Malaysian contractors?", Journal of Design and the Built
Environment, vol. 3, pp. 35-50, 2007.
179
Pharmacognostic Assessment of Polyscias

Fruticosa Leaves in Vietnam
Dao Phan Thi Anh Hue Ha Thi Trang Le Vu Khanh
Ho Chi Minh City University of Ho Chi Minh City University of University of Danang – University
Technology and Education Technology and Education of Science and Education
Ho Chi Minh City, Viet Nam Ho Chi Minh City, Viet Nam Danang City, Viet Nam
daopta@hcmute.edu.vn hueht@hcmute.edu.vn lvktrang@ued.udn.vn
Thanh Le Duc Huong Nguyen Thi Thu Trieu Ly Hai

Research Center of Ginseng and Research Center of Ginseng and Research Center of Ginseng and
Medicinal Materials Ho Chi Minh Medicinal Materials Ho Chi Minh Medicinal Materials Ho Chi Minh
City City City
Ho Chi Minh City, Viet Nam Ho Chi Minh City, Viet Nam Ho Chi Minh City, Viet Nam
* *
lethanh08cnsh@gmail.com huongsam@hotmail.com lhtrieu12csh@gmail.com
Abstract: Polyscias fruticosa leaves have been this system of medicine. The standardized quality control
ubiquitously used to treat various diseases in traditional profile of genuine herbal material comprises pharmacognostic
medicine. This study has tested the quality of P. fruticosa standardization physicochemical and phytochemical
leaves and their extract to prepare different dosage forms in evaluation. Macroscopic and microscopic characters,
traditional medicine. Organoleptic and microscopical physicochemical studies, qualitative and quantitative
features, parameters for pharmacognostic assessment, chemical evaluation, and chromatography of plant materials
including loss on drying, ash values, microbial load, contribute to ascertaining the degree of purity of such
aflatoxins, heavy metal contents, and major chemical material, which will add to their safety and efficacy [3].
constituents were examined in accordance with the guidelines Polyscias fruticosa (L.) Harms is a plant species included
given by the Pharmacopoeia Vietnamica and World Health in the Araliaceae family, one of the most prominent families
Organization. The colour, odour, and taste characteristics of the plant kingdom, being widely used in traditional and
were noted for powdered material and extract of P. fruticosa modern phytotherapy. P. fruticosa has been used as a tonic
leaves. Microscope images, leaf cross-section, and leaf agent based on various pharmacological effects similar to
powder characteristics disclosed good diagnostic properties. some Panax species, such as increasing toughness and
Physicochemical, microbial load, aflatoxin, and heavy metal stimulation of immune system activity, antitoxin, improving
analysis were also achieved. The presence of important fertility, preventing fatigue, increasing work capacity, eating
phytoconstituents like oleanolic acid, which proved valuable and sleeping well [4]. Extracts of P. fruticosa have been
evidence to differentiate drug materials. This study provided reported to contain various potential bioactive compounds
referential information for accurate identification and quality such as saponins, alkaloids, glycosides, polyphenols,
flavonoids, and vitamins (C, B1, B2, B6) which contribute
control of herbal materials from Polyscias fruticosa leaves.
importantly to extensive effects including antidepressant, anti-
stress, improved memory, antioxidant, hepatoprotective,
Keywords: polyscias fruticosa leaves, pharmacognostic
hypolipidemic, and antibacterial effects [5][6][7]. P. fruticosa
evaluation, quality control
has been shown to be a vital source of triterpene saponins
I. INTRODUCTION which were reported to have cytotoxic activity as well as
antibacterial, antifungal, molluscicidal, and wound healing
The usage of herbal drugs has increased with the activities [8][9].
worldwide trend of folks returning to natural remedies.
According to World Health Organization (WHO), above 80% Despite having great medicinal importance and various
of the worldwide population, especially in developing dosage forms prepared from P. fruticosa leaves, research on
countries, understand and utilize traditional medicines for the standardization parameters of P. fruticosa leaves has still
their primary healthcare [1]. WHO has supported, suggested, been limited. Hence, the present work can be an attempt to
and encouraged traditional/herbal remedies development in provide a comprehensive report on the quality control
disease prevention strategies based on the abundant measures and standardization parameters of P. fruticosa
biodiversity of medicinal plants, the relatively low cost, and leaves to consolidate the quality, safety, and effectiveness of
the safety of using plant-derived medicines [2]. Therefore, this medicinal material.
evaluating the health effects as well as establishing
specifications of quality and manufacture of herbal botanicals II. MATERIALS AND METHODS
has been receiving much attention. One of crucial importance A. Plant materials
is implementing documentation and standardization of raw
The leaves of Polyscias fruticosa (L.) Harms were
materials used in herbal medicine for the global compliance of
collected in May 2019 from Nghia Hung District (20°12′N

106°12′E), Nam Dinh Province, Vietnam. Leaves were Thin layer chromatography analysis (TLC finger
randomized without distinguishing between young or old printing): The extract of the leaf powder, the crude extract
leaves on mature plants about 3 years old. These dried leaves solution, and oleanolic acid were made ready for TLC
were ground to a fine powder and passed through a sieve 2 deployment. The emerging color bands were observed,
mm. photographed, and the retention factor (Rf) was determined.
B. Chemicals and reagents K. Quantification of saponins by high-performance liquid
Some major chemicals and reagents were used in this chromatography (HPLC finger printing)
study, such as ethanol (98% v/v) purchased from OPC HPLC fingerprinting was applied to estimate total saponin
Pharmaceutical Company, methanol (HPLC ≥ 99.9%), content using oleanolic acid as a standard with several slight
acetonitrile (HPLC ≥ 99.9%), oleanolic acid (HPLC ≥ 97%) modifications [14]. The total saponin content was calculated
supplied by Sigma-Aldrich® Co. Ltd (USA), and some other and expressed as a percentage of oleanolic acid equivalent per
necessary chemicals and reagents. dried weight through the regression equation was calculated
C. Plant extraction in the form of y = 7509.8x + 4352.2, R2 = 0.9978.
The leaf powder (1.5 kg) was extracted with 45% ethanol L. Data analysis
(v/v) at room temperature for 24 hours in percolator apparatus Experiments were repeated triplicate, and the obtained
until the ratio of material to solvent was 1: 15 (w/v). The liquid results were expressed in terms of mean ± SEM/SD (Standard
extract was collected at a 2 mL/min rate and evaporated using error of the mean/Standard deviation) using MS Excel 2019
a rotary evaporator (60°C, decreased pressure) to get total software.
extract for research. Extraction yield (H%) was calculated by
formula: III. RESULTS AND DISCUSSION
H(%) = ([m × (1 – A)]/[mo × (1 – Ao)]) × 100 (1) M. Organoleptic and microscopic characteristics
In which Ao, mo and A, m were moisture, the weight of The dried powder and crude extract of P. fruticosa leaves
leaf powder and extract (g), respectively. were evaluated for organoleptic characteristics presented in
Table 1.
D. Organoleptic evaluation
Dried leaf powder and leaf extract were observed for Table 1. Some organoleptic features of P. Fruticosa leave
colour, odour, and taste according to the Pharmacopoeia
Observation
Vietnamica 5th Edition. No. 1Characters
Leaf powder Crude extract
E. Microscopic analysis
1 Form Powder Concentrated
Anatomy: Morphological and material powder
composition was observed by a light microscope (Olympus, 2 Colour Dark brown Dark brown
Japan). 3 Odour Characteristic smell Characteristic aroma
F. Loss on drying, total ash, acid-insoluble ash 4 Taste Bitter, slightly acrid More bitter, acrider
Three parameters were determined following the
guidelines given by WHO [10][11]. Microscopy analysis of P. fruticosa leaves was carried out
to identify the composition; observed powdered microscopic
G. Microbial load test characters are shown in Figures 1-3.
The microbial limit study was conducted according to the
standard method described in WHO, ISO, and Pharmacopoeia
Vietnamica V guidelines [3][10]. It integrated total yeast
count, mould count, aerobic microorganisms, and pathogens
like Coliforms, Escherichia coli, Staphylococcus aureus, and
Salmonella spp.
H. Aflatoxins determination
The developed methodology of association of analytical
chemistry official process of the experiment was followed to
assess aflatoxins (aflatoxin B1, aflatoxin B2, aflatoxin G1, and
aflatoxin G2) [12].
I. Heavy metal analysis
Heavy metals including Cadmium (Cd), Arsenic (As), Figure 1. Anatomical features of petiole of P. fruticosa leaf. (A) The
overall image of the transverse section of petiole (×4); (B) 1 – Epidermis, 2
Lead (Pb), Mercury (Hg), Iron (Fe), Copper (Cu), and Zinc – Cortex collenchyma, 3 – Cortex parenchyma (×10); (C) 1 –
(Zn) were analyzed using the ICP-MS technique [13]. Sclerenchymatous endodermis, 2 – Phloem, 3 – Xylem (×10); (D) 1 – Pith
parenchyma, 2 –Secretory tubule (×10); (E) Spherical thorn calcium
J. Qualitative determination of saponins oxalate crystals (×40); (F) Essential oil seeds (×40).
Chemical reactions: The liquid extract of leaf powder and
solution of the crude extract were tested for the presence of
saponins by the following chemical reactions: Foam test and
Liebermann-Burchard’s test.
181
According to Pharmacopoeia Vietnamica V, the specific

limit of Coliforms, total yeast count, total mould count, and
total aerobic microorganisms in plant extracts are 10, 10, 10,
100 CFU/g, respectively, while E. coli, S. aureus, and
Salmonella spp. are not detected [10]. The microbial load
profile demonstrated that pathogenic bacteria, including E.
coli, S. aureus, and Salmonella spp. in the extracted sample
were not present. Others such as Coliforms, total yeast count,
mould count, and aerobic microorganisms were considered
acceptable (under the detection limit of 0.1 CFU/g).
Figure 2. Anatomical features of the leaf blade of P. fruticosa. (A) The
overall image of the transverse section of leaf blade (×4); (B) Superior
Aflatoxins are poisonous secondary metabolites
tendon (×10); (C) Inferior tendon (×10): 1 – Epidermis, 2 – Collenchyma, 3
commonly formed by Aspergillus flavus and A. parasiticus
– Parenchyma, 4 – Phloem, 5 – Xylem, 6 – Secretory tubule; (D) Leaf [15]. Four out of different types of aflatoxins are considered
blade (×10); (E) Spherical thorn calcium oxalate crystals (×40); (F) important and designated as B1, B2, G1, and G2. The
Anomocytic stomata with syncytial cells (×40). International Agency for Research on Cancer (IARC)
classified aflatoxin B1 as class I human carcinogen [16], the
order of toxicity of aflatoxins is B1 > B2 > G1 > G2 [17][18].
According to Pharmacopoeia Vietnamica V, the specific limit
of aflatoxins B1, B2, G1, and G2 in medicinal plants are 2, 4, 4,
and 4 µg/kg, respectively [10]. Analytical results of this study
showed that aflatoxins were not detected in the raw materials
and crude extract of P. fruticosa leaves (under the detection
limit of 0.1 µg/kg). It indicated that P. fruticosa leaf material
and extract were safe to use in their raw form or further into
formulations.
Figure 3. Powder microscopy of P. fruticosa leaf (×40). (A) The epidermal Heavy metal toxicity related to traditional medicines has
fragment containing chloroplasts and stomata; (B) The parenchyma been reported worldwide. Levels of heavy metals in soil and
consisting of polygonal cells of different sizes and containing essential oil agricultural irrigation wastewater could be conglomerated in
particles; (C) Spiral vessel; (D) Annular vessel; (E) Bundle of yarn; (F) medicinal plants. Due to their non-biodegradable properties,
Reticulate vessel; (G) Spherical thorn calcium oxalate crystals. possibility of toxicity to human health may be hazardous or
N. Investigation on quality control parameters reduce bioavailability. The existence of heavy metals can
The physiochemical parameters of raw material and crude interact with essential trace elements with serious health
extract from P. fruticosa leaves were shown in Table 2. outcomes. As specified by WHO, the determined limit of
heavy metals Pb, Cd, As, Hg, Cu, Fe, and Zn in medicinal
Table 2. Some parameters for quality control of material and
extract from P. Fruticosa leaves plants and food are 10, 0.3, 5, 0.5, 20, 20, and 50 mg/kg,
Parameters Leaf powder Leaf extract
respectively [19][20]. After comparison, the presence of
Extraction yield (%) NP 32.50
heavy metals, mainly cadmium (Cd), lead (Pb), arsenic (As),
mercury (Hg), Copper (Cu), and Zinc (Zn) in P. fruticosa leaf
Moisture content (%) 6.85 ± 0.16 15.91 ± 0.13
powder and the extract was observed within the acceptable
Total ash (%) 14.29 ± 0.05 19.61 ± 0.23
limits. Hg content was below the detectable range. The leaf
Acid insoluble ash (%) 0.36 ± 0.06 NP powder's iron (Fe) level was also lower than WHO's
Microbial load (CFU/g) prescribed. However, Fe concentration in the leaf extract was
Coliforms NP <10 higher than the specified limit of medicinal plants.
Escherichia coli NP ND
Previous studies showed that saponins were the main
Staphylococus aureus NP ND constituents of P. fruticosa [14]. In this study, the results
Salmonella spp. NP ND observed from chemical reactions indicated that both leaf
Total yeast count NP <10 powder and extract recorded the presence of saponins (Table
Total mold count NP <10 3). Therefore, these test solutions were further analyzed for
Total aerobic microorganisms NP 70 the existence of saponins by TLC analysis
Aflatoxins (µg/kg) Table 3. The qualitative results of saponins in P. Fruticosa
Aflatoxin B1 ND ND
Leaves
Aflatoxin B2 ND ND
Results
Aflatoxin G1 ND ND Chemical
Solution of leaf Solution of leaf Finding
Aflatoxin G2 ND ND reaction
powder extract
Heavy metals (mg/kg)
Foam test Appeared-foam Appeared-foam +
Cd 0.11 0.02
As 0.21 0.51 Formation of a Formation of a
Liebermann-
violet-colored violet-colored +
Pb 1.10 1.93 Burchard’s test
ring ring
Hg ND ND
(+): the presence of saponins
Fe 6.05 27.91
Cu 7.65 12.09 The successfully developed TLC plate generated different
Zn 9.96 27.03 purple bands under visible light and positions for saponins.
NP: not performed, ND: not detected/below detectable concentrations. The TLC profile indicated the presence of various saponins
182
within the P. fruticosa leaves (Figure 4A, B). The Rf values of having an aroma odor and a more bitter taste. To establish
saponins in solvent systems were shown in Table 4. standard quality control for raw material and extract of P.
Furthermore, the results have shown in Figure 4C, D fruticosa leaves, the pharmacognostic investigations of
illustrated the same colour and Rf value between the oleanolic various parameters have been evaluated as mentioned in
acid standard and the test samples. By HPLC method using Pharmacopoeia Vietnamica V and WHO guidelines. These
oleanolic acid standard (Figure 5), total saponin content of parameters are crucial for detecting drug forgery or improper
material powder and crude extract was determined as 0.98% processing of raw materials [3].
± 0.05 and 2.34% ± 0.01 (Mean ± SD), respectively.
The moisture content of raw medicinal plants is directly
related to the appearance and proliferation of microorganisms.
The lower the moisture level, the higher the stability of the
raw plant medicine, and the drugs’ shelf life also increases
[22]. Considerable values of moisture were found in both the
powder and crude extract P. fruticosa leaves, which met the
standards prescribed by the Pharmacopoeia Vietnamica V
(<13% for raw material and <20% for crude extract), thus the
formulation could be preserved longer and would not easily
be attacked by bacteria, fungi or yeast. Determination of ash
value gives an idea to assess the purity of medicinal plants.
The ash content, the residue remaining after ignition of herbal
Figure 4. The thin-layer chromatography (TLC) pattern of saponins and
oleanolic acid of P. fruticosa leaves. A to D correspond to solvent systems
drug part, basically symbolizes inorganic salts, naturally
of n-butanol: Acetic acid: Water (7: 1: 2 v/v/v), Ethyl acetate: Methanol: occurring in crude herbal drug or remaining to it or
Water (100: 17: 10 v/v/v), Toluene: Ethyl acetate (7: 3 v/v), intentionally added to it, as a part of adulteration. In this study,
Dichloromethane: Methanol (95: 5 v/v). A and B figures indicate the the ash value was observed by total ash and acid insoluble ash,
presence of saponins in P. fruticosa leaves, C and D figures indicate the which both were within acceptable limits of raw materials
presence of oleanolic acid in P. fruticosa leaves. 1, 2, and OA indicate
solutions of leaf powder, leaf extract, and oleanolic acid standard.
(less than 15%) and crude extracts (less than 35%) from
medicinal plants [10].
Table 4. The Rf values on TLC profiling of saponins and oleanolic
acid from P. Fruticosa leaves
With the herbal plants, preparations such as powder
material and crude extract have generally been utilized as
Solvent system Number of spots and their Rf value traditional and alternative therapy treatments. There would be
forever a possibility of microbial load, aflatoxin, and heavy
A 0.17, 0.20, 0.27, 0.33, 0.38, 0.43, 0.52
metal contamination of these crude drugs, which could
B 0.07, 0.17, 0.21, 0.47 produce side effects, reduce efficacy, and even harm human
C 0.60 health. Hence, it is crucial to carry on quality control before
D 0.47 using or manufacturing the natural products to sustain the
suitable quality, safety, and efficacy of final drugs [23].
Raw plant materials could easily be supplanted or Microbial load profile should be under control, aflatoxins
counterfeited; the exact identification and the quality should not be detected or below the detection threshold, while
assurance of the raw material are essential, which ultimately the occurrence of major heavy metals (Cd, As, Pb, Hg)
contribute to the safety and efficacy of the herbal drugs [21]. contamination should be within the acceptable limit.
In the present study, the organoleptic characteristics of However, the Fe level in the investigated extract was above
material powder and extract from P. fruticosa leaves were WHO's permissible limits in medicinal plants. In general,
described, and anatomical features of the petiole and leaf medicinal plants contain unsafe levels of iron. Iron could be
blade of P. fruticosa leaf were also analyzed. These data could toxic depending on the dose and duration of exposure. Fe daily
be helpful to identify and degree of purity of herbal materials. intake is 10-28 mg/day by WHO [20]. Fe level of 27.91 mg/kg
The dried leaf powder of P. fruticosa is dark brown and has extract was detected, it can be safe to use this extract as one
characteristics of smell, bitter taste, and slightly acrid. In person also cannot take 1 kg of extract in a day. Iron also has
contrast, the crude leaf extract is concentrated, dark brown, some key functions in the human body like oxygen supply,
energy production and immunity. Fe deficiency may cause
anemia.
In the pharmacognostic standardization or testing of
medicinal herbs, the qualitative and quantitative
determination of phytochemicals, the main groups of
compounds, the main compounds are the necessary
requirements for quality control of the input materials or
detection of counterfeiting. In the present study, saponins and
oleanolic acid were identified in leaf powder and leaf extract
of P. fruticosa through chemical reactions, thin layer
chromatography (TLC) and high-performance liquid
Figure 5. High-performance liquid chromatograms of extracts from P. chromatography (HPLC) fingerprinting. These methods allow
fruticosa leaves compared to oleanolic acid standard. The retention time rapid detection of the presence or absence of a group of
of oleanolic acid standard, LE, LP peak was at 9.7 min. LP, LE, and STD substances/a compounds in plants [24]. Because the P.
are solutions of leaf powder, the leaf extract of P. fruticosa, and oleanolic
acid standard, respectively.
fruticosa leaves contain many compounds containing
oleanolic acid components, the samples from P. fruticosa
183
leaves have been hydrolyzed in acid to release oleanolic acid. [11] WHO, Quality control methods for medicinal plant materials,
Therefore, the total saponin content was quantified instead of World Health Organization, 1998.
an individual component, oleanolic acid. This quantitative [12] M. W. Trucksess, C. M. Weaver, C. J. Oles, F. S. Fry, G. O.
Noonan, J. M. Betz, and J. I. Rader, “Determination of
result also provides reference data for determining the quality aflatoxins B1, B2, G1, and G2 and ochratoxin A in ginseng and
of the P. fruticosa leaf materials in the future. ginger by multitoxin immunoaffinity column cleanup and
liquid chromatographic quantitation: collaborative study”, J.
IV. CONCLUSION AOAC Int., vol. 91, no. 3, pp. 511–523, 2008.
The analytical results of this study provided an inclusive [13] S. H. Choi, J. Y. Kim, E. M., Choi, M. Y. Lee, J. Y. Yang, G.
H. Lee, K. S. Kim, J. S. Jang, R. E. Russo, J. H. Yoo, G. J.
pharmacognostic profile of P. fruticosa leaves and thereby Kang, and K. S. Park, “Heavy metal determination by
will be helpful for correct identification and authentication of inductively coupled plasma – mass spectrometry (ICP-MS) and
the species for future studies. TLC and HPLC analysis may direct mercury analysis (DMA) and arsenic mapping by
serve as valuable data for the quality control of the raw femtosecond (fs) – laser ablation (LA) ICP-MS in Cereals”,
material and extract from P. fruticosa leaves. The results of Anal. Lett., vol. 52, no. 3, pp. 496–510, 2019.
this study could be employed as appropriate quality control [14] V. T. Tran, T. H. H. Tran, T. D. Nguyen, “Validated high
performance liquid chromatography method for quantification
measurements to make sure the quality, safety, and of a major saponin in Polyscias fruticosa”, J. Multidiscip. Eng.
effectiveness of this medicinal ingredient. Sci. Technol., vol. 3, no. 5, pp. 4880–4882, 2016.
[15] R. Bhat, R. V. Rai, and A.A. Karim, “Mycotoxins in food and
ACKNOWLEDGMENT feed: Present status and future concerns”, Compr. Rev. Food.
The authors express the gratitude to the National Institute Sci. Food Saf., vol. 9, no. 1, pp. 57–81, 2010.
of Medicinal Materials, Vietnam for the financial support [16] International Agency for Research on Cancer, “Some tradional
under grant number 16/2019/HĐ-NVTX-TTS. herbal medicines, some mycotoxins, naphthalene and styrene,”
in: Monograph on the Evaluation of Carcinogenic Risk to
Humans, vol. 82, 2002, Lyon, France.
REFERENCES
[17] M. Quinto, G. Spadaccino, C. Palermo, and D. Centonze,
[1] M. Ekor, “The growing use of herbal medicines: issues relating “Determination of aflatoxins in cereal flours by solid-phase
to adverse reactions and challenges in monitoring safety”, microextraction coupled with liquid chromatography and post-
Front. Pharmacol., vol. 4, p. 177, 2014. column photochemical derivatization-fluorescence detection”,
[2] A. Sofowora, E. Ogunbodede, and A. Onayade. “The role and J. Chromatogr., vol. 1216, no. 49, pp. 8636–8641, 2009.
place of medicinal plants in the strategies for disease [18] S. S. Alwakeel, “The effect of Mycotoxins found in some
prevention”, African J. Tradit. Complement. Altern. Med., vol. herbal plants on biochemical parameters in blood of female
10, no. 5, pp. 210–229, 2013. Albino mice”, Pakistan J. Biol. Sci., vol. 12, no. 8, pp. 637–642,
[3] WHO, Quality assurance of pharmaceuticals: a compendium of 2009.
guidelines and related materials, 2nd ed., vol. 2. Good [19] M. Ayaz, M. Junaid, F. Subhan, F. Ullah, A. Sadiq, S. Ahmad,
manufacturing practices and inspection. World Health M. Imran, Z. Kamal, S. Hussain, and S. M. Shah, “Heavy
Organization, 2007. metals analysis, phytochemical, phytotoxic and anthelmintic
[4] A. Boye, A. OseiOwusu, G. Koffuor, V. Barku, E. Asiamah, investigations of crude methanolic extract, subsequent
and E. Asante, “Assessment of Polyscias fruticosa (L.) Harm fractions and crude saponins from Polygonum hydropiper L.”,
(Araliaceae) leaf extract on male fertility in male Wistar rats”, BMC Complement Altern. Med., vol. 14, no. 1, 465, 2014.
J. Intercult. Ethnopharmacol., vol. 7, no. 1, pp. 45–56, 2018. [20] Joint FAO/WHO Expert Committee on Food Additives, Food
[5] M. P. Nguyen, “Impact of roasting to total phenolic, flavonoid and Agriculture Organization of the United Nations & World
and antioxidant activities in root, bark and leaf of Polyscias Health Organization, Evaluation of certain food additives and
fruticosa”, J. Pharm. Res. Int., vol. 32, no. 2, pp. 13–17, 2020. contaminants: thirty-third report of the Joint FAO/WHO Expert
[6] T. T. H. Nguyen and K. B. Luong, “Research on anti- Committee on Food Additives, World Health Organization,
depressant and anti-stress effects of Polyscias fruticosa”, Geneva, 1989.
Journal of Medicinal Materials, vol. 6, no. (2–3), pp. 84–86, [21] B. S. Nayak and K. N. Patel, “Pharmacognostic studies of the
2001 (in Vietnamese). Jatropha curcas leaves”, Int. J. Pharm. Tech. Res., vol. 2, no.
[7] T. T. H. Nguyen, K. Matsumoto, and H. Watanabe, “Protective 1, pp. 140–143, 2010.
effect of Polyscias fruticosa leaves on social isolation stress- [22] S. Okhale, M. Amanabo, K. A. Jegede, G. O. Egharevba, I.
induced brain tissue damage”, in: Proceeding of The Fourth Muazzam, and O. Kunle, “Phytochemical and
Indochina Conference on Pharmaceutical Science (Pharma pharmacognostic investigation of antidiabetic Scoparia dulcis
IndoChina IV), 2006, pp. 359–362. Ho Chi Minh City Journal Linn Scrophulariaceae whole plant grown in Nigeria”,
of Medicine, Ho Chi Minh City, Vietnam. Researcher, vol. 2, pp. 7–16, 2010.
[8] N. S. Ashmawy, H. A. Gad, M. L. Ashour, S. H. El-Ahmady, [23] G. A. Engwa, P. U. Ferdinand, F. N. Nwalo, and M. N.
and A. N. B. Singab, “The genus Polyscias (Araliaceae): A Unachukwu, M. N., “Mechanism and health effects of heavy
phytochemical and biological review”, J. Herb. Med., vol. 23, metal toxicity in humans, Poisoning in the Modern World -
100377, 2020. New Tricks for an Old Dog?”, IntechOpen, 2019.
[9] T. T. D. Nguyen, Q. D. Nguyen, and T. V. L. Nguyen, “Kinetic [24] A. Beressa, A. Wariyo, G. Chala, and Y. Tefera, “Analytical
study on chlorophyll and antioxidant activity from Polyscias method development for quality control and standardization of
fruticosa (L.) Harms leaves via microwave-assisted medicinal plants: A critical review”, Research & Reviews:
extraction”, Molecules, vol. 26, no. 12, 3761, 2021. Journal of Herbal Science, vol. 10, no. 10, pp. 1-12, 2021.
[10] Ministry of Health, Pharmacopoeia Vietnamica, 5th ed.,
Medical publisher, 2017.
184
Alkali-Activated Slag/Sugarcane Bagasse Ash

Pastes Cured in Room-Air Ambient and in Saturated
Lime Water: A Study on the Compressive Strength
and Shrinkage
Duc-Hien Le My Ngoc-Tra Lam
Faculty of Civil Engineering Department of Civil Engineering
Ton Duc Thang University Ho Chi Minh City Open University
leduchien@tdtu.edu.vn my.lnt@ou.edu.vn
ORCID: 0000-0002-8291-6922
Abstract: Recently, alkali-activated slag (AAS) binder SBA waste for geopolymers [6, 7]. High silica content
has received great attention from researchers, used in place (6075%) is typically found in raw SBA, followed by CaO,
of ordinary Portland cement (OPC). Sugarcane bagasse ash Al2O3, and minor oxides (SO3, K2O,…). Chemical
(SBA) is a final waste of sugar industries, discharged after composition of SBA is similar to that of the common
burning process of bagasse for cogeneration. This study pozzolan coal fly ash, and SBA presents a potential
presents the results of compressive strength and length application as a pozzolan [5].
change of AAS pastes with different levels of SBA Alkali-activated slag (AAS) binder has received great
replacing slag (i.e., 10%, 20%, 30%, 40%) and activator attention in many countries around the world, used in place
moduli (SiO2/Na2O =0.6, 0.8, 1.2). Two curing regimes (i.e., of ordinary Portland cement (OPC). It is a clinker-free
room-air ambient and saturated limewater) were also cement, produced by activation of slag powder (called as
conducted for the alkali-activated pastes. Testing results precursor, source material) with alkali-based chemical
show that the highest strength was examined for mixtures solutions (activator), in which sodium hydroxide and sodium
containing 10 wt.% SBA and 90 wt.% slag; the optimum silicate (water glass) are commonly used [8]. The main
value of silicate modulus providing the highest strength reaction products of AAS are C-A-S-H gels, which directly
moved down from 1.2 to 0.8 as the amount of SBA above affect the performance of the binder. Many previous
20 wt.%. Moreover, SBAslag blends activated by alkali researchers agreed that silica modulus (Ms, SiO2/Na2O molar
solutions with low-silicate modulus could reduce effectively ratio) and Na2O content within activators are the most
the shrinkage in comparison with slag only. Besides, the important parameters, directly affecting AAS properties. For
lowest shrinkage was monitored on the specimens with 20 instance, higher activator modulus and/or alkali
concentration results in better mechanical strength and larger
wt.% SBA among all the mixtures. Furthermore, alkali-
drying shrinkage [9, 10]. Moreover, AAS displays some
activated slagSBA specimens cured in saturated-lime technological and economical advantages over OPC.
water exhibited lower strength and smaller shrinkage than Superior mechanical strength in early ages, lower heat
those cured in room-air ambient. released during hydration, and high durability are reported
for AAS concrete (AASC). Therefore, using AAS to fully
Keywords: sugarcane bagasse ash, alkali-activated slag, replace OPC can be a sustainable concreting technique,
activator, shrinkage, curing regime because AAS composites use industrial wastes as a major
component. Besides, AAS, unlike geopolymers (using a low-
I. INTRODUCTION
calcium precursor like fly ash), can be cured at air ambient,
Nowadays, agricultural waste like sugarcane bagasse, clearly promoting its ecological aspects.
straw, rice husk, etc. can be important resources for
renewable energy production, and residual ashes of which Despite of above benefits, limitations on the uses of AAS
are generated as a waste with an enormous amount. are also reported in literature. Fresh AAS is commonly faced
Sugarcane bagasse ash (SBA) is a final waste of sugar with rapid setting and poor workability that diminishes the
production chains, discharged from combustion of fibrous on-site applications of AAS, especially in densely congested
bagasse in boilers for cogeneration. At present, this agro- reinforcements [9]. In hardened state, high drying shrinkage
waste is dumped increasingly in near areas, causing negative and subsequent formation of micro-cracks for AAS
impacts on the environment and shortage of land for disposal compared to other alkali-activated materials or OPC should
[1]. Previously, SBA utilizations include the use of it as be a noticeable obstacle, reducing long-life durability of
glass-ceramic material production [2], fired-clay brick [3], reinforced concrete structure. In general, shrinkage of AAS
cement replacement in concrete production [4, 5]. Moreover, comprises autogenous shrinkage and drying shrinkage [9].
several researchers have been conducted on utilization of The former shrinkage results from the combination of
chemical reaction and macroscopic process of self-
This research is funded by Vietnam National Foundation for Science and desiccation, relating to the humidity drop due to
Technology Development (NAFOSTED) under grant number 107.01- consumption of water during hydration of slag; and
2020.01.

therefore, it normally develops at a very early age. AAS with adding additives is very few. The behavior of
Meanwhile, the latter performing in volume decrease of alkali-activated slag-fly ash (or alkali-activated slag-bagasse
samples is produced by the chemical reaction with material ash) blends may be more complex than those of alkali-
and moisture exchange with the environment [11]. activated material with pure either slag or fly ash. From this
Consequently, drying shrinkage is usually in accordance viewpoint, the purpose of this study is to investigate the
with weight loss. Drying shrinkage normally occurs when compressive strength and drying shrinkage of alkali-
AAS has more finer porosity, i.e. mesopores (radius of activated slag/bagasse ash when cured in air ambient and in
1.2525 nm), above 80% [12]. Also, the size of the saturated-lime water.
macropores (pore radius of between 25 and 5,000 nm) would
affect the shrinkage too, because it makes an easiness of II. EXPERIMENTAL PLAN
water evaporation from the mesopores [10]. Higher drying A. Materials used
shrinkage of AAS compared to OPC is typically observed,
even though more moisture can be lost on OPC samples. It is Granulated ground blast furnace slag (GBFS), sugarcane
bagasse ash (SBA), sodium hydroxide (NaOH), and water
noted that drying shrinkage of AAS could be 36 times
glass (Na2SiO3) were used for making alkali-activated pastes
greater than that of OPC [13]. As a result, AAS
for experiment. Properties of these materials were described
mortar/concrete commonly shows a potential risk of
as followings:
excessive cracking.
GBFS: an available ground granulated blast furnace slag
The binary/ternary combination of slag with other low-
(Grade 95) in compliance with ASTM C985 [16], collected
aluminosilicate additives such as fly ash, rice-husk ash, or
from a concrete plant was used as source material. Its
bagasse ash for the uses as source materials in alkali-
chemical compositions were presented in Table 1. The
activated material has been studied by earlier scholars, and
Blaine fineness and specific gravity of the slag used were
positive results were noticed. In these cases, co-existence of
4000 cm2/g and 2.89, respectively.
C-A-S-H gels and N-A-S-H gels was mainly identified as the
slag content was above 50% [14]. Aydin [15] stated that SBA: the bagasse ash used in this investigation to
partial replacement of ground granulated blast furnace slag partially GBFS was randomly collected from an opened
(denoted as slag or GBFS) with fly ash (FA) and silica fume dump, nearby a local sugar factory in the South of Vietnam.
(SF) improved effectively drawbacks of AAS, including It was dried at 105110oC for 24 h to remove free water and
obvious reduction of drying shrinkage. Furthermore, using before sieving through the No.100 sieve (opening of 149
combination of slag and sugarcane bagasse ash with different m) to eliminate unburnt coarse substances. Then, the sieved
proportions (100/0, 75/25, 67/33, 50/50) as source material ash was stored in air-tight plastic bags for the investigation.
and alkali-based solution as activator in making alkali- The sieved SBA has the fineness of 4449 cm2/g and the
activated mortar was investigated by Pereira et al. [7]. They specific gravity of 2.165. Tables 1 and 2 present the
suggested that presence of SBA had a marginal influence on properties of the SBA in comparison with those of OPC and
hydration products of AAS; and it consequently did not GBFS.
cause any negative impact on durability of AAS.
Alkali-based activators: A solid NaOH (pure 98%, in
In fact, a number of studies on the shrinkage mechanism pellets) and aqueous water glass (Na2SiO3: 9%Na2O,
of AAS have been reported in literature, while the work of 28%SiO2, 63%H2O; SiO2/Na2O ratio of 3.11) were used.
Table 1. Oxides composition of GBFS and SBA used in this study
Oxides (%) SiO2 Al2O3 Fe2O3 CaO MgO

TiO2 P2O5 SrO SiO3 BaO MnO2 Na2O K2O LOI a
OPC (for reference only) 20.7 4.5 3.3 63 1.8 - - - 2.3 - - 0.1 0.74 2.8
SBA 74.94 6.48 2.14 1.417 1.132 0.398 0.775 0.016 - 0.036 0.058 0.785 2.752 9.07
GBFS 33.65 13.67 1.18 42.11 6.53 - - - 1.87 - - 0.57 0.57 0.58
a.
LOI denotes the loss in ignition.
Table 2. Physical properties of source materials
OPC
Physical properties SBA GBFS
(for reference only)
Fineness (cm2/g) 3470 4449 4000
Specific gravity 3.14 2.165 2.89
Mean particle diameter, D50 (m) - 83 113
Strength activity index (SAI, %):

+ SAI at 7 days - 80 -
+ SAI at 28 days - 75 -
186
Table 3. Mixing proportion and curing regimes for alkali-activated pastes
Curing regimes
Series Mix ID. % GBFS % SBA Ms N, % L/S
A-curing LW-curing
SBA0M0.6 100 - 0.6  
A SBA0M0.8 100 - 0.8  
SBA0M1.2 100 - 1.2  
SBA10M0.6 90 10 0.6  
B SBA10M0.8 90 10 0.8  
SBA10M1.2 90 10 1.2  
SBA20M0.6 80 20 0.6  
C SBA20M0.8 80 20 0.8 6 0.35  
SBA20M1.2 80 20 1.2  
SBA30M0.6 70 30 0.6  
D SBA30M0.8 70 30 0.8  
SBA30M1.2 70 30 1.2  
SBA30M0.6 70 30 0.6  
E SBA30M0.8 70 30 0.8  
SBA30M1.2 70 30 1.2  
Notes: Ms= SiO2/Na2O (Silicate modulus); N (wt. %) = Na2O/source materials (alkali concentration); L/S denotes as water-to-
solid ratio (solid mass = NaOH pellets +GBFS+SBA).
B. Mix proportions accuracy of 0.0025 mm was used for determining

simultaneous length of specimens at the testing periods. The
Regarding mix proportion, the mass of source materials rate of length change (shrinkage) was calculated as the
(GBFS+SBA) for alkali-activated paste was kept at 1400 Equation 1:
kg/m3. Five GBFS/SBA proportions (i.e., 100/0, 90/10,
𝐿𝑖 −𝐿𝑥
80/20, 70/30, 60/40) were selected to develop five paste 𝐿𝑒𝑛𝑔𝑡ℎ 𝑐ℎ𝑎𝑛𝑔𝑒, % = × 100 (1)
groups, namely as A, B, C, D, E, respectively. Moreover, a 𝐿0
10M-NaOH solution and an aqueous water glass were used where, 𝐿i is the initial comparator reading (initial length);
to create three sodium-based activators, changed in silicate 𝐿 x is the comparator reading of shrinkage specimens at a
modulus (Ms =0.6, 0.8, 1.2) with a constant sodium specific age; 𝐿 0 is the nominal gauge length, taken a fixed
concentration (Na2O/source material of 6% by mass). Each value of 285 mm. Each testing measurement was triplicated
proportion of GBFS/SBA was sequentially activated by the from three samples, and the average values were reported.
three mentioned activators. Besides, the distilled water was
additionally added to the mixtures until the acceptable
workability reached and the liquid-solid ratio (L/S) of 0.35
was fixed after several trials. In summary, a series of 15
mixtures (35) was developed, as summarized in Table 3.
C. Sample preparations and testing methods
After the mixing, the paste was cast into 50  50  50 (b)
mm3 cubic modes and 25 25  285 mm3 prismatic molds (a)
with a thorough compactness of alkali-activated paste Figure 1. The cubic specimens for compressive strength test (a)
specimens in accordance with ASTM C109 [17] and ASTM and bar specimens used for length change measurement (b)
C490 [18], respectively. When completion of casting, the
specimens together with molds were covered with plastic
sheets and place in the laboratory room for 24 h before III. RESULTS AND DISCUSSION
demolding and the curing process started. A half number of A. Results of compressive strength
samples was continuously cured at room-air ambient (23
2oC and relative humidity > 90%), denoted as A-curing and The results of compressive strength of different alkali-
the remains were soaked in a lime-saturated water tank, activated mixtures cured in room-air ambient (A-curing) and
denoted as LW-curing until reaching the testing times (see in saturated lime-water (LW-curing) are given in Table 4. It
Fig. 1). can be seen that the compressive strength of alkali-activated
system increased over times as a result of the polymerized
Compressive strength of GBFS/SBA paste was tested on process. The A-curing specimens performs a considerably
the cubic specimens after 3, 7, and 28 days. The compression higher strength than the LW-curing specimens, determined at
load at the loading rate of 0.9 kN/s was applied using a the same ages. For later age, the differences in strength of
Controlled-Compressive Testing Machine. The average cubes between the two curing regime were severely,
result of three samples of each group reported. The prismatic estimated to 3040%. It means that although saturated LW-
bars were used to investigate length change over times (3, 7, curing is commonly preferred for OPC concrete/ mortar, this
14, and 28 days). Length change measurement was carried curing method does not ensure for strength evolution of
out according to ASTM C490 [18]. A modified digital alkali-activated material. Limitation of hydrated reaction as
comparator device with a reference rod and a measuring
187
premature leaching in lime water would be responsible for However, such effect seems to be changed as the SBA
the strength reduction [19]. In addition, it is noted that for percentage increased. For mix series D and E, the highest
mix series A, B, C, the increase of silica modulus from 0.6 to strength was attained on the specimens activated with silicate
0.8 and 1.2 resulted in compressive strength increased, modulus of 0.8. Increasing Ms-value from 0.8 to 1.2 caused a
investigated at all testing ages. The better dissolution of raw 4.315% strength reduction. Above observation
materials in activated solution with higher silicate anions demonstrates that optimum silicate modulus shifted from 1.2
may be attributed to the strength enhancement [20, 21]. to 0.8 when SBA replacement beyond 20%.
Table 4. Results of compressive strength
3 days 7 days 28 days
Series Mix ID.
A-curing LW-curing A-curing LW-curing A-curing LW-curing
SBA0M0.6 41.11 36.86 46.40 39.04 80.46 54.75
A SBA0M0.8 58.36 40.96 79.55 49.52 105.00 65.94
SBA0M1.2 63.99 51.92 95.95 53.09 111.46 67.94
SBA10M0.6 48.73 41.01 52.54 43.69 86.89 59.23
B SBA10M0.8 56.43 46.69 78.56 52.01 110.22 71.82
SBA10M1.2 71.33 53.93 86.47 58.27 114.17 75.55
SBA20M0.6 45.51 33.55 49.06 39.32 69.71 51.92
C SBA20M0.8 53.75 37.82 72.85 45.48 83.61 59.76
SBA20M1.2 59.19 43.99 84.92 53.32 91.03 64.62
SBA30M0.6 41.65 31.07 47.31 37.17 55.75 48.96
D SBA30M0.8 52.74 37.17 72.79 42.99 80.83 53.99
SBA30M1.2 49.82 35.94 67.44 50.33 77.35 54.76
SBA30M0.6 29.04 28.50 45.65 35.75 51.61 39.90
E SBA30M0.8 38.22 31.73 55.41 40.43 72.34 42.70
SBA30M1.2 32.22 29.94 52.34 42.48 61.44 46.16
Moreover, for a given level of silicate modulus, among measured on all specimens at 28 days, cured in air
various mix series, the mix series B incorporated 90% GBFS temperature (A-curing) and in the saturated-lime water
and 10% SBA have the highest compressive strength, (LW-curing), was in range of 0.240.93% and 0.20.8%,
followed by the series A, C, D and E. This results indicates respectively. For alkali-activated binder with pure slag
that 10% SBA in mixture shows the highest Si/Al ratio (100/0), the shrinkage magnitudes of 0.7% (for A-curing)
compared to the other replacement rates. When Si/Al is high, and 0.4% (for LW-curing) were determined at 28 days.
raw materials tend to growingly reacts with Ca(OH)2, These shrinkage values are comparable or even lower than
resulting in strength improvement [22, 23]. Further increase
those of previous measurement, carried out by Ye and
of SBA/GBFS ratio led to gradually decrease compressive
Radlinska (about 0.8%) [24]. Moreover, the lowest
strength. Similar result could be found when fly-slag blends
was used as staring source [20]. It is generally reported that magnitude of length change was estimated on the specimens
the reactivity of SBA or fly ash in alkali solution is not as with 20% SBA replacing GBFS (80/20), regarding either
good as that of slag; hence, compressive strength of alkali- curing regimes or silicate modulus of activator. In
activated GBFS/SBA was reduced along with higher SBA comparison with the 100/0 samples, the shrinkage of the
content. 80/20 samples were decreased by 45% and 35% for A-
curing and LW-curing, averagely. This observation could be
B. Results of drying shrinkage similarly found in a past investigation by Aydin [15],
Effect of GBFS/SBA ratio on the drying shrinkage of explored that shrinkage of AAS was significantly reduced
alkali-activated GBFS/SBA paste. Fig. 2 demonstrates the by FA replacement, mainly due to the stability of the main
influence of GBFS/SBA rate on the shrinkage (length reaction products. Similarly, Moraes [8] confirmed that
change) of the alkali-activated paste specimens, conducted AAS modified with sugarcane straw ash produced a refined
in this study. There exists a tendency that drying shrinkage pore size due to activation of the ash in the alkali medium.
developed progressively with time, resulting from self- However, at higher rate of SBA in the source blends,
desiccation (the consumption of pore water by cement excessive shrinkage was recognized.
hydration) and moisture exchange with environment [11]. It
is noticed a trend that shrinkage of all mixtures would Effect of silica modulus on the drying shrinkage of
alkali-activated GBFS/SBA paste. The influence of
continuously grow after 28 days. Earlier researchers
silicate modulus (Ms=SiO2/Na2O) on the shrinkage of alkali-
revealed that the shrinkage magnitude tended to be stable
activated slag/ bagasse ash are presented in Fig. 3. It is seen
(reaching the constant value) after 40 days [10], or may be
that, shrinkage of the samples was substantially increased
longer. Furthermore, there is an agreement that shrinkage of
with higher SiO2/Na2O ratio within the activator at 3- and 28
AAS is commonly higher than that of OPC (from 36 times,
days, regardless GBFS/SBA proportion. The variation range
in general), due to finer pore structures and lower in
was more noticeable for binder with SBA replacement
stiffness in the matrix [9]. The magnitude of shrinkage
comparing to that without SBA (100/0). For example, for
188
100/0 pastes, when Ms-value increased from 0.6 to 1.2, the alkali-activated GBFS/SBA pastes, a lower shrinkage was
28-day shrinkage increased by 32%; while, that increment shown for specimens with LW-curing when compared to
for other alkali-activated GBFS/SBA were from 61% (for those with A-curing, measured in the testing periods (up to
60/40 sample) to 158% (for 80/20 sample). However, with 28 days). The restraint of evaporation of free water from the
the naked eyes, none of the samples showed visible cracks pores within specimens when soaked in saturated-lime water
caused by shrinkages. may be responsible for that observation. On the contrary, the
90/10 specimens displayed an opposite trend at 28 days. In
Shrinkage of AAS increased significantly with an
fact, the shrinkage of these samples with LW-curing was
increase in activator modulus has been previously reported
about 13% greater than that with A-curing. However, as
[10, 11]. It is believed that increasing activator modulus
discussed in earlier part, the shrinkage of alkali-activated
may influence on the tension surface of pore solutions
based slag would progressively develop after 28 days. Ye
within the cement matrix, and it in-turn causes a
and Radlinska [24] reported that the highest shrinkage of
development in autogenous shrinkage [11]. Moreover, in
AAS system was considerably dependent on the relative
alkali-activated slag/fly ash composite, greater slag content
humidity (RH), and it tends to be attained at a longer time of
and a higher activator modulus led to a higher shrinkage in
curing for implementation of higher RH regime.
general, reported by Gao et al. [25]. And, this tendency
could be reasonable when observed on the present result for Table 5. Ratios of shrinkage magnitude of various specimens
AAS modified with SBA instead of fly ash. In short, both between A-curing and LW-curing
incorporation of low-calcium silicate like fly ash, bagasse
ash into slag and using a low-silicate activator could Mix 3 7 14 28
L/S Ms
series days days days days
effectively reduce the shrinkage. Nevertheless, the shrinkage
mechanism of alkali-activated slag/bagasse ash could be A 0.67 0.63 0.58 0.53
more complicated that needs further consideration. B 0.40 0.47 0.68 1.13
C 0.35 0.6 0.04 0.48 0.57 0.83
Effect of curing regimes on the shrinkage of alkali-
D 0.69 0.54 0.54 0.85
activated GBFS/SBA paste. In this part, effect of curing
regimes on the drying shrinkage was investigated, and the E 0.96 0.73 0.84 0.85
result were tabulated in Table 5. It was observed that most
(a) A-curing specimens (b) LW-curing specimens

Figure 2. Measurement of length change for specimens with changes in GBFS/SBA ratios (Ms =0.6)
Figure 3. Effects of silica modulus on the shrinkage at the ages of 3 and 28 days (A-curing)
189
[8] J. C. B. Moraes et al., "Effect of sugar cane straw ash (SCSA) as

IV. SUMMARY solid precursor and the alkaline activator composition on alkali-
 Three factors including silicate modulus, SBA activated binders based on blast furnace slag (BFS)",
replacement level, and curing regime (A-curing and Construction and Building Materials, vol. 144, pp. 214-224,
2017/07/30/ 2017.
LW-curing) were experimentally confirmed to have [9] P. Awoyera and A. Adesina, "A critical review on application of
significant impact on compressive strength of alkali- alkali activated slag as a sustainable composite binder", Case
activated GBFS/SBA paste. Combination of 10% SBA Studies in Construction Materials, vol. 11, p. e00268,
and 90% GBFS provides the highest strength. The 2019/12/01/ 2019.
optimum value of silicate modulus moves from 1.2 to [10] S. Aydın and B. Baradan, "Effect of activator type and content on
properties of alkali-activated slag mortars", Composites Part B:
0.8 as the SBA replacement above 20%. The LW-
Engineering, vol. 57, pp. 166-172, 2014/02/01/ 2014.
curing considerably decreases the compressive strength [11] X. Hu, C. Shi, Z. Zhang, and Z. Hu, "Autogenous and drying
when comparing to A-curing. shrinkage of alkali-activated slag mortars", vol. 102, no. 8, pp.
4963-4975, 2019.
 The lowest magnitude of shrinkage was determined on [12] H. Maryam, R. Farshad, and R. Aleksandra, "Effect of activator,
the 80/20 specimens, which were decreased by 45% curing and humidity on drying shrinkage of alkali-activated fly
and 35% in average, compared to the 100/0 specimens ash", Green Materials, vol. 7, no. 2, pp. 71-83, 2019.
for the A-curing and LW-curing regimes, respectively. [13] G. Wang and Y. Ma, "Drying shrinkage of alkali-activated fly
However, at higher rate of SBA in blends (30% and ash/slag blended system", Journal of Sustainable Cement-Based
Materials, vol. 7, no. 4, pp. 203-213, 2018/07/04 2018.
40%), excessive shrinkage has been shown. Moreover, [14] I. Ismail, S. A. Bernal, J. L. Provis, R. San Nicolas, S. Hamdan,
drying shrinkage of the alkali-activated samples was and J. S. J. van Deventer, "Modification of phase evolution in
substantially increased along with higher activator alkali-activated blast furnace slag by the incorporation of fly ash",
modulus. It can be concluded that both incorporation of Cement and Concrete Composites, vol. 45, pp. 125-135,
bagasse ash into slag and using a low-silicate activator 2014/01/01/ 2014.
could effectively reduce the shrinkage. [15] S. Aydın, "A ternary optimisation of mineral additives of alkali
activated cement mortars", Construction and Building Materials,
 Curing regimes have a noticeable influence on the vol. 43, pp. 131-138, 2013/06/01/ 2013.
[16] Standard Specification for Slag Cement for Use in Concrete and
shrinkage. Within the testing periods, a lower shrinkage
Mortars, 2010.
was shown for the almost LW-curing specimens [17] ASTM:C109, "Standard Test Method for Compressive Strength
compared to A-curing ones. However, the shrinkage of of Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube
alkali-activated slag/ bagasse ash would progressively Specimens)," ed, 2008.
grow after 28 days, indicating that an investigation of [18] Standard Practice for Use of Apparatus for the Determination of
shrinkage in longer periods would be necessary. Length Change of Hardened Cement Paste, Mortar, and
Concrete, 2011.
[19] M. Komljenović, "Mechanical strength and Young's modulus of
REFERENCES alkali-activated cement-based binders", in Handbook of Alkali-
[1] A. Bahurudeen and M. Santhanam, "Influence of different Activated Cements, Mortars and Concretes, F. Pacheco-Torgal, J.
processing methods on the pozzolanic performance of sugarcane A. Labrincha, C. Leonelli, A. Palomo, and P. Chindaprasirt, Eds.
bagasse ash", Cement and Concrete Composites, vol. 56, pp. 32- Oxford: Woodhead Publishing, 2015, pp. 171-215.
45, 2015/02/01/ 2015. [20] X. Hu, C. Shi, Z. Shi, and L. Zhang, "Compressive strength, pore
[2] N. T. T. Phương, T. Q. Hải, N. V. Hoàn, N. T. Thoa, Đ. T. Hà, structure and chloride transport properties of alkali-activated
and N. M. Hà, "Đặc tính của tro bã mía và sử dụng tro bã mía slag/fly ash mortars", Cement and Concrete Composites, vol. 104,
trong sản xuất gạch ceramic", Tạp chí Khoa học công nghệ, vol. p. 103392, 2019/11/01/ 2019.
45, 2018. [21] D.-H. Le, Y.-N. Sheen, and M. N.-T. Lam, "Potential utilization
[3] S. M. S. Kazmi, S. Abbas, M. A. Saleem, M. J. Munir, and A. of sugarcane bagasse ash for developing alkali-activated
Khitab, "Manufacturing of sustainable clay bricks: Utilization of materials", Journal of Sustainable Cement-Based Materials, pp.
waste sugarcane bagasse and rice husk ashes", Construction and 1-17, 2021.
Building Materials, vol. 120, pp. 29-41, 2016/09/01/ 2016. [22] G. C. Cordeiro, R. D. Toledo Filho, and E. M. R. Fairbairn,
[4] S. Loganayagan, N. Chandra Mohan, and S. Dhivyabharathi, "Effect of calcination temperature on the pozzolanic activity of
"Sugarcane bagasse ash as alternate supplementary cementitious sugar cane bagasse ash", Construction and Building Materials,
material in concrete", Materials Today: Proceedings, vol. 45, pp. vol. 23, no. 10, pp. 3301-3303, 2009/10/01/ 2009.
1004-1007, 2021/01/01/ 2021. [23] G. Sua-Iam and N. Makul, "Utilization of coal- and biomass-fired
[5] M. Jahanzaib Khalil, M. Aslam, and S. Ahmad, "Utilization of ash in the production of self-consolidating concrete: a literature
sugarcane bagasse ash as cement replacement for the production review", Journal of Cleaner Production, vol. 100, pp. 59-76,
of sustainable concrete – A review", Construction and Building 2015/08/01/ 2015.
Materials, vol. 270, p. 121371, 2021/02/08/ 2021. [24] H. Ye and A. Radlińska, "Shrinkage mechanisms of alkali-
[6] J. C. B. Moraes et al., "New use of sugar cane straw ash in alkali- activated slag", Cement and Concrete Research, vol. 88, pp. 126-
activated materials: A silica source for the preparation of the 135, 2016/10/01/ 2016.
alkaline activator", Construction and Building Materials, vol. [25] X. Gao, Q. L. Yu, and H. J. H. Brouwers, "Assessing the porosity
171, pp. 611-621, 2018/05/20/ 2018. and shrinkage of alkali activated slag-fly ash composites designed
[7] A. Pereira et al., "Mechanical and durability properties of alkali- applying a packing model", Construction and Building Materials,
activated mortar based on sugarcane bagasse ash and blast vol. 119, pp. 175-184, 2016/08/30/ 2016.
furnace slag," Ceramics International, vol. 41, no. 10, Part A, pp.
13012-13024, 2015/12/01/ 2015.
190
Production Efficiency Improvement Using Value

Stream Mapping with Simulation: A Case Study in
Vietnam
Xuan-Quang Bach Thanh-Tuan Dang Chia-Nan Wang
Department of Industrial Department of Industrial Department of Industrial
Engineering and Management Engineering and Management Engineering and Management
National Kaohsiung University of National Kaohsiung University of National Kaohsiung University of
Science and Technology Science and Technology Science and Technology
Kaohsiung 80778, Taiwan Kaohsiung 80778, Taiwan Kaohsiung 80778, Taiwan
xuanquang.bach@gmail.com tuandang.ise@gmail.com cn.wang@nkust.edu.tw
Abstract: Lean Manufacturing is a production method that studied the impact of Lean operations practice and design on
companies have long adopted in various industries to manufacturing performance and concluded that the results
eliminate waste and boost efficiency. Due to its benefits, indicate significant gaps in LM practices among different
many plants are transforming their manufacturing model into industries. Likewise, Anand and Rambabu [11] examined the
Lean in Vietnam. Nghia Nippers Corporation (NNC) is one extent to which an organization, among a provided number of
of the biggest Vietnamese nail nipper producers and is other LM organizations, has succeeded in applying Lean and
currently producing in batches with the push method. The their position compared to each other and found that the key
company thus wants to switch its production activities to LM practices have not been sufficiently embraced. Some
advanced LM elements such as pull system, kanban,
Lean Manufacturing with a pull system to increase
production leveling, etc. were even absent.
throughput and machine utilization, and reduce inventory and
lead time. For this case study, to ensure a sustainable result, The purpose of this paper is to showcase that LM tools,
not only a lean pull system was proposed using value stream when implemented properly, can help boost a company’s
mapping to visualize the future state, but simulation was also performance in four major aspects of manufacturing, namely
used to verify if the future state would yield expected production lead time, machine utilization, inventory control,
improvements. The results showed that key manufacturing and total throughput. A factory of Nghia Nippers Corporation
factors were improved substantially: total throughput (NNC) located in Cu Chi, 35 km northwestern from the center
increased one-third, machine utilization went up 25 – 42%, of Ho Chi Minh City, Vietnam, is chosen as the subject of the
production lead time reduced 5 – 7 times, and buffer study. The unique feature of NNC’s production process is that
inventory decreased threefold. the waiting time of WIP is excessively long as the steel parts
must cool down before being passed on to the next step. By
Value Stream Mapping (VSM) and simulation, the current and
Keywords: lean manufacturing, value stream mapping, future states of a model production line are visualized along
work in progress, just-in-time, simulation, Vietnam with their key results quantitatively verified to confirm
improvement.
I. INTRODUCTION
Lean manufacturing (LM) or lean production, is a set of II. LITERATURE REVIEW
methods aiming at the elimination of waste ("Muda") inside a LM has been defined as: “A systematic approach to
manufacturing system [1]. The concept of Lean arose from the identifying and eliminating waste (non-value added activities)
Toyota Production System which was developed by Taiichi through continuous improvement by flowing the product at
Ohno and Eyji Toyoda between 1948 and 1975 [2]. Since its the pull of the customer in pursuit of perfection” [12]. In
birth, LM has been the initiative most major corporations contrast to traditional manufacturing, lean systems focus on
throughout the world try to adopt in the hope of leading the providing the customer exactly what they want at the time they
increasingly competitive global market and staying ahead of want it [13]. In other words, the voice of the customer plays
competitors [3]. Not only is Lean applied in the process the deciding role in the formation of a Lean production system
industries, but nowadays Lean application can also be found [14]. Thus, its targeted achievement is to bring more value to
in a range of other sectors from electronics [4], petroleum [5]
customers while costing fewer resources.
to businesses [6], healthcare, R&D, and finance [7] [8], etc.
Statistically, there are around 25 Lean tools [15]. Other
Despite the fact that LM has become increasingly
widespread, its implementation has proved to be challenging sources would list 12 of them as essential [16], including
in many cases. Shahram [9] described how key areas of cellular manufacturing, Kanban pull system (just-in-time),
manufacturing are evaluated for the appropriateness Heijunka (production leveling), kaizen, etc. It would be
implementing Lean and found that even though opportunities insufficient, however, to discuss all those tools without
for improvement in those areas are identified, there is still a mentioning a core component of Lean: Value Stream Mapping
substantial gap from the lean target. Shahram and Cristian [10] (VSM). As it was invented, VSM is a pencil and paper tool

that helps reveal the flow of material and information as a that automatically generates simulation models based on that
product undergoes the processing steps in the value stream paradigm input.
[13]. In other words, it is a tool for lean management aiming
Even though in the literature much research has been
at providing an in-depth analysis of the current state to identify
carried out about the implementation of Lean into
sources of waste and based on that management will come up
manufacture, to the best of our knowledge, none has been
with a design for the future state where operations
made combining VSM and simulation to examine lead time,
performance is improved through waste elimination. [3]
throughput, machine utilization, and stock inventory within a
analyzed the production flow through VSM and affirmed that
circumstance as novel as NNC’s, where WIP takes an
a large amount of time elapsed during production is waiting
excessive amount of time to handle between processes.
and non-value added and that Lean tools make a positive
Therefore, in this paper, the focus is on developing a
impact on the reduction of waiting time (16.9% for man time
methodology addressing all aforementioned aspects.
and 14.17% for machine time). Likewise, Roberto et al. [17]
redesigned an assembly line through VSM and kanban with III. METHODOLOGY
the milk-run technique and reported that inventory was
reduced twofold while idle times were decreased threefold. As Data collection
Jaideep [18], after applying a framework for process change
in business to examine the implementation of LM and its
outcome asserted that manufacturing batch sizes had shrunk VSM
Current State
Simulation
Current State
from a 30-day lot size to 16 days or less and set up times in
most areas of the plant had been reduced by half compared to
No
3 years before LM.
Answer Verification &
8 questions Validation
Even though LM and VSM are priceless in pinpointing the satisfied?
inefficiencies of the current state and describing the benefits a

plant could gain from a redesigned future state, oftentimes Yes
managers’ decision of whether to apply Lean requires more VSM Simulation

quantifiable evidence than merely what was brought about by Future State Future State
a qualitative tool like VSM [19]. Simultaneously, VSM is also

just a static tool that is incapable of either displaying the No
dynamic behavior or handling the complexity and uncertainty Result better than
Current State?
of a model [20]. In many cases, it is nearly impossible to
predict the fluctuation pattern of inventory level as production
Yes
progresses based on VSM only [21]. Furthermore, a
manufacturing system must often deal with random variation Propose other
scenarios
in customer demand that impacts its inventory requirements,
there is not yet any method available to compute how much is No
sufficient for every case. A common practice of reserving 2

Result better
days of inventory is suggested [13], nonetheless, this has not than current Future
State?
been proved to be the method for validating whether that
amount is too little or too much for a given company on a case- Yes
by-case basis [22].

Accepted
Therefore, many researchers have resorted to simulation as

a tool to bridge this gap in VSM due to its ability to generate Figure 1. Description of methodology
a dynamic view of throughput, lead time, the fluctuation of
buffer, and machine utilization for the current and future states A. Data Collection
to better serve management decision. William et al. [23] Theoretical and practical data are collected for the study.
demonstrated that simulation analysis can be a useful and Theoretical data consists of two categories: planned
important part of VSM and that it can form an integral part of production quantities for different product types from
the toolset. Donatelli and Harris [24] discussed the integration Planning Department and weekly production reports from
of VSM and simulation and determined that they are a natural Operation Department with statistics of throughput yield,
combination and that each enhances the other’s value in the defect rates, comparison between actual throughput and
LM effort. Charles and John [22] explained why Lean needs planned production quantities, and machine uptime. This data
simulation as a means of support and pointed out that Lean’s shows that weekly throughput yield only satisfies 55 – 83% of
deficiencies can be addressed by simulations. Shahram et al. actual demand.
[25] used simulation to verify the feasibility of the design of
Practical data is obtained using a stopwatch while walking
manufacturing cells. While Berna [26] used simulation to
down production lines in a backward fashion from shipping to
understand the concepts of LM including line balancing
the first station as suggested in [13]. From observation, both
versus takt time, pull versus push, Lian and Landeghem [20]
male and female workers that have different cycle times for
took one step further by introducing two new elements to
the same process step are assigned to several similar
examine how VSM can be adapted for use in simulation which
workstations, thus the average value of cycle times for each
are, firstly, a VSM paradigm and secondly, a model generator
process step is calculated and recorded. In addition, since a
192
worker’s productivity varies at different times of day, their simulation model is built, verified, and run. Only after the
performance is also recorded randomly during a working day result of this model satisfies the validation process can the
to ensure the obtained values will reflect the total output the procedure continues to use the data and findings to build the
factory is achieving in reality. Other practical data was taken model for the future state. In the next step, the simulation
during interviews with shop floor supervisors, including model for the future state is built based on its VSM, followed
waiting time of WIP between processes, repairing time, by several other scenarios to identify which one is the best.
frequency of supplier delivery, and the average amount of This result is then accepted as our final solution to the study.
material consumed in a day. Finally, some other The whole methodology can be summarized in Fig.1.
supplementary data such as the history of the working lines, IV. VALUE STREAM MAPPING
workers’ inappropriate habits in handling practice due to the
lack of training, and standardized work was also noted. C. VSM for the Current State
Through direct observation on the floor and interviews with The process of making a nail nipper (both Standard size -
shop floor supervisors, we ensure necessary information is STS and Over and Small size - OVS) is that steel rods will be
gathered sufficiently and the misapprehension about details of bulged, punched, cut, reformed in BPF factory then shipped to
the task, if any, has a chance to be rectified. SFF. After data is collected and analyzed, we draw the VSM
for the current state as illustrated in Fig. 2. It is noticed that the
B. Description of Methodology production lead time to value-added time ratio is high because
After the data collection phase, the VSM for the current between processes WIP spends a substantial amount of time
state is drawn. Next, we come up with the VSM for the future cooling down and waiting, especially after the jaw reforming
state based on an analysis of waste elimination and kaizen step where all WIP is gathered overnight waiting to be shipped
application. Parallelly, from the VSM of the current state, a to SFF the next day.
Production Control
Monthly Daily order
Steel Supplier order SFF
BPF
Steel rod 295 471,400 pairs/month
Floor supervisors - 471,400 “Left”
- 471,400 “Right”
Week plan
Pallet = 500 pcs
7:30am to 4:30pm
4 x daily
1 x daily
Bulging Punching Cutting Jaw reforming Shipping

I I I I Queueing
I
1 hours 45 30 mins 30 mins 17 hours
C/T = 5.5s mins C/T = 5s C/T = 4s C/T = 3.5s
C/O = Ø C/O = Ø C/O = Ø C/O = Ø
OEE = 100% OEE = 45% OEE = 45% OEE = 53%
81,000 sec. avail. 81,000 sec. avail. 81,000 sec. avail. 81,000 sec. avail.
3 shifts 3 shifts 3 shifts 3 shifts
Production lead time

N/A 1 hour 45 mins 30 mins 30 mins 17 hours = 19 hours 45 mins
5.5s 5s 4s 3.5s Value added time

= 18s
Figure 2. VSM for the current state
D. VSM for the Future State  Layout rearranged to cellular manufacturing to allow
for a pull system.
The VSM for the future state as can be seen in Fig. 3 is
created based on waste elimination analysis with the  Takt time (takt) is calculated at 12 seconds, and all
application of kaizen to reduce waiting time and cycle times process step cycle times are controlled to be below
via a combination of processes. It includes other Lean takt. Based on takt, pitch (consistent increment of work
elements such as kanban, supermarkets, JIT, etc. to improve BPF should release to the shop floor to synchronize its
the current state of the production flow. Through designing the production pace with the customer’s demand) is
future state map, we focus on building a chain of computed at 111.6 minutes.
manufacturing processes in which each of them is linked to
the customer by a pull system, ensuring each process only  The pacemaker process has been decided at Shipping.
produces what the customer needs at the pace of demand. The kanban size is calculated at 500 pieces per pallet.
Each Kanban is consistently pulled at pitch.
Altogether, for BPF to achieve this future state, all of these
prerequisites need to be completed:  A continuous flow is introduced by combining
punching and cutting via a process kaizen. This helps
193
eliminate WIP waiting time between punching and  Production leveling ratio between STS and OVS is set
cutting machines thus contributing to the overall lead at 2:1 based on monthly demand forecast.
time reduction.
Compared to the current state, we can see that the lead time
 Machine uptime needs to be enhanced considerably. of the future state has been improved nearly fivefold (19 hours
An extensive discussion has been made with the floor 45 minutes/4 hours 7.5 minutes = 4.78 ≈ 5 times). This
supervisors in BPF and it is agreed that the approach outcome, even though from the static tool VSM, is an obvious
of Total Productive Maintenance can bring about 30% motivation for NNC to restructure its operation to Lean. In the
increase in OEE, resulting in 75% average for next part, for management’s decision, we will utilize
machinery in BPF. simulation as a dynamic tool to examine in particular how,
apart from lead time, the future state will affect throughput,
machine utilization, and buffer inventory in BPF.
Production Control
Monthly Daily order
order BPF
SFF
Steel Supplier
471,400 pairs/month
Steel rod 295
- 471,400 “Left”
Daily order - 471,400 “Right”
Pallet = 500 pcs
500 7:30am to 4:30pm
OXOX
4 x daily
12 x daily
500 500 500 500
500 500
Bulging Punching + Cutting Jaw reforming Shipping
Layout Queueing
I
redesign
Process
Takt = 12s Takt = 12s kaizen Takt = 12s
C/T = 5.5s C/T = 9s C/T = 3.5s
Supermarket
C/O = Ø C/O = Ø C/O = Ø
Uptime
OEE = 100% OEE = 75% OEE = 75%
increased
243,000 sec. avail. 243,000 sec. avail. 243,000 sec. avail.
3 shifts 3 shifts 3 shifts
Production Lead Time =
N/A 1 hours 45 mins 30 mins 112.5 mins 4 hours and 7.5 mins
5.5s 9s 3.5s Value added time = 18s
Figure 3. VSM for the future state
V. SIMULATION MODELS
To measure the extent of benefits the future state of VSM
can bring to BPF, a detailed simulation model is built using
the software Plant Simulation version 12 by Siemens. We start
with a model for the current state whose result is then verified
and validated for credibility. After that, we simulate the future
state to examine its output and then attempt different scenarios
of inventory level to determine the optimum one.
E. Simulation for the Current State Figure 4. Simulation model for the current state
The model parameters include: 1. Throughput
 Daily demand, from BPF’s master production plan
After the simulation run, the throughput we get is 9
 Running time, equal to a working day (3 shifts) pallets.
minus breaks.
 Entities (machines & buffer places in the model) are
set up with corresponding input and output with
cycle times and machine OEE as defined in Fig. 2.
Figure 5. Simulated throughput of the current state
194
Since monthly average demand for BPF is 471,000 probably because to its processing time is the longest. Also
products, we can calculate the average percentage of work noticeable from the chart is that working and waiting have an
BPF can accomplish in a week based on the daily capacity of inverse relationship. The more it works, the less it waits and
one manufacturing cell: vice versa. That is the reason Jaw_reforming waits most
9 × 500 × 3 × 7 among them. We aim to improve their working percentage in
× 100% = 74.459% ≈ 75% the future state.
471,400 ÷ 26 × 7
Compared with the data we’ve got in the collection step 4. Buffer
where actual throughput ranges from 55% to 83% of market
demand, this result from the simulation model lies within the
range of reality, therefore it is accepted.
2. Lead time
Figure 8. Buffer stock analysis of the current state

As Fig. 8 illustrates, buffer gathers most at P1, P3, and P5
where pallets arrive at workstations waiting to be processed.
The most severe stagnation occurs at P1, where pallets
accumulate to 6. This is consistent with observation in reality,
where usually 6 – 7 pallets are seen available at Punching a
Figure 6. Simulated lead time of the current state couple of times in the morning. The dense area of lines
fluctuating quickly between only 0 and 1 (P2, P6, P4,
BPF finishes 9 pallets in 19 hours, which is 2.11 hours per
Gathering, Gathering1) implies a short occupancy of WIP at
pallet. Plus the overnight waiting time of 11 hours, total lead
those places before being pushed to other busier places (P1,
time for each mounts up to 13.11 hours. Compared to
P3, P5), raising questions about the effective use of space for
production lead time specified in VSM of the Current State
these less frequently occupied places.
(Fig. 2), the difference is more than 6 hours. However, our
final goal is to find out the proportion of lead time between F. Simulation for the Future State
the Current State and the Future state to see how well the
Future State can improve BPF’s productivity, therefore we
will need results from the simulation of the Future State in the
next part.
3. Machine utilization
Figure 9. Simulation model for the future state

As discussed in the VSM for the future state, management
will implement TPM to increase OEE to 75%. We run the
model with this parameter and achieve the following results:
1. Throughput
Figure 7. Utilization chart of current state in 100% stacked bars Figure 10. Simulated throughput for the future state
It can be seen from the chart that failure occurs most of Compared to the simulated current state, we can see a
the time. This reflects the low values of OEE in BPF at the dramatic surge of [(12-9)/9] x 100% = 33.33% in throughput
moment. Out of three machines, Punching works the most,
195
of the future state model. Furthermore, it helps BPF to meet make sure Supermarket2 is always occupied. All in all, we can
customer’s requirements as now it can satisfy see that space is better utilized in the future state.
12 × 500 × 3 × 26 VI. RESULT ASSESSMENT AND DISCUSSION
× 100% = 99.28%
471,400
Using VSM and Simulation the improved values of lead
of average total demand. time, machine utilization, inventory level, and throughput are
2. Lead time summarized in below tables:
Table 1. Lead Time Comparison
As production flow runs nonstop, lead time can be
counted by simply dividing the working time available in a Lead time measured by Current State Future State Improvement
VSM 19.75 hrs 4.12 hrs ≈ 5 times
day by the number of finished-good pallets, which equals Simulation 13.11 hrs 1.87 hrs ≈ 7 times
(22.5 ÷ 12) = 1.875 hours. Compared to the lead time
specified in the VSM, the difference is 2 hours 15 minutes. Table 2. Utilization Comparison
We can see that in the dynamic state, the model yields an even Workstations Current State Future State Improvement
shorter lead time than expected. Punching 0.28 0.35 25%
Cutting 0.22 0.28 27.3%
3. Machine utilization Jaw reforming 0.19 0.27 42%
Table 3. Buffer Comparison

Future
Current State Improvement
State
Buffer/Supermarket
8 3 ≈ 3 times
places
Max number of pallets 6 2 3 times
Table 4. Throughput Comparison

Future
Current State Improvement
State
Throughput 9 12 33.33%
It is confirmed by both VSM and Simulation that through

implementing Lean in BPF, NNC can reduce lead time by 5 –
Figure 11. Utilization chart of future state in 100% stacked bars 7 times. Utilization goes up 6 – 8% for machinery in BPF,
meaning a steep rise of 25 – 42% for each of the workstations
Fig. 11, when compared to Fig. 7, reveals a staggering individually. At the same time, buffer was reduced
change in both the working and failure area. Failure time dramatically from the peak of 6 pallets at a time (Fig. 8) to no
decreases significantly owing to the improvement of OEE (to more than 3 pallets (Fig. 12) and that means a substantial space
75%). Interestingly, failure for Cutting decreases to less than improvement where pallets no longer congest in front of
20% of the time, and the other two stations also experience a Punching station.
failure rate reduced by half.
In terms of throughput, even though the future state model
4. Buffer experiences a 33% increase, it only meets 99.28% of the
current average demand. As can be seen in Fig. 11, all
workstations still have a fair amount of idle time, we attempt
to explore the full potential of the future state. The assumption
is that SFF will order continuously, meaning its ordering rate
now synchronizes with the pace at which Jaw_reforming
finishes one pallet. The simulation is run again, and the result
is as follows:
Figure 12. Buffer stock analysis of the future state

Figure 13. Throughput of the full potential future state
After re-zoning buffer places into three supermarkets, we
can see a more stable state of WIP in BPF as shown in Fig. 12. Compared to the current Future State model, throughput
Compared with Fig. 8, the lines fluctuate in a much less erratic increases three pallets, equivalent to 25%. With this yield, the
fashion. Pallets in Supermarket1 only switch between 1 or 2 maximum size of an order BPF could receive can be:
and never reach zero, meaning BPF never runs out of stock to 15 x 500 x 3 x 26 = 585,000 nippers/month.
ship to SFF. Even though stock in Supermarket3 only
fluctuates between 1 and zero, stock in Supermarket2 can only Utilization also goes up 2 – 3% in total for all machines:
reach as little as 1 pallet, which means the pair of Punching
and Cutting can always get WIP (from Bulging) in time to
196
throughput is 18 pallets, increasing 3 pallets in comparison

with that of scenario 0.
Figure 14. Utilization of the full potential future state

The most dramatic change happens in buffer stock at all
places: Figure 17. Total throughput of scenario 2
Again, with [(4 + 4 + 2) – (2 + 3 + 2)] = 3 more bins of
WIP being added, 3 more bins of product have been created.
This result still urges us to investigate further.
 Scenario 3
The final alternative is to set buffer at Supermarket1-2-3 to
4-4-4. The throughput is 22 pallets, increasing 7 pallets
compared to scenario 0.
Figure 15. Buffer of full potential future state

Figure 18. Total throughput of scenario 3
During a working day, buffer in all places only fluctuates
between 1 and 0, signifying a shortage of WIP when BPF According to this scenario, with [(4 + 4 + 4) – (2 + 3 + 2)]
continuously ships to SFF. This situation suggests that = 5 more bins of WIP being added, 7 more bins of product
perhaps the model supposed to be “full potential” is still not are created. The benefit is 7/5 = 1.4 or 140%.
the optimal model and prompts us to assess more scenarios to
The cause of the surge in throughput is owing to the
find out in which one the supermarkets could maintain a
increased number of pallets entering and exiting
buffer level sufficient to sustain an uninterrupted shipment to
Supermarket3. The more pallets are available and ready in the
SFF. This can be done by modifying the buffer size in all
production line at the beginning, the more likely they will be
three supermarkets at the start of the simulation to see how it
processed and passed on to the next stage, contributing to a
will affect the total throughput.
nonstop flow of WIP. The benefit of increasing throughput (7
The current buffer pallets in Supermarket1-2-3 are set at bins) has outweighed the disadvantage of growing buffer (5
2-3-2 (this will be referred to as scenario 0 hereinafter). We bins), therefore this scenario is accepted as the optimal
propose three scenarios where pallets are set at 2-4-2, 4-4-2, model. The maximum size of an order BPF could receive in
and 4-4-4 respectively. As calculated, 4 pallets for each place a month if production takes place in accordance to Lean like
should be maximum, considering factors of average monthly this, is 22 x 500 x 3 x 26 = 858,000 nippers, almost double
demand and the time it would take to fix a workstation in case the current average order.
of machine breakdown. After modification, we run the model
Scenario 3’s stock pattern is as below:
again and achieve the following results for each scenario:
 Scenario 1
Pallets in Supermarket1-2-3 were set to 2-4-2
Figure 16. Total throughput of scenario 1

The throughput is 16 pallets, increasing only 1 pallet
compared to that of scenario 0. One pallet added to
throughput is the result of 1 pallet of WIP added to
Supermarket2. This 1 to 1 ratio prompts us to find out to what Figure 19. Buffer of the future state – third scenario
extent the processing threshold can still tolerate thus this We can see from the chart that buffer in Supermarket3
scenario is not yet accepted. maintains at a high level during the day. This provides the
most continuous flow of WIP in all scenarios to make sure
 Scenario 2
shipping is always ready whenever SFF pulls. Besides, the
Pallets in Supermarket1-2-3 were set to 4-4-2. The chart shows that even though sometimes buffer in
197
Supermarket3 can reach 3 pallets, most of the time it remains [6] C. Michael, O. Tolga and S. Christina, "McKinsey & Company",
at two, which satisfies BPF’s effort in minimizing buffer but 2011. [Online]. Available:
https://www.mckinsey.com/industries/financial-services/our-
at the same time ensuring a maximum possible throughput. insights/using-lean-in-wholesale-financial-services.
Working rate of each workstation of this full capacity [7] J. P. Womack, A. P. Byrne, O. J. Fiume, G. S. Kaplan and J. Toussaint,
"Improvement, Institute for Healthcare", 2005. [Online]. Available:
scenario also skyrockets, increasing 12 – 16% in total https://www.entnet.org/wp-
compared to scenario 0: content/uploads/files/GoingLeaninHealthCareWhitePaper-3.pdf.
[8] B. Simon, D. Christophe and G. Deepak, "Rethinking Lean: Beyond
the Shop Floor", 2009.
[9] T. Shahram and B. Lismar, "Application of constrained management
and lean manufacturing in developing best practices for productivity
improvement in an auto-assemble plant", International Journal of
Productivity and Performance Management, vol. 55, no. 3/4, pp. 332-
345, 2006.
Figure 20. Full utilization of JIT model [10] T. Shahram and M. Cristian, "The impact of lean operations on the
Chinese manufacturing performance", Journal of Manufacturing
Lead time also plummets to (22.5 ÷ 22) = 1.02 hours for Technology Management, vol. 22, no. 2, pp. 223-240, 2011.
one pallet, an increase of [(22.5 ÷ 15)] ÷ 1.02 = 1.47 or 47% [11] G. Anand and K. Rambabu, "Application of benchmarking for
compared to scenario 0. assessing the lean manufacturing implementation", Benchmarking: An
International Journal, vol. 16, no. 2, pp. 274-308, 2009.
VII. CONCLUSION [12] NIST/MEP, Principles of lean manufacturing with live simulation,
Gaithersburg, MD, 2000.
The company of NNC has met with challenges in
improving its efficiency in major aspects of manufacturing, [13] M. Rother and J. Shook, Learning to See: Value Stream Mapping to
Add Value and Eliminate Muda, Brookline, MA: The Lean Enterprise
thus it strived to reorient production activities towards Lean Institute, Inc., 1999.
Manufacturing. Considering that the implementation of LM [14] B. B. Charles, R. C. Terry and L. M. Edward, "Transformation from
into each company has its unique features, a tailor-made Batch to Lean Manufacturing: The Performance Issues", Engineering
methodology especially for NNC’s situation has been Management Journal, vol. 18, no. 2, pp. 3-14, 2006.
developed. In this paper, the current situation of NNC is [15] Vorne, "Top 25 Lean Tools", Vorne Industries Inc., 2016. [Online].
Available: https://www.leanproduction.com/top-25-lean-tools/.
examined and analyzed first to form a clear problem [Accessed 8 2 2022].
statement. Next, the VSM for the current and future state of [16] A. Muspratt, "12 Essential Lean Six Sigma concepts and tools",
NNC are drawn, based on which the simulation models are Process Excellence Network, 16 9 2018. [Online]. Available:
built to confirm the benefits gained by applying LM. Further https://www.processexcellencenetwork.com/lean-six-sigma-
business-performance/articles/12-essential-lean-concepts-and-tools.
to the initial results, other alternatives are then tested to finally [Accessed 8 2 2022].
determine an optimal state for NNC. [17] A. Roberto, C. Roque and M. P. Marta, "Redesigning an assembly line
through lean manufacturing tools", International Journal of Advance
The results showed that with the implementation of LM, Manufacturing Technology, vol. 43, pp. 949-958, 2009.
key factors of manufacturing were improved substantially. [18] M. Jaideep, "A business process change framework for examining
Production lead time was reduced 5 – 7 times, total throughput lean manufacturing: a case study", Industrial Management and Data
increased one-third, machine utilization went up 25 – 42%, System, vol. 103, no. 5, pp. 339-346, 2003.
and buffer inventory decreased threefold. If exploited to its [19] R. B. Detty and J. C. Yingling, "Quantifying benefits of conversion to
full capacity, the future state can even yield benefits better lean manufacturing with discrete event simulation: a case study",
International Journal of Production Research, vol. 38, no. 2, p. 429 –
than that. 445, 2000.
The developed methodology can be utilized in other [20] Y. H. Lian and H. Van Landeghem, "Analysing the effects of Lean
Manufacturing using a value stream mapping-based simulation
companies or sectors that have the same features as NNC does, generator", International Journal of Production Research, vol. 45, no.
meaning the steel industry or other industrial branches that 13, p. 3037 – 3058, 2007.
have similar processing manners, e.g. WIP between processes [21] T. McDonald, E. M. Van Anken and A. F. Rentes, "Utilizing
has to wait to cool down and no WIP is processed overnight simulation to enhance Value Stream Mapping: A manufacturing case
application", International Journal of Logistics Research and
but in its shift only. Applications, vol. 5, no. 2, p. 213 – 232, 2002.
REFERENCES [22] R. S. Charles and H. M. John, "Why lean needs simulation", in WSC
'06: Proceedings of the 38th conference on Winter simulation, 2006.
[1] J. P. Womack, D. T. Jones and D. Roos, The machine that changed the [23] G. S. William, N. M. Thomas and M. V. A. Eileen, "Equipment
world, Free Press, 1990. replacement decisions and lean manufacturing", Robotics and
[2] "Strategos: Consultants - Engineers - Strategists", Strategos, Inc., Computer Integrated Manufacturing, vol. 18, pp. 255-265, 2002.
[Online]. Available: https://strategosinc.com/RESOURCES/01- [24] A. J. Donatelli and G. A. Harris, "Combining value stream mapping
General/toyota_production.htm. [Accessed 7 2 2022]. and discrete event simulation", in Proceedings of the Huntsville
[3] A. Rahani and a.-A. Muhammad, "Production Flow Analysis through Simulation Conference, Huntsville, 2001.
Value Stream Mapping: A Lean Manufacturing Process Case Study", [25] T. Shahram, S. C. David, L. Jochen and W. D. James, "Simulation and
Procedia Engineering, vol. 41, p. 1727 – 1734, 2012. Production planning for manufacturing cells", in WSC 98, 1998.
[4] M. Yogesh, G. Chandramohan and G. Thomas, "Lean Manufacturing [26] J. S. Bernard, "Simulation as a Tool in Understanding the Concepts of
in Electronics & Electrical Manufacturing Industry in India", Lean Manufacturing", SIMULATION, vol. 80, no. 3, pp. 171-175,
International Journal of Scientific & Engineering Research, vol. 5, 2004.
no. 12, pp. 101-103, 2014.
[5] S. Rohan, "Lean Manufacturing in the Oil and Gas Industry", Auburn
University, Auburn, 2011.
198
Study on the Effect of Mixing Ratio of Biodiesel

Fuel Made from Animal Fat on Exhaust Emissions
of the Fishing Vessel’s Diesel Engines
Pham Dinh Trung Mai Duc Nghia Ho Duc Tuan
Faculty of Information Technology Faculty of Mechanical Engineering Faculty of Transportation
Yersin University Air Force Officer’s College Engineering
Da Lat, Vietnam Nha Trang, Vietnam Nha Trang University
pre@yersin.edu.vn nghiamaiduc@gmail.com Nha Trang, Vietnam
tuanhd@ntu.edu.vn
Abstract: Faced with the challenges of environmental outstanding advantage in terms of excess products from basa
protection and the depletion of petroleum fuel resources, the fat, and catfish fat, collectively known as fish fat compared
production and application of biofuels for internal to a number of countries in the world [1].
combustion engines are urgent. Among them, biodiesel needs This is an abundant source of raw materials for biodiesel
to be used as the fuel for diesel engines because they are production, both contributing positively to socio-economic
environmentally friendly and a renewable fuel source. development and respond the requirements of environmental
However, biodiesel is made from diverse sources of raw protection.
materials, so when applied to different engines, it is necessary
to have appropriate mixing ratios because each engine type Through actual survey in the Mekong Delta, biodiesel
will have private fuel injection system parameters. This paper B100 derived from fish fat has been studied by some
studies the combustion and emission characteristics of a Vietnamese enterprises such as Minh Tu Company, Agrifish
fishing vessel's Cummins NTA855 diesel engine, when using An Giang, Southwest Company, etc, production research the
biodiesel fuel produced from fish fat, by numerical standards of TCVN 7717, ASTM D 6751 of the US and EN
simulation method, using KIVA - 3V software. The results 14214 of the EU, exported to markets such as Singapore,
show that at the ratios of B5 (5% biodiesel B100, 95% diesel Japan [13-14, 16].
fuel) and B10 (10% biodiesel B100, 90% diesel fuel), there Therefore, biodiesel produced from fish fat will be an
are lower pollutant emissions than petroleum diesel fuel abundant alternative fuel source, especially with the
(DO). When increasing the mixing ratio up to B15 (15% increasing number of engines for boats. However, the diesel
biodiesel B100, 85% diesel fuel), the formation of the engines used as the main engine of fishing boats are very
combustion mixture decreases, and soot emissions increase. diverse, with different fuel injection systems, so it is
Based on the research results, determine the appropriate necessary to choose the most popular engine for research.
mixing ratio (B5, B10) origin from fish fat, used as an Survey data on types of diesel engines as main engines
alternative fuel for traditional fuel of the fishing vessel's installed on fishing vessels are presented in Table 1 and Table
diesel engines, when using do without to adjust the fuel 2.
injection system.
From the table data, Cummins engines are commonly
Keywords: biodiesel, fishing vessel, diesel engine, used for fishing vessels, so reducing the time and cost before
Cummins NTA855, exhaust emissions, fish fat experimental research, simulation by software is being
widely used, including software to simulate the combustion
I. INTRODUCTION process of internal combustion engine the KIVA-3V [3-4].
To reduce harmful emissions due to the combustion of KIVA-3V simulation software, developed over the years at
petroleum fuel from internal combustion engines to the Los Alamos National Laboratory, USA, and used by many
environment, including combustion products of diesel research centers around the world, is the CFD software that
engines, mainly soot and NOx, in addition to the optimal simulates 2D and 3D turbulent fluid flows, such as fuel
solution for the injection process, high-pressure injection, injection, and has chemical reactions such as simulation of
multi-stage injection. The production and application of combustion of an internal combustion engine [2].
biofuel for internal combustion engines are essential [1, 9, Table 1. Survey of fishing vessels in some provinces of Vietnam
12].
The total Number Ratio
Biodiesel is a form of biofuel, derived from animal and o
N Province
number of Popular of by
vegetable oils and fats, contains oxygen and low sulfur survey fishing engine type engines, provinc
vessels, pcs pcs e, %
content, burns cleanly, and reduces environmental pollution 1 Da Nang 210 Cummins 84 40
[11, 15]. Quang
2 160 Cummins 55 34.38
Vietnam, with the advantage of more than 3.260 km of Nam
coastline and the vast Mekong Delta region, which is an

The total Number Ratio Table 4. Properties of biodiesel fuel and DO

o number of Popular of by
N Province
survey fishing engine type engines, provinc Cetane
Density, Viscosity at Calorific
vessels, pcs pcs e, % Fuel number,
g/cm3 400c, mm2/s value, kcal/kg
Quang CN
3 1.802 Cummins 835 46.34 D0 0.8300 52.1 2.419 10915
Ngai
Binh B5 0.8360 52.4 2.740 10850
4 2.006 Cummins 1034 51.55 B10 0.8389 52.5 2.848 10845
Dinh
B15 0.8474 52.7 2.963 10837
5 Phu Yen 411 Daewoo 131 3.87
Khanh (Quatest 3, Ho Chi Minh City, Vietnam)
6 634 Cummins 226 35.65
Hoa
Binh
Basis of the simulation:
7 1.032 Hino 551 53.39
Thuan The basis of the simulation process by KIVA-3V software
Ninh is based on the conservation equations [2, 7]:
8 145 Mitsubishi 42 28.97
Thuan
Total 6.400 2.958 Conservation of mass:
(Binh Dinh Province Maritime Branch, Vietnam)
m ( mui )   ( m /  ) 
+ =  D  + m + m
s c (1)
Table 2. Types of engines commonly used as the main engine of t xi xi  xi 
fishing vessel
Number in the total Momentum equation:
o Ratio,
N Survey engine type number of engines
%
(  u j )  p
( )
surveyed, pcs
+  ui u j −  ij = − +  Fjs +  g j (2)
1 Cummins 2619 40.92 t xi x j
2 Hino 1211 18.92
Conservation of energy:
3 Mitsubishi 635 9.92
 T T   2T   ( m /  ) 
4 Yanmar 468 7.31 cp 
 t
+ =k 2 +
xi  xi xi
 D


h
m
m
xi



(3)
5 Komatsu 450 7.04
+  + Q s + Q c
6 Daewoo 430 6.72
Turbulent flow model:
7 Hyundai 246 3.84
KIVA-3V software uses the k- RNG turbulent model
8 Other types 196 3.06
[2-3]:
9 Doosan 37 0.58
 k 2
+  (  uk ) = −  k u +  : u
10 Isuzu 94 1.47 t 3 (4)
11 Niigata 14 0.22 +  ( k k ) −  + W s
Total 6.400 100%  2 2 k 

+  (  u ) = −  C1 − C3 + C C u  u
(Binh Dinh Province Maritime Branch, Vietnam) t  3 3  
+  (  ) (5)
II. MATERIALS AND RESEARCH METHODS

( 1
+ C − C ) : u − C2  + CsW s 
In this study, the paper uses KIVA-3V software to k 
simulate the effects of biodiesel fuel on combustion and
emissions of Cummins NTA855 diesel engines. Engine In equation (5):
parameters are shown in Table 3. Biodiesel fuel derived from
fish fat: B5 (5% biodiesel B100 and 95% diesel oil), B10 −1 + 2C − 3m ( m − 1) + ( −1)8 6C C 
1 1 2  
(10% biodiesel B100 and 90% diesel oil), B15 (15% C =
3 3
biodiesel B100 and 85% diesel oil) are shown in Table 4. (6)
 
 1 −  /  
Table 3. Engine and simulation parameters [17] C =  0  , = S k
 1 +  3 3

Parameters Value
Engine/ Chamber type Cummins NTA 855/Unified In the above equations:
Number of cylinders 6 T – temperature, K; t – time, s; xi – coordinates, m; p –
Cylinder diameter, mm 140 pressure, N/m2; ρ – density of liquid, kg/m3; ρm – density of a
Piston stroke, mm 152 substance m, kg/m3; ui, uj – velocity of the fluid element, m/s;
tij – components of the stress tensor, N/m2; μ – dynamic
Compression ratio 14.1:1
viscosity, kg/ms; cp - isostatic specific heat, J/kgK; hm –
The engine speed of simulation, rpm 1200 enthalpy, J/kgK;  ms - the density of substance m when
200
spraying, kg/m3;  mc - density of substance m when burning, is higher than that of B15. Fuel B15 has the highest viscosity,
kg/m3; F js - the rate of momentum due to the spray, kgm/s; so the amount of evaporation fuel is reduced compared to
other fuels, leading to a decrease in combustion pressure in
Q s - mass component due to injection, kg/m3s; Q c - mass the cylinder.
composition due to combustion, kg/m3s; ws - Source terms
interaction with the spray, kg/m3s; k - turbulent kinetic
energy, kJ/kg;  - dissipation rate of turbulent kinetic energy
K, kJ/kgs;  - is the dependence ratio S; S - mean strain
magnitude; model constants: m1=0.5 và m2=1.4; Cs = 1.5; C
= 0.0845; C1 = 1.42; C2 = 1.68; k =  = 1.39; 0 = 4.38 và 
= 0.012 [2].
The fire model used in the simulation is the Shell model
[10].
The emission model used is the "multi-step" soot
formation model of Kazakov. A and D.E.Foster [8] and the
model of NOx formation according to the Zel'dovich
mechanism [3, 5-6].
The simulation grid model is built according to the Figure 2. Variation of combustion pressure of fuels
parameters, and the size of the engine's combustion chamber
is shown in Fig. 1 (1/4 of the combustion chamber). Temperature variation and heat release rate: In Fig. 3, at
the beginning of the combustion process, the heat release rate
is great, causing the pressure in the cylinder to increase
rapidly, when the rate of heat release decreases, the
combustion temperature also decreases. The maximum heat
release rate of fuel B5 and B10 is nearly equal to the
maximum heat of fuel DO. In the case of fuel B15 with the
lowest combustion temperature, as shown in Fig. 4, the heat
release rate is also lower than that of other fuels.
Figure 1. Cummins NTA855 engine combustion chamber

simulation grid
III. RESEARCH RESULTS AND DISCUSSION

The density of fuel increases with the mixing ratio,
leading to an increase in the spray penetration and the fuel
particle size. However, because the viscosity and especially
the surface tension of biodiesel fuel is greater than that of
diesel (DO), it hinders the development of the jet. In addition,
parameters such as cetane number, and calorific value are Figure 3. The heat release rate of fuels
factors affecting the process of mixture formation,
combustion, and engine emissions:
Variation of combustion pressure in the cylinder: Fig. 2
shows the variation of combustion pressure of the fuels.
Diesel fuel has the highest maximum combustion pressure,
which is due to the increased amount of evaporation fuel,
leading to an increase in combustion. When changing the
proportion of biodiesel in the fuel from 5% of B100 (B5) to
15% of B100 (B15), the cetane index increases with the
increase in the ratio, causing the combustion ignition time to
decrease accordingly. The change in ignition time can be seen
on the graph at the beginning of the increase in pressure. For
fuels B5 and B10, the process of injection and mixture
formation is similar to that of DO fuel, because the mixture
is more homogeneous. So the maximum combustion pressure Figure 4. Combustion temperatures of fuels
201
Because the fuels have different viscosity, calorific value,

and cetane number, the process of mixture formation and fuel
combustion is also different, which is also the cause of
different exhaust emissions. The temperature distribution of
fuels in the combustion chamber of a diesel engine is shown
in Fig. 5.
Figure 6. NOx emissions of fuels
Figure 7. Soot emissions of fuels
Figure 8. Comparison of the maximum emissions of biodiesel and

diesel fuel
Figure 5. Distribution of combustion temperatures of fuels in the
engine combustion chamber at 200 ATDC IV. CONCLUSION AND FUTURE WORK
Exhaust gas emissions: The combustion temperature of A. Conclusion
fuel DO, B5, and B10 is high, so NOx emissions increase. The
combustion temperature of B15 fuel is low, so NOx emissions Using KIVA-3V software simulated pressure,
are reduced because NOx emissions are formed at high combustion temperature in the cylinder, and the level of soot,
temperatures, NOx emission rules are shown in Fig. 6. On the NOx emissions of Cummins NTA855 diesel engine when
graph of Fig. 7, soot emission of fuel DO and B15 is the using biodiesel fuel from fish fat with different mixing ratios.
highest because DO fuel contains a lot of sulfur, leading to a Limit the time and cost before the experimental setup. The
large amount of soot formed. B15 fuel has a low combustion simulation results show that the soot emission of the engine
temperature, so soot oxidation is reduced because the soot is much reduced when using B5 and B10 fuel.
emission is the difference between the mass of soot formed B. Future Work
and the mass of soot oxidized at high temperature. Compare
the emission results of the fuels shown in Fig. 8. The experimental research on diesel engines of the fishing
vessels using fuel B5 and B10.
202
ACKNOWLEDGMENT [10] S. Kong and R. Reitz, Multidimensional Modeling of Diesel

Ignition and Combustion Using a Multistep Kinetics Model,
For the successful implementation of this research, we ASME Transactions, 1995.
would like to express our sincere thanks to the Faculty of [11] S Sinha, A K Agarwal, Experimental Investigation of The
Information Technology, Yersin University, Da Lat, Combustion Characteristics of Biodiesel (Rice-Bran Oil
Vietnam. Methyl Ester) - Fuelled Direct-Injection Transportation Diesel
Engine, Engineers, Part D. Journal of Automobile
REFERENCES Engineering, First Published August 1, 2017.
[1] Phan Van Quan, “Pilot Application of Catfish Oil for Inland [12] Yonwaba Sinuka, “Performance Testing of a Diesel Engine
Waterway Vessels Install The Diesel Engine in The Mekong Running on Varying Blends of Jatropha Oil, Waste Cooking
Delta”, Project Summary Report under the National Target Oil and Diesel Fuel”, Degree of Doctor of Philosophy,
Program Copes with Climate Change, University of Mechanical Engineering, Renewable Energy in The Faculty
Transportation, Ho Chi Minh City, 2015. of Engineering at The Cape Peninsula University of
Technology, 2016.
[2] Amsden, KIVA-3V Release -Los Alamos LA-UR-99-915,
1999. [13] VASEP. (2021, Mar. 17). Sản xuất biodiesel từ mỡ cá tra
[Online]. Available: http://vasep.com.vn/san-pham-xuat-
[3] A. Lakshminarayanan, Yogesh V. Aghav, Modelling Diesel
khau/ca-tra/nguyen-lieu/san-xuat-biodiesel-tu-mo-ca-tra-
Combustion, Springer Dordrecht Heidelberg London New
21334.html.
York, ISSN 0941-5122, 2010.
[14] MOST. (2020, Aug. 18). Hoàn thiện công nghệ sản xuất liên
[4] Carsten Baumgarten, Mixture Formation in Internal
tục Diesel sinh học gốc B100 từ nguồn nguyên liệu axit béo
Combustion Engines, Springer. ISBN3540308350, 2010.
phế thải và dầu hạt Jatropha Curcas [Online]. Available:
[5] K. Mollenhauer, H. Tschoeke, Handbook of Diesel Engines, https://www.most.gov.vn/vn/tin-tuc/18205/hoan-thien-cong-
DOI 10.1007/978-3-540-89083-6, Springer-Verlag Berlin nghe-san-xuat-lien-tuc-diesel-sinh-hoc-goc-b100-tu-nguon-
Heidelberg 2010. nguyen-lieu-axit-beo-phe-thai-va-dau-hat-jatropha-
[6] Han, Z., and R.D.Reizt, “A Temperature Wall Function curcas.aspx.
Formulation for Variable-Density Turbulence Flow with [15] VNEEP. (2012, Oct. 12). Triển vọng nhiên liệu sinh học từ
Application to Engine Convective heat Transfer Modeling”. dầu phế thải [Online]. Available:
Submitted to International Journal of Heat and Mass Transfer, http://tietkiemnangluong.com.vn/tin-tuc/khoa-hoc-cong-
1995. nghe/t13518/trien-vong-nhien-lieu-sinh-hoc-tu-dau-phe-
[7] Haiying Li, “CFD Modelling Study of Sprays and Combustion thai.html
of Gasoline and DMF in Direct Injection Gasoline Engines”, [16] Department of Science and Technology - MOIT. (2021, Mar,
A thesis of Doctor of Philosophy, University of Birmingham 01). Sản xuất biodiesel với giá thành thấp: Một bước ngoặt
Research Archive, 2013. mới [Online]. Available: https://khoahocphattrien.vn/cong-
[8] Kazakov, A., and D.E.Foster, “Modeling of Soot Formation nghe/san-xuat-biodiesel-voi-gia-thanh-thap-mot-buoc-
During DI Diesel Combustion Using a Multi-Step ngoat.html.
Phenomenological Model”. SAE Paper No. 982463, 1998.
[17] ARTSTC. Cummins NTA855 diesel engines [Online].
[9] Mai Duc Nghia, Pham Dinh Trung, “Investigations of The Available: https://www.artstc.com/marine/nt855-diesel-
Combustion and Exhaust Emission Process of The Fishing engine.html.
Vessel's Diesel Engines Using Biodiesel Made From Fish Fat
by Avl - Fire Software”, International Liberty
Interdisciplinary Studies Conference, Manhattan, New York,
January 16-17, 2022.
203
A Study on Mechanical Properties of Ca-alginate

Hydrogels
Thanh Tan Nguyen Van Tron Tran
Faculty of Mechanical Engineering Faculty of Mechanical Engineering
Education Education
Ho Chi Minh City 71307, Vietnam Ho Chi Minh City 71307, Vietnam
tannt@hcmute.edu.vn trontv@hcmute.edu.vn
Long Nhut-Phi Nguyen Nguyen Thi My Le

Faculty of Mechanical Engineering Faculty of Chemical and Food Technology
Education Education
Ho Chi Minh City 71307, Vietnam Ho Chi Minh City 71307, Vietnam
longnnp@hcmute.edu.vn lentm@hcmute.edu.vn
Abstract: Containing a three-dimensional cross-linked fields [3,8,10–12]. For instance, polyacrylamide hydrogels
network of hydrophilic polymers, hydrogels can absorb and coated with hydrophobic substances were employed for
retain a large amount of water. They show great potential fabricating oil-water separation devices, which rapidly and
for applications in both biomedical and engineering fields. efficiently collected spilled oils with a wide range of
Among the numerous developed hydrogels, calcium cross- viscosities [10]. In another report [8], biocompatible and
linked alginate (Ca-alginate) hydrogels have gained biodegradable poly(disulfide) hydrogels exhibiting extremely
particular attention due to their fascinating properties such high self-healing rate in multiple environments could be used
as high-water content, biocompatibility, biodegradability, as a biological adhesive material.
and ease of fabrication. Recently, Ca-alginate hydrogels Alginate extracted from brown seaweeds is a negatively
have been extensively used in biomedical applications such charged polymer that is composed of mannuronic acid (M
as scaffolding, drug delivery, and wound treatment. unit) and guluronic acid (G unit) [1]. It has biocompatible,
Moreover, the gels are also regarded as a promising material biodegradable, and high water-soluble abilities [13‒15]. The
for replacing traditional plastics for underwater uses to save G block of alginate easily cross-links with multivalent ions
fossil fuels. To afford these applications, Ca-alginate for forming “egg-box” structure for gelation [1,14,15].
hydrogels have to exhibit rational and tunable mechanical Among the multivalent ions, calcium is the most widely
properties, including stiffness, strength, and toughness. To used cation (Ca2+) applied to fabricate ionic alginate gels
systematically evaluate the mechanical properties of Ca- [16]. Due to its biological capabilities, alginate has been
alginate hydrogels, we prepare various Ca-alginate employed extensively for preparing alginate-based
hydrogels via a diffusion method by varying concentrations hydrogels cross-linked by Ca2+ for medical applications
of Ca2+ ion from 0.1 M to 2.5 M and alginate polymer from such as wound dressing [17], drug delivery systems [13],
2 wt% to 5 wt%. The mechanical properties of the water- and tissue engineering [15,18]. Moreover, considered a
equilibrated hydrogels are determined by tensile tests. The relatively low-cost compared to other polymeric materials,
results indicate that the gels with a water content of >91 alginate is a prominent polymer for fabricating water-based
wt% exhibit the highest average Young’s modulus, tensile substances that are alternative to fossil-based materials to
strength, and work of extension of 1090.4 ± 123.8 kPa, 566 save fossil fuels and the environment [4,19]. Therefore,
± 23.6 kPa, and 641.9 ± 40.6 kJ/m3, respectively. many attempts have been made recently to improve the
mechanical properties of alginate-based hydrogels by
Keywords: hydrogels, alginate, stiffness, strength, developing new fabricating approaches [1,15,20]. For
toughness example, Z. Suo and co-workers successfully developed
highly stretchable and tough hydrogels fabricated by
I. INTRODUCTION combining two networks of Ca-alginate and polyacrylamide
Hydrogels are three-dimensional networks of cross- in a system [1]. The developed hydrogels achieved
linked hydrophilic polymers that can absorb and retain large remarkable mechanical properties with tensile strength of
quantities of water or biological liquids. They have attracted over 200 kPa, Young’s modulus of over 100 kPa, and
particular attention in recent years due to their excellent toughness of ~ 9,000 J/m2 [1]. From the above discussion, it
flexibility and strength [1,2], rapid self-recovery rate [3,4], is found that Ca-alginate-based hydrogels have high
outstanding biocompatibility and biodegradability [5,6], potential for various applications. To afford this, they should
extraordinary self-healing ability [7,8], and high electrical exhibit rational and tunable mechanical properties. Although
and thermal conductivity [3,9]. With these numerous many types of research related to Ca-alginate hydrogels
promising properties, hydrogels are one of the best candidate have been done to date, the studies systematically providing
materials for applications in both biomedical and engineering the mechanical properties of Ca-alginate hydrogels are

hardly found in the literature. Therefore, a systematical

evaluation of the mechanical properties of this material is
still necessary.
To overcome this issue, we systematically synthesized a
series of Ca-alginate hydrogels via a simple diffusion method
(Fig. 1). The gels were obtained by varying concentrations of
Ca2+ (from 0.1 M to 2.5 M) and alginate (from 2 wt% to 5 Figure 2. Fabrication process of the water-equilibrated Ca-
wt%). The mechanical properties of the water-equilibrated alginate/0.5/4 hydrogel. (a) 4wt% Na-alginate aq. solution was
hydrogels are determined by tensile tests. poured into a reaction mold until it was around half-filled. Then
0.5 M CaCl2 aq. solution was inserted from the open end. (b) The
initial hydrogel was obtained when complete gelation. (c) The
initial hydrogel was immersed under 0.5 M CaCl2 solution for one
day to ensure completely cross-linking and subsequently placed
underwater for around two days with multiple water replacements
for washing uncross-linked ions and polymers. (d) The final
equilibrium hydrogel was obtained for characterization.
C. Water Content
The water content of the water-equilibrated Ca-
alginate/x/y hydrogel was determined using the weight of wet
(Ww) and dried (Wd) hydrogel. After removal from water, the
wet specimen was carefully wiped with tissue paper to
absorb any excess water remaining on its surface and then
weighed to obtain Ww. Subsequently, the specimen was
completely dried at 120C for around twenty-four hours and
Figure 1. Schematic representations of Ca-alginate/x/y hydrogel Wd was measured. The water content was calculated as
follows:
A. Materials  Water content (wt%) = [(Ww − Wd)/Ww] × 100 (1)
Sodium alginate (Na-alginate, viscosity 80–120 cP) was D. Mechanical Property Characterization
purchased from FUJIFILM Wako Pure Chemical
The mechanical properties of the Ca-alginate/x/y
Corporation, Japan. Calcium chloride (CaCl2) was bought
hydrogels were determined by tensile test which was carried
from Samchun Pure Chemical Co., Ltd., Republic of Korea.
out in ambient conditions (temperature: ~ 28C, humidity: ~
All chemicals were used as received without further 65%) using a tensile strength tester (PT-1699vdo, PRO
purification. Distilled water was used for gel preparation. TEST, Taiwan) and a 50-kgf loadcell. The Ca-alginate/x/y
B. Preparation of Ca-alginate Hydrogels hydrogels with a thickness of ~ 2 mm were cut into
rectangular samples (width: ~ 10 mm, length: ~ 60 mm)
The calcium alginate [Ca-alginate/x/y] hydrogels, (inset of Fig. 3a), which are commonly employed for testing
where x, y denote the initial concentration of Ca2+ (M) and the mechanical properties of gel materials [1,2,4,8]. Both
weight percent of Na-alginate (wt%), respectively, were ends of the sample were mounted along the length direction
prepared by the following process. First, the predetermined in the tester while maintaining an initial length ~ 12 mm
amount of Na-alginate was dissolved in water to obtain a between the clamps. All tensile tests were carried out by
homogeneous solution. Next, a reaction mold with an upper- pulling the upper clamp upward with a constant deformation
end open was prepared by sandwiching two glass plates (150 speed of 1000% per min until the sample fractured. Three
mm × 120 mm), separated using a 3 mm spacer. Then the measurements were performed on each gel. To obtain
Na-alginate aqueous solution was inserted into the mold until Young’s modulus values, the tensile stress-strain curves were
it was around half-filled. Subsequently, CaCl2 aqueous linearly fitted over a strain range of ~ 2–10%.
solution was inserted into the mold from the open end, and
this solution was frequently replaced by the fresh one (Fig. III. RESULTS AND DISCUSSION
2a). Ca2+ ions downwards diffused through the Na-alginate
In this study, the Ca-alginate/x/y hydrogels were
solution for gelation. After complete gelation (Fig. 2b), the
fabricated via a facial diffusion method in which Ca2+
initial Ca-alginate hydrogel was taken out of the mold and
diffused through a Na-alginate aqueous solution for forming
immersed under the corresponding CaCl2 aqueous solution
physical ionic cross-links with G block of alginate for
for one day for completely cross-linking. After that, the gel
gelation. This method has been widely used for preparing
was placed underwater for around two days with multiple
various ionic cross-linked hydrogels with different cations
water replacements for washing uncross-linked ions and
[3,8,15]. As demonstrated, the mechanical properties of these
polymers (Fig. 2c). Finally, the water-equilibrated hydrogel
classified hydrogels have been strongly affected by the
was obtained and used for characterization (Fig. 2d).
concentration of cation and polymer [3,8,15]. Therefore, the
effect of Ca2+ concentration on mechanical properties of our
hydrogels was firstly evaluated. To do this, a series of the
Ca-alginate/x/4 hydrogels were prepared by varying Ca2+
concentration from 0.1 M to 2.5 M at a fixed 4 wt% of Na-
alginate aqueous solution. The mechanical properties of
205
water-equilibrated hydrogels were determined by tensile tests To evaluate the effect of alginate concentration on the
with a specimen showed in the inset of Fig. 3a. The results mechanical properties of the hydrogels, another series of the
indicated that an increase in Ca2+ concentration from 0.1 to Ca-alginate/0.5/y hydrogels were fabricated by increasing the
2.5 M led to an increase in the average Young’s modulus and Na-alginate from 2 wt% to 5 wt% at a fixed 0.5 M CaCl2
tensile strength of the gels from 574.9 ± 137.8 to 1090.4 ± concentration. It was found that when increasing alginate
123.8 kPa and 399.1 ± 35 to 566 ± 23.6 kPa (Fig. 3a,b and concentration, the mechanical properties of the hydrogels
Table 1), respectively, while average work of extension first including the average Young’s modulus, tensile strength, and
increased from 415.2 ± 37.1 to 508.5 ± 30.6 kJ/m3 and then work of extension significantly increased from 201 ± 18.9 to
decreased to 475.9 ± 185.4 kJ/m3 due to a decrease in tensile 708.2 ± 133.8 kPa, 169.1 ± 52 to 542 ± 90 kPa, and 166.9 ±
strain (Fig. 3c and Table 1). Moreover, the water content of 46.3 to 641.9 ± 40.9 kJ/m3 (Fig. 4a‒c and Table 2),
the gels decreased together with an increase in Ca2+ respectively. The results implied that an increase in density
concentration (Fig. 3d and Table 1). The results were of the backbone polymer maintained the gel network's
consistent with previous reports [3,8]. Low Ca2+ homogenous structure and helped retain its stretchability.
concentration is insufficient for forming stable ionic cross- Therefore, an increase in alginate concentration led to an
link, whereas high Ca2+ concentration has a disruptive effect increase in both Young’s modulus and tensile strength and
in which forming inhomogeneous structure consisting of work of extension of the hydrogel. However, the water
both low and high ionic cross-links in the gel network [8]. content of the hydrogels decreased with an increase in Na-
Among them, 0.5 M Ca2+ concentration resulted in the alginate concentration (Fig. 4d and Table 2). Particularly, the
formation of the Ca-alginate/0.5/4 hydrogel with a water Ca-alginate/0.5/2 hydrogel exhibited a water content of 96.9
content of 94.9 ± 1 wt% exhibiting the highest average work ± 0.4 wt%, while it was 93.3 ± 0.4 wt% for Ca-alginate/0.5/5
of extension (508.5 ± 30.6 kJ/m3) and remarkable average hydrogel (Fig. 4d and Table 2).
Young’s modulus (702.2 ± 150.1 kPa) and tensile strength
(494 ± 54.4 kPa). Thus this Ca2+ concentration was selected
for further study.
Figure 4. Effect of Na-alginate concentration on the mechanical

properties and water content of the Ca-alginate/0.5/y hydrogels (y
= 2, 3, 4, and 5 wt%). (a) Typical tensile stress-strain curve of the
Figure 3. Effect of Ca2+ concentration on the mechanical resulted hydrogels. (b) Young’s modulus and tensile strength, (c)
properties and water content of the Ca-alginate/x/4 hydrogels (x = work of extension, and (d) water content of the hydrogels. In plots
0.1, 0.5, and 2.5 M). (a) Typical tensile stress-strain curves of the (b‒d), error bars indicate mean absolute deviations (n = 3).
hydrogels. The inset showed a tensile specimen of the Ca-
alginate/0.5/4 hydrogel with ~ 60 mm × ~ 10 mm × ~ 2 mm
Table 2. Mechanical properties and water content of the Ca-
(length × width × thickness). (b) Young’s modulus and tensile alginate/0.5/y hydrogels
strength, (c) work of extension, and (d) water content of the
hydrogels. In plots (b‒d), error bars indicate mean absolute Sample
Na-alginate concentration (wt%)
deviations (n = 3). 2 3 4 5
Young’s
modulus 201 ± 18.9 386.4 ± 116.4 702.2 ± 150.1 708.2 ± 133.8
Table 1. Mechanical properties and water content of the Ca- (kPa)
alginate/x/4 hydrogels
Tensile
Ca2+ concentration (M) strength 169.1 ± 52 354.1 ± 53.9 494 ± 54.4 542 ± 90
Sample (kPa)
0.1 0.5 2.5
Young’s modulus Tensile
574.9 ± 137.8 702.2 ± 150.1 1090.4 ± 123.8 strain 153 ± 12 160 ± 16 171 ± 24 191 ± 21
(kPa)
Tensile strength (%)
399.1 ± 35 494 ± 54.4 566 ± 23.6 Work of
(kPa) 166.9 ±
Tensile strain extension 338.7 ± 105.4 508.5 ± 30.6 641.9 ± 40.9
172 ± 26 171 ± 24 130 ± 40 46.3
(%) (kJ/m3)
Work of extension Water
415.2 ± 37.1 508.5 ± 30.6 475.9 ± 185.4 content 96.9 ± 0.4 95.7 ± 0.5 94.9 ± 1 93.3 ± 0.4
(kJ/m3)
Water content (wt%)
95.2 ± 0.2 94.9 ± 1 91.5 ± 0.3
(wt%)
206
IV. CONCLUSION [8] V. T. Tran, Md. T. I. Mredha, J. Y. Na, J.-K. Seon, J. Cui, and
I. Jeon, “Multifunctional poly(disulfide) hydrogels with
In this research, we systematically fabricated various Ca- extremely fast self-healing ability and degradability”, Chem.
alginate hydrogels via a facile diffusion method by varying Eng. J., vol. 394, 124941, August 2020.
concentrations of Ca2+ and alginate and evaluated their [9] V. T. Tran, Md. T. I. Mredha, Y. Lee, M. Todo, H. So, E.
mechanical properties by tensile test. The results indicated Jeong, W. Park, and I. Jeon, “Electrically, Thermally, and
that increasing Ca2+ concentration led to a continuous Mechanically Anisotropic Gels with a Wide Operational
Temperature Range”, Adv. Funct. Mater., vol. 32, 2110177,
increase in Young’s modulus and tensile strength of the April 2022.
hydrogel, whereas work of extension first increased and then [10] V. T. Tran, X. Xu, Md. T. I. Mredha, J. Cui, J. J. Vlassak, and
decreased due to decreasing stretchability of the gel, which is I. Jeon, “Hydrogel bowls for cleaning oil spills on water”,
ascribed to forming inhomogeneous gel structure at high Water Res., vol. 145, pp. 640-649, November 2018.
Ca2+ concentration. By contrast, increasing alginate [11] H. H. Le, V. T. Tran, Md. T. I. Mredha, J. Y. Na, J.-K. Seon,
concentration increased both Young’s modulus and tensile and I. Jeon, “Thin-film hydrogels with superior stiffness,
strength and work of extension of the hydrogel because its strength, and stretchability”, Extreme Mech. Lett., vol. 37,
100720, May 2020.
homogeneous structure remained, which helped retain its
[12] S. Hong, D. Sycks, H. F. Chan, S. Lin, G. P. Lopez, F. Guilak,
stretchability. The results provide the systematical evaluation K. W. Leong, and X. Zhao, “3D Printing of Highly
of the mechanical properties of the Ca-alginate hydrogels Stretchable and Tough Hydrogels into Complex, Cellularized
and more insight into the effect of Ca2+ and alginate Structures”, Adv. Mater., vol. 27, pp. 4035–4040, July 2015.
concentrations on their mechanical properties. [13] F. Abasalizadeh, S. V. Moghaddam, E. Alizadeh, E. Akbari,
E. Kashani, S. M. B. Fazljou, M. Torbati, and A. Akbarzadeh,
ACKNOWLEDGMENT “Alginate-based hydrogels as drug delivery vehicles in cancer
treatment and their applications in wound dressing and 3D
This work was supported by Ho Chi Minh City bioprinting”, J. Biol. Eng., vol. 14, 8, March 2020.
University of Technology and Education, Ho Chi Minh City, [14] R. Teshima, Y. Kawano, T. Hanawa, and A. Kikuchi,
Vietnam. “Preparation and evaluation of physicochemical properties of
novel alkaline calcium alginate hydrogels with carbonated
REFERENCES water”, Polym. Adv. Technol., vol. 31, pp. 3032–3038,
December 2020.
[15] Md. T. I. Mredha, V. T. Tran, S.-G. Jeong, J.-K. Seon, and I.
[1] J.-Y. Sun, X. Zhao, W. R. K. Illeperuma, O. Chaudhuri, K. H. Jeon, “A diffusion-driven fabrication technique for
Oh, D. J. Mooney, J. J. Vlassak, and Z. Suo, “Highly anisotropic tubular hydrogels”, Soft Matter, vol. 14, pp. 7706–
stretchable and tough hydrogels”, Nature, vol. 489, pp. 133– 7713, October 2018.
136, September 2012.
[16] H. Zhang, J. Cheng, and Q. Ao, “Preparation of Alginate-
[2] Md. T. I. Mredha, S. K. Pathak, V. T. Tran, J. Cui, and I. Based Biomaterials and Their Applications in Biomedicine”,
Jeon, “Hydrogels with superior mechanical properties from Mar. Drugs, vol. 19, 264, May 2021.
the synergistic effect in hydrophobic–hydrophilic
copolymers”, Chem. Eng. J., vol. 362, pp. 325–338, April [17] M. C. Straccia, G. G. d’Ayala, I. Romano, A. Oliva, and P.
2019. Laurienzo, “Alginate hydrogels coated with chitosan for
wound dressing”, Mar. Drugs, vol. 13, pp. 2890–2908, May
[3] V. T. Tran, Md. T. I. Mredha, S. K. Pathak, H. Yoon, J. Cui, 2015.
and I. Jeon, “Conductive Tough Hydrogels with a Staggered
Ion-Coordinating Structure for High Self-Recovery Rate”, [18] K. Y. Lee and D. J. Mooney, “Alginate: Properties and
ACS Appl. Mater. Interfaces, vol. 11, pp. 24598–24608, July biomedical applications”, Prog. Polym. Sci., vol. 37, pp. 106–
2019. 126, January 2012.
[4] V. T. Tran, Md. T. I. Mredha, and I. Jeon, “High-water- [19] Q. Wang, J. L. Mynar, M. Yoshida, E. Lee, M. Lee, K. Okuro,
content hydrogels exhibiting superior stiffness, strength, and K. Kinbara, and T. Aida, “High-water-content mouldable
toughness”, Extreme Mech. Lett., vol. 37, 100691, May 2020. hydrogels by mixing clay and a dendritic molecular binder”,
Nature, vol. 463, pp. 339–343, January 2010.
[5] B. Yang, Y. Zhang, X. Zhang, L. Tao, S. Li, and Y. Wei,
“Facilely prepared inexpensive and biocompatible self- [20] M. T. I. Mredha, Y. Z. Guo, T. Nonoyama, T. Nakajima, T.
healing hydrogel: A new injectable cell therapy carrier”, Kurokawa, and J. P. Gong, “A Facile Method to Fabricate
Polym. Chem., vol 3, pp. 3235–3238, August 2012. Anisotropic Hydrogels with Perfectly Aligned Hierarchical
Fibrous Structures”, Adv. Mater., vol. 30, 1704937, March
[6] F.-Y. Hsieh, L. Tao, Y. Wei, and S.-h. Hsu, “A novel 2018.
biodegradable self-healing hydrogel to induce blood capillary
formation”, NPG Asia Mater., vol. 9, e363, March 2017.
[7] I. Jeon, J. Cui, W. R. K. Illeperuma, J. Aizenberg, and J. J.
Vlassak, “Extremely Stretchable and Fast Self-Healing
Hydrogels”, Adv. Mater., vol. 28, pp. 4678–4683, June 2016.
207
Mechanical Characteristics of PBT Based Blend

Hoang-Khang Lu Ngoc Tran-Nhu Nguyen Huy Huynh-Nhat Do
Faculty of Mechanical Engineering Faculty of Mechanical Engineering Faculty of Mechanical Engineering
HCMC University of Technology HCMC University of Technology HCMC University of Technology
and Education and Education and Education
18104020@student.hcmute.edu.vn 18104035@student.hcmute.edu.vn 18104016@student.hcmute.edu.vn
Van-Huong Hoang Van-Thuc Nguyen Nga Thi-Hong Pham

huonghv@hcmute.edu.vn nvthuc@hcmute.edu.vn hongnga@hcmute.edu.vn
Van-Tron Tran Long Nhut-Phi Nguyen Thanh-Tan Nguyen

trontv@hcmute.edu.vn longnnp@hcmute.edu.vn tannt@hcmute.edu.vn
Abstract: Polybutylene terephthalate (PBT) is known to I. INTRODUCTION

be a brittle material and is limited in applications where PBT is widely used due to its high value in strength,
impact strength is required. To improve the impact strength hardness, toughness, and good dielectric properties [1]. PBT
of PBT, in this report, we conduct a study to mix PBT with is also a brittle material with poor impact strength, so it is not
Thermoplastic Polyurethane (TPU) 20% and Polycarbonate widely used in products requiring high-impact strength. In
(PC) 20% by weight to improve the impact strength of PBT. contrast, thermoplastic polyurethanes (TPUs) have good
Prepare samples by injection molding in proportions of impact toughness and low-temperature flexibility. TPU is a
100% PBT, 20% TPU, and 20% PC blends. These samples linear copolymer consisting of soft and hard segments
are then tested for flexural strength, tensile strength, and separated by microscopic molecules.
impact toughness. The results show that the 20%TPU
Recently, to improve the durability of PBT, many authors
composite is 15% higher than the neat PBT and has an have mixed it with other resins [2,3]. For example, Tehran et
impact toughness of 4.23 kJ/m2. However, its tensile and al. [4] studied to improve PBT/TPU tensile strength. The
flexural strengths are 28.38 MPa and 56.24 MPa, about 23% result indicated that was system containing PBT/TPU/clay
and 16% lower, respectively, followed by a neat PBT. nanocomposite (80/20/3) has excellent tensile and impact
Furthermore, the 20% PC blend exhibits an impact strength properties by adding clay nanoparticles to the PBT/TPU
of 3.37 kJ/m2, decreasing 7.9% compared to a neat PBT. composite. In another report by Yanping [5], the addition of
While its tensile strength increased by about 42%, and the TPU and PBT improved the strength of the composite and
flexural strength of composite increased by 47.23%. In the preparation by melting their mixture. Besides, adding
general, increased toughness leads to a decrease in tensile carbon nanotubes (CNTs) to the PBT/TPU composite
and flexural strength and vice versa. SEM images significantly improved tensile strength, flexural strength, and
demonstrate the mixed microstructure of 20% TPU and 20% elastic modulus [6]. Thermodynamic analysis shows that
PC before and after fracture to be more clearly demonstrated incorporating CNTs into PBT/TPU increases the transition
through SEM images to study the adhesion between TPU, temperature of the glass. In another study, Palanivelu et al.
PC, and PBT matrices and explain the fracture mechanism [7] demonstrated that adding rubber TPU to PBT increased
broken. This study provides a deeper insight into the impact strength and reduced tensile and flexural strengths of
influence of TPU and PC on the mechanical properties of the PBT/TPU composite. Huang et al. [8] have shown that
PBT composite. carbon fiber can evenly distribute in TPU/PBT matrix and
has no agglomeration phenomenon. The study was carried
Keywords: polybutylene terephthalate, thermoplastic out by adding carbon fiber to TPU/PBT composite. Research
by Hale et al. [9] demonstrated that the combination of
polyurethane, PBT/TPU blend, microstructure, impact
methyl methacrylate, glycidyl methacrylate, ethyl acrylate
toughness
terpolymer, and polytyl methacrylate-glycidyl methacrylate-
ethyl acrylate terpolymer into PBT as a reactive coupler
could improve low-temperature Izod impact resistance. The

fracture properties depend on the mixing order of the

composite components. However, the extent of the effect of
PBT still needs to be investigated further, especially in the
EBT mix.
In this report, we investigate the mechanical
characteristics of PBT-based blends, which are prepared by
adding TPU or PC to PBT. An injection molding technique
prepares standard samples for flexural, tensile, and impact
tests. SEM also analyzes the samples to reveal the fracture
surface microstructure.
II. MATERIALS AND METHOD
The experiments used PBT from POCAN B4225 Lanxess a) PBT sample
Germany, TPU code TPU- F-X9190, China. The various
PBT/TPU composition was mixed in weight percent, as
displayed in Table 1. They were dried at 120°C for 6 to 8
hours to ensure their humidity which is less than 0.03%.
Table 1. Compositions of the samples (wt.%)
Sample TPU (wt.%) PC (wt.%) PBT (wt.%)
PBT 0 0 100
20TPU 20 - 80
20PC - 20 80
The sample employed for testing were repaired following

the Flexural test ASTM D790, the Tensile test ASTM D638, b) 20TPU
and the Charpy ISO 179-1 standards. The Universal Testing
Machine was used for Flexural and tensile tests. For the
Charpy impact testing, the Tinius Olsen IT504 was used. The
sample's microstructure was observed using a high-resolution
SEM microscope HITACHI S-4800. The basic parameters of
SEM are the resolution of secondary electronic image 1.0
nm, 1.4 nm, accelerated voltage reducer); 2.0 nm (1 kV, WD
= 1.5 nm, conventional model); high magnification HM 100-
800000 times; magnification: LM 20-2000 times.
Fig. 1 shows the impact samples of neat PBT, 20 TPU,
and 20TPU blends prepared by injection molding. The
samples obtained from neat PBT and the 20TPU blends have
opaque colors. c) 20PC
Figure 2. The FESEM micrographs of (a) neat PBT, (b)

20TPU blend, and (c) 20PC blend before testing
B. The tensile strength testing

Fig. 3 shows the stress-strain curves of the tensile
sample. Compared to other samples, the neat PBT samples
Figure 1. Surface morphology
have relatively high tensile strength and low fracture strain
A. Microstructure before testing due to their brittle characteristic. Adding 20% TPU results
in a lower tensile strength value as it decreases from 37.13
The FESEM microstructures of these samples before
MPa to 13.51 MPa. However, the 20TPU sample has a
testing are shown in Fig. 2. The PBT sample has
greatly higher elongation (~7 times) than neat PBT. It means
homogenous structures, as shown in Fig. 2a. On the neat
PBT sample surface, there are sheets and ripple matrices. that adding 20% TPU could improve the fracture strain of
However, the structures of the 20TPU and the 20PC blends the PBT blend. For the 20PC blend, the strain curve was
show two seprate regions in which the dispersed droplets (for 12.5 times higher than the neat PBT. Interestingly, the
TPU phases) separated from the PBT matrix, as shown in tensile strength value of the 20PC is also higher than the
Fig. 2b and Fig. 2c. The adhesion between TPU and PBT in neat PBT sample. Therefore, adding PC can improve the
the 20TPU blend is not evidently found as well as PC and fracture strain and the tensile strength of the PBT samples.
PBT, as the surface of the particles is prominent.
209
2022 6th International Conference on Green Technology
PBT1 and Sustainable Development (GTSD)
50 PBT2
PBT3
PBT4
40 PBT5 80
64.46
Stress (MPa)
Tensile Strength (MPa)

30
60
20
40 37.13
10
20 13.51
0
0.0 0.2 0.4 0.6 0.8 1.0
20TPU1
20 Strain (%)
20TPU2
0
20TPU3 PBT 20TPU 20PC
a) PBT sample 20TPU4
20TPU5 Figure 4. The average tensile strength of test samples
15
Stress (MPa)
C. The flexural strength testing

PBT1
10 80 PBT2
PBT3
PBT4
PBT5
60
5
Stress (N/mm2)
40
0
0 1 2 3 4
20
Strain (%)
b) 20TPU
0
0 1 2 3
Strain (%)
a) PBT sample
20TPU1
30 20TPU2
20TPU3
20TPU4
25 20TPU5
Stress (N/mm2)
20
15
10
c) 20PC
5
Figure 3. Tensile strain stress chart 0
Fig. 4 shows the average tensile strength of these 0 2 4 6 8 10 12
samples. The tensile strength of the neat PBT, 20TPU, and Strain (%)
20PC is 37.13, 13.51, and 64.46 MPa, respectively. The b) 20TPU
results show that adding 20% TPU leads to a 63.6%
decrease in tensile strength, while adding 20% PC results in
a 42% increase in tensile strength. This result may be due to
the glass transition temperatures of PBT and TPU, 55-65oC
and -44oC, respectively. So PBT has high tensile strength
and low elongation. In contrast, the neat TPU displays
opposing properties at these temperatures. This result shows
that when mixing PBT with TPU, the glass transition
temperature of the mixture should be lower than neat PBT.
Therefore, the 20TPU blend has higher elongation and
lower tensile strength. Ahad et al. [10] also showed that the
tensile strength and elastic modulus of the PBT/TPU
composite decreased with increasing TPU content.
Furthermore, the presence of TPU elastic domains in PBT
c) 20PC
acts as a stress concentrator, resulting in lower stress
Figure 5. Flexural strain stress chart
distribution of the composite than the neat PBT.
210
The flexural stress-strain curves of the compositions are strength of the neat PBT, 20TPU blend, and 20PC blend are
shown in Fig. 5. It is found that the PBT exhibits the highest 3.66, 7.47, and 3.37 kJ/m2, respectively. It shows that when
strength, whereas its fracture strain is the lowest. Adding adding 20TPU content, the mixture has an increase in
20% TPU causes a decrease in flexural strength and an impact strength (104.1%), while adding 20PC content, the
increased fracture strain (~11%). For the 20PC blend, the impact strength of the blend decreases (7.9%). The results
elongation at break value is 4.3 times higher than the neat show an opposition between PBT/TPU and PBT/PC blends,
PBT, pointing out that PC can also significantly improve the an increase in the TPU content leads to an increased impact
fracture strain of the PBT. Similar to the tensile strength toughness of the mixture. However, adding more PC to PBT
result, adding 20% PC to PBT also give rise to the flexural resulted in a slight decrease in the impact toughness. These
strength from about 70 MPa to 95 MPa. results are consistent with Ahad et al. report [12].
The poor glass transition temperature compatibility of E. Microstructure of fracture surface
TPU and PBT may have made the average flexural strength
of the 20TPU composite lower than that of the PBT of 21.39
MPa and 66.89 MPa, respectively. According to Huang et
al. [7], this decrease in flexural strength is due to TPU
disrupting the structural regularity of PBT. Therefore, the
crystallinity of the compositions is reduced as the TPU
content increases, leading to a reduction of the flexural
strength. However, the flexural strength of 20TPU is higher
than that of the neat TPU (17.75 MPa). This result showed
that, in the PBT matrix, adding a small amount of TPU
resulted in a slight decrease in flexural strength, which is
consistent with the previous report [11]. Notably, adding
20PC content significantly increases flexural strength
(47.23%) due to the higher strength of the PC material.
120 a) PBT sample
98.48
100
Flexural Strength (MPa)
80
66.89
60
40
21.39
20
0
PBT 20TPU 20PC
Figure 6. Flexural strain stress chart
D. Charpy Impact Testing b) 20TPU
10
7.47
Impact Strength (kJ/m2)
6 3.37
4 3.66
0
PBT 20TPU 20PC
Figure 7. The impact strength of the PBT, PBT/20% TPU,
c) 20PC
and PBT/20% PC
Figure 8. The microstructure of (a) neat PBT, (b) 20TPU
The impact strength of PBT, PBT/20% TPU, and
blend and (c) 20PC blend
PBT/20% PC samples is shown in Fig.7. The impact
211
During fracturing, the crack propagates from the initial wollastonite”. Transactions of the Indian Institute of Metals,
crack zone to the crack propagation zone and then to the vol. 64, no. 1, pp. 127-132, 2011.
rapid fracture zone. The SEM picture of the fracture surface [4] Ata Chalabi Tehran, Karim Shelesh-Nezhad, Pouya Faraji
of the neat PBT sample after the impact test is shown in Fig. Kalajahi et al., “A Study of the Effect of TPU and Clay
8a. Radical crack lines are formed in the crack propagation Nanoparticles on the Mechanical Behavior of PBT-Based
and rapid fracture zones. This phenomenon is similar to Du Nanocomposites”, Mechanics of Advanced Composite
et al. report [11]. The 20TPU blend also fractured similarly, Structures, vol. 4, no. 3, pp. 179-186, 2016.
as presented in Fig. 8b. Additionally, the low value of the [5] H. Yanping, H. Yang, H. Zhang, Z. Mo. Miscibility,
20PC blend is due to the poor adhesion between them, as “Crystallization behaviors and toughening mechanism of poly
illustrated in Fig. 8c. (butylene terephthalate)/thermoplastic polyurethane blends”,
Fibers and Polymers, vol. 19, pp. 1-10, 2018.
F. Discussion
[6] Ata Chalabi Tehran, Karim Shelesh-Nezhad, Farshid Javidi
The tensile and flexural strengths of the 20TPU blend are Barazandeh. “Mechanical and thermal properties of TPU-
lower than those of the neat PBT because the glass transition toughened PBT/CNT nanocomposites”, Journals of
temperature of the PBT/TPU blend is lower than that of the Thermoplastic Composite Materials, vol. 32, no. 6, pp. 815-
neat PBT sample. Moreover, the impact toughness values are 830, 2018.
3.66, 7.47, and 3.37 kJ/m2, corresponding to neat PBT, the [7] Palanivelu, Psivaraman, Mdasaratha Reddy, “Studies on
20TPU blend, and the 20PC blend. This result implies that thermoplastic polyurethane toughened poly (butylene
adding TPU to the PBT-based blend could enhance the terephthalate) blends”, Polymer Testing, 21 (3), pp. 345-351,
impact toughness of the material. The impact toughness of 2002.
the PBT/PC blend is lower than the neat PBT. However, this [8] Jintao Huang, Huanyu Liu, Xiang Lu et al. “Thermal and
blend's tensile and flexural strengths are higher than the neat mechanical properties of TPU/PBT reinforced by carbon
PBT. Therefore, the added PC improves the tensile and fiber”, AIP Conference Proceedings, vol. 1713, no. 1, 2016.
flexural strengths of the neat PBT. [9] B. W. Hale, H. Keskkula, D. R. Paul. “Effect of crosslinking
The SEM microstructure of the sample shows that the reactions and order of mixing on properties of compatibilized
PBT/ABS blends”, Polymer, vol.40, no. 13, pp. 3665-3676,
20TPU sample has a blend surface reaction. This surface 1999.
increases the adhesive compatibility between the two phases
[10] Nor Azwin Ahad, “A Recent blend of thermoplastic
of TPU and PBT, improving the impact toughness.
polyurethane (TPU)”, IOP Conference Series: Materials
IV. CONCLUSION Science and Engineering, vol. 957, 2020
[11] Jintao Huang, Xiang Lu, Guizhen Zhang et al., “Study on the
This study indicates that adding 20 % TPU to PBT leads rheological, thermal and mechanical properties of
to an increase in impact toughness and a decrease in tensile thermoplastic polyurethane/poly (butylene terephthalate)
and flexural strengths. The 20TPU blend has 7.47 kJ/m2 of blends”, Polymer Testing, vol. 36, pp. 69-74, 2014.
impact toughness, 13.51 MPa of tensile strength, and 21.39 [12] Nor Azwin Ahad, “A Recent blend of thermoplastic
MPa of flexural strength. The 20PC blend obtains 3.37 kJ/m2 polyurethane (TPU)”, Materials Science and Engineering, vol.
of impact toughness, 64.46 MPa of tensile strength, and 957, pp. 3-8, 2021.
98.48 MPa of flexural strength. In comparison, those values
[13] P. Du, B. Xue, Y. Song et al, “Fracture surface characteristics
of the neat PBT sample have 3.66 kJ/m2, 37.13 MPa, and
and impact properties of poly(butylene terephthalate)”,
66.89 MPa, respectively. The SEM microstructure of the Polymer Bulletin, vol. 64, no. 2, pp. 185-196, 2010.
sample shows the 20TPU sample has a blend surface
reaction that increases the adhesive compatibility between
the two phases of TPU and PBT. Through these results, it is
possible to recommend that adding TPU could improve the
impact strength of PBT while preserving the advantages of
high tensile and flexural strength of the PBT/TPU blend.
ACKNOWLEDGMENT
This work belongs to the project grant No: SV2022-97,
funded by Ho Chi Minh City University of Technology and
Education, Vietnam.
REFERENCES
[1] M A Mikitaev, T A Borukaev, G E Zaikov “Polybutilene
Terephthalate (PBT), Synthesis and Properties”, vol. 1, pp.10-
13, 2006
[2] Stoyko Fakirov, “Handbook of Thermoplastic Polymers:
Homopolymers, Copolymers, Blends, and Composites”. Asia-
Pacific Journal of Chemical Engineering, chapter. 2, pp. 59-
103, 2002.
[3] Gauri S. Deshmukh, D. R. Peshwe, S. U. Pathak et al.
“Evaluation of mechanical and thermal properties of Poly
(butylene terephthalate) (PBT) composites reinforced with
212
An Evaluation of Corporation Social Responsibility

Performance for Vietnamese Contractors
Nguyen Van Minh Ha Duy Khanh
Department of Civil Engineering Department of Civil Engineering
Education Education
nguyenvanminh@hcmute.edu.vn khanhhd@hcmute.edu.vn
Soo Yong Kim Chu Viet Cuong

Department of Civil Engineering Academy of Managers for Construction and Cities
Pukyong National University Ho Chi Minh City, Vietnam
Busan, Korea chuvietcuong@gmail.com
kims@pknu.ac.kr
Abstract: Corporate social responsibility (CSR) plays an During the last decade, CSR has actively attracted serious
important role in different business sectors. Although CSR attention from scholars worldwide particularly “reason for
practices can be discretely found in organizational activities, CSR implementation” and “CSR identification” [1], [6].
this topic has not received adequate attention from the However, very few if any have exclusively examined the
construction industry. Meanwhile, an evaluation of the extent to which CSR practices be implemented in
importance level of CSR provides the leaders with a clear construction organizations, particularly the contractors.
understanding of CSR performance within their While CSR has not gotten much consideration, CSR activities
organizations, which is very important, especially when can be intentionally or unintentionally implemented in every
making further CSR policies. This study, therefore, aims to organization. The CSR practices can be discretely found in
explore CSR performance by developing a reliable and organizational activities, such as training schemes, post-
practical tool that can be applied to evaluate the importance construction services, the use of local products and services,
level of CSR. This study employed data collected from and the application of environmental management systems
professional respondents to evaluate CSR activities which [7]. Moreover, the cultural relativist perspective explains that
were grouped into four categories namely stakeholder people’s judgments will always depend on their social,
relation CSR, philanthropic CSR, environmental CSR, and cultural, political, and economic context [8]. Therefore,
ethical CSR. Using the fuzzy synthesis evaluation (FSE) different businesses in different geographical areas may lead
method, the findings showed that stakeholder CSR is the most to different CSR standards [9].
important category, followed by ethical CSR, philanthropic
This study, therefore, aims to explore CSR performance
CSR, and environmental CSR. The proposed prioritization
by developing a reliable and practical tool that can be applied
methodology is a robust tool that allows the researchers to
to evaluate the importance level of CSR with a focus on the
CSR prioritization.
Vietnam construction industry. Specifically, the CSR
Keywords: Corporate social responsibility (CSR), practices of the Vietnamese contractors will be identified and
construction, Fuzzy synthesis evaluation (FSE), Vietnam, categorized. The importance level of each CSR category will
sustainability be evaluated and prioritized.
I. INTRODUCTION II. LITERATURE REVIEW

Corporate social responsibility (CSR) has been widely A. Corporation social responsibility performance in the
embraced across business and academia because of its construction industry
potential in promoting a more sustainable world [1]. CSR The idea of CSR is to view how an organization integrates
implementation enhances organizational performance [2] and stakeholders’ interests and social values to reinforce its
contractor competitiveness [3]. Specifically, CSR has relations with society. CSR is defined as “the firm’s
positive effects on an organization's financial performance, consideration of, and response to, issues beyond the narrow
corporate image, and stakeholders' satisfaction [4]. However, economic, technical, and legal requirements of the firm to
this topic has not received adequate attention from the accomplish social and environmental benefits along with the
construction industry [5]. Studying CSR in the construction traditional economic gains and to fairly balance the needs
industry, thus, becomes a topic worth focusing on both and concerns of various stakeholders - all those who impact,
industry professionals and academia. are impacted by, or have a legitimate interest in the
company’s actions and performance” [6].

Construction activities bring negative effects to the four parts. First, economic responsibility refers to a return on
environment and human health. Contamination of soil, air, investment, new products and new investments achievement,
and water may be delivered as a typical environmental and new job opportunities. Second, legal responsibility refers
consequence of construction activities. Waste generated in to the expectations that organizations obey legal frameworks
construction projects creates tremendous pressure on the when doing business. Third, ethical responsibility refers to
limited landfill capacities as well as creates long-term adverse the communities' and societies' expectations in organizational
environmental effects which may include land deterioration, business, which is not regulated by the laws. Finally,
resource depletion, various forms of pollution, and discretionary responsibility refers to philanthropic activities
hindrances to economic and social development [10]. of organizations, such as education contributions or training
Construction workers worldwide are required to work in an schemes. However, the concept of stakeholders is a
environment that exposes them to poor physical conditions fundamental foundation of CSR [15] and environmental
making minor injuries and work-related ill-health (WRI) protection also plays a vital role in CSR performance [13].
commonplace [11]. Research conducted in the UK, Germany, Therefore, a previous study [16] revisited Carroll’s four-part
the United States of America, Finland, and Singapore found pyramid and investigated CSR under four categories, namely
that works relating to construction increase instances of stakeholder relation, philanthropic, environmental, and
asthma, cancer, respiratory diseases, musculoskeletal ethical CSR.
disorders (MSD), skin neoplasia, contact dermatitis (CD),
In summary, the above review highlights the importance
mesothelioma, lung cancer, pneumoconiosis, and other
of CSR performance in the construction industry. However,
benign pleural diseases [12].
for CSR performance evaluation, previous studies merely
The above discussions highlight the need for CSR focused on empirical and conceptual. More importantly, no
performance to achieve sustainable development in the study was found to investigate the evaluation of CSR
construction industry. However, the industry is different in performance, particularly from the contractors' perspective.
comparison to others such as the unique feature of its project- This study, therefore, aims to bridge this research gap by
based nature [6]. Therefore, CSR performance in proving industrial professionals and scholars with a reliable
construction organizations becomes complex and less and practical tool for evaluating CSR performance within
comprehensive, which is place-based, dynamic, and flexible contractors' perspective
than in other industries [1]. While numerous benefits are
associated with CSR performance, it is acknowledged that III. RESEARCH METHODOLOGY
CSR in the construction industry is unconceptualized and still C. Identification of CSR activities
in the early stage of development [9]. Although many
The first research stage is to identify CSR activities
companies have followed international environmental
implemented in the Vietnam construction industry. An initial
management standards (EMS) ISO 14000, the practical
list of CSR activities was extracted from the thorough
implementation of several construction firms has commonly
literature review of [7]. Accordingly, 69 CSR activities were
shown a waste of resources and inefficient resource
determined through content analysis and interview methods.
utilization [6].
Three experts who have many years of working in
B. Corporation Social Responsibility assessment models construction organizations were invited to discuss and revise
Although CSR performance evaluation is considered one the initial list. Through the interviews, each expert was
of the mainstream CSR studies [13], most of the research is provided the 69 activities and was asked to determine which
conceptually based on implementation [1], [6]. While the appropriate CSR activities according to their working
non-economic impact is considered sorted and determinable, experience. Finally, 31 CSR activities were renamed,
the economic impact is often controversially debated among finalized, and grouped under stakeholder CSR, philanthropic
scholars. It is emphasized that CSR performance can also be CSR, environmental CSR, and ethical CSR and it is presented
evaluated through a specific framework, which can capture in Table 1.
stakeholder interests and transparently weight [1]. CSR D. Questionnaire survey and participants
practices can be also categorized into six groups by factor
Based on the identified CSR activities, a survey-based
analysis [7]. However, this study did not provide the
approach was employed to examine CSR performance. A
weighting scheme for the identified CSR categories. In
questionnaire survey was designed in which the respondents
summary, a previous study [1] specified a guideline to
were asked to evaluate the 31 CSR activities according to a
evaluate CSR performance. First, CSR activities that affect
seven-point Likert scale. Finally, 252 valid questionnaires
the interests of project-related stakeholders should be listed
were deemed valid and used for further analysis.
and reasonably grouped as a framework. These activities
should incorporate both organizational and project levels at E. Fuzzy synthetic evaluation
which construction businesses operate. Second, an equitable Fuzzy synthetic evaluation (FSE) is a branch of the fuzzy
and transparent weighting scheme should be assigned and set theory developed by Zadeh [17], [18]. FSE targets to
incorporated into the proposed framework. Finally, the generate a synthetic evaluation of research objectives with
framework is used to quantify economic benefits and societal multiple criteria in a fuzzy decision environment. FSE is a
well-being derivable from CSR implementation. prominent method for multi-criteria evaluation of research
The fundamental theoretical constructs of CSR were problems, which contain ambiguous and uncertain
proposed [14], which divided business responsibility into information. By using FSE, uncertainty from the differences
among participants could be reduced. FSE can be adopted the
214
heuristic knowledge of expert panelists in the evaluation of IV. RESULTS AND DISCUSSIONS
problems.
F. Results
This study, therefore, employs this method to evaluate the Regarding the collected data and calculation processes,
extent to which corporate social responsibility performance weightings and membership functions for 31 CSR
has been implemented in construction companies. Two‐level performance variables and four categories were estimated.
factors were formed before employing fuzzy modeling. The Table 2 presents the calculation of weightings and
first level consists of four categories (stakeholder relation membership functions for 15 CSR performance variables of
CSR, philanthropic CSR, environmental CSR, and ethical ‘stakeholder relation CSR’ categories.
CSR), and the second level involves the 31 selected variables.
The FSE procedure in this paper follows commonly-agreed
processes proposed by previous studies [17-18].
Table 2. Weightings and membership functions
Table 1. Corporation social responsibility
215
Consequently, the important level for each CSR category CSR performance to attract potential customers is a crucial
is calculated and illustrated in Table 3. matter that organizations may consider.
c) Environmental CSR: Environmental CSR
Table 3. Important level of CSR performance by categories
emphasizes the impact of business performance on both
living and non-living natural systems, including ecosystems,
land, air, and water in the environment. It is investigated that
environmental practice is one of the critical CSR dimensions
embedded in the Chinese construction industry [7]. It is also
acknowledged environmental and social CSR are equally
important [19]. However, philanthropic and human activities
receive more attention from Malaysian property developers
Table 3 shows the importance of each CSR category. than environmental CSR [20]. Meanwhile, environmental
Accordingly, stakeholder CSR ranks in the highest position CSR ranks the lowest among CSR categories in the
with a value of 5.699, followed by ethical CSR, Australian context [21]. In Taiwan, while the real-estate
environmental CSR, and philanthropic CSR with indices developers emphasize pre-construction environmental
5.599, 4.949, and 4.770, respectively. evaluations and green building designs, the contractor's
The results were then validated through discussions with emphasis is on environmental influence and impact on
CSR experts in Vietnamese construction companies. Finally, biodiversity at construction sites [22]. Overall, the
the prioritization is reasonable and appropriate in the research environmental dimension of CSR incorporates practices such
context. Discussions of these categories are presented as as (1) applying environmental management systems, (2)
follows. minimizing undesirable business impacts on the local
community, (3) promoting environmental training schemes,
G. Discussions and (4) encouraging customers to use friendly-
a) Stakeholder CSR: A previous study [16] divided environmentally products.
stakeholders into internal and external groups. Internal d) Philanthropic CSR: It is argued that CSR practices
stakeholders or the employee is a crucial factor in CSR originated from a philanthropic perspective [23]. Corporate
performance. The involvement of employees largely philanthropy covers all forms of giving in business, which
determines the success of CSR implementation. Therefore, embraces voluntary or discretionary practices and may not be
CSR practices which directly related to employees receive a responsibility in a literal sense. The public normally expects
much attention from the practitioners. Establishing a humane businesses to acquire philanthropic responsibility that is not
salary and welfare system, building an effective mechanism required by laws. In that sense, ethical motivation may lead
for employer-employee communication and negation, businesses to do what is right for society [9]. Therefore, the
applying an equal employment system and establishing an relationship between corporate philanthropy and other CSR
employee promotion mechanism, and taking care of categories becomes complicated and multidimensional.
employee's families and helping them to achieve work-life Philanthropic responsibility is not only empirically
balance are all potential to improving contractor interrelated but also conceptually independent of the other
competitiveness. Meanwhile, external stakeholders consist of dimensions in CSR performance. A previous study
customers, shareholders, suppliers and partners, competitors, highlighted a positive link between philanthropic practices
and the government [16]. Although external stakeholders do and organizational performance [24]. Supporting the
not control the organizational resources, they can regulate the development of the local community, encouraging and
opinions of the public towards an organization CSR supporting employees to engage in activities operated by
performance. Among CSR activities within external social charities, prioritizing the use of local products and
stakeholders, customers receive much attention with a fairly services, and promoting community and social employment
high mean value of (1) improving business performance to are highly recommended philanthropic practices.
meet customers' demand on quality, safety, and
environmental aspects, (2) delivering proper post- V. CONCLUSION
construction services to customers, and (3) establishing an
This research has studied CSR performance with two
effective communication mechanism with customers.
derived objectives: to identify and categorized CSR activities
b) Ethical CSR: CSR performance under ethical of Vietnamese contractors and to evaluate the importance
principles such as trustworthiness, trust, and cooperativeness level of each CSR category. This study employed data
can significantly contribute to competitive advantage for an collected from 252 professional respondents from a
organization's business. Moreover, corporate ethics programs questionnaire survey to evaluate 31 CSR activities which
have positive impacts on employees' job satisfaction [1], [6]. were grouped into four categories namely stakeholder
Therefore, ethical CSR performance is considered an relation CSR, philanthropic CSR, environmental CSR, and
influential factor in construction. However, it is witnessed ethical CSR. Using the FSE method, the findings showed that
that while most mid-size companies in Turkey acknowledge stakeholder CSR is the most important category, followed by
ethical and economic drivers in CSR implementation, ethical ethical CSR, philanthropic CSR, and environmental CSR.
CSR is not the basic foundation in CSR adoption but
From an academic perspective, the research findings will
economic orientation [19]. In sumary, achieving good ethical
expand the current body of knowledge in CSR management.
The proposed prioritization methodology is a robust tool that
216
allows the researchers to CSR prioritization. In terms of [11] S. J. Stocks, S. Turner, R. Mcnamee, M. Carder, L. Hussey,
practice and management, evaluation of the importance level and R. M. Agius, “Occupation and work-related ill-health in
of CSR provides the leaders with a clear understanding of UK construction workers”, Occup. Med. (Chic. Ill)., vol. 61,
CSR performance within their organizations, which is very no. 6, pp. 407–415, 2011.
[12] R. Sauni, P. Oksa, S. Huikko, P. Roto, and J. Uitti, “Increased
important, especially when making further CSR policies. risk of asthma among Finnish construction workers”, Occup.
Furthermore, this research provides valuable CSR activities Med. (Chic. Ill)., vol. 53, no. 8, pp. 527–531, 2003.
to construction companies that intend to follow a sustainable [13] Q. Zhang, B. L. Oo, and B. T. H. Lim, “Linking corporate
development path. social responsibility (CSR) practices and organizational
performance in the construction industry: A resource
ACKNOWLEDGMENT collaboration network”, Resour. Conserv. Recycl., vol. 179, p.
The authors would like to express their sincere gratitude to 106113, Apr. 2022.
Ho Chi Minh City University of Technology and Education [14] A. B. Carroll, No Title, vol. 4. 1973, pp. 497–505.
for supporting this research. [15] A. L. Friedman and S. Miles, “Developing stakeholder
theory”, J. Manag. Stud., vol. 39, no. 1, pp. 1–21, 2002.
REFERENCES [16] Y. Yu and Y. Choi, “Stakeholder pressure and CSR adoption:
[1] B. Xia, A. Olanipekun, Q. Chen, L. Xie, and Y. Liu, The mediating role of organizational culture for Chinese
“Conceptualising the state of the art of corporate social companies”, Soc. Sci. J., vol. 53, no. 2, pp. 226–235, Jun.
responsibility (CSR) in the construction industry and its nexus 2016.
to sustainable development”, J. Clean. Prod., vol. 195, pp. [17] R. Osei-Kyei, A. P. C. Chan, and E. E. Ameyaw, “A fuzzy
340–353, Sep. 2018. synthetic evaluation analysis of operational management
[2] L. Suganthi, “Examining the relationship between corporate critical success factors for public-private partnership
social responsibility, performance, employees’ pro- infrastructure projects”, Benchmarking, vol. 24, no. 7, pp.
environmental behavior at work with green practices as 2092–2112, 2017.
mediator”, J. Clean. Prod., vol. 232, pp. 739–750, Sep. 2019. [18] M. N. Gebremeskel, S. Y. Kim, L. D. Thuc, and M. V Nguyen,
[3] G. Watts, A. Dainty, and S. Fernie, “Making sense of CSR in “Forming a driving index for implementing public-private
construction: Do contractor and client perceptions align?”, partnership projects in emerging economy: Ethiopian
Proc. 31st Annu. Assoc. Res. Constr. Manag. Conf. ARCOM perception”, Eng. Constr. Archit. Manag., 2020.
2015, no. September, pp. 197–206, 2015. [19] L. R. Barnes and N. Croker, “The relevance of the ISO26000
[4] A. Martos-Pedrero, F. J. Cortés-García, and D. Jiménez- social responsibility issues to the Hong Kong Construction
Castillo, “The relationship between social responsibility and Industry”, Australas. J. Constr. Econ. Build., vol. 13, no. 3,
business performance: An analysis of the agri-food sector of pp. 37–50, 2013.
southeast Spain”, vol. 11, no. 22, Nov. 2019. [20] S. Y. Lee Hong, M. Ismail, T. S. Yin, S. Y. L. Hong, M.
[5] X. Lin, C. M. F. Ho, and G. Q. P. Shen, “Research on Ismail, and T. S. Yin, “Corporate social responsibility in
corporate social responsibility in the construction context: a Malaysia housing development - The developer’s
critical review and future directions”, Int. J. Constr. Manag., perspective”, Pacific Rim Prop. Res. J., vol. 14, no. 2, pp.
vol. 18, no. 5, pp. 394–404, Sep. 2018. 177–198, 2008.
[6] Z. Y. Zhao, X. J. Zhao, K. Davidson, and J. Zuo, “A corporate [21] E. A. M. Bevan and P. Yung, “Implementation of corporate
social responsibility indicator system for construction social responsibility in Australian construction SMEs”, Eng.
enterprises”, J. Clean. Prod., vol. 29–30, pp. 277–289, 2012. Constr. Archit. Manag., vol. 22, no. 3, pp. 295–311, 2015.
[7] W. Jiang and J. K. W. W. Wong, “Key activity areas of [22] C. F. Huang, W. H. Lu, T. T. Lin, and E. J. Wu, “The current
corporate social responsibility (CSR) in the construction conditions of CSR implementation in construction industry: A
industry: A study of China”, J. Clean. Prod., vol. 113, pp. lesson from Taiwan”, Appl. Ecol. Environ. Res., vol. 15, no.
850–860, Feb. 2016. 2, pp. 67–80, 2017.
[8] I. C. Jarvie, “Cultural Relativism, Enciclopedia Delle Scienze [23] D. Ulutaş Duman, H. Giritli, P. McDermott, U. D. Duman, H.
Sociali.” 1995. Giritli, and P. McDermott, “Corporate social responsibility in
[9] M. Loosemore and B. T. H. Lim, “Mapping corporate social construction industry: A comparative study between UK and
responsibility strategies in the construction and engineering Turkey”, Built Environ. Proj. Asset Manag., vol. 6, no. 2, pp.
industry”, Constr. Manag. Econ., vol. 36, no. 2, pp. 67–82, 218–231, 2016.
Feb. 2018. [24] A. Lindgreen, V. Swaen, and W. Johnston, “The supporting
[10] S. Y. Kim, M. V. Nguyen, and V. T. Luu, “A performance function of marketing in corporate social responsibility”,
evaluation framework for construction and demolition waste Corp. Reput. Rev., vol. 12, no. 2, pp. 120–139, 2009.
management: stakeholder perspectives”, Eng. Constr. Archit.
Manag., vol. 27, no. 10, pp. 3189–3213, 2020.
217
Local Waste Seeds as Organic-Based Coagulant

Aids in Water and Wastewater Treatment
Nhung Thi-Tuyet Hoang Anh Thi-Kim Tran Luu Hong Quang
Faculty of Chemical and Food Faculty of Chemical and Food Chung Yuan Christian University
Technology Technology Taoyuan City, Taiwan
Ho Chi Minh City University of Ho Chi Minh City University of hongquang97@gmail.com
Technology and Education Technology and Education
nhunghtt@hcmute.edu.vn anhttk@hcmute.edu.vn
Abstract: The coagulation process is one of the effective have antimicrobial properties that are safe for human health
methods to remove suspended and colloidal materials as [8]. Natural coagulants have been used for many years in
well as colors. In the coagulation process, chemical developing countries in traditional water treatment methods
coagulants have been used because of their high efficiency [7]. Almost natural coagulants are cheap, simple to use,
in wastewater treatment. However, the drawbacks of these shear-stable, environmentally friendly, and rapidly
coagulants are high costs, substantial sludge volumes, and biodegradable compared to inorganic coagulants[9]. These
the link between Alzheimer's disease and the residuals of materials were extracted from plant seeds, leaves, or roots.
Aluminum after water treatment. Therefore, the minimum These natural coagulants are concerned because of no
chemical coagulants in water treatment were studied by danger to human health and are less expensive than
conventional chemicals [10]. In general, the four
adding natural coagulant aids. In this study, Local waste
mechanisms of particle coagulation include (1) sweep
seeds such as avocado seed extracts (AVSE) and ambarella flocculation, (2) double-layer compression, (3) adsorption
seed Extracts (AMSE) were used as coagulant aids in and interparticle bridging, and (4) adsorption and charge
combination with Poly Aluminum Chloride (PAC). Jar tests neutralization. Among these, natural coagulants, namely
were performed to examine the effectiveness of AVSE and tannins, are related to third and fourth mechanisms due to
AMSE in turbidity removal under different pH solutions (5, their long-chained structure, which upsurges the availability
6, and 7) with the PAC as a coagulant. At 3 mg PAC/L, the of unoccupied sites [11]. Among the common and
turbidity removal percentage was 98% compared to 87% established natural coagulants are dragon fruit foliage plant
with and without adding AVSE or AMSE. The results help [12], chestnut and acorn [13], Tamarindus indica and Litchi
reduce 7 mg PAC/L for water treatment. This process will chinensis Seeds [14], Moringa oleifera [15], etc.
save operational costs and reduce the health risk associated
An obstacle to applying natural coagulants on a large
with residuals of PAC in water.
scale may be widely due to limited sources. A few
researchers use the natural extract to produce aids for
Keywords: natural coagulant aid, avocado seeds,
chemical coagulants to enhance the pollutant’s removal
ambarella seeds, coagulation [16]. In this study, an experiment was undertaken to analyze
the efficiency of seed extracts of avocado seed extracts and
I. INTRODUCTION
ambarella seed extracts to support the coagulation of the
The coagulation technology is one of the most effective Poly Aluminum chloride (PAC) to meet the standard of
methods to remove suspended solids, colors or chemical Vietnam supply water. Furthermore, the optimum usage of
oxygen demand in water and wastewater [1] [2]. Coagulant two coagulant aids will reduce the cost and residual effect of
chemicals with charges opposite suspended solids are added chemicals coagulants.
to the water to neutralize the negative charges on non- II. EXPERIMENTAL
settleable solids [3]. Chemical coagulants such as aluminum
sulfate (alum) and poly aluminum chloride (PAC) [4] might A. Materials and wastewater for coagulation
be the best choices for water and wastewater pretreatment.
In this study, artificial turbid water was used to analyze
Although their highly effective treatment, they still have
fruit seeds' efficiencies in coagulation. Artificial turbid
several drawbacks, such as being pH-dependent and high
water was prepared by stirring slowly two solutions for 1
residual aluminum concentrations in water and sludge.
hour: 10 g/L of hydrazine sulfate powder (NH2.NH2.H2SO4)
Residuals were proved to link to neurodegenerative diseases
(China) and 10g/L of hexa – methylene – tetramine powder
such as Alzheimer’s as well as neurotoxic and exhibited
(C6H12N4) (China) and then let stand for 24 h to allow for
carcinogenic effects [5]. The threat of residuals to water
complete hydration of formazin with a standard turbidity
resources has become a significant problem in many
solution of 4000 NTU.
countries worldwide [6]. Furthermore, the cost of those
chemical coagulants may be quite high, increasing the cost B. Preparation of the crude extract of fruit seeds
of treatment and stabilizing the organic materials in sludge The fruit's seeds were collected, washed, and dried in the
[7]. oven at 105°C for 6 hours. Then, they were pre-grinded into
To overcome the drawbacks of using chemical powder. Seed powder was stored in a cool environment
coagulants, natural coagulants are proven to coagulate and (5oC). The obtained powder (1 g) was mixed with 200 mL

of solvent (NaOH 0.1N) [10] and then shaken at 150 rpm for higher pH of water, the higher the turbidity efficiency. The
15 minutes to extract the active ingredients of the fruit pH of the water was adjusted to various pH values with the
seeds. These active components were filtered by a paper addition of H2SO4 or NaOH and then analyzed for residual
filter (0.45 µm). The crude extracts of fruit seeds were used turbidity after 30-minute - sedimentation. In this study, the
for the wastewater coagulation in this study without any effect of higher pH was not observed because the PAC
further preparation. precipitates led to a decrease of coagulation [17]. With final
pH values of 5.0, 6.0, and 7.0 [Fig. 3-5] there is a decrease in
effluent water's turbidity with increasing PAC dosages.
Figure 1. Procedure for preparation of the crude extract of fruit

seeds
C. Coagulation experiments Figure 2. Effect of pH on turbidity coagulation of PAC

Turbidity removal by coagulants from fruit seeds was
carried out by jar test machine (JLT 6 Velp Scientifica). Six
beakers (jars) with 500mL artificial water or natural
wastewater. The crude extract of fruit seeds was added to
these beakers and then rapidly stirred at 100 rpm for 2
minutes and slowly stirred at 40 rpm in 5 minutes, 30 rpm in
10 minutes, 20 rpm in 10 minutes. After slowly stirring, the
flocs were settled for 30 min [10]. The water turbidity was
analyzed by Spectrophotometer (Hitachi U2910) at 450 nm.
The equation gives the efficiency of turbidity removal:
(1)
Where as Figure 3. Effect of PAC dosage on turbidity removal at pH = 5
Ti is the turbidity of influent water, mg/L
Te is the turbidity of effluent water, mg/L
The turbidity removal efficiencies were determined by
coagulation with poly aluminum chloride (PAC) in the case
of pH and coagulant dose. The effect of pH was analyzed by
various pH of the synthetic water in the range of 3-8, and
the coagulation test was performed by jar test procedure at
room temperature with initial turbidity of 200 NTU. The
effect of the coagulant dose (1 – 20 mg/L) for the reducing
turbidity process was studied by fixing the optimal pH, the
initial turbidity of 200 NTU. The coagulation aids from
avocado seed extracts (AVSE) and ambarella seed Extracts
(AMSE) were added with dosages from 5 to 60 mg/L at
various pH.
Figure 4. Effect of PAC dosage on turbidity removal at pH = 6
D. Coagulation of polyaluminum chloride (PAC)
Polyaluminum Chloride (PAC) is a prepolymerized
Al(III) chemical with an amount of hydrolysis and polymeric
species. Coagulation efficiency with PAC was investigated
by measuring the turbidity of influent and effluent water at
various pH and dosages. Fig. 2 shows the effects of PAC
coagulant on the turbidity efficiency at different final pH
values from 3 to 8. pH is an important coagulation factor.
The best coagulation ability is observed at the optimum pH
range with a given dosage. The type of colloid in the water
also affects the pH range for efficient coagulation. The Figure 5. Effect of PAC dosage on turbidity removal at pH = 7
219
The dose of coagulant is one of the most significant nearly 100% at dose of PAC 3 mg/L and AVSE 30 mg/L.
factors in determining coagulation treatment. The optimal The turbidity in effluent water increases when the amount of
coagulant dosage means the minimum coagulant used in AVSE is more than 35 mg/L. In conclusion, the best
coagulation treatment with the highest efficiency. The removal was obtained by using 25 ppm of AVSE combined
optimal dosage of PAC is determined according to the with 3 mg/L PAC and it got more than 97.5% from the
change of coagulant dosages at a fixed pH value. Therefore, initial turbidity of 200 NTU.
the coagulation efficiencies with different PAC dosages
were observed at the pH of 5.0, 6.0, and 7.0. With the initial
turbidity of 200 NTU, the efficiency increases from 64 to
98% from pH from 3 to 8 and the effluent turbidity of water
only met the Vietnamese standard (QCVN 02:2009/BYT) at
pH = 8 with 4.22 NTU. The turbidity efficiency of
coagulation met the standard of Vietnam with the optimum
dose of 15 (pH 5.0) (Fig. 3) and 12 (pH 6.0) (Fig.4) and 10
(pH 7.0) (Fig.5) mg/L. Generally, the PAC concentration ≤
10 mg/L is not enough to remove the turbidity to meet the
Vietnamese standard at pH 5 to 7. To reduce the amount of
PAC to limit the residues in effluent water, the coagulant aid
was added to increase the efficiency instead of increasing
the dose of PAC. There is a difference in efficiency when
the PAC dosage increase from 1 to 3 mg/L. At higher
dosage of PAC, the turbidity slightly decreases until more Figure 7. Effect of AVSE as coagulant aid on turbidity removal
than 10 mg/L to meet the Vietnamese standard. Therefore, 3 with PAC = 3 mg/L
mg PAC/L is the optimal dosage for combining coagulant
aids from Avocado and Ambarella seeds. The turbidity removal of PAC with the coagulant aids
AMSE. Unlike AVSE, there was an extreme difference in
the pH of the water. The optimum dosage of AMSE is 15
mg/L, the same with AVSE at pH =7. However, the higher
dosage of AMSE at pH =6 was 25 mg/L, more than AVSE.
At pH = 5, the support of AMSE is not clear, and hard to
meet the standard of Vietnam, 45 mg/L. In conclusion, by
using AMSE, the maximum turbidity reduction (99.5%) was
obtained for 0.3 mg/L PAC and 20 mg/L of AMSE at pH =
7. Ambarella seed has functional groups such as –OH,
–COOH, and –NH. When dissolved in NaOH solvent,
Ambarella seed will create several anions COO-, OH-, NH-
[19].
Figure 6. Effect of PAC dosage on turbidity removal at pH = 5,

6,7
E. Effect of natural coagulant aids to PAC

The removal of turbidity performance using various
doses of coagulant aids is shown in Figures 7 and 8. The
seed extracts from Avocado seed were dosed from 0 to 60
mg/L along with three various pH of water (5, 6, 7). After
using 10 mg/L extract from Avocado seed as a coagulant
aid, turbidity was reduced from 23, 22 and 15 NTU to 20,
14 and 7 corresponding to pH of water 5, 6 and 7,
respectively. With 15 mg AVSE per liter and pH=7, the
efficiency increased 5 times and met the Vietnamese
standard with 3 NTU. At pH =6, the dose of AVSE is Figure 8. Effect of AMSE as coagulant aid on turbidity removal
required more than pH = 7, 20 mg/L to meet the standard. In with PAC = 3 mg/L
acid conditions, the support of coagulation aid also meets
the standard, but at a higher dose, 30 mg/L. Avocado seed IV. CONCLUSION
starch contains only the O-H functional groups from a The main objective of this study was to reduce the
phenolic compound called Dopamine (3,4- chemical coagulant and meet the Vietnamese standard
dihydroxyphenylalanine), which has the same coagulation (QCVN 02:2009/BYT) for water supply. By using
effect as tannin. When added into an alkaline solution polyaluminum chloride (PAC) to coagulate turbidity in range
(NaOH solvent), the O-H bonds in avocado seeds were of pH from 3 to 8 and dosage of PAC from 1 to 20 mg/L.
broken and then the H+ ion interacts with negative ions in With the initial turbidity of 200 NTU, the turbidity of
water, increasing the coagulation efficiency [18]. Under coagulation met the standard of Vietnam with the optimum
neutral conditions, the percentage of turbidity removal was dose of PAC 15 mg/L (pH 6.0) and 12 mg/L (pH 7.0) and 10
220
mg/L (pH 8.0). To reduce the PAC dosage, AVSE and Remove Turbidity from Institutional Wastewater", 05/10
AMSE were added as coagulant aids. The amount of PAC 2020.
was fixed at 3 mg/L and the coagulant aids were changed [10] C. S. Lee, J. Robinson, and M. F. Chong, "A review on
from 5 to 60 mg/L. The optimum natural coagulant aids for application of flocculants in wastewater treatment", Process
AVSE were 25 ppm with 97.5% from the initial turbidity of Safety and Environmental Protection, vol. 92, no. 6, pp. 489-
200 NTU in range pH of 5 to 7. The optimum dosage of 508, 2014/11/01/ 2014.
AMSE was the same with AVSE at pH = 6 and 7 but not [11] A. Ibrahim, A. Z. Yaser, and J. Lamaming, "Synthesising
tannin-based coagulants for water and wastewater
efficiency at pH = 5.
application: A review", Journal of Environmental Chemical
REFERENCES Engineering, vol. 9, no. 1, p. 105007, 2021/02/01/ 2021.
[12] J. Idris, A. Md. Som, M. Musa, K. Hamid, R. Husen, and M.
[1] S. Ansari, J. Alavi, and Z. M. Yaseen, "Performance of full- Muhd Rodhi, "Dragon Fruit Foliage Plant-Based Coagulant
scale coagulation-flocculation/DAF as a pre-treatment for Treatment of Concentrated Latex Effluent: Comparison
technology for biodegradability enhancement of high of Treatment with Ferric Sulfate", Journal of Chemistry, vol.
strength wastepaper-recycling wastewater", Environmental 2013, 01/01 2013.
Science and Pollution Research, vol. 25, no. 34, pp. 33978- [13] M. Sciban, M. Klasnja, M. Antov, and B. Skrbic, "Removal
33991, 2018/12/01 2018. of Water Turbidity by Natural Coagulants obtained from
[2] Z. M. Yaseen et al., "Laundry wastewater treatment using a Chestnut and Acorn", Bioresource Technology, vol. 100, pp.
combination of sand filter, bio-char and teff straw media," 6639-43, 08/01 2009.
Scientific Reports, vol. 9, no. 1, p. 18709, 2019/12/10 2019. [14] M. M. Rahman et al., "Removal of Turbidity from the River
[3] J. Edzwald, "Coagulation in Drinking Water Treatment: Water usingTamarindusindica and Litchi chinensis Seeds as
Particles, Organics and Coagulants", Water Science and Natural Coagulant", International Journal of Environmental
Technology, vol. 27, pp. 21-35, 06/01 1993. Protection and Policy, vol. 3, pp. 13-20, 02/14 2015.
[4] G. Vijayaraghavan, T. Sivakumar, and A. V. Kumar, [15] K. A. Ghebremichael, K. R. Gunaratna, H. Henriksson, H.
"Application of Plant Based Coagulants for Waste Water Brumer, and G. Dalhammar, "A simple purification and
Treatment", 2011. activity assay of the coagulant protein from Moringa oleifera
[5] G. Muthuraman and S. Sasikala, "Removal of turbidity from seed", Water Research, vol. 39, no. 11, pp. 2338-2344,
drinking water using natural coagulants", Journal of 2005/06/01/ 2005.
Industrial and Engineering Chemistry, vol. 20, no. 4, pp. [16] M. Ramal, A. Jalal, M. Sahab, and Z. Yaseen, "River water
1727-1731, 2014/07/25/ 2014. turbidity removal using new natural coagulant aids: case
[6] W. H. M. Wan Mohtar, K. N. Abdul Maulud, N. S. study of Euphrates River, Iraq", Water Supply, 12/20 2021.
Muhammad, S. Sharil, and Z. M. Yaseen, "Spatial and [17] R. D. Letterman, S. G. Vanderbrook, and P.
temporal risk quotient based river assessment for water Sricharoenchaikit, "Electrophoretic mobility measurements
resources management", Environmental Pollution, vol. 248, in coagulation with aluminum salts", Journal (American
pp. 133-144, 2019/05/01/ 2019. Water Works Association), vol. 74, no. 1, pp. 44-51, 1982.
[7] M. T. F. de Souza et al., "Extraction and use of Cereus [18] J.-R. Jeon, E.-J. Kim, Y.-M. Kim, K. Murugesan, J.-H. Kim,
peruvianus cactus mucilage in the treatment of textile and Y.-S. Chang, "Use of grape seed and its natural
effluents", Journal of the Taiwan Institute of Chemical polyphenol extracts as a natural organic coagulant for
Engineers, vol. 67, pp. 174-183, 2016/10/01/ 2016. removal of cationic dyes", Chemosphere, vol. 77, no. 8, pp.
[8] G. Chhipi-Shrestha, N. Beauchamp, R. Sadiq, and M. 1090-1098, 2009/11/01/ 2009.
Rodriguez, "Investigating unregulated disinfection byproduct [19] S. Ishak, N. Ismail, M. Azemi, and H. Ahmad, "Some
reduction efficiencies in drinking waters using zirconium physical and chemical properties of ambarella (Spondias
oxychloride, a novel coagulant", Journal of Water Process cytherea Sonn.) at three different stages of maturity", Journal
Engineering, vol. 37, p. 101496, 2020/10/01/ 2020. of Food Composition and Analysis, vol. 18, pp. 819-827,
[9] V. Kumar, N. Othman, and S. Mohd Asharuddin, "Partial 12/01 2005.
Replacement of Alum by Using Natural Coagulant Aid to
221
The Implementation of 5C’s in Online-Foreign Language

Teaching for Vietnamese Students in the 4.0 Era
Chau Le Thi Bao Nhu Vo Hoang Nhu Nhi Ho Yen
Faculty of Foreign Languages Faculty of Foreign Languages Faculty of Foreign Languages
Nha Trang University Nha Trang University Nha Trang University
Nha Trang, Vietnam Nha Trang, Vietnam Nha Trang, Vietnam
lbchau@ntu.edu.vn nhuvhn@ntu.edu.vn nhihy@ntu.edu.vn
Abstract: The concept of the 5C’s in foreign language

education (Communication, Cultures, Comparison,
Connections, and Communities) invented by the American
Council for the Teaching of Foreign Languages (ACTFL) has
become the target and the development tendency of foreign
language teaching in the twenty-first century. In the 4.0 era, the
approach of utilizing 5C’s in online foreign language teaching
has opened numerous potential gates. This study focuses on
discussing and evaluating the outcomes of implementing the
5C’s in producing the teaching contents, forming the
methodologies, and designing activities in online foreign
language learning for Vietnamese students aiming to enhance
the learners’ interaction and communication skills, the abilities
to compare and contrast the differences in cultures, and the
practical application in authentic communication. The Action
Research approach was the main research methodology
including plan, act, observe and reflect. The findings of the
study revealed that the 5C’s model could affect each type of
language in different ways, and the most improved aspect of Figure 1. The 5C’s [2]
the learning outcomes after utilizing this teaching method was Connection: Learners can connect the knowledge they
communication. From the implementation, a variety of teaching have learned to their own experiences or to other fields.
approaches, which can improve language teaching in the 4.0
era, are mentioned. Communities: Learners can apply knowledge learned in
the outside environments or communities as well as in real
Keywords: the implementation of 5C’s, online-foreign life.
language teaching, Vietnamese students, the 4.0 era Currently, the theory of 5C’s is widely applied to foreign
language teaching around the world. For example, in
I. INTRODUCTION Chinese, 5C’s is applied in setting teaching objectives,
In 1999, the 5C’s concept invented by ACTFL included designing methods, compiling textbooks, evaluating courses,
five main standards: Communication, Cultures, training teachers, etc. [3].
Comparisons, Connections, and Communities [1]. These From the five dimensions provided in the analysis, the
standards were believed to have a close-knit relationship real picture of language teaching has been revealed precisely
leading to an interdependent aspect becoming the main goal through the appearance of various relevant factors affecting
as well as the global trend of foreign language teaching and the features of education. Zare Behtash & Banaruee (2017)
learning in the 21st century [2-3]. mentioned that the bridge of “old” approaches in education
According to Li (2013), the 5C’s theoretical framework had been changed, and it had the trajectory of focusing on
is interpreted as in the following explanation [2]. not only educating “a person”. The essence of language
teaching methodology following the “5C’s” has drawn the
Communication: Learners’ communication ability can be relationship between language and globalization. In
improved by focusing on listening, speaking, reading, and particular, numerous studies concerning the implementation
writing skills, emphasizing interaction, cooperation, and of “5C’s” in teaching one of the most crucial global
exchange in teaching and learning activities. languages (the English language) have been becoming the
Cultures: Learners can understand the cultural foundation of language teaching in the world [4]. In a study
characteristics of the language they learn. by Chen Lijuan in 2021, he inferred the aspect of utilizing
the “5C’s” in teaching the English language to middle school
Comparisons: Learners can perceive and compare the students, and the orientation of achieving the English
similarities and differences between different languages and competencies through the 5C’s model [5]. Zare Behtash &
cultures. Banaruee (2017), on the other hand, proposed research
regarding the English teaching materials utilized the
approaches of 5C’s, affirming the essential function of the

mentioned standards in the English teaching methodology at Nha Trang University instructed by the researchers from
framework [4]. In addition, in the era of technology September 2021 to February 2022.
development, a variety of innovative approaches to changing
the features of language teaching have been considered, and Observation: Observing 883 students in twenty Chinese
the combination of 5C’s and this new way of teaching turned and English classes of various levels including A2, B1 and
out to be remarkable attention. Gandhi & Monika (2018) also B2 (Table 1).
considered the tendency as the main content of their research Reflection: Evaluating the benefits and limits or
focusing on exploring “the role of emerging technologies in drawbacks of using 5C’s and suggesting some solutions.
standardizing language teaching and learning through 5C’s”
[6]. Table 1. Classes utilizing 5C’s
In the global circumstance of encountering various Languages Number of Number of
Levels
challenges, the outstanding development of technology has students classes
partially maintained and escalated foreign language teaching Chinese 155 5 A2
and learning through the approaches of distance learning or English (for non-English
347 7 B1
online learning and other beneficial learning apps. In majored students)
particular, the relationship between the model of 5C’s and English (for English
381 8 B2
online-foreign language learning has also become one of the majored students)
most prevalent considerations. In the book by Heather Parris Total 883 20 3
and Lisa M. Estrada – Digital Age Teaching for English Common European Framework of Reference for Languages (CEFR)
Learners, the authors already mentioned some relevant
aspects of this issue, and they inferred the student-centered
DATELs approach as an effective tendency of improving the III. FINDINGS AND DISCUSSION
5C’s in the virtual English learning process [7]. Additionally, The action of the research included the process of
other studies also concentrated on investigating the designing teaching activities (Table 2), the researchers
relationship between online foreign language learning and observed the online teaching process to evaluate the
the effects of 5C’s. For instance, in an article published in the effectiveness and the limitations of the 5C’s standards.
Journal of Education and Educational Development, Natalya
Kuznetsova and Kamal Ahmed Soomro propose the Web 2.0 A. Analyzing the utilization of 5C’s in online foreign
practices in the foreign language learning process, and they language teaching
examined the improvement of the 5C’s of their students in
learning Spanish, German, Russian, French, Arabic, English, 1) Communication:
and Italian [8]. However, in Vietnam, studies regarding Compared with other standards, “Communication” was
similar aspects have been considered and developed. utilized with relatively high efficiency in online-foreign
Following this tendency, the study investigating the process language teaching classes at Nha Trang University. A2 level
of organizing online foreign language teaching utilizing 5C’s students in Chinese classes developed their communication
for students of Nha Trang University can be seen as a new ability through activities like practicing conversations in
approach to contributing to the era of digital or virtual pairs, discussing in groups the different topics under the
language teaching in Vietnam. instruction of the lecturer on Zoom, and doing online
homework on Google Meet. However, reading and writing
skills were not really improved through direct online
II. METHODOLOGY learning, so the students were required to accomplish some
The target subject of this research concentrates on the extra exercises or practice essay writing on other online
implementation of 5C’s on online-foreign language teaching modes.
in Vietnam. The main issue in this discussion was the
questions concerning the approaches of utilizing the model Concerning the non-English major classes, the
of 5C’s and the effects of this implementation on various implementation of paired conversations and group
languages and learners. The scope of this research was presentations helped to better students’ 4 main skills:
carried out on students learning Chinese and English listening, speaking, reading, and writing. Having students
language at Nha Trang University. In particular, all of the present the new words in groups is an act of enhancing their
learners came from some elementary Chinese classes (level vocabulary which is a very important factor in
A2), intermediate English classes (level B1) and English communication. The wide range of vocabulary could
majored students (Syntax classes - level B2). This research’s definitely result in the improvement of listening, speaking,
method was mostly in the model of Action Research reading, and writing skills. Moreover, under the instruction
including four steps: Plan-Act-Observe-Reflect [9-11]. of the lecturer, break-out rooms were created on Zoom for
students to have conversations in pairs. This activity focuses
Plan: Clarifying the research’s subject focusing on the on speaking skills, and it is an effective way to bring comfort
implementation of 5C’s into online foreign language to students. As only their partners and the lecturer could hear
teaching for students of English and Chinese classes at Nha the conversations, they were less afraid of making mistakes.
Trang University and suggesting the approaches in designing Besides, they could exchange their knowledge or provide
teaching activities following the 5C’s. feedback to each other while they were in the break-out
rooms.
Action: Designing some teaching activities utilizing the
theories of 5C’s in teaching Chinese language and English Despite that, there remained some drawbacks in the
language online through LMS, Google Meet and Zoom for application of the “Communication” standard. First, it was
English majored students and non-English majored students hard to evaluate the students’ body language only by seeing
223
their facial expressions. Some even “read” but not presented observation, the learners were interested in this kind of
the information to the class. Second, passive students activity but the “Cultures” standard was performed
sometimes turned their mics off during the speaking sections. appropriately in some specific topics and levels of the
learners.
Online teaching has the trajectory of encountering a
variety of difficulties in designing some types of Non-English majored classes aimed to provide
communication activities. This circumstance is completely knowledge of a variety of cultures. Different means of
true in the case of online-foreign language classes of the transport, festivals, sports, and so on from both English and
researchers. However, the flexibility of understanding and non-English speaking cultures were introduced to the
implementing the 5C’s standards in designing some students. Hence, the cultures of English-speaking countries
appropriate teaching methodologies might reduce the barriers were not discussed much in those online classes.
to direct communicative features. In particular, English
majored students were required to participate in group In this research, English majored students of Syntax
working and panel-discussion activities, which were classes were chosen to utilize 5C’s in online teaching. As a
designed by utilizing some tools available on Zoom. matter of fact, the discovering aspects regarding culture in
Following this, the students were instructed to work in the Syntax obviously found numerous difficulties in transferring
break-out rooms for group work, and the lecturer could also cultural features to students. Nevertheless, the researcher
observe each room separately without the need of asking for realized that it was also extremely vital to combine culture in
permission. Besides, when conducting the panel discussion, her teaching process due to the fact that students could “feel”
the instructor or lecturer did change the role of operating the and “touch” the features of the sentence structures existing in
Zoom room for one student who was assigned to be the the English language system. This explains the reason why
moderator, and other students with the role of panelists the lecturers always mentioned the cultural aspects through
joined in the discussion by listening to the instructions. the habits of using English sentence structures, which also
Through these two activities, students had the opportunities represented precisely the characteristic of English-native
to communicate with each other, and they could certainly speakers.
gain some benefits from exchanging their knowledge. Through the combination of language and culture,
However, from the observation of the researcher, some students were able to be comprehensible in English and
students were quite passive, and they tended to “escape” American cultures, providing some necessary knowledge in
learning some cultural subjects.
from the discussion to seek safe zones, leading to ineffective
results in online learning. It was obvious that students could 3) Comparisons:
have the habit of hiding from answering the questions during There are certain similarities between Chinese and
online learning due to the fact that they enjoyed doing some Vietnamese, thus, Vietnamese students learning Chinese at A
other distracting things when being alone. These features level could compare the two countries as well as their
explained the reason why the lecturer did make lots of efforts languages while studying vocabulary, grammar, or
in managing the classes through a set of group-working discussing and exchanging cultures to differentiate ways of
rules, which already reduced the distractions of students, greetings, addressing somebody, cuisine, and climate
although some of the students did not try their best to follow between Vietnam and China. However, since the students
them. were still at the elementary level, they mainly employed
From a positive point of view, most of the students sentence-making or translating methods to see the
already received some great outcomes from this type of similarities and differences between the two languages, and
implementation. Some of them were changed from being to discuss the cultural characteristics of the two countries, the
passive to being more energetic thanks to the activity of students used Vietnamese - their mother tongue.
online panel discussion. The researcher realized that As previously said, different cultures were included in
communicative activities could help students enhance their the study program of non-English majored classes.
psychological aspects related to opening their minds and Therefore, few comparisons could be made between
changing their thoughts. In particular, they tended to express Vietnamese and English-speaking cultures. Nevertheless, the
their ideas more comfortably compared to offline differences in pronunciation between English and
classrooms, explaining the fact that students were more Vietnamese were acknowledged such as the differences in
confident when they did not have to perform their language pronouncing vowels, consonants, final sounds, and the
abilities directly. number of syllables.
2) Cultures: Syntax is a subject requiring lots of comparisons in
Chinese classes at A2 level learn basic communication learning, so if an English language learner cannot clarify the
topics which are close to everyday life. Therefore, the differences between his or her mother language and English,
“Cultures” standard is mainly reflected in topics such as he or she may face some problems regarding the habits of
greetings, weather, travel, food, and so on. The lecturer language use. Based on the features of the subject, the
introduced the cultural features while explaining the researcher also designed some comparing activities in her
vocabulary using photos and videos in online classes. After lesson plans emphasizing activating the awareness of using
being introduced to some cultural knowledge, the learners correct English sentence structures. In detail, the learners
had to participate in group discussion activities to enhance were required to observe the structures of two languages and
their skills. Additionally, some Chinese speakers were also indicated their thoughts via an analysis report consisting of
invited to join the online classrooms providing opportunities the differences and similarities of the syntactic aspects.
for students to get engaged in learning Chinese culture and in These kinds of reports were completed in group working
practicing their language skills. From the lecturer’s activities in the break-out rooms, and they were submitted
224
onto a designed file of the LMS of the researchers’ university After being instructed to compare the two languages,
called E-learning. students could have the ability to diminish their bad habits in
English language production.
Table 2. Class activities designed utilizing the 5C’s theory

5C’s Chinese classes Non-English majored classes English majored classes
Communication - Conversation: students talked in pairs on - Conversation: students talked in pairs on - Group work: students were
Zoom about the topics greetings, study, Zoom about lifestyle, competitions, instructed to participate in some
weather, shopping, food and entertainment. transport, challenges, the environment, activities involving discussion,
- Group work: students discussed the 6 topics and stages in life. panel discussion on Zoom.
and fulfilled the tasks provided by the lecturer - Presentation: students presented in -Presentation: students were
such as revisions, presentations, recording groups vocabulary related to the 6 topics required to deliver or make
videos, etc. using Google Meet and designed mini-games for the classes. presentations on Zoom.
Cultures - The lecturer introduced and instructed
learners to discuss some ways of greetings or
The lecturer introduced different means of
features concerning the climates, cuisines, or
transport, festivals, sports, and so on from Introducing the features of culture
some tourist destinations in China.
different cultures through listening and through language use.
- The lecturer designed at least one Zoom
reading activities.
meeting with Chinese speakers via online
classrooms.
Comparisons - Students were instructed to translate, make - Students acknowledged the differences
Students were instructed to clarify
sentences, and they would be given in pronunciation between English and
the differences and similarities in
opportunities to compare the Chinese language Vietnamese through listening activities.
the syntactical aspects of the two
and the Vietnamese language. - Students were instructed to discuss on
languages, and the concept of UG
- Students were instructed to discuss on Zoom Zoom the similarities and differences
was introduced to reveal the
the features regarding the differences between between countries regarding means of
language features.
Chinese and Vietnamese cultures. transport, festivals, sports, and so on.
Connection Through the discussions and group-working Through the useful tools designed
Students would be able to utilize their
assignments, students would be able to express on the LMS, English majored
language knowledge in learning
their perspectives and connect with their lives students are always provided some
specialized subjects or doing research
to perform their language ability in various opportunities to join some projects
from activities like having conversations,
daily topics; for instance, movies, music, travel connecting the language theories
presenting vocabulary, doing quizzes, etc.
or cuisine… and authentic practices.
Communities - Students utilized what they learned to have Students utilized their language
Students were encouraged to participate in
after-class group revisions on Google Meet. knowledge in supporting the
some events organized by the NTU
- Students participated in some Chinese clubs lecturers in designing some relevant
English Club.
and Chinese language competitions. activities.
4) Connection: about an order, and invite someone. Doing grammar quizzes

This standard requires students to connect with other helped to reinforce students’ grammar which is helpful for
fields in the process of learning languages. For Chinese their research writing as well. The topic of lifestyle,
classes at A2 level, when studying the topics like however, could not reach high efficiency.
entertainment, students could relate their background Syntax always has a close-knit relationship with other
knowledge to music, films or hospitality. In particular, language skills including speaking and writing, and the
students were instructed to have conversations and purposes of being comprehensible in syntactic aspects are to
discussions in the online classes through the break-out room satisfy the competencies of language communication. During
tool on Zoom, so they could express themselves and share the online teaching process, the integrated activities were
their knowledge of related fields. However, with some topics quite hard to implement; however, through the useful tools
like greetings, studying and shopping, it is difficult to designed on the LMS, students were given the opportunities
achieve high efficiency. to conduct some authentic assignments. In this regard, the
Conversations, presentations, quizzes, etc. are effective assignments were designed as mini-projects in which the
activities to help non-English majored students build a bridge students were required to examine some structures’ mistakes
connecting the English language with their specialized in one passage, or they also participated in carrying out some
knowledge. Those activities designed for the topics of surveys investigating the common syntactical errors made by
competitions, transport, challenges, the environment, and other English majored students, and they had to suggest
stages in life could be linked to hospitality, marketing, some solutions.
business, etc. For example, when learning the topic A deep understanding of syntactic aspects can help
competitions students were asked to analyze and discuss the students produce language more precisely, and the
information needed for an advertisement, the characteristics assignments could already build the bridge to connect what
of a good advertisement, and common writing mistakes in an they were taught and what they will apply.
advertisement. Students majoring in marketing could utilize
their specialized knowledge and involve themselves actively 5) Communities:
in those discussions. The topics of transport and challenges Compared with other standards, “Communities” is more
could be related to hospitality. The topics of environment difficult to achieve when conducting online classes because
and stages in life created connections with not only students students have limited opportunities to practice
majoring in the environmental field but also those majoring communicating outside the classroom. In the case of students
in business since they were provided chances to practice in A2 level Chinese classes at Nha Trang University, the
phoning about an order, write formal sentences in emails “Communities” standard was promoted mainly through
225
after–class group revisions on Google Meet. In addition, improving their language communication, language and
some students participated in the Chinese Club organized by culture comparison, and connection. The two categories
the University. regarding culture and communities were not developed
appropriately in this case because of the features of the
Similar to Chinese students, non-English majored subject applied. Syntax classes were the scope of this
students were encouraged to participate in Nha Trang research, so this certainly required lots of specific
University English Club events. However, not many of them characteristics of teaching. The most significant effective
joined the events as they were not confident with their activities in these types of classes mainly focused on
English ability. discussion and self-improvement, and it explained why only
During the online Syntax courses, the students were some standards in 5C’s were developed thoroughly.
instructed to join a Writing Correction Group where they had
to correct the paragraphs or essays for English majored Table 3. The evaluation of 5C’s from the lecturers’ perspectives
freshmen. In addition, they also supported their lecturers to 5C’s Commu- Cul- Compari- Connect- Commu-
prepare some writing exercises, which were utilized in nication tures sons ion nities
teaching non-English majored students. After implementing Chinese
+++ ++ ++ ++ +
these activities in teaching, some students complained that classes
they did not have enough time to correct, and they thought Non-
English
that it was not necessary to get involved in writing correction majored
+++ + + ++ +
or writing preparation. However, others believed that these classes
activities could raise their awareness and the habit of writing. English
majored +++ + +++ +++ +
From the analysis provided, the researchers can firstly classes
arrive at a conclusion that the implementation of 5C’s on Based on the observations through the teaching process
online-foreign language teaching possessed some positive +++ Extremely effective
aspects and limits due to the specific features of the ++ Effective
+ Normal
languages, the subjects’ content, teaching activities, learners’
level and other relevant teaching affected factors.
2) The evaluation of 5C’s from the learners’
B. Evaluating the implementation of 5C’s in online-foreign perspectives:
language teaching a) The evaluations of the learners on the effects of
1) The evaluation of 5C’s from the lecturers’ 5C’s on online learning:
perspectives: The researchers also collected the data concerning the
The evaluation of the implementation of 5C’s in online- effects of 5C’s from the perspectives of the learners. 883
foreign language teaching (including Chinese and English students (Chinese classes, English majored classes, and non-
language) for non-English majored students and English English majored classes) based on their language abilities
majored students at Nha Trang University is presented in after being instructed by the model of 5C’s, and they
Table 3 (the evaluation from the lecturers’ perspectives). themselves evaluated each standard separately.
The first glance at the table revealed the facts of A2-
Chinese level classes’ students that they had the tendency to
develop their listening and speaking abilities through some
online activities, so “Communication” was evaluated
perfectly. “Culture” and “Comparison” also possessed some
successful features at some topics in the lessons.
“Connection” followed as the third rank in the evaluation due
to some relevant factors including the learners’ interests.
“Communities”, in other words, was like the authentic
practice of the Chinese language outside of the classrooms,
and the students mostly participated in some online-group
working activities or other relevant online activities to
review the topics of the lessons that they already learned.
As for the non-English majored students,
“Communication” was found most effective since a number
of class activities were employed with focuses on developing Figure 2. The evaluations of the learners on the effects of 5C’s on
speaking skills. Connection was in the second place because online learning
students could connect and apply what they learned to other
specialized subjects at university. “Cultures”, “Comparisons” From the bar chart (Fig. 2) provided, although the 5C’s
and “Communities” were the three least effective standards standards were implemented into the teaching processes of
due to the fact that a variety of cultures from both English- three specific categories of classes with different learners’
speaking and non-English-speaking countries were discussed characteristics and number of students, the
during the course, and students were afraid of judgments “Communication” standard was evaluated as the most
when participating in English-speaking communities. improved ability with 97.3% for Chinese classes, 98.5% for
non-English majored classes and 98.7% for English majored
In terms of the effects of 5C’s on the online-English classes. These figures revealed the fact that it was quite
majored classes, the students received lots of benefits in
226
successful in enhancing the communicative aspects of the

learners.
However, the trajectory of “Cultures” received another
approach of evaluation in which only students of Chinese
classes (level A2) and students of non-English majored
classes could witness the improvement (with 58.4% and
44.3% respectively) due to the fact that most of the lessons in
the courses were designed by following some familiar daily
topic compared to specific features of English majored
classes creating the low improvement of the standard (with
only 39.6%). In stark contrast, students of English majored
classes experienced lots of positive changes when being
Figure 3. The students’ satisfactions of 5C’s
instructed to compare the two languages during the teaching
process, being at 93.6%, which was higher than that of non-
English majored students and students of A2-Chinese classes In 846 evaluations of students regarding the
(with 47.1% and 65.8% respectively). There is a fact that implementation of 5C’s on their online learning processes,
English majored classes (Syntax was the subject of these 67.3% of them expressed ideas of being extremely satisfied,
classes) possessed a variety of specific differences in the 27.7% belonged to the satisfied group, 4.2% of them were
knowledge of the language and the approach to transferring quite satisfied with what they were taught, and the
knowledge to students. Obviously, Syntax is a subject percentage of discontent was at 0.7%, and 0.1%. Based on
requiring the acquisition of English language structures, and the proportion of the data, it may arrive at the conclusion that
the ability to produce precise English sentences without the most of the students highly evaluated the 5C’s due to the
interference of the Vietnamese language’s habits. As a result, great effects of teacher-student interaction and student-
the instructors already designed a host of relevant activities student interaction in online classrooms. Although the rate of
while teaching, so the learners already received their targets dissatisfaction was low, it still reflected the fact that some
of the learning process. students were not really enjoying themselves, or they might
encounter some difficulties and obstacles in following the
In terms of “Connection”, English majored students were
5C’s teaching model.
also evaluated perfectly in their improvement (98.4%), while
students of A2-Chinese level classes and non-English From the analysis above, the researchers can arrive at a
majored classes received quite a low improvement (54.4% conclusion that the implementation of 5C’s can have the
and 61.1%). From the figures, the researchers affirmed that ability to create lots of benefits in improving the aspects of
“Connection” depended on the subjects’ features and the communication of the learners. “Cultures”, “Comparisons”
intention of the lecturers. For instance, in the case of and “Connection” are three standards depending on the
teaching Syntax, the lecturers tended to create lots of features of specific subjects and teaching approaches, so they
relevant activities in order to stimulate the ability of students received quite different results. However, “Communities”
to utilize what they were taught in their language abilities became the least efficient standard due to the characteristics
(which mostly concentrated on improving writing and of online learning.
speaking skills), so the Connection was promoted
significantly. IV. CONCLUSION
With regards to the “Communities”, these three In the 4.0 era, online-foreign language teaching has
categories of classes achieved quite similar results, which become one of the major tendencies globally. Following this,
were below 50% in total. There was a fact that online classes lots of studies concerning the mentioned were conducted,
could not encourage students to engage in some kinds of and they tended to focus on analyzing some relevant factors
community activities, and what students did was only focus including the teaching tools, curriculum, time, space, and
on doing some required assignments or projects, which were even appropriate teaching models in specific classes.
believed to be the “community” features in some aspects.
The results were quite similar to what the researchers could This research concentrated on discovering the effects of
observe. implementing the standards of 5C’s in online teaching for
883 students of Chinese classes, non-English majored classes
b) The students’ satisfaction with 5C’s: and English majored classes at a university in Vietnam. The
The students’ evaluations of the satisfaction of learning researchers followed the action research model, and the
by following the 5C’s model were collected through an observations of real circumstances of the teaching and
online survey of 883 students. In detail, 846 survey results learning processes were carried out to get the efficiency of
(149 results of Chinese classes’ students, 323 results of non- 5C’s from the perspectives of the lecturers and the learners.
English majored classes’ students, 374 results of English From the collected data, the effects of the 5C’s depend on
majored classes’ students) were acceptable. The survey was the features of each language and the teaching activities
designed based on 5 criteria of evaluation including designed by different lecturers. However, students of three
extremely satisfied (ES), satisfied (S), quite satisfied (QS), categories of classes selected Communication as the most
not satisfied (NS), extremely not satisfied (ENS). The pie improved aspect after implementing the 5C’s model. In
chart below shows the relevant data (Fig. 3.). terms of Cultures, Comparisons, and Connection, these
standards received different outcomes due to the variety of
language and the purposes of teaching and learning, and the
227
Communities became the least improved standard. There is a Nha Trang University, who already contributed greatly to the
fact that the effectiveness of connecting to the communities success of our work.
is hard to realize in online teaching compared to offline
teaching processes. REFERENCES
As a result, to enhance the English language abilities of [1] American Council on the Teaching of Foreign Languages,
students in the circumstance of online learning, the lecturers Standards for Foreign Language Learning in the 21st Century.
can flexibly implement each standard of 5C’s in specific Yonkers, NY, 1999.
language classes. In particular, with students in this research, [2] Li Xiufen. AP 中文 5C 教学理论、策略与实务 (in Chinese).
Taipei, Xin Xuelin, 2013.
it is more appropriate to utilize these standards differently.
For instance: [3] Zhang Jinlan. 華語文課程與教學設計(in Chinese). Taipei,
Wenguang, 2016.
Implementing Communication for students of all classes. [4] Zare Behtash, E. & Banaruee, H. “Critical Evaluation of the
New Headway Advanced and the ILI Advanced Series: A
Implementing the Communication, Cultures, Comparison of Curricular Components and CLT Objectives
Comparisons to design the major topics for Chinese lessons Based on ACTFL”. International Journal of Applied
to enhance the communicative competence of the students, Linguistics and English Literature, Vol. 6, No. 5 p. 182, 2017.
help them understand the cultures, and have the ability to [5] Lijuan, C. “Cultivation of Middle School Students’ English
Core Competencies From American 5C Standards”. US-China
clarify the differences between the Chinese and Vietnamese Foreign Language, Vol. 19, No. 9, pp. 242-247, 2021.
languages. [6] Gandhi, S. & Monika, J. “The Role of Emerging
Implementing the Communication and Connection in Technologies in Standardizing Language Teaching and
Learning through 5C's”. International Journal of Emerging
designing class activities for non-English majored students Technologies in Engineering Research (IJETER) Volume 6,
utilizing different types of online applications, systems or Issue 4, 2018.
social networks to promote their language knowledge as well [7] Parris, H. and Estrada, L.M. (2019). “Digital Age Teaching
as the ability to connect what they have learned to other for English Learners”. In The Handbook of TESOL in K-12,
fields. L.C. Oliveira (Ed.).
https://doi.org/10.1002/9781119421702.ch10
Implementing the Communication, Comparisons and [8] Kuznetsova, Natalya & Soomro, Kamal. (2019). “Students’
Connection in English majored classes (especially Syntax). Out-of-class Web 2.0 Practices in Foreign Language
Learning”. Journal of Education and Educational
In Vietnam, the tendency of conducting studies regarding Development. 6. 10.22555/joeed.v6i1.2713.
the implementation of 5C’s on foreign language teaching is [9] Kemmis, S & Mc Targat, R. The Action Research Planner.
not prevalent, so this research aimed to contribute to the Victoria, Deakin University Press, 1982.
improvement of qualities in foreign language teaching and [10] Carr, W & Kemmis, S. Becoming Critical: Education,
learning in the 4.0 era. Knowledge and Action Research. Palmer Press, London, 1986
[11] McTaggart, R. Participatory Action Research. New York:
State University of New York Press, 1997.
ACKNOWLEDGMENT
This paper and the research behind it would not have
been possible without the support of lecturers and students at
228
Antibacterial Activity of Aqueous Extracts from

Marine Sponges Found in Vietnam’s Sea
Huynh Nguyen Duy Bao Nguyen Khac Bat
Faculty of Food Technology Research Institute for Marine Fisheries
Nha Trang University Hai Phong, Vietnam
Nha Trang, Vietnam nkbat2005@gmail.com
hndbao@ntu.edu.vn
microorganisms. The previous study reported that more than

Abstract: In recent years, the emergence of new infectious
diseases, the resurgence of previously controlled infections, and 201 species of sponges are found in the sea of Vietnam [7].
the rise in bacterial resistance have necessitated research into In which, some species have been interested in researching
the development of new antibacterials. Marine sponges have and exploiting bioactive compounds [2][8][9]. However, the
been known as one of the abundant sources of bioactive research on bioactive compounds of sponges in Vietnam's
compounds from marine creatures. However, they have not sea and their application is still limited, so this valuable
been effectively exploited and utilized in Vietnam. The present marine resource has not been exploited and used effectively.
study was to investigate the antibacterial activity of aqueous The present study, therefore, was to investigate the
extracts from marine sponges found in Vietnam’s sea. All the antibacterial activity and preliminary identify some
extracts had antibacterial activity against both Gram-positive bioactive compounds in aqueous extracts from Vietnamese
(Bacillus cereus, Staphylococcus aureus) and Gram-negative marine sponges in order to find potential species for use in
(Escherichia coli, Salmonella typhimurium) pathogenic strains. the extraction of antibacterial substances applied in the
The extracts from Spheciospongia sp. exhibited the highest pharmaceutical and food industries.
antibacterial activity against both Bacillus cereus and
Staphylococcus aureus, whereas the extracts from Clathria II. MATERIALS AND METHODS
(Thalysias) reinwardti exhibited the highest antibacterial A. Materials
activity against Escherichia coli. Extracts from Xestospongia
testudinaria, Hyrtios erecta, and Ircinia mutans inhibited Marine sponges were collected from the sea of Vietnam at
Salmonella typhimurium more effectively. The antibacterial Phu Quoc, Co To, and Hai Van - Son Tra (Table 1 and Fig.
activity of the extracts varied depending on the marine sponge's 1). The sponges were immediately chilled after collection and
habitat. These findings indicate that marine sponges could be a divided into 2 parts. One part was transported to the Research
potential source for antibacterial research and acquisition. Institute for Marine Fisheries (Hai Phong, Vietnam) for
species identification, and the other part was transported to
Keywords: marine sponges, antibacterial activity, Nha Trang University's laboratory. The sponges were kept
bioactive compounds frozen at -20oC in the laboratory for this study.
I. INTRODUCTION B. Identification of sponge species
Scientists have found over 5,000 different species of Marine sponges were identified according to the guide of
sponges, but it is believed that there are over 8,000 different Hooper [10], using two forms of histological preparation: a
species of sponges on Earth [1]. Marine sponges are ranked spicule preparation (for those species with a mineral
at the top of the list for their natural sources of bioactive skeleton) to determine the diversity and geometry of
compounds and their potential for medicinal uses due to the spicules in the skeleton; and a perpendicular section through
diversity of chemical structures of the metabolites they the sponge tissue to determine the structure of the skeleton,
contain. Recent studies have revealed that marine sponges the water-canal system, and other aspects of histology.
contain bioactive chemicals that are antibacterial, anti-
inflammatory, anti-tuberculosis, anti-cancer, antifungal, Table 1. Sampling locations of sponge species in Vietnam's sea
antimalarial, antiviral, HIV resistance agents, and No. Species Location Code
antioxidants [2][3]. Marine sponges are classified as a group 1 Xestospongia testudinaria Phu Quoc PQ
of natural resources containing a number of antibacterial 2 Gelliodes fibulata Phu Quoc PQ
substances. Extracts from different marine sponge species 3 Hyrtios erecta Phu Quoc PQ
4 Ecionemia acervus Phu Quoc PQ
were found to be active against a wide range of pathogenic
5 Petrosia sp. Phu Quoc PQ
bacteria [4]. The bioactive compounds in sponge extracts 6 Clathria (Thalysias) reinwardti Phu Quoc PQ
have been isolated and characterized, such as phenolic 7 Ircinia mutans Phu Quoc PQ
compounds, alkylpiperidine, bromopyrrole, alkaloids, 8 Callyspongia (Toxochalina) sp. Co To CT
polyketides, pyrroloiminoquinone, sesquiterpene-quinones, 9 Haliclona sp. Co To CT
sesquiterpene-hydroquinones, terpenoids, peptides, 10 Dysidea fragilis Co To CT
11 Ircinia mutans Co To CT
and proteins [5][6]. 12 Biemna sp. Co To CT
Natural conditions in Vietnam are favorable for the 13 Spongia sp. Co To CT
14 Spheciospongia sp. Co To CT
growth of marine sponges and their associated 15 Xestospongia testudinaria Hai Van – Son Tra HV-ST

vigorously for 2 minutes. If foam appears stable for 10

minutes, identify the presence of saponins.
Sterols: 1 ml of concentrated acid H2SO4 was added
along the wall of a test tube containing 0.5 ml of sponge
extract in chloroform. A reddish brown color appearing in
the chloroform layer indicates the presence of sterols.
Flavonoids: The sponge extract was reacted with a few
drops of lead acetate solution. The formation of a yellow
precipitate indicates the presence of flavonoids.
Glycosides: A solution of glacial acetic acid (2.0 ml)
containing 1 drop of 2% ferric chloride was mixed with 5 ml
of sponge extract and 0.5 ml of H2SO4 concentrated. A
brown ring formed between the layers indicates the presence
Figure 1. Marine sponges collected in Vietnam’s sea of glycosides.
C. Preparation of extracts from marine sponges Polyphenols: The sponge extract was reacted with a few
drops of ferric chloride solution. The appearance of a blue
Marine sponge extracts were prepared by adapting our
black color indicates the presence of polyphenols.
previously developed procedure [11] with slight
modifications. Briefly, the sponges were individually Alkaloids: The sponge extract was reacted with Hager’s
ground with a Panasonic blender model MX-GM1011 reagent, a strong picric acid solution. The appearance of a
(Panasonic Malaysia), and 100 g of the ground material was yellow-colored precipitate indicates the presence of
separately extracted with 200 mL of double-distilled water alkaloids.
at 30 ± 2°C for 15 min by ultrasound-assisted extraction F. Statistical analysis
technology (frequency 20 kHz). The extract was obtained by
All of the experiments were carried out in triplicate. A
centrifuging the mixture at 3,000  g for 20 minutes at 4°C
one-way ANOVA and Tukey's Multiple Comparisons of
and then filtering it through Whatman grade 1 filter paper.
Means were used to determine the statistical difference
The volume of the extract was raised to 200 mL by the
between the samples at P < 0.05 using R software version
addition of double-distilled water. Consequently, 1 mL of
4.1.2 (http://cran.R-project.org).
each extract was equivalent to 0.5 g of the sponge.
D. Evaluation of antibacterial activity III. RESULTS AND DISCUSSION
Antibacterial activity was evaluated using the well Antimicrobial activity of aqueous extracts from marine
diffusion method on Trypticase soy agar (TSA, Sigma- sponges against S. aureus as shown in Fig. 2.
Aldrich, USA) as described by Bao et al. [12]. Four species
of foodborne pathogenic bacteria were used as references
for the antibacterial assay of the extracts, including Bacillus
cereus, Escherichia coli, Salmonella typhimurium, and
Staphylococcus aureus. All tested microorganisms were
supplied by the Microbiology Laboratory, Center for
Experiments and Practices, Nha Trang University. Briefly,
TSA agar plates were inoculated with bacterial strains under
aseptic conditions, and wells (diameter = 5 mm) were filled
with 30 µL of the extracts or 30 μL of double-distilled water
for control. The plates were incubated at 37 oC for 24 hours
under aerobic conditions. After the incubation period, the
diameter of the growth inhibition zone around the wells was
measured. The inhibition zones were reported in millimetres
(mm).
Figure 2. Antimicrobial activity of aqueous extracts from marine
E. Identification of bioactive compounds sponges against Staphylococcus aureus (PQ: Phu Quoc; CT: Co
To; HV-ST: Hai Van – Son Tra). Data are presented as mean ± SD
Bioactive compounds in the extracts from marine
(n = 3).
sponges were identified on a preliminary basis according to
the procedures of Harborne [13]. Among the tested extracts, those from Callyspongia
Protein: The sponge extract was reacted with a few (Toxochalina) sp. (CT) and Dysidea fragilis (CT) did not
drops of Millon’s reagent, which was then heated gently. A show antibacterial activity against S. aureus. The extracts
reddish-brown coloration or precipitate indicates the from Xestospongia testudinaria (PQ), Clathria (Thalysias)
presence of protein. reinwardti (PQ), Biemna sp. (CT), Spongia sp. (CT), and
Saponins: A test tube containing a mixture of 0.5 mL of Spheciospongia sp. (CT) had remarkably higher (P < 0.05)
sponge extract and 2 mL of distilled water was shaken antibacterial activity against S. aureus than those from
others. The extract from Spheciospongia sp. (CT) had the
230
highest antibacterial activity against S. aureus. The results antibacterial activity against S. typhimurium than the extract
revealed that the extract from X. testudinaria collected at from X. testudinaria (HV-ST), and the extract from I.
Phu Quoc (PQ) had much stronger antibacterial activity mutans (PQ) had significantly lower antibacterial activity
against S. aureus than the extract from X. testudinaria against S. typhimurium than the extract from I. mutans (CT).
collected at Hai Van – Son Tra (HV-ST), while the extract
from I. mutans collected at Phu Quoc (PQ) had lower
antibacterial activity against S. aureus than the extract from
I. mutans collected at Co To (CT).
Figure 5. Antimicrobial activity of aqueous extracts from marine

sponges against Escherichia coli (PQ: Phu Quoc; CT: Co To; HV-
ST: Hai Van – Son Tra). Data are presented as mean ± SD (n = 3).
Figure 3. Antimicrobial activity of aqueous extracts from marine As shown in Fig. 5, the antibacterial activity of the
sponges against Bacillus cereus (PQ: Phu Quoc; CT: Co To; HV- marine sponge extracts tested against E. coli was
ST: Hai Van – Son Tra). Data are presented as mean ± SD (n = 3). considerably strong, with the exception of extracts from G.
fibulata (PQ), Petrosia sp. (PQ), Spheciospongia sp. (CT)
Fig. 3 shows the antimicrobial activity of aqueous and X. testudinaria (HV-ST). The antibacterial activity of
extracts of marine sponges against B. cereus. Antibacterial the extract from X. testudinaria (PQ) against E. coli was
activity against B. cereus was found in all of the marine likewise higher than that of the extract from X. testudinaria
sponge extracts tested. When compared to other extracts, the (HV-ST), and the antibacterial activity of the extract from I.
extracts from I. mutans and Spheciospongia sp. (CT) mutans (PQ) against E. coli was lower than that of the
demonstrated much stronger antibacterial activity against B. extract from I. mutans (CT).
cereus. The results also found that the extract from X.
Beesoo et al. [14] reported that the crude extract from
testudinaria (PQ) had much stronger antibacterial activity
Spheciospongia sp. displayed antibacterial activity against S.
against B. cereus than the extract from X. testudinaria (HV-
aureus and B. cereus but not against E. coli. However, the
ST), and the extract from I. mutans (PQ) had higher
present study found the aqueous extract from
antibacterial activity against B. cereus than the extract from
Spheciospongia sp. collected at Co To, Vietnam had
I. mutans (CT).
antibacterial activity against S. aureus, B. cereus, E. coli.
This is probably because habitat influences the content and
composition of bioactive compounds in sponges.
Table 2. Preliminarily identified bioactive compounds

Bioactive compounds
No. Species
Pr Sa St Fl
1 Xestospongia testudinaria (PQ) + + + +
2 Gelliodes fibulata (PQ) + - + +
3 Hyrtios erecta (PQ) + + + +
4 Ecionemia acervus (PQ) + + + +
5 Petrosia sp. (PQ) + + + +
6 Clathria (Thalysias) reinwardti (PQ) + + + +
7 Ircinia mutans (PQ) + + + +
Figure 4. Antimicrobial activity of aqueous extracts from marine 8 Callyspongia (Toxochalina) sp. (CT) + - + +
sponges against Salmonella typhimurium (PQ: Phu Quoc; CT: Co 9 Haliclona sp. (CT) + + + +
To; HV-ST: Hai Van – Son Tra). Data are presented as mean ± SD 10 Dysidea fragilis (CT) + + + +
(n = 3). 11 Ircinia mutans (CT) + + + +
12 Biemna sp. (CT) + + + +
The marine sponge extracts tested showed considerably
13 Spongia sp. (CT) + + + +
strong antibacterial activity against S. typhimurium with the
14 Spheciospongia sp. (CT) + + + +
exception of extracts from Gelliodes fibulata (PQ), Petrosia
15 Xestospongia testudinaria (HV-ST) + - + +
sp. (PQ) and X. testudinaria (HV-ST) (Fig. 4). Similarly, the Protein (Pr); Saponins (Sa); Sterols (St); Flavonoids (Fl); Presence (+); Absence (-).
extract from X. testudinaria (PQ) had significantly higher
231
Table 2. Preliminarily identified bioactive compounds (cont.) [2] M. F. Mehbub, J. Lei, C. Franco, and W. Zhang, “Marine
sponge derived natural products between 2001 and 2010:
Bioactive compounds
No. Species Trends and opportunities for discovery of bioactives”,
Gl Po Al Marine Drugs, vol. 12, no. 8. pp. 4539–4577, 2014.
1 Xestospongia testudinaria (PQ) - + + [3] R. Pallela and H. Ehrlich, Marine sponges:
2 Gelliodes fibulata (PQ) + + + Chemicobiological and biomedical applications. 2016.
3 Hyrtios erecta (PQ) + + + [4] P. R. Marinho, G. R. S. Muricy, M. F. L. Silva, M. G. de
4 Ecionemia acervus (PQ) + + + Marval, and M. S. Laport, “Antibiotic-resistant bacteria
5 Petrosia sp. (PQ) + + + inhibited by extracts and fractions from Brazilian marine
6 Clathria (Thalysias) reinwardti (PQ) + + + sponges”, Rev. Bras. Farmacogn., vol. 20, no. 2, pp. 267–
275, 2010.
7 Ircinia mutans (PQ) - + +
[5] M. Laport, O. Santos, and G. Muricy, “Marine Sponges:
8 Callyspongia (Toxochalina) sp. (CT) - + +
Potential Sources of New Antimicrobial Drugs”, Curr.
9 Haliclona sp. (CT) + + + Pharm. Biotechnol., vol. 10, no. 1, pp. 86–105, 2009.
10 Dysidea fragilis (CT) + + + [6] K. S. Singh and M. S. Majik, “Bioactive alkaloids from
11 Ircinia mutans (CT) - + + marine sponges”, in Marine Sponges: Chemicobiological
12 Biemna sp. (CT) + + + and Biomedical Applications, 2016, pp. 257–286.
13 Spongia sp. (CT) + + + [7] T. M. Quang, “A review of the diversity of Sponges
14 Spheciospongia sp. (CT) + + + (Porifera) in Vietnam”, The 2nd International workshop on
15 Xestospongia testudinaria (HV-ST) - + +
Marine Bioresources of Viet Nam. pp. 109–115, 2013.
Glycosides (Gl); Polyphenols (Po); Alkaloids (Al); Presence (+); Absence (-).
[8] M. S. Kumar and A. K. Pal, “Investigation of bioactivity of
extracts of Marine Sponge, Spongosorites halichondrioides
Table 2 presents the bioactive compounds in the extracts (Dendy, 1905) from western coastal areas of India”, Asian
Pac. J. Trop. Biomed., vol. 2, no. 3 SUPPL., 2012.
from marine sponges that were identified on a preliminary [9] S. K. Kim and P. Dewapriya, “Bioactive Compounds from
basis. Protein, sterols, flavonoids, polyphenols, and Marine Sponges and Their Symbiotic Microbes. A Potential
alkaloids are presented in all of the extracts. These Source of Nutraceuticals”, in Advances in Food and
compounds are well known to have antibacterial activity Nutrition Research, vol. 65, 2012, pp. 137–151.
[15][16]. Therefore, all of the extracts had antibacterial [10] J. N. a. Hooper, “‘Sponguide’. Guide To Sponge Collection
activity against both Gram-positive and Gram-negative and Identification”, Order A J. Theory Ordered Sets Its
pathogenic strains. Saponins were not detected in the Appl., 2003.
extracts from G. fibulata (PQ), Callyspongia (Toxochalina) [11] H. N. D. Bao and N. K. Bat, “Antioxidant activity of extracts
sp. (CT), or X. testudinaria (HV-ST). Glycosides were also from selected marine sponge species in the sea of Vietnam”,
Sci. Technol. J. Agric. Rural Dev., no. 24/2018, pp. 61–68,
not detected in the extracts from Callyspongia 2018.
(Toxochalina) sp. (CT), X. testudinaria (PQ), X. [12] H. N. Duy Bao, P. Thi Hien, and V. Le Quyen, “Antioxidant
testudinaria (HV-ST), I. mutans (PQ), or I. mutans (CT). and antimicrobial activities of hydrophilic extracts from
shallot and garlic bulbs, and their effects on round scad
IV. CONCLUSION during ICED storage”, Ann. Univ. Dunarea Jos Galati,
Fascicle VI Food Technol., vol. 44, no. 1, pp. 26–44, 2020.
The present study found that the aqueous extracts from [13] J. B. Harborne, “Phytochemical Methods A Guide To
Vietnamese marine sponges had antibacterial activity Modern Techniques Of Plant Analysis, Third Edition”,
against both Gram-positive and Gram-negative pathogenic Chapman Hall, p. 58, 1998.
strains. The antibacterial activity of the extracts varied [14] R. Beesoo, R. Bhagooli, V. S. Neergheen-Bhujun, W. W. Li,
depending on the sponge species and habitat in which they A. Kagansky, and T. Bahorun, “Antibacterial and antibiotic
thrived. These findings indicate that marine sponges living potentiating activities of tropical marine sponge extracts”,
in Vietnam’s sea could be a promising source for Comp. Biochem. Physiol. Part - C Toxicol. Pharmacol., vol.
196, pp. 81–90, 2017.
antibacterial research and acquisition.
[15] D. S. Bhakuni and D. S. Rawat, Bioactive marine natural
REFERENCES products. 2005.
[16] A. Kanazawa, “Sterols in marine invertebrates”, Fisheries
[1] K. Chairman, A. J. A. R. Singh, and G. Alagumuthu, Science, vol. 67, no. 6. pp. 997–1007, 2001.
“Cytotoxic and antioxidant activity of selected marine
sponges”, Asian Pacific J. Trop. Dis., vol. 2, no. 3, pp. 234–
238, 2012.
232
Digital Competence of University Students:

Developing Information and Data Literacy for IT
Students at Ho Chi Minh City University of
Anh Tho Mai Quynh Trang Mai
Faculty of Information Technology Faculty of Fashion and Tourism
Education Education
https://orcid.org/0000-0001-9159-2379 mqtrang@hcmute.edu.vn
Abstract: As one of the five competence areas of the better in digital learning [7], [8], thereby more successful
DigComp framework, information and data literacy is crucial learning and work in environments increasingly influenced by
to university students and comprises the following technologies [9], [10]. Digital competence is therefore
competences: (1.1) browsing, searching, and filtering, essential to students in the 21st century [11].
information and digital content; (1.2) evaluating data, The issue of digital competence development has thus
information, and digital content; and (1.3) managing data, attracted the attention of scholars, practitioners, and
information, and digital content. Surveys of information and policymakers in hope of equipping citizens with sufficient
data literacy of information technology students at Ho Chi competences for their participation in digital society [12] and
Minh City University of Technology and Education competences needed for lifelong learning [2].
(HCMUTE) and students at three other Vietnamese
Digital competence development is also one of the
universities found students’ high levels of browsing,
educational priorities [13]. A review of 126 papers on the
searching, and filtering data but intermediate levels of
digital competence of university students [11] revealed lower-
evaluating and managing data. The study presented findings than-expected levels of proficiency despite being born into a
of the application of the four-step process to intentional digital era. It is essential that universities increase their
information search in a Web Programming course for 88 students’ digital competence in the 21st century [8],
third-year IT students at HCMUTE. Empirical results showed contributing to their successful learning and at the same time
that students achieved advanced levels in information and meeting new job requirements [5].
data literacy after having received instructions on information
seeking. Research findings provide insights into helping There are two main approaches to developing students’
students develop information and data literacy by instructing digital competence [11]. From the macro approach, (1)
them to apply the four-step process, subsequently enhancing research into competence development solutions was carried
out systematically at different levels of higher education,
their digital competence.
including national-level policies of digital competence
development specified by university-level goals and activities
Keywords: digital competence, DigComp framework,
[14]–[16]; (2) infrastructure of universities and strategic
digital competence development, information and data
leadership [14], [16]; and (3) systematic innovation of
literacy, university students curriculum and integration of digital competence development
I. INTRODUCTION content into the curriculum [17]–[19].
Digital competence and development of students’ Studies from the micro approach reported on specific
competence measures lecturers can use, including the widespread and
diverse technological application in teaching and training
Digital competence is regarded as a set of technological activities to increase students’ exposure to technologies [20].
skills used to optimize daily [1], and “the confident, critical In particular, lecturers use e-learning systems in combination
and responsible use of the technologies from the society of with digital tools for teaching activities [21], [22], implement
information for work, entertainment and education” [2, p9]. positive teaching with the help of virtual classroom tools [23],
It is one of the eight key competences for lifelong learning and design specialized processes or modules to develop digital
[2], and a crucial factor that helps ensure studies and learning competence [24].
outcomes of students in current learning contexts [4], [5]. While the development of students’ digital competence
Possessing high levels of digital competence heightens from the macro approach is at national or university levels,
students’ readiness to use digital technologies in learning [6], lecturers’ attempts are from the micro approach, which was
facilitating their abilities to interpret, understand and perform also the focus of the present study where students were

233
instructed to seek information purposefully in a bid to increase 350 IT students at Ho Chi Minh City, University of
their digital competence. Technology and Education (HCMUTE), and an actual test
with 3,467 students from three Vietnamese universities. The
DigComp framework and information and data three participating universities include an autonomous, public
literacy university (HCMUTE), a non-autonomous, public university
Analyzing digital competence frameworks in practice, (Ho Chi Minh City Nong Lam University - HCMNLU), and a
Anuska Ferrari (2012) presented the definition of digital private university Ho Chi Minh City University of Foreign
competence as a set of knowledge, skills, and attitudes Languages & Information Technology - HUFLIT).
(including abilities, strategies, values, and perceptions)
required when using information and communication Table 2. Information and data literacy of students
technologies and digital tools to complete tasks, solve Phase 1 Competence Mean SD
problems, communicate, manage information, collaborate, Pilot test (350 IT 1.1 7.08 2.97
create and share content, build knowledge proactively, students at 1.2 4.64 2.83
flexibly, creatively, effectively, adequately, ethically and HCMUTE) 1.3 4.99 3.31
appropriately for work, learning, entertainment, and social life Phase 2 Unive HCMUTE HCMNLU HUFLIT
[25]. The definition was used by the European Commission to rsity (1336) (1021) (1110)
develop the DigComp framework. UNESCO (2018) found Comp
Mean SD Mean SD Mean SD
etence
competences in 47 frameworks of different countries and Actual
test (3467 1.1 6.12 3.03 5.40 3.48 5.50 3.49
regions are reflected in Europe’s DigComp [26]. students) 4.03 2.64 3.36 2.72 3.83 2.94
1.2
Among the three versions of DigComp, DigComp 2.1 1.3 4.43 3.04 3.71 3.02 3.99 3.15
[27], was upgraded by the Institute for Prospective
As shown in Table 2, IT students at HCMUTE possessed
Technological Studies (IPTS) from DigComp 2.0 [28] and
advanced levels at competence 1.1 (Browsing, searching, and
DigComp [1]. DigComp, developed from 15 frameworks
filtering data, information, and digital content), but lower
previously designed, is the most updated and comprehensive
intermediate levels at competences 1.2 (Evaluating data,
framework [29], [30]. It was aimed at developing policies of
information, and digital content) and 1.3 (Managing data,
digital competence and used as a basis for educational and
information, and digital content). Students at all three
training initiatives [28].
universities also performed better at competence 1.1 than
The DigComp framework comprises 21 competences of competences 1.2 and 1.3.
knowledge, skills, and attitudes in five areas, namely (1)
As shared by students in follow-up interviews, although
Information and data literacy, (2) Communication and
they were familiar with using Google for information search,
collaboration, (3) Digital content creation, (4) Safety, and (5)
they evaluated and managed information mainly with their
Problem-solving. Competences in the first area, the focus of
own experience and did not have formal instructions.
the present study, are shown in Table 1.
Therefore, the study aimed to present the findings by
Table 1. The Information and data literacy competence instructing students to apply the four-step information search
Dimension Sub-skills Definition process to help them improve information and data literacy as
well as digital competence.
1.1 To articulate information needs, to
Browsing, search for data, information, and II. METHOD
searching, content in digital environments, to
access and navigate between Research Design
and filtering
data, them. To create and update The experimental method was used in this research with a
information, personal search strategies. treatment group (TG) and a control group (CG), and a pretest-
and digital posttest control group design.
content
Research Sample
1.2 To analyze, compare and critically
Information Evaluating evaluate the credibility and Eighty-eight IT students enrolling for the Web
and data data, reliability of sources of data, programming course in semester one of the 2021-2022
literacy information, information, and digital content. academic year at HCMUTE were put into two classes based
and digital To analyze, interpret and critically on their course registration. The treatment group consisted of
content evaluate the data, information, 44 students in the class WEPR330479_21_1_01CLC, whereas
and digital content the other 44 students in the control group were in the class
1.3 To organize, store and retrieve WEPR330479_21_1_02CLC.
Managing data, information, and content in
data, digital environments. To organize All students participated in the digital competence
information, and process them in a structured assessment at the beginning of the semester. However, 37 out
and digital environment. of 44 students in the treatment group completed the
content assessment to be eligible for the next phase of the study,
whereas the number of students in the control group was 32.
Vietnamese students’ levels of proficiency in
information and data literacy After taking the digital competence assessment, students
of the treatment group received instructions for applying the
The study was conducted in two phases, a pilot test of the four-step process of intentional information search, no
digital competence assessment tool with the participation of instructions of the kind were provided for the control group.
234
Research Instrument and Procedure Steps Activities Lecturer’s assistance/Digital tools

The digital competence assessment tool Depending on students’ levels of
The study used MATPlatform, a web-based digital proficiency, the lecturer equips
students with the necessary search
competence assessment tool developed from findings of the skills by selecting the right tools.
ERASMUS+ 2016 project [31], [32]. The tool, available on
https://nanglucso.hcmute.edu.vn, measured students’ digital • I-Search model [38]
competence in both groups before and after the intervention. • Available search engines with their
Procedures to instruct students’ intentional information features, how they work, and
when to use which engines:
search
Google (Google Scholar,
Previous research showed that asking students to find Google’s preferences, Google’s
information on the Web might not deliver good learning Advanced), Yahoo!, Bing, Ask,
results [33], [34]. When students search for information, they and AOLSearch
Develop
tend to search for predetermined answers and are not • Directories: Yahoo!, About, and
search
comprehending or reflect on the meaning of what they have Use
strategies to
LookSmart
found; they are simply searching for the one answer that the strategies
to search
identify • Invisible Web: free (AskERIC và
teacher is looking for [34]. Information seeking without appropriate FindArticles) and paid (EBSCO,
specific purposes would result in a meaningless search. the Web
sources of GALE, ProQuest…)
However, purposeful information search requires a process of information
four steps: (a) plan, (b) use strategies to search the Web, (c) • Feeds: websites, blog/Twitter
feeds, YouTube videos, or
evaluate, and (d) triangulate sources [35].
podcasts
• Personalized information search,
management and sharing:
Bookmarking, online tools:
Diigo (www.diigo.com), Zotero
(firefox extension), eThemes
(http://ethemes.missouri.edu/the
mes), TrackStar
(http://trackstar.4teachers.org),
and NoodleTools (http://
Figure 1. The four-step research process noodletools.com/)
As a result, students were instructed to use the four-step Students are asked to save information
process of intentional information search [35]. The process Evaluate the found in a Google Drive folder.
was covered in manuals of meaningful learning with authenticity,
technology [36], [37]. Detailed instructions were provided to The lecturer models the analysis of a
accuracy, website and asks guiding questions to
help students understand every step and apply the process to and sources help students determine the reliability
meet the course’s requirements. Within the class, students of of a website.
worked in groups to develop a specific website application. To Evaluate information.
do this, students need to search, examine and assess at least • www.lib.berkeley.edu/TeachingLi
Analyze
six similar websites. The list of groups and topics was b/Guides/Internet//Evaluate.ht
information
accessible on a Google Sheets document file. Reports and relevancy ml (to evaluate resources)
presentation recordings of all groups were also accessible in a and • www.easywhois.com (to look up
Google Drive folder for reference and feedback. reliability. domains and verify website
authenticity)
Table 3 showed procedures to instruct students to apply
the four-step process, facilitating the information search Identify at Information after being evaluated and
requirements of the course. least two verified is saved in another folder by
Triangulat each group on Google Drive.
sources of
e sources
Table 3. Steps to instruct students in intentional information search information
to verify.
Steps Activities Lecturer’s assistance/Digital tools
Selecting search terms might be Validity and reliability

Identify
search challenging for students. The lecturer MATPlatform is used to assess the digital competence of
purposes, models questions such as who, where,
when, and what to identify the right
both treatment and control groups before and after the
Plan search intervention. It was developed from the tool of Bartolomé et
terms, and search terms.
al. (2021) [31], which was tested by 21 European experts.
types of
information.
Search terms are developed into search MATPlatform’s content validity was tested by eight experts in
strings to be used on search engines. IT (4), education (3), and quality assurance (1), who are
Vietnamese lecturers with research interests in digital
competence. The tool was also found to be valid and reliable,
using Cronbach’s alpha, in the pilot test with 350 IT students
and the actual test with 3,467 students from three universities
(See Table 4).
235
Table 4. Reliability of MATPlatform instrument Table 6. Summary Independent-Samples T-Test between two
groups
Stages of research Number of Cronbach’s alpha
students Std. Mean
Crit- Sig. (2-
Pilot test 350 0.951 Group N M S.D. Error Differe
eria tailed)
Mean -nce
Actual test 3467 0.945
TG 37 6.43 2.977 0.489
Data analysis 1.1 -1.320 0.063
CG 32 7.75 2.782 0.492
The data analysis was carried out via IBM SPSS 25.0
package software. Mean values were calculated for both TG 37 3.89 2.612 0.429
groups before and after the intervention. Paired-samples t-test 1.2 -0.360 0.568
CG 32 4.25 2.578 0.456
and summary independent-samples t-test were used to
compare the means of two samples. TG 37 4.54 2.745 0.451
1.3 0.040 0.954
III. RESULT CG 32 4.50 2.973 0.526
Pretest results
Table 5 showed the mean and standard deviation values Posttest results
concerning information and data literacy of the treatment
Mean and standard deviation values after the intervention
group and the control group. Before the intervention, the mean
of the two groups were shown in Table 7.
values of the treatment group in all three competences, 1.1,
1.2, and 1.3, were smaller than those of the control group. Table 7. Post-test data of TG and CG groups
Table 5. Pre-test data of TG and CG groups Criteria Group N Mean S.D. Min Max
Criteria Group N M S.D. Min Max
TG 37 7.89 2.447 2 10
TG 37 6.43 2.977 0 10 1.1
CG 32 7.88 2.537 1 10
1.1
CG 32 7.75 2.782 0 10
TG 37 5.78 2.760 1 10
TG 37 3.89 2.612 0 10 1.2
1.2 CG 32 4.34 2.535 1 10
CG 32 4.25 2.578 0 10
TG 37 6.35 2.463 1 10
TG 37 4.54 2.745 0 10 1.3
1.3 CG 32 4.72 2.842 1 10
CG 32 4.50 2.973 0 10
Paired-samples t-test (see Table 8) was used to compare
Digital competence between two groups in each the mean ranges of each pair of variables, i.e. competences
competence of information and data literacy was tested with a 1.1, 1.2, and 1.3 before and after the intervention of the two
summary independent-samples t-test with 95% confidence. groups. The range of the mean depended on standard
Hartley’s test for homogeneity of variance was used. deviation, standard error of the variables, and the reliability of
 H0: TG variance = CG variance the test.
In the treatment group, in all three competences 1.1, 1.2,
 H1: TG variance # CG variance
and 1.3, sig. values were < 0.05, allowing for the rejection of
As sig. values of all three competences are higher than 5%, H0, which means there was a statistically significant
with 0.3501 (competence 1.1), 0.4729 (competence 1.2) and difference in the mean values between the two variables. The
0.3181 (competence 1.3), H0 could not be rejected. Sig. values negative mean values indicated the higher mean values
(two-tailed) of summary independent-samples t-test of the compared to the previous test, and thereby students’ higher
three competences with equal variances assumed were shown levels of competence after information search instructions
in Table 6. Hypothesis testing for the mean was as follows. with 95% confidence.
 H0: TG mean = CG mean In the control group, sig. values in all three competences
were > 0.05, which means that H0 was accepted. Although the
 H1: TG mean # CG mean mean values were negative, indicating the increase in
Sig. values (two-tailed) in all three competences were information and data literacy of the control group, there was
higher than 5% with 0.063, 0.568, and 0.954 respectively, H0 no statistically significant difference between the two
was thus not rejected. Therefore, the mean values of the two variables at the 5% level.
groups were equal.
It could be hence concluded that levels of information and
data literacy of the treatment and control groups before
intervention are equal with 95% confidence.
236
Table 8. Before and After The Intervention of Paired Samples Test universities’ assistance in their learning process [41]. Study
results also confirmed previous research that students need
Paired Samples Test lecturers’ help with technological application in learning,
Criteria N Mean S.D. Std. Sig.(2-
otherwise their application without instructions would appear
to be less effective or even detrimental to the learning process
Error tailed) [42]–[46].
Mean This process needs to be expanded to other classes of the
IT faculty as well as other faculties at HCMUTE to further
TG – Before and After The Intervention
support its effectiveness. Long-term, observational studies
Pair 1.1 37 -1.459 1.070 0.176 0.000 should therefore be conducted in the future to closely examine
the effects of technological instructions and integration into
Pair 1.2 37 -1.892 0.875 0.144 0.000 courses to increase students’ digital competence.
Professional demands of lecturers have required their ability to
Pair 1.3 37 -1.811 0.967 0.159 0.000 aid students' digital competence development. As the 2016 Horizon
report [47] recommended, digital competence remains a challenge
CG – Before and After The Intervention for higher education, and projects aimed at developing students’
digital competence are essential to increase their learning motivation
Pair 1.1 32 -0.125 0.421 0.074 0.103 [9] in today’s increasingly digitized educational environment.
Pair 1.2 32 -0.094 0.296 0.052 0.083 REFERENCES
[1] A. Ferrari, “DIGCOMP: A Framework for Developing and
Pair 1.3 32 -0.219 0.659 0.117 0.070 Understanding Digital Competence in Europe”, 2013, doi:
10.2788/52966.
Summary independent-samples t-test continued to be used [2] European Commission, “Proposal for a COUNCIL
to examine the competence difference after the intervention of RECOMMENDATION on Key Competences for LifeLong
Learning”, 2018, [Online]. Available:
the two groups (see Table 9). http://www.theeuropeanlibrary.org/tel4/record/200000467889
8.
Table 9. The difference in information and data literacy between [3] S. Kluzer and L. Pujol Priego, DigComp into action - Get
TG and CG after the intervention inspired, make it happen. A user guide to the European Digital
Competence Framework. (JRC Science for Policy Report).
Std. Mean Publications Office of the European Union. 10.2760/112945,
Crit Gro Sig. (2-
N Mean S.D. Error Differe 2018.
eria up tailed) [4] T. Yu, “Examining construct validity of the student online
Mean nce
learning readiness (SOLR) instrument using confirmatory
factor analysis”, Online Learning Journal, vol. 22, no. 4. pp.
TG 37 -1.495 1.070 0.176 277–288, 2018, doi: 10.24059/olj.v22i4.1297.
1.1 -1.370 0.000
CG 32 -0.125 0.421 0.074 [5] M. Florence, S. Brandy, and F. Claudia, “Examining student
perception of readiness for online learning: Importance and
TG 37 -1.892 0.875 0.144 confidence”, Online Learn. J., vol. 24, no. 2, pp. 38–58, 2020,
1.2 -1.798 0.000 doi: 10.24059/olj.v24i2.2053.
CG 32 -0.094 0.296 0.052 [6] H. J. Kim, A. J. Hong, and H. D. Song, “The relationships of
family, perceived digital competence and attitude, and learning
TG 37 -1.811 0.967 0.159 agility in sustainable student engagement in higher education”,
1.3 -1.592 0.000 Sustainability (Switzerland), vol. 10, no. 12. 2018, doi:
CG 32 -0.219 0.659 0.117 10.3390/su10124635.
[7] A. A. Mosa, M. Naz’ri bin Mahrin, and R. Ibrrahim,
“Technological Aspects of E-Learning Readiness in Higher
As sig values (2-tailed) were smaller than 5% with 0.000 Education: A Review of the Literature”, Computer and
in all competences, H0 was rejected. This means there was a Information Science, vol. 9, no. 1. p. 113, 2016, doi:
significant difference in the three competences between the 10.5539/cis.v9n1p113.
treatment and control groups after intervention with 95% [8] E. López-Meneses, F. M. Sirignano, E. Vázquez-Cano, and J.
confidence. It could be concluded that students of the M. Ramírez-Hurtado, “University students’ digital competence
in three areas of the DigCom 2.1 model: A comparative study
treatment group had higher levels of information and data at three European universities”, Australas. J. Educ. Technol.,
literacy. vol. 36, no. 3, pp. 69–88, 2020, doi: 10.14742/AJET.5583.
[9] N. Bergdahl, J. Nouri, and U. Fors, “Disengagement,
IV. DISCUSSION AND CONCLUSION engagement and digital skills in technology-enhanced
Findings showed that with instructions for intentional learning”, Educ. Inf. Technol., vol. 25, no. 2, pp. 957–983,
2020, doi: 10.1007/s10639-019-09998-w.
information search, students’ performance in competences
1.1, 1.2, and 1.3 of the competence area of information and [10] T. He, C. Zhu, and F. Questier, “Predicting digital informal
learning: an empirical study among Chinese University
data literacy was better. This supports the possibility of students”, Asia Pacific Educ. Rev., vol. 19, no. 1, pp. 79–90,
lecturers helping students develop their digital competence via 2018, doi: 10.1007/s12564-018-9517-x.
instructions in courses they are in charge of. The integration [11] A. Sánchez-Caballé, M. Gisbert-Cervera, and F. Esteve-Mon,
of digital competence into the curriculum is believed to be “The digital competence of university students: a systematic
positive due to its relevance to courses [18], [39]. Students literature review”, Aloma, vol. 38, no. 1, pp. 63–74, 2020.
prefer the integration of digital competence content into their [12] C. Iordache, I. Mariën, and D. Baelden, “Developing digital
existing courses in the curriculum [40], rather than taking skills and competences: A quick-scan analysis of 13 digital
courses specializing in digital competence and want
237
literacy models”, Ital. J. Sociol. Educ., vol. 9, no. 1, pp. 6–30, [30] UNESCO, “A Global Framework of Reference on Digital
2017, doi: 10.14658/pupj-ijse-2017-1-2. Literacy”, Information Paper, vol. 51, no. 51. pp. 1–146, 2018.
[13] L. Slavova and K. Garov, “Increasing the digital competences [31] J. Bartolomé, P. Garaizar, and X. Larrucea, A Pragmatic
of students”, Mat. Informatics, vol. 62, no. 1, pp. 42–51, 2019. Approach for Evaluating and Accrediting Digital Competence
[14] A.-L. Petersen, “Teachers’ Perceptions of Principals’ ICT of Digital Profiles : A Case Study of Entrepreneurs and Remote
Leadership”, Contemporary Educational Technology, vol. 5, Workers, no. 0123456789. Springer Netherlands, 2021.
no. 4. 2014, doi: 10.30935/cedtech/6132. [32] J. Bartolomé et al., “Developing a Digital Competence
[15] M. Soby, “Digital competence - a password to a new Assessment and Accreditation Platform for Digital Profiles”,
interdisciplinary field”, Nord. J. Digit. Lit., 2015. no. November, 2018, doi: 10.21125/inted.2018.0888.
[16] B. Newland and F. Handley, “Developing the digital literacies [33] J. Schacter, G. K. W. K. Chung, and A. Dorr, “Children’s
of academic staff: An institutional approach”, Research in internet searching on complex problems: Performance and
Learning Technology, vol. 24. 2016, doi: process analyses”, J. Am. Soc. Inf. Sci., vol. 49, no. 9, pp. 840–
10.3402/rlt.v24.31501. 849, 1998, doi: 10.1002/(sici)1097-
4571(199807)49:9<840::aid-asi9>3.3.co;2-4.
[17] F. M. Røkenes and R. J. Krumsvik, “Prepared to teach ESL
with ICT? A study of digital competence in Norwegian teacher [34] R. Fidel, “A visit to the information mall: Web searching
education”, Computers and Education, vol. 97. pp. 1–20, 2016, behavior of high school students”, J. Am. Soc. Inf. Sci., vol. 50,
doi: 10.1016/j.compedu.2016.02.014. no. 1, pp. 24–37, 1999, doi: 10.1002/(SICI)1097-
4571(1999)50:1<24::AID-ASI5>3.0.CO;2-W.
[18] A. I. Starčič, M. Cotic, I. Solomonides, and M. Volk,
“Engaging preservice primary and preprimary school teachers [35] S. Colaric and D. Jonassen, “Information Equals Knowledge,
in digital storytelling for the teaching and learning of Searching Equals Learning, and Hyperlinking Is Good
mathematics”, British Journal of Educational Technology, vol. Instruction”, Comput. Sch., no. February 2015, pp. 37–41,
47, no. 1. pp. 29–50, 2016, doi: 10.1111/bjet.12253. 2001, doi: 10.1300/J025v17n03.
[19] A. Sánchez-Caballé, M. Gisbert-Cervera, and F. Esteve-Món, [36] J. L. Howland, D. H. Jonassen, and R. M. Marra, Meaningful
“Integrating digital competence in higher education curricula: Learning with Technology Pearson New International Edition.
An institutional analysis”, Educar, vol. 57, no. 1. pp. 241–258, 2014.
2021, doi: 10.5565/REV/EDUCAR.1174. [37] J. L. Howland, D. H. Jonassen, and R. M. Marra, Meaningful
[20] F. Guzmán-Simón, E. García-Jiménez, and I. López-Cobo, learning with technology. (4th ed.). Columbus, OH: Merrill/
“Undergraduate students’ perspectives on digital competence Prentice-Hall., 2012.
and academic literacy in a Spanish University”, Comput. [38] J. Tallman and M. Joyce, “What’s new with the I-Search
Human Behav., vol. 74, pp. 196–204, 2017, doi: research/writing process? Paper presented at the”, 2005.
10.1016/j.chb.2017.04.040. [39] J. A. English, “A Digital Literacy Initiative in Honors:
[21] M. F. Tretinjak and V. Andelic, “Digital Competences for Perceptions of Students and Instructors about Its Impact on
Teachers: Classroom Practice”, 2016 39th Int. Conv. Inf. Learning and Pedagogy”, J. Natl. Coll. Honor. Counc., vol. 17,
Commun. Technol. Electron. Microelectron. MIPRO 2016 - no. 2, pp. 125–155, 2016.
Proc., pp. 807–811, 2016, doi: 10.1109/MIPRO.2016.7522250. [40] M. Hall, I. Nix, and K. Baker, “Student experiences and
[22] H. J. Kim, A. J. Hong, and H. D. Song, “The roles of academic perceptions of digital literacy skills development: Engaging
engagement and digital readiness in students’ achievements in learners by design?”, Electron. J. e-Learning, vol. 11, no. 3, pp.
university e-learning environments”, International Journal of 207–225, 2013.
Educational Technology in Higher Education, vol. 16, no. 1. [41] T. Hallaq, “Evaluating Online Media Literacy in Higher
2019, doi: 10.1186/s41239-019-0152-3. Education: Validity and Reliability of the Digital Online Media
[23] C. Romero-García, O. Buzón-García, and P. de Paz-Lugo, Literacy Assessment (DOMLA)”, Journal of Media Literacy
“Improving future teachers’ digital competence using active Education, vol. 8, no. 1. pp. 62–84, 2016, [Online]. Available:
methodologies”, Sustain., vol. 12, no. 18, 2020, doi: www.jmle.org.
10.3390/SU12187798. [42] K. Aesaert, J. Voogt, E. Kuiper, and J. van Braak, “Accuracy
[24] R. Romero-Tena, R. Barragán-Sánchez, C. Llorente-Cejudo, and bias of ICT self-efficacy: An empirical study into students’
and A. Palacios-Rodríguez, “The challenge of initial training over- and underestimation of their ICT competences”, Comput.
for early childhood teachers. A cross sectional study of their Human Behav., vol. 75, pp. 92–102, 2017, doi:
digital competences”, Sustainability (Switzerland), vol. 12, no. 10.1016/j.chb.2017.05.010.
11. 2020, doi: 10.3390/su12114782. [43] N. Bergdahl, U. Fors, P. Hernwall, and O. Knutsson, “The use
[25] A. Ferrari, “Digital Competence in Practice: An Analysis of of learning technologies and student engagement in learning
Frameworks”, Jt. Res. Cent. Eur. Comm., p. 91, 2012, doi: activities”, Nord. J. Digit. Lit., vol. 13, no. 2, pp. 113–130,
10.2791/82116. 2018, doi: 10.18261/ISSN.1891-943X-2018-02-04.
[26] X. Jashari, B. Fetaji, A. Nussbaumer, and C. Gütl, “Assessing [44] O. E. Hatlevik, G. B. Guomundsdóttir, and M. Loi, “Digital
Digital Skills and Competencies for Different Groups and diversity among upper secondary students: A multilevel
Devising a Conceptual Model to Support Teaching and analysis of the relationship between cultural capital, self-
Training”, 2021, pp. 982–995. efficacy, strategic use of information and digital competence”,
[27] S. Carretero, R. Vuorikari, and Y. Punie, “DigComp 2.1: The Comput. Educ., vol. 81, pp. 345–353, 2015, doi:
Digital Competence Framework for Citizens. With eight 10.1016/j.compedu.2014.10.019.
proficiency levels and examples of use”, Publications Office of [45] F. Goldhammer, G. Gniewosz, and J. Zylka, “ICT Engagement
the European Union. p. 48, 2017, doi: 10.2760/38842. in Learning Environments”, pp. 331–351, 2016, doi:
[28] R. Vuorikari, Y. Punie, S. Carretero, and L. Van Den Brande, 10.1007/978-3-319-45357-6_13.
“DigComp 2.0: The Digital Competence Framework for [46] L. Hietajärvi, K. Salmela-Aro, H. Tuominen, K. Hakkarainen,
Citizens. Update Phase 1: The Conceptual Reference Model”, and K. Lonka, “Beyond screen time: Multidimensionality of
EU Commission JRC Technical Reports. pp. 1–40, 2016, doi: socio-digital participation and relations to academic well-being
10.2791/11517. in three educational phases”, Comput. Human Behav., vol. 93,
[29] F. Siddiq, O. E. Hatlevik, R. V. Olsen, I. Throndsen, and R. no. June 2018, pp. 13–24, 2019, doi:
Scherer, “Taking a future perspective by learning from the past 10.1016/j.chb.2018.11.049.
- A systematic review of assessment instruments that aim to [47] L. Johnson, S. Adams Becker, M. Cummins, V. Estrada, A.
measure primary and secondary school students’ ICT literacy”, Freeman, and C. Hall, Horizon Report - 2016 Higher Education
Educational Research Review, vol. 19. pp. 58–84, 2016, doi: Edition. 2016.
10.1016/j.edurev.2016.05.002.
238
A Simple Design Method For Piled Raft

Foundations
Tong Nguyen Nhat Nguyen Le Anh Dat Nguyen Thanh
Faculty of Civil Engineering Faculty of Civil Engineering Faculty of Civil Engineering
HCMC University of Technology and HCMC University of Technology and HCMC University of Technology and
Education Education Education
tongn@hcmute.edu.vn 17149234@student.hcmute.edu.vn 17149193@student.hcmute.edu.vn
reduce the settlement. Increasing the number of piles depends

Abstract: There have been many previous studies on on reducing the settlement requirement. If the required
analyzing the behavior of the piled raft foundation system by settlement reduction is quite large, the number of piles will
numerical and empirical methods. However, most are quite increase significantly, then a part of the load transfers from the
complex and do not provide a simple enough procedure to raft to the piles due to the increased stiffness of the pile group.
help engineers. In this paper, the objective is to analyze the The key point is how a group of piles is supposed to work to
piled raft foundation by a combination of the settlement- either reduce settlement or take a part of the load from the raft,
based approach and the plate-on-spring approach. The thereby reducing the pressure of the raft and the required
proposed analytical method to estimate the settlement of the settlement. This is a complex problem based on soil-structure
raft and pile group is in the Vietnamese standards TCVN interaction.
9362:2012 and TCVN 10304:2014. The authors employed an
Unlike the classical design for a fully loaded group of
iterative procedure to determine a load sharing factor between piles, where only the pile-soil interaction is concerned to
the raft and pile group. In addition, a plate-on-spring ensure the mobilization of load capacity, the design of the
approach was used to determine pile reactions and the piled raft system needs to pay attention to four types of
internal forces of the raft. Simultaneously, a three- interactions: raft-soil interaction, pile-soil interaction, raft-pile
dimensional piled raft foundation model on PLAXIS 3D interaction, and pile-pile interaction. If the primary concern of
software was used for verification. The results of this paper piled raft foundation design is the settlement as specified in
have to indicate clearly. the requirement to reduce settlement of the raft when piles are
added, then the focus of the piled raft foundation design
Keywords: piled raft foundation, pile bearing capacity,
should be to find this settlement. Rolf Katzenbach and
settlement-based approach, plate-on-spring approach
Deepankar Choudhury (2013) [3] proposed a load distribution
coefficient between raft and group of piles pr, which is a
I. INTRODUCTION
coefficient that depends on settlement. With this coefficient,
In big cities, when the demand for housing is increasing, the pile-raft interaction is simplified. The piled raft foundation
while the land bank for construction is getting less and less, is analyzed as an unpiled raft and a group of piles. The analysis
construction works are forced to be built in areas with of the raft settlement is a fundamental problem in geotechnical
unfavorable ground conditions for the design of the engineering. The difficulty is that there are almost no studies
foundation. Simultaneously, the land bank for housing is also that completely solve the contact pressure distribution below
limited by the land bank for transport infrastructure, which the raft bottom. While the settlement analysis of the pile group
forces buildings to be built higher, increasing the load on the was clearly shown by Fellenius (2021) [4]. Here, the problem
ground. Therefore, the price of the foundation components of pile settlement analysis concerns the behavior of a single
will increase, contributing to pushing up housing prices even pile with the surrounding soil and the superposition of stresses
higher. Engineers must consider not only structural safety but below the pile toe in a large group of piles, known as the
also the cost. Normally, for such large loads, pile foundations equivalent foundation settlement. Furthermore, the pile-soil
are required. A raft foundation is also an option, but it is only interaction is also affected by the presence of surrounding
suitable for medium-load structures on the good underlying piles. Therefore, the settlement analysis of the pile group must
soil. A different type of foundation is required to satisfy both include all three factors mentioned above: the interaction of
structural safety and economic requirement. the single pile with the surrounding soil, the influence of
The pile raft foundation system is a type of combined neighboring piles on this interaction, and the superposition of
foundation often proposed in the first studies of Burland stresses under the pile toe.
(1995) [1], Poulos (2001) [2]. The first form is that piles are In this paper, the authors propose the process of analyzing
used to reduce settlement while the raft bears the entire load. the piled raft foundation by the settlement-based approach. An
This solution limits the number of piles while reducing the iterative procedure has been proposed, and the loop stop
settlement of the raft and thus taking advantage of the load- condition is the similarity of the raft and pile group settlements
bearing capacity of the ground below the bottom of the raft, so that the assumed load sharing factor between the raft and
reducing the cost of the foundation structure. But the problem pile group is confirmed. The raft is considered as a uniformly
is that with larger loads, if only the raft is fully loaded, loaded rectangular area resting elastic half-infinite mass. Its
sometimes it is still not impossible to ensure the solution to settlement is estimated by the partial settlement summation
reduce settlement as the original. When the settlement method [5]. While the method of settlement of piles is taken
increases, it is imperative to increase the number of piles to according to section 7.4, Vietnamese standard TCVN

10304:2014 [6]. Note that the settlement of the pile is closely analytical based on SPT data. Some symbols in Table 1 have
related to the mobilization of the pile bearing capacity. the following meanings: Rc,k is the characteristic value of pile
Therefore, instead of finding the mobilization of pile bearing capacity, Rc,d is the design bearing capacity of the pile, Rstr is
capacity according to settlement to completely ensure the the pile structural capacity and k is the reliability coefficient.
settlement-based approach, the authors still forecast the
ultimate bearing capacity of the pile by using different Table 1. The bearing capacity of pile
classical methods specified in section 7.2.3 and annexes G.2
and G.3 of standard TCVN 10304:2014 [6]. From that, Pile capacity
Rc,k k Rc,d Rstr
consider the design bearing capacity of the pile as the Method 1 Method 2 Method 3
maximum load that needs to be mobilized to achieve the (kN) (kN) (kN) (kN) - (kN) (kN)
required settlement.
10361 18549 13621 10361 1.4 7401 10242
Analysis and design of the piled raft foundation are not
only limited to settlement analysis but also related to behavior
analysis of the raft. Therefore, the next goal that the authors
aim to develop is to propose the plate-on-spring approach. The
stiffness of these springs is calculated based on the settlement
analysis procedure above.
A three-dimensional model of the piled raft foundation has
been built independently on PLAXIS 3D software [7] for
verification. C.M. Brian B. Sheil and Bryan A. McCabe
(2012) [8] highlights the ability of embedded piles to predict
single pile and pile group responses with reasonable accuracy.
Smulders, S. Hosseini, and R.B.J. Brinkgreve (2019) [9]
showed that embedded beams are an efficient way to model
multiple piles for complex practical applications. In this paper,
the authors employ the embedded beam piles to create a model
for the piled raft foundation.
II. METHODOLOGY AND MODELING
A piled raft foundation is investigated as shown in Fig. 1.
The foundation consists of 8 x 9 piles with a distance between
the piles of 4.8 m x 4.5 m. The raft has dimensions BR x LR x
HR = 37 x 39.4 x 2.3 m. The bored pile has a diameter d = 1
Figure 2. Soil profile and soil parameters
m, and a design length l = 55 m. The soil profile includes
layers shown in Fig. 2. The loads from columns and an
elevator core are transmitted to the foundation as shown in A. Analytical procedure of piled raft foundation
Fig. 1. These total vertical loads are about 377310 kN. Start
Assumed load
sharing factor
Calculation of settlement of Calculation of raft settlement

the ith pile in the group, spi at each pile position, sRi
NO
spi sRi
YES
Final settlements: spi & sRi
Calculation of spring Calculation of spring

stiffness of the pile, Kp,i stiffness of subsoil, KR,i
Analysis of internal forces for

piled raft foundation
Figure 1. Foundation plan and pile configuration End
The pile bearing capacity is predicted according to TCVN Figure 3. The analytical procedure for piled raft foundation
10304:2014 [6] with three other methods: the analytical
method based on the soil index properties, the analytical The analytical procedure for the piled raft foundation with the
method based on the soil strength properties, and the settlement-based approach is shown in Fig. 3.
240
First, the load sharing factor is assumed (This factor is the of error to ensure that the stopping condition can be fulfilled.
ratio of the load carried by the raft divided by the total load This margin of error is not more than 10%-15%.
acting on the piled raft system). Its value is in the range of 10% Fourth, the calculation scheme of the foundation-subsoil
- 30% (Priyanka Bhartiya1, Tanusree Chakraborty and system is established as Fig. 6. Pile springs are estimated from
Dipanjan Basu, 2020) [10]. Second, the settlement of the ith the settlement of each pile as calculated above and the design
pile in the group is estimated according to the procedure in bearing capacity of the pile. The spring stiffness of the subsoil
Fig. 4, while the raft settlement at each pile position is is estimated through a variation of the Vesic formula. Modulus
calculated as shown in Fig. 5. Es is the modulus of soil layers within the compression zone
In this analysis of raft settlement, the raft is considered a of the foundation (the zone corresponding to the depth at
rigid rectangular area with a uniform vertical loading on semi- which the additional vertical stress is less than 10-20% of the
infinite mass. The principle of superposition is employed to primary stress). CSI SAFE software [12] is used to model this
estimate the stresses beneath the points under consideration. foundation-subsoil, thus determining pile forces, soil surface
A rectangular raft (BR = 2B1 & LR = 2L1) can be considered pressure, and internal force of the raft.
rigid if: (Poulos, 1991) [11]
r r is the element which has s
vo(i,r)  10Ds zi,r if E 5 Mpa
S Ri   Si , j
 B12 L1 Es 8 j 1 r is the element which has s vo(i,r)  5Ds zi,r if E  5 Mpa
 (1)
D 1  s2   L1 
1/ 2
U
  i , j 2 2 W
 B1  Si , j   Ds zi , j  h i , j i , j  1   0.8
E0i , j 1 
Where
ER H R3 Ds zi ,( j 1)  Ds zi , j
D (2) Ds zi , j 
12 1  
V
2 2 i
R
Ds zi , j  ki , j  pgl
X
G1 2 
 Gi  hi kn  2.82  3.87  2.18 2
ki , j  k W , X   k W  U , X   k W , X  V   k W  U , X  V 
h i where k  m, n   sign  mn  k  m , n 
sign  mn   1 when mn  0
k Gl k L sign  mn   1 mn  0
  0.17 ln n 1
'
a  0.17 ln( n )
' when
2G2 d d k  m , n  is the influence factor beneath the corner of
a rectangular mxn
' E p Ap Figure 5. A procedure for calculating settlement of the raft at

1 
 '
a ' different positions
 
2
G1 L
1 
3 i a j Slab element
2.12   4
  Ni 1 
s  Ni  
a
3 a
G1  l 1  2.12   4 j j
n Nj kv G1l kv G1l
S pi  s ( N i )    i , j  i , j  0.17 ln if 1 Subsoil
j 1 G1l 2G2 a 2G2 a
elastic supports
kv G1l KR,i = ks,R *AR,i
i, j  0 if 1 Pile elastic supports
2G2 a Rc , d
K pi 
Compression zone
S pi
In estimating pile settlement, the load Ni acting on each 0 
pile isFigure 4. Procedure
assumed to be thefor calculating
same. This canthe be
settlement of piles
acceptable due to 1 E r 1
the rigid raft condition where moments are neglected. Ni’s ks,R   s 2
B 1s
 i
1 
2
value is the total vertical load divided by the number of piles. 1 Es  i 1
2 
B    BR
r
i
E
3
G1 is the shear modulus of the soil layers around the pile shaft.
5 i 1 3
G2 is the shear modulus of the soil layers below the pile toe, i
the extent of which is shown in Figure 2, taken within 0.5l, i is the area of the stress graph of element i
from depth l to 1.5l from the top of the pile. The settlement of
pile is the sum between the settlement of a single pile s(Ni) Figure 6. Calculation scheme of the foundation-subsoil system
and the settlement due to the interference of piles si,j in the B. Plaxis 3D model for verification
group.
The model of piled raft foundation is set up in Plaxis 3D
Third, an iterative procedure is done with the loop stop for verification. Soil constitutive model is the Mohr-Coulomb
condition being the similarity between the settlement of ith model. The stiffness and strength parameters of the soil are
pile in the group (Spi) and the raft settlement at each pile defined in terms of effective stress properties. The raft is
position (SRi). Of course, there is not a perfect similarity modeled as a plate. The piles are modeled as embedded
between settlement Spi and SRi. There is an acceptable margin beams.
241
Table 2. Soil parameters of the plaxis 3d model 1D 5.2 5.3 1.9 6D 7.7 8.5 9.4
Paramete Layer Layer Layer Layer Layer Layer 1E 5.2 5.3 1.9 6E 7.7 8.5 9.4
rs 1a 1b 2 3 4 5
1F 4.9 5.1 3.9 6F 7.3 8.1 9.9
Soil Type CH CH CL SM CL SM
1G 4.4 4.6 4.3 6G 6.4 7.1 9.9
Soil
MC MC MC MC MC MC 1H 3.6 4 10.0 6G 5.3 5.4 1.9
Model
Material Draine Draine Draine Draine Draine Draine 2A 4.4 4.8 8.3 7A 4.9 5.1 3.9
Type d d d d d d 2B 5.4 6.1 11.5 7B 6.0 6.7 10.4
γunsat 2C 6.0 6.9 13.0 7C 6.8 7.7 11.7
15.10 16.60 19.60 21.00 20.28 21.30
(kN/m3)
2D 6.4 7.3 12.3 7D 7.2 8.1 11.1
γsat
15.28 16.80 19.79 21.28 20.44 21.54
(kN/m3) 2E 6.4 7.3 12.3 7E 7.2 8.1 11.1
c’ 2F 6.0 6.9 13.0 7F 6.8 7.7 11.7
7.20 10.50 31.00 8.00 42.48 9.92
(kN/m2)
2G 5.4 6.1 11.5 7G 6.0 6.7 10.4
φ’ (0) 3.48 4.73 14.18 32.62 15.06 33.68
2H 4.4 4.8 8.3 7H 4.9 5.1 3.9
Eoed
11230 11230 17210 33210 27500 47570 3A 4.9 5.1 3.9 8A 4.4 4.6 4.3
(kN/m2)
3B 6.0 6.7 10.4 8B 5.4 5.9 8.5
 0.35 0.35 0.35 0.30 0.35 0.35
3C 6.8 7.7 11.7 8C 6.0 6.7 10.4
Embedded pile bearing capacity is an input to the analysis
and not a result. The input of the bearing capacity parameters 3D 7.2 8.1 11.1 8D 6.4 7.0 8.6
is maximum skin resistance Tmax and base resistance, Fmax. 3E 7.2 8.1 11.1 8E 6.4 7.0 8.6
The skin resistance profile is defined in many different ways. 3F 6.8 7.7 11.7 8F 6.0 6.7 10.4
In this paper, the authors employ the uniform “linear”
resistance profile of skin. Since the piles lie in many soil 3G 6.0 6.7 10.4 8G 5.4 5.9 8.5
layers, each layer uses a different skin resistance profile. The 3H 4.9 5.1 3.9 8H 4.4 4.6 4.3
Ttop, max, Tbot, max and Fmax are derived from the single pile 4A 5.3 5.4 1.9 9A 3.6 4.0 10.0
bearing capacity by minimum value of three methods in Table
1. 4B 6.4 7.1 9.9 9B 4.4 4.6 4.3
Table 3. Embedded pile properties 4C 7.3 8.1 9.9 9C 4.9 5.1 3.9
4D 7.7 8.5 9.4 9D 5.2 5.3 1.9
Layer Layer Layer Layer Layer Layer
Layers
1a 1b 2 3 4 5 4E 7.7 8.5 9.4 9E 5.2 5.3 1.9
Ttop, max 4F 7.3 8.1 9.9 9F 4.9 5.1 3.9
0 8.64 277.68 475.2 360 1020
kN/m
4G 6.4 7.1 9.9 9G 4.4 4.6 4.3
Tbot, max
0 8.64 2778.68 475.2 360 1020
kN/m 4H 5.3 5.4 1.9 9H 3.6 4.0 10.0
F max 5A 5.4 5.5 1.8 5E 7.9 8.7 9.2
0 0 0 0 0 3636
kN 5B 6.6 7.3 9.6 5F 7.5 8.3 9.6
III. RESULTS 5C 7.5 8.3 9.6 5G 6.6 7.3 9.6

5D 7.9 8.7 9.2 5H 5.4 5.5 1.8
C. Settlement of piles and raft from the analytical method
The settlement of ith pile in the group and the raft Errors between the settlement of piles and rafts at different
settlement at each pile position from the analytical procedure points are within the allowable margin (15%). The settlement
proposed above are shown in Table 4. of rafts at different points tends to be larger than that of piles,
ranging from 1.8% to 13.0%, with an average of 8%. It is
The authors have built a subroutine in Python language to noted that an appropriate load sharing factor is to give equal
solve the iterative problem. Through the algorithm, the settlement of raft and piles at the same point, but that is
authors have found a load sharing factor of 0.21. This is a unlikely due to assumptions that simplify the pile-raft
suitable value so that the settlement of piles in the group and interaction. This factor is best for a slightly larger raft
the settlement of the raft at each pile position are considered settlement. In Fig. 7, the settlement shape of the raft and pile
to be the same. group is similar, largest in the middle and smaller at the edges.
Table 4. Settlement of piles and rafts from the analytical methods At positions near the edges, the settlement of the raft and piles
is almost the same. However, further into the middle of the
Settlement Settlement foundation, there is a separation between the settlement of the
Point Errors Point Errors
Pile Raft Pile Raft raft and the settlement of the pile. At these positions, the
cm cm % cm cm % settlement of the raft is much larger than that of the piles,
1A 3.6 4 10.0 6A 5.3 5.4 1.9 especially in the central positions. This may be due to the
compaction of the soil under the raft bottom due to the
1B 4.4 4.6 4.3 6B 6.4 7.1 9.9 presence of piles, especially in the middle areas of the raft
1C 4.9 5.1 3.9 6C 7.3 8.1 9.9 where the proposed method estimates the settlement of the raft
when considering the soil without any effect of piles.
242
Table 5. Pile spring stiffness and soil spring stiffness
Spring stiffness of piles
Point Kpi (kN/m) Point Kpi (kN/m)
1A, 9A, 1H, 9H 211457 3C, 7C, 3F, 7F 112136
1B, 9B, 1G, 9G 176214 3D, 7D, 3E, 7E 107261
1C, 9C,1F, 9F 154188 4A, 6A, 4H, 6H 148020
1D, 9D, 1E, 9E 148020 4B, 6B, 4G, 6G 119371
2A,8A, 2H, 8H 176214 4C,6C, 4F, 6F 105729

Point
1A 1B 1C 1D 1E 1F 1G 1H 2B,8B, 2G, 8G 145118 4D, 6D, 4E, 6E 98680
0,0
2C, 8C, 2F, 8F 127603 5A, 5H 145118
Settlement (cm)
2,0 Pile - AM
2D, 8D, 2E, 8E 121328 5B, 5G 117476
4,0 Raft - AM
3A, 7A, 3H, 7H 157468 5C, 5F 102792
6,0
3B, 7B, 3G, 7G 127603 5D, 5E 97382
8,0
Point Soil Subgrade ks,R (kN/m3) 1729
2A 2B 2C 2D 2E 2F 2G 2H
0,0 E. Verification for the settlement-based approach
Settlement (cm)
2,0 Pile - AM
4,0 Raft - AM
6,0
8,0
Point
3A 3B 3C 3D 3E 3F 3G 3H
0,0
Settlement (cm)
2,0 Pile - AM
4,0 Raft - AM
6,0
8,0
Point Point
4A 4B 4C 4D 4E 4F 4G 4H
0,0 1A 1B 1C 1D 1E 1F 1G 1H
0,0
Settlement (cm)
2,0 Pile - AM
1,0
4,0 Raft - AM Plaxis 3D
Settlement (cm)
2,0
6,0 3,0 Pile - AM
8,0 4,0 Raft - AM

5,0
Point 6,0
5A 5B 5C 5D 5E 5F 5G 5H
7,0
0,0
8,0
Settlement (cm)
2,0 Pile - AM
Points
4,0 Raft - AM 5A 5B 5C 5D 5E 5F 5G 5H
6,0 0,0
1,0
8,0 2,0 Plaxis 3D
Settlement (cm)
3,0
Pile - AM
4,0
Figure 7. Settlement of the raft and piles at different points Raft - AM
5,0
according to analytical methods
6,0
D. Pile spring stiffness and soil spring stiffness. 7,0
8,0
Based on the calculation scheme in Fig. 6, the results of 9,0
pile spring stiffness and soil spring stiffness are shown in
Table 5. Figure 8. Settlements from analytical methods and Plaxis 3D
model
243
Table 6. Errors in the settlement between the Plaxis 3D model and Hence, the difference between the settlement of the two
analytical method methods at different points may be due to the error of the
Errors Errors analytical method. However, the settlement shape of the
Plaxis Plaxis Plaxis 3D model tends to be larger on the elevator side, while
Point 3D- 3D- Point 3D- 3D-
3D 3D
Pile Raft Pile Raft on the opposite side is less. This is in contrast to the symmetric
cm % % cm % % settlement shape in the analytical method. Therefore,
1A 4.3 -16.3 -7 6A 7 -24.3 1.9 differences in the settlement shape of the two methods may be
1B 5.1 -13.7 -9.8 6B 7.6 -15.8 9.9 due to the assumption of the analytical method that the contact
1C 5.5 -10.9 -7.3 6C 8.1 -9.9 9.9
pressure at the raft bottom is uniform and the forces acting on
each pile are the same. While it is possible that the actual load
1D 5.7 -8.8 -7 6D 8.2 -6.1 9.4
distribution on the piles is not the same due to the contribution
1E 5.5 -5.5 -3.6 6E 7.8 -1.3 9.4
of the moments from the elevator.
1F 4.9 0 4.1 6F 7 4.3 9.9
1G 4.0 10 15 6G 6 6.7 9.9 F. Verification for the plate-on-spring approach
1H 3.0 20 33.3 6G 4.9 8.2 1.9
2A 5.2 -15.4 -7.7 7A 6.7 -26.9 3.9 Table 7. Pile reaction from CSI-SAFE model and plaxis 3D model
2B 5.9 -8.5 3.4 7B 7.2 -16.7 10.4 Pile reaction Pile reaction
2C 6.4 -6.3 7.8 7C 7.7 -11.7 11.7 Points Plaxis CSI- Points Plaxis CSI-
Error Error
2D 6.6 -3 10.6 7D 7.7 -6.5 11.1 3D SAFE 3D SAFE
2E 6.3 1.6 15.9 7E 7.4 -2.7 11.1 kN kN % kN kN %
2F 5.7 5.3 21.1 7F 6.7 1.5 11.7 1A 6123 4903 -19.9 6A 6578 5737 -12.8
2G 4.8 12.5 27.1 7G 5.7 5.3 10.4 1B 5267 5008 -4.9 6B 4170 5093 22.1
2H 3.8 15.8 26.3 7H 4.7 4.3 3.9 1C 5384 4926 -8.5 6C 4150 4847 16.8
3A 6 -18.3 -15 8A 6 -26.7 4.3
1D 5373 4899 -8.8 6D 4179 4575 9.5
3B 6.8 -11.8 -1.5 8B 6.6 -18.2 8.5
1E 5353 4757 -11.1 6E 3786 4298 13.5
3C 7.3 -6.8 5.5 8C 7 -14.3 10.4
1F 4767 4373 -8.3 6F 3205 4038 26
3D 7.4 -2.7 9.5 8D 7.1 -9.9 8.6
3E 7.1 1.4 14.1 8E 6.8 -5.9 8.6 1G 4002 4002 0 6G 2913 3746 28.6
3F 6.4 6.3 20.3 8F 6.1 -1.6 10.4 1H 4124 3514 -14.8 6H 3767 3613 -4.1
3G 5.4 11.1 24.1 8G 5.2 3.8 8.5 2A 5283 4882 -7.6 7A 6521 5875 -9.9
3H 4.4 11.4 15.9 8H 4.2 4.8 4.3 2B 3556 4748 33.5 7B 3888 5180 33.2
4A 6.7 -20.9 -19.4 9A 5.4 -33.3 10.0 2C 3539 4614 30.4 7C 3861 4839 25.3
4B 7.4 -13.5 -4.1 9B 5.9 -25.4 4.3
2D 3909 4515 15.5 7D 3765 4674 24.1
4C 7.9 -7.6 2.5 9C 6.3 -22.2 3.9
2E 3576 4349 21.6 7E 3512 4420 25.9
4D 8.1 -4.9 4.9 9D 6.4 -18.8 1.9
2F 3231 4067 25.9 7F 3291 4108 24.8
4E 7.7 0 10.4 9E 6.1 -14.8 1.9
4F 6.9 5.8 17.4 9F 5.5 -10.9 3.9 2G 2793 3797 35.9 7G 2922 3860 32.1
4G 5.9 8.5 20.3 9G 4.6 -4.3 4.3 2H 3510 3554 1.3 7H 3858 3777 -2.1
4H 4.8 10.4 12.5 9H 3.7 -2.7 10.0 3A 5545 5124 -7.6 8A 6656 6097 -8.4
5A 7.0 -22.9 -21.4 5E 8.0 -1.3 8.7 3B 3739 4787 28 8B 4106 5468 33.2
5B 7.7 -14.3 -5.2 5F 7.1 5.6 16.9 3C 3838 4605 20 8C 3872 5091 31.5
5C 8.2 -8.5 1.2 5G 6.1 8.2 19.7 3D 3733 4498 20.5 8D 3844 4873 26.8
5D 8.4 -6 3.6 5H 4.9 10.2 12.2
3E 3524 4298 22 8E 3659 4642 26.9
Despite the simplified assumptions in the proposed 3F 3163 3938 24.5 8F 3322 4344 30.8
method, there is a similarity in settlement shape between the 3G 2799 3704 32.3 8G 3182 4125 29.6
analytical results and those obtained from the Plaxis 3D 3H 3695 3558 -3.7 8H 3872 3961 2.3
model. As shown in Fig. 8, the settlement from the Plaxis 3D
model is approximately equal to the pile settlement from the 4A 6241 5375 -13.9 9A 7728 6782 -12.2
analytical method but slightly less than that of the raft if 4B 4176 4899 17.3 9B 6570 6143 -6.5
considering the same point. Of course, there are exceptions in 4C 4367 4732 8.4 9C 6502 5705 -12.3
some points, but that is the trend. If the errors of the analytical 4D 4314 4513 4.6 9D 6367 5516 -13.4
method are  15%, then the errors between the analytical 4E 4095 4268 4.2 9E 6184 5266 -14.8
results of the raft and piles compared with that of the Plaxis
3D model at most points are within this error margin (see 4F 3381 3978 17.7 9F 5689 4925 -13.4
Table 6). There are still a few points that exceed the 15% 4G 2917 3683 26.3 9G 4941 4637 -6.2
margin (15/72 points exceed the margin of error, most of them 4H 3856 3564 -7.6 9H 5502 4389 -20.2
are near the edges). The load sharing factor from the analytical 5A 6581 5601 -14.9 5E 4078 4346 6.6
method and the Plaxis 3D model are 0.21 and 0.24,
5B 4217 5054 19.8 5F 3382 3980 17.7
respectively. The difference in the load sharing coefficient
between the analytical method and the Plaxis 3D model is 5C 4367 4784 9.5 5G 2888 3723 28.9
14%, which is still within the error margin of the analytical. 5D 4420 4614 4.4 5H 3810 3555 -6.7
244
pile head is the same. The consequence shown in Figure 8 is a

difference in settlement shape between the 3D Plaxis model
and the analytical method.
Moment M11
a. Plaxis 3D b. CSI-SAFE
7000
CSI-SAFE Plaxis 3D
6000 Moment M22
Pile reaction kN
5000 a. Plaxis 3D b. CSI-SAFE

4000 Distance, m
3000 0,00 10,00 20,00 30,00 40,00
-2.000
2000
Moment, kNm/m
0
1000
0 2.000
1A 1B 1C 1D 1E 1F 1G 1H 4.000
Point PLAXIS 3D
6.000
CSI-SAFE
7000 8.000
CSI-SAFE Plaxis 3D
6000 Bending moment of the raft at axis 1 of the foundation
Pile reaction kN
5000 Distance, m
4000 0,00 10,00 20,00 30,00 40,00
-2.000
3000 0
Moment kNm/m
2000 2.000
1000 4.000
6.000
0
5A 5B 5C 5D 5E 5F 5G 5H 8.000 PLAXIS 3D
10.000 CSI-SAFE
Point
12.000
Figure 9. Pile reaction from the CSI-SAFE model and the
Plaxis 3D model Bending moment of the raft at axis 5 of the foundation
Table 7 shows the results of pile reaction from the CSI- Figure 10. Bending moment of the raft
SAFE model and the Plaxis 3D model. The force distribution
between the two models is nearly linear. For the piles near the Fig. 10 shows the bending moment values of the raft from
edge, the pile reactions from the Plaxis 3D model are larger the Plaxis 3D model and the CSI-SAFE model. The shape of
than those from the CSI-SAFE model, while the inside piles the moment diagram is the same in both models. The bending
tend to be opposite. There is a difference in the pile reaction moments of the raft at positions near the edge are
values between the two models. Error range is from 1.3% to approximately the same between the two CSI-SAFE and
35.9%, mean 19.5%. The load sharing factor of the CSI-SAFE Plaxis 3D models. However, in the middle of the raft, the
model is 0.21, less than that of the Plaxis 3D model. This magnitudes of the bending moments differ greatly between the
means that the total pile reactions in the CSI-SAFE model are two models. The maximum bending moment value of the
5.5% larger than that of the 3D Plaxis model (see more Plaxis 3D model is about 34.5% larger than that of the CSI-
comments in section C). As mentioned, the average error of SAFE model. The error of load sharing factor between the
the CSI-SAFE model compared with the 3D Plaxis model is Plaxis 3D model and the CSI-SAFE model is 14%, which
19.5%, which means it exceeds the 15% error margin of the means the raft of the PLAXIS 3D model has a larger load than
analytical method. This 5% difference may be due to a the raft of the CSI-SAFE model by 14%. This error value is
deviation in the initial assumption that the force acting on the calculated on the average of all positions. Based on Table 7,
245
the maximum error of pile reaction between the two Plaxis 3D With the error in the settlement analysis of  15% and the
and CSI-SAFE models is 35.9%. It conforms to the maximum error in force analysis of  30%, the proposed method has
error range for bending moment values between the two certain limitations. However, it will be appropriate in the early
models. This means that this difference belongs to the inherent stage of the design process. Although the article used a
error of the method. relatively complex load system, the foundation shape is still
simple rectangular. In future studies, the authors hope to be
IV. CONCLUSION AND DISCUSSION able to apply the proposed method to any shapes.
With the settlement-based approach, the interaction
between the raft and piles is reduced to the load sharing factor, REFERENCES
then the settlement of the piled raft system will be estimated
easily. The error of the method is in the range of  15%.
[1] J. Burland, "Piles as Settlement Reducers", 18th Italian
However, the proposed method is limited only when the raft
Congress, Pavia, Italy, 1995.
is considered rigid, which means ensuring equation (1). This
is a condition related to the assumption of contact pressure at [2] H. G. Poulos, "Piled raft foundation: design and applications",
the bottom of the raft and the distribution of the force acting Geotechnique, pp. 95-113, 2001.
on the piles. The settlement of the raft at positions in the [3] Choudhury, Rolf Katzenbach, ISSMGE Combined pile-raft
middle of the foundation tends to be larger than that of piles, foundation guideline, Darmstadt, Germany: Technische
so it is necessary to find a correction method to account for the Universität Darmstadt Institute and Laboratory of
compaction of the soil at these positions when taking into Geotechnics, 2013.
account the effect of piles. [4] Fellenius, "Basics of foundation design", Revised Electronic
Edition, www.Fellenius.net, 534p, British Columbia, 2021.
With the plate-on-spring approach, the paper provides
engineers with a simple tool to solve the soil-structure [5] Vietnamese Standard TCVN 9362:2012, "Specifications for
design of foundation for buildings and structures", Hanoi,
interaction. The error of the proposed method is  30% when 2012.
determining pile reactions and internal forces of the raft. In the
design, limiting the maximum pile reaction and the maximum [6] Vietnamese standard TCVN 10304:2014, "Pile Foundation –
Design Standard", Hanoi, 2014.
bending moment of the raft is very important. The proposed
method has a defect in determining a maximum margin in the [7] Bentley Systems, Incorporated, "PLAXIS 3D tutorial Manual
values of the pile reaction and bending moment of the raft. CONNECT edition V20", 2019.
However, if the model's error range is determined, engineers [8] C.M. Brian B. Sheil and Bryan A. McCabe, "Predictions of
can estimate the maximum difference appropriately before friction pile group response using embedded piles in
selecting more in-depth analysis methods. PLAXIS", in The 3rd IInternational Conference on New
Developments in Soil Mechanics and Geotechnical
The settlement-based approach and the plate-on-spring Engineering, Nicosia, Cyprus, 2012.
approach are mutually supportive. Through the settlement- [9] Smulders, S. Hosseini, R.B.J. Brinkgreve, "Improved
based approach, there is a physical meaning to the embedded beam with interaction surface", in Proceedings of
determination of spring stiffness. With the plate-on-spring the XVII ECSMGE-2019: Geotechnical Engineering
approach, it is better to support the determination of the pile foundation of the future, Icelandic Geotechnical Society
reaction and the internal force of the raft. There is no physical (IGS), 2019.
basis for using the plate-on-spring approach to determine the
[10] Priyanka Bhartiya1, Tanusree Chakraborty, Dipanjan Basu,
settlement of the piled-raft system without a settlement-based "Settlement Estimation of Piled Rafts for Initial Design",
approach. The connection of these two approaches is based on Journal of Geotechnical and Geoenvironmental Engineering,
reducing the pile-raft interaction to the load sharing factor. vol. 146, no. 2, p. 04019127, 2020.
Such a reduction faces the assumption of contact pressure at
[11] Poulos, Elastic solutions for soil and rock mechanics, Sydney:
the bottom of the raft and the distribution of loads from the
Center for geotechnical research, university of Sydney, 1991.
superstructure to the foundation. Of course, there will be a
redistribution of forces between the raft and the piles [12] Computer and Structures, Inc, "CSI analysis reference manual
depending on the stiffness of the raft and the stiffness of the for SAP2000, ETABS, SAFE", Computer and Structures, Inc,
1995 University Avenue Berkeley, California 94704 USA,
pile group. However, if the raft is rigid, the redistribution can
2020.
be neglected. That is the key assumption of the proposed
method.
246
Evaluating Performance of Petroleum Industry

Using Data Envelopment Analysis: A Case Study in
Vietnam
Kristofer Neal C. Imperial Chia-Nan Wang
National Kaohsiung University of Science and National Kaohsiung University of Science and
Technology Technology
Kaohsiung City, Taiwan Kaohsiung City, Taiwan
kristofer.neal@gmail.com cn.wang@nkust.edu.tw
Thanh-Tuan Dang Nguyen Ngoc Hiep

National Kaohsiung University of Science and National Kaohsiung University of Science and
Kaohsiung City, Taiwan Kaohsiung City, Taiwan
tuandang.ise@gmail.com ngocnguyen71193@gmail.com
Abstract: This study focuses on Vietnam’s petroleum use in vehicles and some countries use it to produce energy.
industry by evaluating its operational performance and Despite the high demand for petroleum, it has become a
analyzing the companies’ technological efficiency and commodity that even countries have a lot of sources of oil
technical efficiency. Among the 39 Petroleum companies in traded it on the world market. That created employment
Vietnam, 11 companies were chosen by the average amount opportunities in the petroleum industry. With the increase in
of assets within 4 years. Data on the industry were obtained the demand, oil firms made a lot strategy to sell their product
from Vietnam Stock Market. In analyzing the performance of to the consumers that they would offer expensive petroleum
the oil companies during the period of 2016–2019, the oil with high quality and vice versa with cheaper petroleum
Malmquist Productivity Index was used to determine that oil with low quality of the product.
achieved the highest performance in terms of the input and In Southeast Asia, Vietnam become known as an
output variables. By utilizing a variety of variables such as important oil and natural gas producer that boosted the
total assets, liabilities, owners’ equity, gross profit, and exploration activities and foreign firms start to invest and
revenue the authors managed to determine which companies cooperate in the oil and gas sector, which lead to introduced
that manage to attain their operational performance in the limited market reforms to support the energy industry. For the
span of 4 years. The results indicate that Petroleum Logistics past few decades, the country’s rapid economic growth,
and Investment (PLA) achieved the highest among other industrialization, and export market expansion have spurred
Petroleum companies followed by Thai Duong Petrol Joint domestic energy consumption. The neighboring countries of
Stock Company (TDG) and HFC Petroleum Corporation Vietnam, which includes the Philippines, Malaysia, China,
(HFC). Meanwhile, PetroVietnam Technical Services Taiwan, and Brunei claimed sovereignty over all or parts of
the Spratly Islands but China claims the islands are part of the
Corporation (PVS) and Phuong Dong Petroleum Tourism
South China Sea. Meanwhile, Vietnam called for an
JSC (PDC) have the lowest result amongst the companies
agreement with the neighboring countries to have a joint
during the studied period. This paper can be a beneficial exploration of oil and natural gas resources in the region. But
reference to oil firms for the policymakers, investors, with the unease of Vietnam towards China tensions were
development, and oil companies management. Applying the escalated with the territorial disputes that China violated that
DEA Malmquist Productivity Index model can also give a Vietnam began to accuse during 2016.
significant addition to measuring the organizational
performance of a company. The second-largest crude oil reserves in East Asia are
Vietnam with 4.4 billion barrels or 630 million tons. The
Keywords: Data Envelopment Analysis (DEA), largest oil-producing fields in the country were Hang Ngoc,
Malmquist Productivity Index (MPI), Petroleum Industry, Bach Ho, Dai Hung, Su Tu Den, and Rang Dong they
catch-up efficiency, frontier-shift, productivity index maintain an average volume of 340,000 barrels per day. In
2016, Vietnam consumes oil 470,000 barrels per day (B/d) but
I. INTRODUCTION due to the oil demand, the consumption of oil gradually rose
in 2019 by 528,000 barrels per day (B/d). Meanwhile, on the
Petroleum Industry plays a vital role in any country and it
side of the production of Oil in Vietnam managed to produce
is considered one of the largest contributors to the economy of
317,000 B/d in 2016. However, there was a decrease in
one country. It is a known fact that oil is widely known for its
production that the country only supplied 236,000 B/d in

2019. The oil companies in Vietnam with the highest total Lukoil, BP, and Turpas. Meanwhile, the enterprises of
assets were PetroVietnam Technical Services Corporation China’s petroleum technological innovation efficiency are
(PVS), Vietnam National Petroleum Group (PLX), and rising indicating the input and output efficiency improving.
Petrolimex Petrochemical Corporation – JSC (PLC). Of those Nonetheless, as time passes by the efficiency of technological
three mentioned, Vietnam National Petroleum Group has the innovation is not balanced and continuously growing which
highest revenue gathered in 4 years the best year was 2018 it suggests the oil companies must work hard to enhance their
accumulated up to 191,932,078 VND or equivalent to 8,347 technological innovation efficiency Wang et al. [15] for
USD. Despite these major companies in Vietnam, Petroleum instance, if the investment is too high in technology research
Logistic Service and Investment JSC (PLA) was leading in and development it would result in the output having
operational performance efficiency among other companies insufficient technological innovation. Consequently, the
using Data Envelopment Analysis and Malmquist companies will be affected by international oil prices and it
Productivity Index Model. cause unstable economic conditions. In Vietnam's economy,
the petroleum industry plays a critical role that attracts a large
At present, Petroleum companies have become a necessity quantity of manpower. Still, the companies are using an
to consumers, especially to automobile owners and immense number of high technology equipment. However,
companies. But in this day of age, Petroleum has become less most workers are still operating with manually controlled
used because most automobile companies are offering electric equipment. Health is one of the top concerns because of the
automobiles to protect the environment from air pollution or harsh and toxic working environment for this reason the
other greenhouse gases. That also includes some countries like requirements need to improve the working environment,
Norway that aim to become the first country to stop using laborer health and avert occupational diseases in the
fossil fuel-powered cars and switch to fully electric vehicles. petroleum industry [16].
This paper includes five parts. The 1st part is the Toshiyuki et al. proposed an approach to analyzing the
introduction. The 2nd part indicates some previous studies performance of the petroleum industry under private and
related to performance assessment. The 3rd part introduces public ownership. Having analyzed the performance, they
materials and methods regarding the theory of the Malmquist, identified two findings. One of the two findings was national
and the Pearson correlation coefficient. The 4th part presents oil firms under private ownership performed poorly while the
the results and discussion. The last section discusses the public ownership did better in terms of unified efficiency and
conclusion contribution and, shortcomings of the research.
unified efficiency under natural disposability. Despite the
II. LITERATURE REVIEW increasing trend in the unified efficiency of the performance
of international oil firms. While on the other hand, the second
Data Envelopment Analysis is a technique applied in finding was international oil corporation surpasses the
mathematical programming posed by Charnes et al. [9]. national oil firms in terms of unified efficiency under
Making such improvements in measuring the productivity of managerial disposability. As a result, international oil
a certain industry. Caves et al. introduced the Malmquist companies must consider the environmental performance
productivity index (MPI) as an effective way for productivity more vigorously than national oil companies at the same time
measurement in DEA. This also concludes that MPI is strengthening their operation sizes and providing new
favorable when it comes to calculating the productiveness of technology so they can meet the international standard on
an industry [10]. MPI determines the relative performance of CO2 emission [17]. Wang et al. [18] analyze the Public Sector
DMU at different duration of time using the technology of a Undertakings that India’s crude oil production did not exhibit
base period [11]. Eller et al. showed the empirical evidence any substantial increase for the last 10 years despite the
by correlating the National Oil Companies (NOCs) and growth of the refining capacity by 20% within 5 years.
private international oil companies (IOC) that NOCs are less Therefore, using DEA and Grey Model Theory for the
efficient than IOCs in the revenue efficiency of the evaluation of the performance of 14 public sector
companies [12]. In the late 1960s Organization of the undertakings resulted in 3 companies Gas Authority of India
Petroleum Exporting Countries (OPEC) was created and Limited (GAIL), Chennai Petroleum Corporation Limited
signed an agreement by Iraq, the Islamic Republic of Iran, (CPCL), and Oil India Limited (OIL) are the most influential
Kuwait, Saudi Arabia, and Venezuela. The purpose of OPEC amongst another public sector. Also, in the sector of
is to regulate and consolidate the petroleum policies that Agriculture and Forestry Malmquist Productivity Index and
provided safeguard with oil markets and to assure having an Grey Model were used by Wang et al. [19] to analyze the
efficient, economic, and regular supply of petroleum to the productivity performance of 10 agroforestry enterprises in
customer, substantial income to producers, and having a Vietnam that also includes assessing the current and
return of investment on those that invest in the petroleum predicting the future performance of each firm. Using the
industry [13]. Lu et al. evaluated and predicted the petroleum environmental indicator to conduct the analysis of eco-
companies in China that show the national petroleum efficiency by obtaining the sustainability reports of each firm
companies affiliated with OPEC performed poorly, while the in 2008 GLOBAL 500 that includes the data of the discharge
international petroleum companies operated better [13]. The of pollutants to the environment then DEA was performed to
petroleum companies in Poland and Turkey Öztürk et al. [14] achieve the overall technical efficiency. Ismail et al. [20]
conducted and compared the efficiency of both countries in conducted this study on the eco-efficiency and economic
terms of oil companies by obtaining the financial statements scores with the result revealing that technical efficiency and
and the official annual reports of the companies using MPI in eco-efficiency have a weak positive relationship. Halkos et
order to assess the technical efficiency changes. There are 2 al. [21] conducted a study with 42 countries on the relation
Polish national companies Orlen and Lotos have the greatest between oil consumption and economic efficiency in the
efficiency changes from 2005 to 2018 amongst other period of 1986 – 2006 and also used DEA window analysis
companies such as Shell, Exxonmobil, Chevron, Total, that revealed eco-efficiency and oil consumption have a
248
connection by using the generalized method of moments Step 2- Inputs and outputs collection: Total Assets,
(GMM) shows an inverted U-shape relationship which Liabilities, and Owners' equity are taken as inputs and output
resulted in countries’ economic efficiency grew higher with includes Gross profit and Revenue. The input and output
inflating of oil consumption that indicates the consumption variables selected are very important because they will affect
of oil is the main driver of urbanization and industrialization. the outcome.
In the rankings of Energy Intelligence’s Petroleum
Intelligence Weekly from 2000 to 2010, data was gathered Step 3- Pearson Correlation method: Correlation shows
and it was divided into four classes based on ownership. the interrelationship between all the variables and their impact
Ohene-Asere et al. [22] defined the 4 classes as state on each other. This proves that the relationship between the
multinationals (MNOC), private multinationals (MIOC), input and output variables is isotonic.
private locals (LIOC), and location–state locals (LNOC) in Step 4- DEA Models: The DEA models are extensively
which evaluating the relative efficiency of using DEA with a used to estimate the efficiency and performance of the DMUs.
bootstrap approach with the outcome of private firms (IOCs) The Malmquist DEA is used to test the productivity change
are a lot more efficient that state firms (NOCs) with regards over time and can discuss individual changes in efficiency and
in technical and SE at 1 percent significance level. Ike et al. technology.
[23] assessed the productivity and efficiency of the world’s
largest national oil companies (NOC) and international oil Step 5- Results, discussion: The results of the catch-up
companies (IOC) in the period of 2003–2010 by using a index, the frontier-shift, and the Malmquist index of the
random-effects regression model for the second stage DMUs are measured, discussed.
analysis in relation to the environmental factors that have an Step 6- Conclusion: The last step is to analyze the results
impact on efficiency level and productivity that led to a result from the findings and then conclude.
the Organization of the Petroleum Exporting Countries
(OPEC) that big IOCs outperforms than NOCs. In Saudi
Arabia’s Petrochemical Industry, Alidrisi et al. [24] used the
Data Envelopment Analysis Multidimensional Scaling
(DEA-MDS) hybrid approach to be able to identify the
performance with unbiased positioning between firms that
also gives importance to regularly monitoring the
performance and efficiency of each sector in the companies
that would lead to success to the economy of the country. In
2008, Mehmet Ertürk at el. [25] analyzed 38 Turkish natural
gas distribution using the non-parametric method DEA in
detecting criteria that affect the efficiency of the companies
and also to find probable characteristics of the inefficiency of
firms in which two common characteristics were found. First,
low scales of operation in inefficient companies, and
secondly, not utilizing the network effectively in some
immature companies would result that public firms showing
more efficiency than private firms.
A review of these studies provides evidence that the
Malmquist Productivity Index (MPI) is an excellent
instrument for calculating the productivity changes of
industries, countries, and organizations by taking into
account a comprehensive assessment both of the internal as
well as the external performance aspects. This Malmquist
model is a suitable method of study to carry out the most Figure 1. Procedure of Research
thorough assessment of the operational efficiency of oil
companies. For this purpose, to analyze the data, the authors B. Malmquist Productivity Index (MPI)
decided to conduct the Malmquist Productivity Index model. The MPI values measure variations in total factor
productivity of DMUs in two periods, and they are correlated
III. MATERIALS AND METHODS with changes in technical efficiency (catch-up index) and
A. Research Process technological change (frontier-shift index). Any change in
technical efficiency can be related to the DMUs ability to
The authors analyzed the top 11 petroleum companies in improve or deteriorate efficiency. In contrast, technological
Vietnam from 2016 to 2019, by applying the Malmquist change indicates any progress in technology development and
productivity index model in the DEA method. The research innovation of DMUs between periods 1 to 2 [27].
process is presented in Fig. 1 as follows
The DMUi was identified by the authors as (𝑎1𝑖 , 𝑏𝑖1 ) for
Step 1- DMUs collection: The information about
Vietnam's petrol companies and collecting their relevant period 1 and (𝑎𝑖2 , 𝑏𝑖2 ) for period 2. The efficiency score of the
information. There are 11 Vietnamese petrol companies DMUi(𝑎1𝑖 , 𝑏𝑖1 )𝑡1 is measured by the technological frontier 𝑡2 :
selected in this study. The data sources were collected on the 𝑑 𝑡2 ((𝑎𝑖 , 𝑏𝑖 )𝑡1 ); 𝑡1 = 1, 2 and 𝑡2 = 1, 2).
stock market (Vietstock.vn) from 2016 to 2019.
249
To compute the catch-up index (C), frontier-shift index Table 1. List of 11 DMU oil companies in Vietnam
(F), and Malmquist Index (MI), the following formulas can be DMU List of Companies
applied:
DMU1 PVS PetroVietnam Technica Services Corporation
𝑑 2 ((𝑎𝑖 , 𝑏𝑖 )2 )
𝐶= 1 (1) DMU2 TDG ThaiDuong Petrol Join Stock Company
𝑑 ((𝑎𝑖 , 𝑏𝑖 )1 )
DMU3 PLX Viet Nam National Petroleum Group
1
1 1 1 2
𝑑 ((𝑎𝑖 , 𝑏𝑖 ) ) 𝑑 ((𝑎𝑖 , 𝑏𝑖 ) ) 2
(2) DMU4 PDC Phuong Dong Petroleum Tourism JSC
𝐹=[ × ]
𝑑 2 ((𝑎𝑖 , 𝑏𝑖 )1 ) 𝑑 2 ((𝑎𝑖 , 𝑏𝑖 )2 ) DMU5 HFC HFC Petroleum Corporation
2 2
𝑑 ((𝑎𝑖 , 𝑏𝑖 ) ) DMU6 COM Materials – Petroleum Joint Stock Company
𝑀𝐼 = 𝐶 𝑥 𝐹 =
𝑑1 ((𝑎𝑖 , 𝑏𝑖 )1 ) DMU7 PEQ Petrolimex Equipment Joint Stock Company
𝑑1 ((𝑎𝑖 , 𝑏𝑖 )1 )
×[ 2 (3) DMU8 POB
Thai Binh PetroVietNam Oil Joint Stock
𝑑 ((𝑎𝑖 , 𝑏𝑖 )1 ) Company
1
DMU9 PMS Petroleum Mechanical Stock Company
𝑑1 ((𝑎𝑖 , 𝑏𝑖 )2 ) 2
× 2 ] Petrolimex Petrochemical Corporation – JSC
𝑑 ((𝑎𝑖 , 𝑏𝑖 )2 ) DMU10 PLC
1 DMU11 PLA Petroleum Logistic Service and Investment JSC

𝑑1 ((𝑎𝑖 , 𝑏𝑖 )2 ) 𝑑 2 ((𝑎𝑖 , 𝑏𝑖 )2 ) 2 (4)
𝑀𝐼 = [ 1 × ]
𝑑 ((𝑎𝑖 , 𝑏𝑖 )1 ) 𝑑 2 ((𝑎𝑖 , 𝑏𝑖 )1 ) 2. Choosing the Input and Output Factors
According to the above equations, the total factor This study aims to estimate the performance to improve
productivity (TFP) of the DMUs represents increases or the business of decision-making strategies. Selecting inputs
decreases in their technical and technological innovation and outputs is not only important but also must be consistent
efficiency.C, F, and MI values that show progress, stability, or with the DEA resolution program. The findings are
regress from period 1 to 2 are higher, lower, or above 1, significantly influenced by the inputs and outputs. The authors
respectively, when the value C, F, or MI is greater, lower, or therefore opted to select three inputs and two outputs:
greater than 1, respectively. Inputs:
C. Pearson Correlation Coefficient 1. Total Asset (TA): all the assets or items of value a business
Throughout many previous researches, the Pearson owns (as cash, inventory, equipment, tools, etc.)
correlation is typically utilized. It describes the linear
2. Liabilities (LI): reflects the total value of amounts owed 's
dependency of two variables or sets of data and has a value
a long-term business.
between -1 and +1, with +1 denoting a total positive linear
correlation, 0 denoting no linear correlation, and -1 denoting 3. Owners’ equity (OE): Equity is the capital resources
a total negative linear correlation [28]. owned by the business owner
The following computation represents the Pearson’s (r) Outputs:
correlation coefficient equation for the two variables (x) and
(y): 1. Gross profit (GP): the amount of profit a company earns
after subtracting costs related to the production and sale
∑𝑛𝑖=1(𝑥𝑖 − 𝑥)(𝑦𝑖 − 𝑦) of products, or costs related to the provision of a services
𝑟𝑥𝑦 = (5) company.
√∑𝑛𝑖=1(𝑥𝑖 − 𝑥)2 ∑𝑛𝑖=1(𝑦𝑖 − 𝑦)2
2. Revenue (RE): It is the amount of money the business
where n is the sample size; xi, yi represents the individual receives from its customers before operating expenses
1 and taxes are deducted.
sample points indicated with i; Moreover, 𝑥̅ = ∑𝑛1 𝑥𝑖 is the
𝑛
mean of the sample, which is similar for 𝑦̅. There are above five financial indicators important for the
activity business with a company. The inputs show the capital
Input and output data must be positively correlated since
situation and operate in that period of the enterprise. Along
homogeneity and isotonicity constitute mainly two DEA data
with the values of the input factors, the output factors also
parameters.
need to be enhanced considering it reflects the outcome of the
IV. RESULTS AND DISCUSSION business production and operational processes. The inputs and
outputs must be an isotonic condition. That is the rationale for
D. Data Collection the authors’ selection of research variables.
1. DMU Selection E. Catch-Up Index (Technical Efficiency)
After researching the situation of the oil and gas industry Improvements in technical, technological, and total factor
in Vietnam, oil and gas are one of the parts industries directly productivity are among the components of the Malmquist
affected by development economically. Considering 39 productivity index that are utilized in performance
companies on the stock market, the authors have selected 11 measurement. Findings of efficiency change are presented by
major companies to conduct research. In this paper, below is the authors. The catch-up index depicted in Table 2 and Fig 2
the data of 11 of Vietnam's top petroleum companies (Table serves as an expression of the technical effective
1) [29-32]. modifications of the DMUs.
250
DMUs' technical efficacy tended to rise in 2018–2019 advancement and innovation, petroleum businesses in
after declining in the years 2017–2018. The majority (7 Vietnam exhibit a common tendency, rising in the second
DMUS: DMU2, 3, 5, 6, 7, 9, 11) attained technical efficiency phase and falling in the remaining two periods.
in the period (2016–2019), as shown by the average catch-up
= 1.028759 (>1). On average, DMU4 (PDC) had the lowest G. Malmquist Index (Total Productivity Change)
and least-stable efficiency performance, while DMU11 (PLA) In this paper, the DEA-Malmquist model is used to
had the highest and most steady technical performance. estimate the efficiency performance of the top 11 petroleum
companies in Vietnam.
7 of the 11 DMUs obtained technical efficiency with a
catch-up index greater than 1 during the 2016-2017 period. Details on the Malmquist productivity index of the DMUs
DMU11 (PLA) which had a value of 2.724938, had the through 2016 to 2019 are provided in Table 4 and Fig. 4
highest technical efficiency of any DMU. On the other hand, Overall, it is obvious that the MPI of most DMUs tended to
DMU8 (POB) had the lowest score of 0.522344. The DMUs increase from 2016–2017 to 2017–2018 and decreased in the
did accomplish progressive technological efficiency in 2017 – 2018–2019 period.
2018 (DMU 3, 5, 7, and 8, with catch-up scores over 1). DMU On average, DMU11 has achieved the best performance in
4 (PDC), which received a score of 0.465134, was the least terms of technical and technological efficiency during the
effective oil company during this time. While DMU3 (PLX) period. The DMU1 is the lowest of 11 DMUs. The average
experienced a remarkable increase in technical efficiency, MPI of all DMUs is 1.0447638 from 2016 to 2019. While the
rising to the position of the period’s most technically efficient maximum average MPI is 1.630 (DMU11), the minimum
producer with a score of 1.235678. average MPI is 0.797, the difference is 0.8326.
The petroleum businesses significantly increased their In 2016–2017, DMU1-PetroVietnam, DMU2-ThaiDuong
technical efficiency in the following period, 2018–2019, after Petrol, DMU5- HFC Petroleum, DMU6-Materials -
their poor performance in the previous period. Only three out Petroleum, DMU11-Petroleum Logistic achieved progress in
of eleven organizations (DMU1, 3, and 8) have catch-up total factor productivity. DMU11-Petroleum Logistic is the
values below one, according to the findings. best score with 2.5536, while DMU4-Phuong Dong
F. Frontier-Shift Index (Technological Change) Petroleum obtained a 0.55489 productivity score. The average
The efficiency frontiers of DMUs throughout two periods MPI of all DMUs is 1.1286 in this period.
are measured using the frontier-shift index. Table 3 In 2017–2018, most of the DMUs performed more
demonstrates that the technological efficiency of petroleum efficiently with the MPI gather than 1, namely DMU1-
businesses improved in the year 2017–2018 and dropped in PetroVietnam, DMU3-Viet Nam National Petroleum, DMU5-
the year 2018–2019. HFC Petroleum, DMU7-Petrolimex Equipment, DMU8-Thai
The remainder manufacturers (8 out of 11) struggled to Binh PetroVietNam, DMU11-Petroleum Logistic achieved a
make technological advancements in the first quarter of 2016– performance during in the period. During 2017-2018
2017, with the exception of DMU2, DMU5, and DMU8. compared with 2016-2017 DMU8 rapidly increased from
However, the petroleum businesses worked hard to innovate 0.7710 to 1.269, while DMU11 considerably decreased from
2.5536 to 1.04866. The average MPI of all DMUs is 1.0172.
technology in the following period (2017–2018) and saw
success; all of their frontier-shift indicators were higher than Only five DMUs (DMU2, DMU4, DMU9, DMU10, and
1. They were unable to continue making gains during the DMU11) improved overall productivity throughout the 2018–
following period (2018–2019), as all of their frontier-shift 2019 period, with six DMUs performing inefficiently and
indicators fell. having MPIs below 1. DMU1-PetroVietnam enormously go
It can be noted in Fig. 3 the petroleum businesses did not down from 1.1169 to 0.24225 is one of the DMUs with the
accomplish technological advancement over the period of lowest score in this period. On the other hand, DMU2, DMU4,
2016–2017. Only three of the top petroleum businesses, DMU9, and DMU10 improved scores higher than 1. The
DMU2 (TDG), DMU5 (HFC), and DMU8 (POB), earned a average MPI of all DMUs declined to 0.98845.
frontier-shift index that was larger than 1, indicating that these The last column in Table 4 lists the MPI change score of
companies' technological and innovative advancements have DMUs in the whole period (2016–2019). The values range
not improved significantly and still have many restrictions. from 0.748 to 1.630. The year-on-year productivity changes
Manufacturers chose to engage in technological innovation from 2016 to 2019 are found quite stable. Among these
following a period of subpar technological performance from improved companies, DMU9 improved the most. On the
2016 to 2017, which led to the achievement of technological overall average, the number of improved companies decrease
efficiency in the years 2017 to 2018, particularly for DMU1, to 6 from 4. After this period, DMU1 had a big recession while
DMU4, and DMU10. Due to the rapid pace of technological others improved (from 1.033 to 0.242 productivity).
advancement, manufacturers were unable to retain their gains
Table 2. Results of Catch-up index
and even experienced a significant decline in the years that
followed. Catch-up 2016=>2017 2017=>2018 2018=>2019 Average
The top petroleum companies' frontier-shift scores for the DMU1 1.238298 0.814380 0.238776 0.763818
years 2018–2019 are displayed in Fig. 3. DMU6 was the DMU2 1.286354 0.704701 1.214168 1.068408
manufacturer with the lowest efficiency throughout this time
(0.916963). DMU1, 2, 3, 8, and 9 are often the only DMU3 0.961230 1.235678 0.877957 1.024955
manufacturers with excellent performance (>1). Since no DMU4 0.577361 0.465134 1.120626 0.721040
company demonstrated consistent technological efficiency or
DMU5 1.134859 1.002153 1.003127 1.046713
assumed any leadership positions in the race for technological
251
Catch-up 2016=>2017 2017=>2018 2018=>2019 Average

DMU6 1.629596 0.658560 1.052756 1.113637
DMU7 0.963714 1.146300 1.006007 1.038674
DMU8 0.522344 1.229080 0.894065 0.881830
DMU9 1.013691 0.825076 1.255864 1.031543
DMU10 1.008699 0.785138 1.096304 0.963380
DMU11 2.724938 0.940575 1.321525 1.662346
Average 1.187371 0.891525 1.007380 1.028759
Max 2.724938 1.235678 1.321525 1.662346
Min 0.522344 0.465134 0.238776 0.721040
SD 0.596024 0.245888 0.290920 0.245234
Figure 2. Each DMU’s technical efficiency changes
Table 3. Results of Frontier-Shift index
Frontier 2016=>2017 2017=>2018 2018=>2019 Average

DMU1 0.834337 1.371571 1.014588 1.073499
DMU2 1.007376 1.128522 1.012656 1.049518
DMU3 0.969671 1.023190 1.019249 1.004037
DMU4 0.961086 1.320877 0.960985 1.080983
DMU5 1.016231 1.152357 0.974257 1.047615
DMU6 0.893741 1.187575 0.916963 0.999426
DMU7 0.889845 1.082417 0.965549 0.979270
DMU8 1.476075 1.032857 1.062930 1.190620
DMU9 0.914814 1.104700 1.010182 1.009899
DMU10 0.872066 1.262207 0.919857 1.018044
Figure 3. Each DMU’s Technological changes
DMU11 0.937148 1.114924 0.974664 1.008912
Average 0.979308 1.161927 0.984716 1.041984
Max 1.476075 1.371571 1.062930 1.190620
Min 0.834337 1.023190 0.916963 0.979270
SD 0.174107 0.113512 0.044238 0.058801
Table 4. Results of Malmquist Productivity index
MPI 2016=>2017 2017=>2018 2018=>2019 Average

DMU1 1.033158 1.116979 0.242259 0.797465
DMU2 1.295843 0.795271 1.229535 1.106883
DMU3 0.932077 1.264334 0.894856 1.030422
DMU4 0.554894 0.614385 1.076905 0.748728

DMU5 1.153279 1.154838 0.977303 1.095140 Figure 4. Total factor productivity change
DMU6 1.456437 0.782089 0.965338 1.067955 V. CONCLUSION

DMU7 0.857556 1.240775 0.971350 1.023227 In this study, the DEA model (i.e. Malmquist model) was
DMU8 0.771019 1.269464 0.950328 0.996937
applied to assess the performance of the top 11 petroleum
companies in Vietnam for the period from 2016 to 2019. In
DMU9 0.927339 0.911462 1.268651 1.035817 this model, based on the main financial indicators, the inputs
DMU10 0.879652 0.991007 1.008443 0.959701
are Asset, Liabilities, and Owners' equity, while Gross profit
and Revenue are selected as output variables. The model result
DMU11 2.553670 1.048670 1.288044 1.630128 shows that the efficiency scores of banks tended to fluctuate
Average 1.128629 1.017207 0.988456 1.044764 sharply over the considered periods. All of the companies
achieved improvements in this research period. DMU11-
Max 2.553670 1.269464 1.288044 1.630128 Petroleum Logistics had the best performance throughout all
Min 0.554894 0.614385 0.242259 0.748728 periods. This company also had a surprising improvement that
achieved a productivity score in this period. DMU1-
SD 0.533625 0.219828 0.283298 0.225533 PetroVietnam had the most unstable performance. The
252
DMU11 also failed to maintain its productivity indexes network model, Sustainable Computing: Informatics and
achieved during the period. Systems, Vol. 27, 2020, 100397, ISSN 2210-5379,
[14] Öztürk, M., Trojak, M. and Gancarczyk, M. (n.d.) “Efficiency
The two DMUs with the highest technological change comparison of petroleum companies by using Data
indices are DMU2-ThaiDuong Petrol and DMU9-Petroleum Envelopment Analysis.”
Mechanical, and only these two DMUs average MPIs higher [15] Wang, Y., Zhu, Z. & Liu, Z. Evaluation of technological
than 1. This suggests that for the analyzed period, these DMUs innovation efficiency of petroleum companies based on BCC–
Malmquist index model. J Petrol Explor Prod
are the best and most reliable in the petroleum businesses' Technol 9, 2405–2416 (2019).
industry. [16] Linh, H. X., & Hanh, V. V. (2002, January 1). Present Working
The contribution of the paper is to provide an evaluation Environment Conditions and Measures of Improvement in
Vietnam Petroleum Industry. Society of Petroleum Engineers.
and comparison of the business of eleven companies. For this doi:10.2118/73953-MS
industry to succeed, the study findings contribute to increased [17] Toshiyuki Sueyoshi, Mika Goto, Data envelopment analysis
corporate productivity, economic growth, and technological for environmental assessment: Comparison between public and
innovation breakthroughs. The results are anticipated to private ownership in petroleum industry, European Journal of
provide useful information and reflect the current state of the Operational Research, Vol. 216, No. 3, 2012, pp. 668-678
oil and gas business, according to the authors. This paper is to [18] Analyzing PSU’s performance: A case from ministry of
support the decision-makers to be aware of their petroleum and natural gas of India CN Wang, LC Lin, D
Murugesan Mathematical Problems in Engineering 2013
organization’s position to apprehend overall efficiency and
productivity. This method provides an intensive guideline in [19] Applying data envelopment analysis and grey model for the
productivity evaluation of Vietnamese agroforestry industry
the decision-making process with other industries. CN Wang, HS Lin, HP Hsu, VT Le, TF Lin Sustainability 8
(11), 1139
However, the limitation of this paper, this study only
considers some specific input and output variables, future [20] Zubaidah Ismail, Jie Chuin Tai, Keen Kuan Kong, Kim Hing
Law, Sharif Moniruzzaman Shirazi, Ramlee Karim, using data
studies should consider more input and output variables (cost, envelopment analysis in comparing the environmental
investment, total capital, etc.) to give a higher resolution on performance and technical efficiency of selected companies in
the performance scores. Due to the lack of annual reports, their global petroleum operations, Measurement, Vol. 46, No.
therefore, the overview of companies didn't show detail for the 9, 2013, pp. 3401-3413,
research in the future should give a more detailed overall [21] Halkos, G.E. and Tzeremes, N.G. (2011), “Oil consumption
view. Comparing the results to those obtained from other and economic efficiency: a comparative analysis of advanced,
developing and emerging economies”, Ecological Economics,
approaches can be performed in future research. Vol. 70, No. 7, pp. 1354-1362.
[22] Ohene-Asare, K., Turkson, C. and Afful-Dadzie, A. (2017),
REFERENCES "Multinational operation, ownership and efficiency differences
[1] "EIA – Electricity Data". www.eia.gov. Archived from the in the international oil industry", Energy Economics, Vol. 68,
original on July 10, 2017. Retrieved April 18, 2017. October, pp. 303-312.
[2] "Top Oil Producers and Consumers". InfoPlease. Archived [23] Ike, C.B. and Lee, H. (2014), “Measurement of the efficiency
from the original on April 25, 2017. Retrieved March 18, 2018. and productivity of national oil companies and its
[3] "US soon to leapfrog Saudis, Russia as top oil determinants”, Geosystem Engineering, Vol. 17, No. 1, pp. 1-
producer". www.abqjournal.com. The Associated Press. 10.
Retrieved October 6, 2018. [24] Alidrisi H, Aydin ME, Bafail AO, Abdulal R, Karuvatt SA.
[4] "The United States is now the largest global crude oil Monitoring the Performance of Petrochemical Organizations in
producer". www.eia.gov. Today in Energy – U.S. Energy Saudi Arabia Using Data Envelopment Analysis. Mathematics.
Information Administration (EIA). Retrieved October 6, 2018. 2019; 7(6):519.
[5] "Canada's oil sands survive, but can't thrive in a $50 oil [25] Mehmet Ertürk, Serap Türüt-Aşık, Efficiency analysis of
world". Reuters. October 18, 2017. Retrieved April 5, 2020. Turkish natural gas distribution companies by using data
envelopment analysis method, Energy Policy, Vol. 39, No. 3,
[6] "Crude Oil Forecast | Canadian Association of Petroleum 2011, pp. 1426-1438.
Producers". CAPP. Retrieved April 5, 2020.
[26] Wang, C. N., Nguyen, N. A. T., Dang, T. T., & Trinh, T. T. Q.
[7] Vietstock. [Online] available athttps://vietstock.vn[Accessed (2021). A decision support model for measuring technological
2020]. progress and productivity growth: The case of commercial
[8] "IHS Markit: Canadian oil sands production to be ~1M barrels banks in Vietnam. Axioms, 10(3), 131.
higher by 2030 but with lower annual growth; boosted by [27] Wang, C. N., Le, T. Q., Yu, C. H., Ling, H. C., & Dang, T. T.
deterioration in Venezuela". Green Car Congress. (2022). Strategic Environmental Assessment of Land
Retrieved April 5, 2020. Transportation: An Application of DEA with Undesirable
[9] Charnes, A.; Cooper, W.W.; Rhodes, E. Measuring the Output Approach. Sustainability, 14(2), 972.
efficiency of decision making units. Eur. J. Oper. Res 1978, 2, [28] EIA. (2017). Analysis - Energy Sector Highlights [Online].
pp. 429–444. [CrossRef] Available:https://www.eia.gov/international/analysis/country/
[10] D.W. Caves, L.R. Christensen, W.E. Diewert The economic VNM
theory of index numbers and the measurement of input, output, [29] AUSTRADE. (2017). Oil and gas to Vietnam-Trends and
and productivity Econometrica, 50(6) (1982), pp. 1393-1414 opportunities [Online]. Available:
[11] R. Färe, S. Grosskopf, B. Lindgren, P. Roos Productivity https://www.austrade.gov.au/australian/export/export-
change in Swedish pharmacies 1980–1989: a nonparametric markets/countries/vietnam/industries/oil-and-
Malmquist approach Journal of Productivity Analysis, 3 gas#:~:text=The%20majority%20of%20Vietnam's%20oil,to
(1992), pp. 85-102 %20run%20ahead%20of%20production.
[12] S.L. Eller, P.R. Hartley, K.B. Medlock Empirical evidence on [30] Oil & Gas Vietnam. (2021). Industry Facts [Online].
the operational efficiency of national oil companies Empir. Available: http://oilgasvietnam.com/industry-fact.html
Econ., 40(3) (2011), pp. 623-643 [31] BP. (2022). Statistical Review of World Energy [Online].
[13] Li Lu, Jijun Zhang, Fan Yang, Yongjian Zhang, Evaluation and Availble: https://www.bp.com/en/global/corporate/energy-
prediction on total factor productivity of Chinese petroleum economics/statistical-review-of-world-energy.
companies via three-stage DEA model and time series neural
253
Fire Resistance Properties and Geopolymer Coating

Van Su Le* Van Vu Nguyen Artem Sharko
Department of Material Science Department of Material Science Department of Material Science
Technical University of Liberec Technical University of Liberec Technical University of Liberec
Liberec, Czech Republic Liberec, Czech Republic Liberec, Czech Republic
su.le.van@tul.cz nguyen.van.vu@tul.cz artem.sharko@tul.cz
Doan Hung Tran Petr Louda Piotr Łoś

Department of Material Science Department of Material Science Department of Material Science
Nha Trang University Technical University of Liberec Technical University of Liberec
Nha Trang, Vietnam Liberec, Czech Republic Liberec, Czech Republic
doanhung.tran@ntu.vn petr.louda@tul.cz Piotr.los@tul.cz
Thang Xiem Nguyen Stanisław Mitura Katarzyna Buczkowska

Department of Material Science Medical Faculty Department of Material Science
Faculty of Civil Engineering President Stanisław Wojciechowski Faculty of Mechanical Engineering
Nha Trang University State University of Applied Sciences Technical University of Liberec
Nha Trang, Vietnam in Kalisz, Liberec, Czech Republic
thangxiem.nguyen@ntu.vn Kalisz, Poland katarzyna.buczkowska@tul.cz
mitura.stanislaw@kut.pl
Abstract: This paper describes the procedure of applying inorganic crystals occur in geopolymers. Zeolites consist of a
porous geopolymer to a steel base using a spraying method. (Si - O - Al) polymer framework. Unlike zeolite,
The main goal is to protect materials from environmental geopolymers are amorphous and not crystalline. The reaction
influences. Researchers found that one of the key properties results in the formation of tetrahedral SiO4 and AlO4
of the geopolymer used is its foaming porosity. This frameworks, which are linked by oxygen atoms. Tetrahedral
contributes to the physical and mechanical properties of the frameworks are connected by extensive covalent bonds. This
material, such as sound absorption, fire resistance, and high means that the geopolymer structure appears as a dense
temperature resistance. A patent-protected application of amorphous phase composed of a semi-crystalline three-
geopolymer coatings to a surface of any shape with any dimensional aluminosilicate microstructure [2].
configuration is described. Geopolymers can be applied by
Compared to conventional Portland cement, geopolymers
spraying onto different surfaces and at different angles. The
have better mechanical properties, low permeability,
advantage of this technology is that it can be applied to any
chemical resistance to aggressive media, and long-term fire
smooth or rough surface. Geopolymer composite is an
resistance. Likewise, geopolymers have a high indicator of
environmentally friendly alternative to concrete. The authors
environmental sustainability due to their low CO2 emissions
present a series of empirical examples that complement their
during production, as well as their low cost and durability [3].
previous work investigating the physical and mechanical
Researchers are interested in investigating the properties of
properties and applications of foamed geopolymer
geopolymers due to their practical application for heat
composites. After being coated with a foamed geopolymer
absorption, sound absorption [4], and fire protection [5].
layer, carbon steel plates' fire resistance has been
Geopolymer composites are used in membranes and
substantially improved. The fire-resistance times for carbon
membrane support in construction equipment as catalysts,
steel plates with a 6 mm coating thickness are around 80
adsorbents, and fillers [6], [7].
minutes. A foamed geopolymer with a suitable coating has
great potential for surface coating. Research has focused on improving the above-mentioned
properties and manufacturing processes of geopolymer
Keywords: spraying, metal, technologies, geopolymer composites. One of the most promising geopolymer
composite, application, fire-resistance, foaming composites is highly porous geopolymer foam [8]–[11].
I. INTRODUCTION A geopolymer foam is lightweight and resistant to acidic
A geopolymer was first described by J. Davidovich [1], environments; it also has lower thermal conductivity and a
who presented it as a new class of inorganic polymers lower sintering temperature. It also has a low density, which
obtained by polycondensation aluminosilicate materials in a is associated with strength. Geopolymer foam is enhanced
strongly alkaline medium. with aluminum-based metal powder, inorganic basalt fibers,
and glass fibers. Additionally, a variety of foaming agents are
Geopolymers form covalently bonded, inorganic used for the synthesis of low-density geopolymers, which are
crystalline networks. Covalently bonded networks of applied to solid surfaces. Using coarse fly ash, quartz sand,

and silica fume as fillers, it was found that potassium-based improved significantly when samples were heated to 1100°C.
geopolymers retain their structure completely after a fire test Cilla et al.[26] showed that the temperature treatment resulted
without any significant macroscopic damage [13]. in a significant improvement in the compressive strength
through sintering and the formation of crystalline phases. It
Foaming of the geopolymer reduces its mechanical
is protected from high temperatures by a geopolymer and
properties but increases its insulating properties. Geopolymer
does not lose its mechanical properties even at temperatures
foam composites are a superior alternative to mineral wool,
up to 1000°C. However, Kürklü [21] found that the coast of
glass, and foam concrete. Geopolymers can be used to make
fly ash-based geopolymer lost its mechanical properties after
fire-resistant coatings, thermal and acoustic insulators.
reaching a temperature above 500-600°C.
We have studied the effects of high temperatures on
Sarker et al. [26] investigated the fire resistance of steel-
geopolymers by changing their physical and mechanical
reinforced fly ash geopolymer and OPC concrete.
properties, apparent density, water absorption, weight loss,
Geopolymer concrete elements have shown greater fire
drying shrinkage, compressive strength, and flexural strength
resistance than OPC counterparts. Peng et al. [28]reported
[4], [13]–[17]. In this work, the effect of high temperatures
that alkali-activated amorphous framework structures
on geopolymer coatings was studied to determine their fire
transform into smooth ceramics after processing at high
resistance capability.
temperatures. Consequently, these geopolymers are
Geopolymers can usually withstand fire. The oxygen characterized by a stable porous structure and exceptional fire
flammability index is a measure of fire resistance. When resistance. Shuai et al. [29] studied the fire resistance of
geopolymers are exposed to flame, combustion zones of phosphoric acid-based geopolymers and found that they had
thermo-oxidative destruction are observed along with the relatively low thermal conductivity and good fire resistance.
destruction of the surface layer of the polymer and intense Additionally, acid-based geopolymers have been found to
heat and mass transfer, as well as zones of thermodynamic maintain structural stability and excellent mechanical
destruction of the inner layers of the geopolymer. Thermal- properties at high temperatures [27], [30], [31].
oxidative degradation occurs at temperatures below 250°C.
In this article, we examine the fire resistance of steel
The fire resistance of polymers can be estimated by plates coated with a protective foamed geopolymer layer to
several parameters, such as self-burning time or the flame ensure their thermal stability. In addition, the article explains
propagation speed. The heat resistance of geopolymer is its the process of spraying geopolymer onto the surface of the
ability to withstand the effects of heat. Polymers have an steel to be solved. Emphasis is placed on the use of
obvious, but complex relationship between heat resistance geopolymer composites.
and fire resistance. To determine the fire resistance limit, you
need to know the strength, thermophysical properties, and II. METHODS, TECHNOLOGIES AND MATERIALS
other properties of the material. Design solutions are aimed at improving the durability of
materials. These materials have been applied to various types
Many studies have been conducted on the effects of
of surfaces in the form of coatings that do not require
elevated temperatures on geopolymer materials [18]–[24]. At
formwork or sharping during the curing process.
high temperatures, geopolymers lose their mechanical
Geopolymers are being developed to study the effects of
strength[18]–[21],[25]. Chithambaram et al. [19] reported
elevated temperature on their physical, mechanical, and
that the compressive strength of geopolymer mortar
strength properties. Carbon steel was used as a fire protection
decreased and the mass loss increased when the sample was
material. A carbon steel plate of 500mm x 500mm x 50mm
exposed to the temperature of 1000°C. Yang et al. [20]
carbon steel with a 6mm thick layer of geopolymer foam on
showed a reduction in the compressive strength and young’s
the steel substrates was used for the fire resistance test.
modulus of geopolymer made from red mud slurry and class
F fly ash with the temperature increased to 1000°C. Kürklü A geopolymer mixture was used as the coating material.
[21] stated that geopolymer mortars based on coarse fly ash This consisted of Baucis Lk geopolymer based on metakaolin
lost around 58% of their strength at 1000°C. However, with an alkaline potassium solution, reinforced with chopped
several studies on geopolymer foams (GFs) have shown that basalt fiber for mechanical strength, and aluminum powders
GFs exposed to high temperatures have good strength for foam generation. The component ratio of foam
retention and low thermal shrinkage [22]–[24]. František et geopolymers varied in parametric studies. Our previous work
al. [22] reported that the mechanical properties of fly ash- provided the formulas for the geopolymers (see Table 1).
based GFs were improved and stable at 1000°C over one-year Further information can be found in "Study on Temperature-
period studied. Hlavácek et al. [23] showed that the Dependent Properties and Fire Resistance of Metakaolin-
compressive strength of alkali-activated fly ash GFs Based Geopolymer Foams" [13].
Table 1. Mix proportions of geopolymer foam
Binder Aluminum Sand/Binder Basalt Fiber/Binder

Powder/Binder
Baucis lk Activator
1 0.9 0.008 0.5 0.26
255
Geopolymer coatings were applied to a carbon steel panel During the fire resistance tests, the heating process was
by spraying (see Figure 1). The steps involved in preparing documented by distortion and deformation of the surface. The
the geopolymer mix were as follows: First, cement-based infrared thermometer Voltcraft IR 650-16D was used to
metakaolin and an alkaline potassium activator (in a 5:4 ratio measure the temperature. The furnace flame temperature was
according to the manufacturer) were thoroughly mixed for determined according to ISO 834.
five minutes. Second, sand and fibers were added to the
geopolymer mortar and mixed for five minutes. The T  T0  345 log 10 (8t  1)
aluminum powder was added at the final stage and mixed at where t – time in minutes, T0 – starting temperature in degrees
high speed for one minute resulting in a foaming effect on the Celsius.
geomortar. The mixture was then sprayed onto the steel plate
and cured at room temperature for 28 days. The result of this III. ANALYSIS, PROCESSING AND
process is shown in figure 2, the foam geopolymer layer of DOCUMENTATION OF RESEARCH RESULTS
6mm thickness is firmly seated in the steel plate.
Figure 1. Coasting geopolymer on a steel plate using a nebulizer

The tests were carried out on a 500 x 500 x 5 mm steel
plate, covered with a 6 mm thick layer of foam geopolymer.
A thermocouple was attached both to the outside and inside
of the test object and the chimney of the natural gas furnace.
The experimental setup is shown in Figure 3. For more details
on the experimental setup, the readers are referred to [32], Figure 3. Experimental setup of the fire-resistance study
[33]
The temperature dependence of the fire resistance is
Interface between 2 layers sprayed shown in Figure 4. Initially, the surface temperature of the
geopolymer foam samples exposed to the flame of a gas burner rapidly
increased up to 20 minutes, and then saturation occurred. The
fire resistance of coated steel samples is determined by the
thickness of the coating. The sample (with a coating thickness
of 6 mm) withstood the fire for a period of 80 minutes, while
the steel plate without the coating layer withstood only 9
minutes. The fire resistance of samples is also affected by the
thermal conductivity of the coatings as well as their thickness.
Figure 2. Experimental sample
Figure 4. Temperature-time diagram of the sample test
256
On the Temperature-Time Curve from the 20th to the 80th support of the Institutional Endowment for the Long-Term
minute, the temperature inside the furnace and the Conceptual Development of Research Institutes, as provided by the
Ministry of Education, Youth and Sports of the Czech Republic in
temperature outside the test sample were almost constant. As
the year 2022.
the heat outside the plate is exposed to the environment, it is
lost. In the next study, the author will investigate the behavior Conflict of Interest and other Ethics Statements: The authors
of the Temperature-Time curve as a function of the test certify that they have no affiliations with or involvement in any
organization or entity with any financial or non-financial interest in
conditions. On the outside, the test sample is covered with an the subject matter or materials discussed in this manuscript: “fire
insulating material. Figure 5 depicts the sample after testing. resistance properties and geopolymer coating”
REFERENCES
[1] J. Davidovits, “Geopolymers and geopolymeric materials”,
Journal of thermal analysis, vol. 35, no. 2, pp. 429–441, 1989.
[2] J. Davidovits, “Years of successes and failures in geopolymer
applications. Market trends and potential breakthroughs”, in
Geopolymer 2002 Conference, 2002, vol. 28, p. 29.
[3] B. Singh, G. Ishwarya, M. Gupta, and S. K. Bhattacharyya,
“Geopolymer concrete: A review of some recent
developments”, Construction and Building Materials, vol. 85,
pp. 78–90, 2015.
Figure 5. Pictures showing surfaces of geopolymer composite [4] V. V. Nguyen et al., “Low-Density Geopolymer Composites
before (left) and after (right) the test for the Construction Industry”, Polymers 2022, vol. 14, no. 2,
p. 304, Jan. 2022, doi: 10.3390/POLYM14020304.
The composition of the geopolymer changed during [5] Z. Zhang, J. L. Provis, A. Reid, and H. Wang, “Geopolymer
destruction and combustion, as well as diffusion and heat foam concrete: An emerging material for sustainable
construction”, Construction and Building Materials, vol. 56,
transfer conditions. As the temperature increases, the number pp. 113–127, 2014.
and size of microcracks appear on the surface of the material. [6] C. Bai and P. Colombo, “Processing, properties and
This reduces the thermal conductivity of the geopolymer applications of highly porous geopolymers: A review”,
composites. Ceramics International, vol. 44, no. 14, pp. 16103–16118,
2018.
IV. CONCLUSION [7] R. M. Novais, R. C. Pullar, and J. A. Labrincha, “Geopolymer
The use of an alkaline-based geopolymer material ensures foams: An overview of recent advancements”, Progress in
the performance of the protective coating. The liquefied Materials Science, vol. 109, p. 100621, 2020.
[8] Z. Abdollahnejad, F. Pacheco-Torgal, T. Félix, W. Tahri, and
material is applied under pressure by means of the nebulizer.
J. B. Aguiar, “Mix design, properties and cost analysis of fly
With the presented method, a substrate of any configuration ash-based geopolymer foam”, Construction and Building
and shape can be coated effectively. Geopolymers can also Materials, vol. 80, pp. 18–30, 2015.
be sprayed on different surfaces from different angles, [9] V. Ducman and L. Korat, “Characterization of geopolymer
allowing different operations to be performed. fly-ash based foams obtained with the addition of Al powder
or H2O2 as foaming agents”, Materials Characterization, vol.
With the spraying method presented in this article, the 113, pp. 207–213, 2016.
coating layer adheres firmly to metal plates due to its high [10] R. M. Novais, G. Ascensão, L. H. Buruberri, L. Senff, and J.
porosity. As a result of the empirical procedure given in the A. Labrincha, “Influence of blowing agent on the fresh-and
paper, it is possible to produce a porous geopolymer layer hardened-state properties of lightweight geopolymers”,
with excellent thermal insulation, noise reduction, and fire Materials & Design, vol. 108, pp. 551–559, 2016.
resistance properties. A foam geopolymer layer withstood [11] C. Bai and P. Colombo, “High-porosity geopolymer
temperatures of up to 1000oC for 80 minutes without losing membrane supports by peroxide route with the addition of egg
its mechanical properties. white as surfactant”, Ceramics International, vol. 43, no. 2,
pp. 2267–2273, 2017, doi: 10.1016/j.ceramint.2016.10.205.
Geopolymer spray coating has a wide range of [12] K. Sakkas, D. Panias, P. P. Nomikos, and A. I. Sofianos,
applications. A substrate of any thickness and roughness can “Potassium based geopolymer for passive fire protection of
be coated with this method. It is the proposed technology and concrete tunnels linings”, Tunnelling and Underground Space
apparatus for its implementation that provide an effective Technology, vol. 43, pp. 148–156, 2014.
[13] V. S. Le et al., “Study on Temperature-Dependent Properties
method for studying the parameters and composition of and Fire Resistance of Metakaolin-Based Geopolymer
geopolymer materials under various operational influences. Foams”, undefined, vol. 12, no. 12, pp. 1–18, Dec. 2020, doi:
Author Contributions: Writing—review and editing, the authors 10.3390/POLYM12122994.
have read and agreed to the published version of the manuscript. [14] T. D. Hung, P. Louda, D. Kroisová, O. Bortnovsky, and N. T.
Xiem, “New generation of geopolymer composite for fire-
Funding: This publication was written at the Technical University resistance”, in Advances in Composite Materials-Analysis of
of Liberec, Faculty of Mechanical Engineering with the support of Natural and Man-Made Materials, InTech, 2011.
the Institutional Endowment for the Long-Term Conceptual [15] V. S. Le et al., “Thermal conductivity of reinforced
Development of Research Institutes, as provided by the Ministry of geopolymer foams”, Ceramics-Silikaty, 2019.
Education, Youth and Sports of the Czech Republic in the year 2022. [16] D. H. Tran et al., “Effect of curing temperature on flexural
Acknowledgments: This publication was written at the Technical properties of silica-based geopolymer-carbon reinforced
University of Liberec, Faculty of Mechanical Engineering with the
257
composite”, Manufacturing Engineering, vol. 37, no. 2, pp. [25] V. S. Le et al., “Mechanical properties of geopolymer foam at
492–497, 2009. high temperature”, Science and Engineering of Composite
[17] X. Thang, P. Louda, and D. Kroisova, “Thermophysical Materials, vol. 27, no. 1, pp. 129–138, 2020, doi:
properties of woven fabrics reinforced geopolymer https://doi.org/10.1515/secm-2020-0013.
composites”, World Journal of Engineering, vol. 10, no. 2, pp. [26] P. K. Sarker and S. Mcbeath, “Fire endurance of steel
139–144, 2013. reinforced fly ash geopolymer concrete elements”,
[18] J. Ye, W. Zhang, and D. Shi, “Effect of elevated temperature Construction and Building Materials, vol. 90, pp. 91–98,
on the properties of geopolymer synthesized from calcined 2015.
ore-dressing tailing of bauxite and ground-granulated blast [27] B. Kozub, P. Bazan, D. Mierzwiński, and K. Korniejenko,
furnace slag”, Construction and Building Materials, vol. 69, “Fly-Ash-Based Geopolymers Reinforced by Melamine
pp. 41–48, 2014. Fibers”, Materials 2021, vol. 14, no. 2, p. 400, Jan. 2021, doi:
[19] S. J. Chithambaram, S. Kumar, and M. M. Prasad, “Thermo- 10.3390/MA14020400.
mechanical characteristics of geopolymer mortar”, [28] X. Peng, H. Li, Q. Shuai, and L. Wang, “Fire Resistance of
Construction and Building Materials, vol. 213, pp. 100–108, Alkali Activated Geopolymer Foams Produced from
2019. Metakaolin and Na2O2”, Materials, vol. 13, no. 3, p. 535,
[20] Z. Yang, R. Mocadlo, M. Zhao, R. D. Sisson Jr, M. Tao, and 2020.
J. Liang, “Preparation of a geopolymer from red mud slurry [29] Q. Shuai et al., “Fire resistance of phosphoric acid-based
and class F fly ash and its behavior at elevated temperatures”, geopolymer foams fabricated from metakaolin and hydrogen
Construction and Building Materials, vol. 221, pp. 308–317, peroxide”, Materials Letters, vol. 263, p. 127228, 2020.
2019. [30] L. Liu, X. Cui, Y. He, S. Liu, and S. Gong, “The phase
[21] G. Kürklü, “The effect of high temperature on the design of evolution of phosphoric acid-based geopolymers at elevated
blast furnace slag and coarse fly ash-based geopolymer temperatures”, Materials Letters, vol. 66, no. 1, pp. 10–12,
mortar”, Composites Part B: Engineering, vol. 92, pp. 9–18, 2012.
2016. [31] L. Le-Ping, C. Xue-Min, Q. Shu-Heng, Y. Jun-Li, and Z. Lin,
[22] Š. František, Š. Rostislav, T. Zdeněk, S. Petr, Š. Vít, and Z. C. “Preparation of phosphoric acid-based porous geopolymers”,
Zuzana, “Preparation and properties of fly ashbased Applied Clay Science, vol. 50, no. 4, pp. 600–603, 2010.
geopolymer foams”, Ceramics-Silikaty, vol. 58, no. 3, pp. [32] V. S. Le et al., “Study on temperature-dependent properties
188–197, 2014. and fire resistance of metakaolin-based geopolymer foams”,
[23] P. Hlaváček, V. Šmilauer, F. Škvára, L. Kopecký, and R. Šulc, Polymers (Basel), vol. 12, no. 12, pp. 1–18, Dec. 2020, doi:
“Inorganic foams made from alkali-activated fly ash: 10.3390/POLYM12122994.
Mechanical, chemical and physical properties”, J Eur Ceram [33] V. S. Le and P. Louda, “Research of curing time and
Soc, vol. 35, no. 2, pp. 703–709, 2015, doi: temperature-dependent strengths and fire resistance of
10.1016/j.jeurceramsoc.2014.08.024. geopolymer foam coated on an aluminum plate”, Coatings,
[24] M. S. Cilla, M. D. de Mello Innocentini, M. R. Morelli, and P. vol. 11, no. 1, pp. 1–13, Jan. 2021, doi:
Colombo, “Geopolymer foams obtained by the 10.3390/COATINGS11010087.
saponification/peroxide/gelcasting combined route using
different soap foam precursors”, Journal of the American
Ceramic Society, vol. 100, no. 8, pp. 3440–3450, 2017.
258
Effects of Intake Air Temperature on Power and

Emission Characteristics of the HCCI Engine
Fueled with the Blends of 15% Ethanol and 85%
Petrol Fuels
Minh Xuan Le Thanh Tuan Nguyen*
Automotive Engineering Dept. Automotive Engineering Dept.
Dong A University Nha Trang University
Da Nang City, Vietnam Nha Trang City, Khanh Hoa, Vietnam
xuanlm@donga.edu.vn nguyenthanhtuan@ntu.edu.vn
(*Corresponding author)
Abstract: This paper presents the research results on the announcements on using alternative fuels for HCCI engines
influence of intake air temperature on the power and emission are still limited.
concentration of the HCCI engine converted from the HCCI engines can be run with a high excess air
naturally aspired diesel engine. The fuel used is a mixture of coefficient or a dilute fuel mixture, resulting in low fuel
15% ethanol and 80% gasoline (E15). The air temperature is consumption. At the same time, HCCI also produces lower
heated through an electric heater in the intake manifold and emission concentrations [12, 13]. In the HCCI engine, the
varies from ambient temperature to 90°C. Experiments were mixture is homogeneously mixed and brought to the
performed with a single-cylinder, four-stroke, HCCI engine combustion chamber. The fuel mixture ignites itself like a
speed at 1300 rpm. The results show that when adjusting the compression-ignition engine.
electric heater from atmospheric temperature to 70°C, the
One of the current new technologies is the study of setting
brake power increases; as the intake air temperature
up an engine operating on the HCCI combustion principle.
increases, the brake power tends to decrease gradually.
The scientific community first recognized the engine running
Specific fuel consumption and NOx emissions tend to on this principle as the CAI combustion model research work.
increase at higher temperatures of 70°C to 90°C. It has also (Control auto ignition) was implemented by Onishi and
been found that CO and HC emissions first increase and then Noguchi in 1979. After the first work of Onishi and Noguchi,
begin to decrease after adjusting the heater temperature to a trend of engine research and development exploded. Using
70°C and 80°C, respectively. this fire model, typically Honda, with the first CAI engine
produced and installed a 2-stroke ARC motorcycle engine in
Keywords: intake air, temperature, HCCI engine, power, cars. In 1983, Najt ran the first four-stroke HCCI engine.
emission concentration Stockinger achieves HCCI combustion in four-cylinder and
four-stroke engines with limited engine load and speed [13–
I. INTRODUCTION 15]. In the mid-1990s, research began to be more systematic
Most air pollution in cities is caused by exhaust gases and focused on the applicability of the HCCI combustion
from internal combustion engines in motor vehicles. model to diesel engines in automobiles. Several methods
Therefore, more and more countries worldwide apply stricter have been applied for the fuel to burn according to the HCCI
standards for vehicles participating in traffic. In Vietnam, a mechanism, such as using exhaust gas recirculation,
developing country, the emission standard for cars is Euro 4 increasing the compression ratio, and heating the intake air.
standard. This standard will be at a higher level soon [1, 2]. HCCI combustion is initiated by spontaneously igniting the
With the pressure of pollution problems and oil reserves charge mixture at multiple points in the combustion chamber.
dwindling, a lot of research has been done on internal Exothermic is faster than spark ignition [16]. Some studies
combustion engines. This focuses on the research direction of have shown that, at full load, a dilute fuel mixture prevents
using alternative fuels in two main principles: LPG, CNG, the temperature rise in the cylinder and slows down the
Biogas... for gasoline engines [3–5], and vegetable oil and chemical reaction to complete combustion. As a result, partial
diesel engines [6, 7]. For alternative fuels for gasoline discharge occurs, and CO and HC emissions increase [12,
engines, these studies have offered many solutions using fuel 17].
supply systems to increase the number of features to
overcome the disadvantages of LPG fuel [8, 9] or solutions to Ethanol has higher anti-knock properties than gasoline.
reduce pollution in each field with the intervention of the When the engine works with HCCI mode, it can work with a
combustion process, change the ignition advance angle to high compression ratio. Along with that, ethanol has a high
limit the fire retardation time of fuels such as CNG, biogas, self-ignition temperature and a longer burning time than
etc [10, 11]. For diesel engines that use vegetable oil, many gasoline. Mixing gasoline and ethanol fuel will slow down the
solutions from the fuel supply system reduce the fuel's combustion process, leading to a decrease in peak pressure
viscosity and improve combustion efficiency. However, the and a reduction in engine power. With the HCCI engine built
research on HCCI is still entirely new, especially the from a primitive single-cylinder diesel engine, we have
proven experimentally that a compression ratio of 14 is the

most suitable when using a mixture of gasoline and E15 fuel. Table 2. Fuel properties of gasoline and ethanol [17]
With this compression, the engine works stably in the speed
Properties Gasoline E15
range from 1000 rpm to 1600 rpm. However, the engine power
is significantly reduced when compared to the HCCI engine Chemical formula C4–C12 C2H5OH
using gasoline fuel. To compensate for the disadvantages of Molecular weight 95–120 46
using gasoline mixed with alcohol. The primary purpose of
this paper is to study the influence of intake air temperature on LHV (MJ/kg) 42.8 26.8
performance parameters and concentrations of toxic Stoichiometric A/F (kg/kg) 14.7 9
substances when using a single-cylinder HCCI engine that has
Density (kg/m3) 742 790
been converted from a pure diesel engine. Fuel is a mixture of
15% ethanol and 80% petrol. Boiling point (◦C at 1 atm) 25–215 78
Flash point (◦C at 1 atm) −40 13

II. EXPERIMENTAL SETUP PROCEDURE
Auto-ignition temperature (◦C) ∼300 420
A. Experimental Apparatus
Latent heat of evaporation (kJ/kg) 380–500 904
This study converted the naturally aspirated, single-
cylinder diesel engine D28 to an HCCI engine running on E15 Reid vapour pressure (kPa) 60–90 17
gasoline. Engine reworked fuel delivery system with fuel Octane number (RON) 91 106
injectors in the intake manifold. At the same time, there is a
blower installed to push air into the combustion chamber in Composition (C, H and O) (%wt) 86, 14, 0 52, 13, 35
the intake manifold, similar to the turbocharger installed on Laminar flame speed (m/s atm at 25◦C) 0.33 0.41
the engine. The rotary blower is regulated with the intake air
pressure maintained at 3.5 bar. In the intake manifold, An
The schematic diagram of the test system is shown in
electric air heater is installed. The heater can be adjusted the
Figure 2. The motor is connected to the POWER TEST
temperature to serve the experimental process. The engine can
500HP power test bench. Use class K thermistors to measure
also change the compression ratio when adjusting two screws
intake air, fuel, and coolant temperatures. The cooling water
with a diameter of 26 mm drilled and installed into the
temperature is controlled within the limit of 75 oC. The
combustion chamber through the threaded hole on the engine
amount of fuel consumed is measured by a flow meter. The
mount. After conversion, the engine is mounted on an iron
injection time is controlled by an electronic controller
frame for adjustment and operation as shown in Fig. 1 before
designed and connected to the engine speed sensor. Exhaust
being completed and put on the test platform. The engine
gas measurement equipment is used EKOS MOD - 9000.
parameters before and after the transformation are shown in
Table 1. Gasoline and ethanol fuel properties are shown in
Table 2.
Figure 1. Diesel engine after being transferred to HCCI engine Figure 2. Schematic diagram of the homogeneous charge
compression ignition (HCCI) engine test
Table 1. The spirated engine parameters and after conversion to
HCCI
B. Eperimental Procedure
Engine CI, Single cylinder HCCI, Single cylinder When the engine starts working, to start the HCCI engine
Max power 17.2 (kW) - at low speed and warm-up, the engine initially uses spark
plug ignition mounted on the top of the engine cover. After
Max engine speed 2200 (rpm) -
running when stabilizing at low speed, the engine will cut off
Bore x stroke 125 x 120 (mm) 125 x 120 (mm) the ignition source to run on the HCCI principle.
Displacement 1,472 (l) 1,472 (l) The experiment was conducted with the same amount of
Compression ratio 17 14 fuel injected into the combustion chamber of the cycles. This
amount of fuel is calculated at ambient temperature
Ignition advance angle 21 (degree) -
conditions. The air excess coefficient is equal to 1. Fuel used
Specific fuel consumption 257 (g/kW.h) - E15 at the rate of 15% ethanol and 85% gasoline A95 by
volume. The engine rotation speed and compression ratio
260
remained the same in the experimental modes with values of As a result, incomplete combustion occurs. In addition, the
1300rpm and 14, respectively. This is the value that is the flame cannot enter the crevices of the piston and piston rings.
most suitable for this HCCI engine in the previous study [18]. So the flame will go out mainly when the engine is operating
The atmospheric temperature when experimenting was 30oC. with a smaller charge mixture. It can also be concluded that
The steady-state tests were repeated at least twice to ensure the combustion response is improved, and combustion occurs
that the results were repeatable within the experimental faster above the intake air temperature of 80°C. Thus, HC
uncertainties. emissions are reduced. In addition, the auto-ignition
temperature is reached earlier as the intake air temperature is
III. RESULT AND DISCUSSION increased, and the tendency to incomplete combustion is
C. Effects of intake air temperature on power and specific reduced due to faster explosion in the combustion chamber.
fuel consumption Therefore, the process of auto-ignition can be achieved more
quickly.
Figure 3 shows the experimental results showing the
influence of the intake air temperature on the engine's power
and specific fuel consumption at 1300 rpm. From the graph,
it can be seen that the engine power will increase when
adjusting the heater's temperature. The engine power will
decrease gradually until the heating device temperature is
adjusted to 80°C, 90°C; the engine power will decrease
progressively. The maximum brake power is 14.2 kW when
the heater temperature is adjusted to 70°C. While the specific
fuel consumption tends to decrease gradually as the intake air
temperature increases, the lowest value of fuel consumption
rate is 235 g/kWh measured in the test mode of adjusting the
heater at 70°C and 80°C. When increasing the temperature to
90 °C, fuel consumption tends to increase. As the intake air
temperature increases, combustion takes place faster and the
peak pressure increases, increasing the power (Temperature
adjusts from atmospheric temperature to 70°C. But when the
intake air temperature increases, the intake coefficient
decreases due to a reduction in air density (Adjustable heating Figure 4. Effects of intake air temperature on HC concentration
device more significant than 70°C). This is the cause CO emissions increased with the increase of intake air
affecting the drop in engine power when the intake air temperature and then decreased at higher intake air
temperature is increased high. temperatures. CO is an intermediate product mainly
generated from low-temperature combustion reactions and
oxidized to CO2 in high-temperature combustion reactions.
The in-cylinder gas temperature increased at high air
temperatures, and chemical reactions improved. Thus, CO
was oxidized, and CO2 formed.
Figure 3. Effects of intake air temperature on HCCI engine

performance
D. Effects of intake air temperature on emission
characteristic
Fig. 4-6 shows the change in the concentration of HC, CO
and NOx substances when changing the heater's temperature
in the intake manifold. When adjusting the heater from Figure 5. Effects of intake air temperature on CO concentration
atmospheric temperature to 80°C in the intake air temperature
As shown in Fig. 6, NOx emissions first decreased with
range, HC emissions increase as the intake air temperature
increased intake air temperature, while they tended to
increases. The cause of the rise in HC emissions is the low
increase at high intake air temperatures. Warmer intake air
volumetric efficiency due to insufficient airflow to the engine
temperatures accelerated the chemical reactions and
at high intake air temperatures. One of the most important
increased the in-cylinder gas temperature at the end of
reasons for increasing HC emissions is the richer charge
combustion. Thus, NOx emissions increased.
mixture. The engine will operate with a richer mixture with
constant fuel injection, part of the fuel that cannot be oxidized.
261
in Vietnam”. In: XLI. International Scientific Conference of

Czech and Slovak University Departments and Institutions
Dealing With the Research of Combustion Engines, Czech
Republic, 2010.
[5] N. T. Tuan and N. P. Dong, “Theoretical and experimental
study of an injector of LPG liquid phase injection system”,
Energy Sustain. Dev, 2021, 63, pp. 103–112.
[6] A. T. Hoang, M. Tabatabaei, M. Aghbashlo, A. P. Carlucci,
A. I. Ölçer, A. T. Le, A. Ghassemi, “Rice bran oil-based
biodiesel as a promising renewable fuel alternative to
petrodiesel: A review”, Renew. Sustain. Energy Rev, 2021,
135, 110204
[7] N. P. Dong, N. T. Tuan, R. Prochazka, “Performance
parameters reevaluate and predict the fuel consumption of
cummin engine running on CNG-Diesel Duel fuel by GT-
Power software”, 2021 Int. Conf. Syst. Sci. Eng. ICSSE 2021.
pp. 283–288.
[8] T. N. Thanh, D. N. Phu, “A Study of CNG Fuel System Uses
Mixer for Engine of the Suzuki Viva Motorcycle”, Int. J.
Mech. Eng. Robot. Res, 2022, 11, pp. 37–42.
Figure 6. Effects of intake air temperature on NOx concentration
[9] N. T. Tuan, L. M. Xuan, L. T. Hieu, D. Ph Tho, N. P. Dong,
IV. CONCLUSION “Research using the CNG fuel system from the petrol fuel
system for the honda wave engine”, J. Tech. Educ. Sci 2021,
The effects of intake air temperature performance and pp. 69–75.
exhaust emissions were investigated in a modified single- [10] N. T. Tuan, N. P. Dong, “Improving performance and
cylinder, four-stroke, port injection gasoline HCCI engine reducing emissions from a gasoline and liquefied petroleum
gas bi-fuel system based on a motorcycle engine fuel injection
fueled with 15% ethanol and 85% gasoline fuels. The tests system”, Energy Sustain. Dev, 2022, 67, pp. 93–101.
were performed at the engine speed of 1300 rpm and different [11] N. T. Tuan, “Movement and Vaporization of the Single
intake air temperatures (from ambient to 90°C). HCCI Liquefied Petroleum Gas Droplet Into the Intake Manifold”,
combustion was achieved at low intake temperatures at λ = 1 Int. J. Adv. Res. Eng. Technol, 2020, 11, pp. 714–719.
excessive air coefficient. The specific fuel consumption and [12] C. Cinar, A. Uyumaz, H. Solmaz, F. Sahin, S. Polat, E.
NO emissions increased at higher intake air temperatures. Yilmaz, “Effects of intake air temperature on combustion,
Maximum brake power was measured at 14.2 kW. It was seen performance and emission characteristics of a HCCI engine
fueled with the blends of 20% n-heptane and 80% isooctane
that brake power decreased as the intake air temperature was fuels”, Fuel Process. Technol. 2015, 130, pp. 275–281.
heated at high temperatures. CO emissions first increased and [13] P. Iodice, G. Langella, A. Amoresano, “Ethanol in gasoline
then started to decline with increased intake of air fuel blends: Effect on fuel consumption and engine out
temperature. It is hoped that this study contributes to emissions of SI engines in cold operating conditions”, Appl.
determining the effects of the intake air temperature and Therm. Eng, 2018, 130, pp. 1081–1089.
mixture composition on HCCI combustion. In this way, [14] R. K. Maurya and A. K. Agarwal, “Experimental investigation
alternative fuels can be used to understand the impacts on on the effect of intake air temperature and air–fuel ratio on
cycle-to-cycle variations of HCCI combustion and
HCCI combustion better. performance parameters”, Appl. Energy, 2011, 88, pp. 1153–
1163.
REFERENCES [15] M. K. Mohammed, H. H. Balla, Z. M. H. Dulaimi, Z. S.
[1] N. P. Dong, Laurin. J, N. T. Tuan, “Combustion of Natural Kareem, M. S. Al-Zuhairy, “Effect of ethanol-gasoline blends
Gas in Engines for Heavy-Duty Vehicles”, 51th International on SI engine performance and emissions”, Case Stud. Therm.
Scientific Conference of Czech and Slovak University Eng, 2021, 25.
Departments and Institutions Dealing with the Research of [16] D. H. Qi and C. F. Lee, “Combustion and emissions behaviour
Internal Combustion Engines, 2020. for ethanol–gasoline-blended fuels in a multipoint electronic
[2] P. D. Nguyen, T. T. Nguyen, “Research, Development and fuel injection engine”, International Journal of Sustainable
operation of Gas Engines in Viet Nam”, International Energy, 2014, 35, pp. 323–338.
Scientific Conference of Czech and Slovak Universities and [17] J. E. Tibaquirá, J. I. Huertas, S. Ospina, L. F. Quirama, J. E.
Institutions Dealing with Motor Vehicles and Internal Niño, “The Effect of Using Ethanol-Gasoline Blends on the
Combustion Engines Research, 2019, pp. 219–224. Mechanical, Energy and Environmental Performance of In-
[3] N. T. Tuan, “Compressed natural gas (CNG) as an alternative Use Vehicles”, Energies 2018, Vol. 11, pp. 221- 232.
fuel for internal combustion engine, problems and propose [18] L. M. Xuan, D. V. Dung, N. T. Tuan, “Research to convert the
technical solutions for growth CNG vehicles in Vietnam”, compression ignition engine to the HCCI engine using E20
XLI. International scientific conference of Czech and Slovak petrol fuel”, Journal of Technology education science, 2022,
university departments and institutions dealing with the 70.
research of combustion engines, Czech Republic, 2010.
[4] T. T. Nguyen, “Applicability and development LPG vehicles
262
Linear Viscoelastic Characterization of an SMA

Mixture Using Dynamic Indirect Tensile Test
H. T. Tai Nguyen Hong Ha Mai Van Hien Nguyen
Dept. of Transportation Dept. of Bridges and Roads, Post-graduate student, Institute of
Engineering, Faculty of Civil Institute of Civil Engineering Civil Engineering
Engineering Ho Chi Minh City University of Ho Chi Minh City University of
Ho Chi Minh City University of Transportation Transportation
Technology and Education Ho Chi Minh City, Vietnam, Ho Chi Minh City, Vietnam
Ho Chi Minh City, Vietnam, ha.mh@ut.edu.vn
tainht@hcmute.edu.vn
Abstract: Stone mastic asphalt (SMA) is a gap-graded Therefore, in this paper, the linear viscoelastic (LVE)
mixture known for its high resistance to rutting and fatigue behavior of an SMA mixture modified with a PE-based
cracking; thus, it is a promising material for preventing additive by the dry process was carried out with the aid of the
premature distresses of flexible pavements subjected to dynamic indirect tensile test.
highly-trafficked and hot climatic conditions in Vietnam. In
order to adequately design a pavement structure comprised of
an SMA layer, its mechanical behavior should be well A. Materials
examined. This paper reports a preliminary study on the This study investigates the LVE properties of an SMA
linear viscoelastic (LVE) behavior of an SMA mixture 12.5 mixture produced with a 60/70 pen bitumen modified by
modified with a PE-based additive via the dry process with a PE-based high modulus additive using the dry process. The
the aid of the dynamic indirect tensile test. The results of this amount of the additive was fixed at 2.5% of the total weight
work will provide a full understanding of the material’s of the mixture. The coarse aggregates and fine aggregates are
behavior, with which the stress and strain of SMA pavement andesite ones, which were gathered from the South of
can be evaluated. Vietnam. The bitumen used is a 60/70 pen bitumen, which is
equivalent to a PG 64 bitumen. The properties of aggregates
Keywords: stone mastic asphalt (SMA), linear and bitumen are as those presented in [3].
viscoelasticity, complex modulus, indirect tensile test, flexible
Marshall method was used for mix design, wherein 2 x 75
pavement blows instead of 2 x 50 blows were used to compact the
I. INTRODUCTION sample considering the issue of overloading in Vietnam. The
gradation curve was designed based on the limits prescribed
Rutting has been recognized as the deadliest distress of in AASHTO M 325 [8] for SMA of 12.5 mm of nominal
pavement structures as traffic safety will be drastically maximum aggregate size (NMAS) as shown in Figure 1. The
reduced once ruts occur at the pavement surface. Therefore, optimal binder content (OBC) was found 6.5% (of the total
rutting distress has drawn much attention from practitioners weight of the mixture). The engineering properties of the dry
and researchers around the world. Many attempts made been process modified SMA 12.5 at OBC were characterized
made in Vietnam to prevent rutting distress, and some presented in Table 1. In what follows, only samples produced
promising results have been gained thanks to the development with the OBC were considered for linear viscoelastic (LVE)
of prediction methods, improvement in material performance, characterization.
and construction techniques [1,2,3,4].
B. Testing methods used for LVE characterization
Compared to rutting, fatigue cracking occurs much later.
However, once cracks appear in the pavement surface, water The LVE of the dry process modified SMA 12.5 was
can penetrate and thus weaken the integrity of the beneath investigated following the standard dynamic modulus test
layers, which may accelerate the damage to the pavement procedure [9]. However, the indirect tensile test was
surface hence the reduction of traffic safety. Thus, asphalt considered to facilitate the preparation of test samples. The
materials should be designed to have enough resistance to both test set-up was illustrated in Figure 2.
rutting and fatigue cracking [5,6]. Let F0 be the magnitude of the applied force signal and
Stone Mastic Asphalt (SMA) is a gap-graded mixture that U 0 be the magnitude of the horizontal displacement signal
allows a large amount of binder content while maintaining a measured by LVDTs, the dynamic modulus of the sample can
very good resistance to rutting due to the high level of be computed as follows [10]:
interlocking between coarse aggregates [7]. Thus, SMA is a
promising material for preventing premature pavement F0
distresses of flexible pavements subjected to highly-trafficked  E*  (  0.27)  
and hot climatic conditions like those in Vietnam. In order to U 0  H
adequately design the SMA pavement, the mechanical
properties of SMA mixtures should be well understood.

where  is the Poisson ratio of the material, and H is the

height of the sample. If there is no data about the Poisson
ratio, a value of 0.35 can be used as suggested in [10].
Table 1. Fundamental properties of the dry process modified SMA

12.5
Property Value
Maximum specific gravity 2.385
Bulk specific gravity 2.318
Air void (%) 2.815
Void in aggregate (%) 19.25
Marshall stability (kN) 7.819

Figure 2. Test set-up for LVE characterization using the dynamic
indirect tensile test. (1) Two LVDTs. (2) Load platens. (3)
Marshall compacted sample. (4) Dummy sample for temperature
measurement
D. Phase angle
The obtained phase angles were shifted using the same
shift factors determined above and are presented in Figure 5.
It was found that the shifted phase angles do not create a single
smooth curve due to large measurement errors.
Compared with the compression test, the phase angle in
the indirect tensile test is more difficult to be measured, which
can be explained by the fact that the stress and strain are not
homogeneously distributed in the sample. In addition, the
Figure 1. Gradation design used for SMA 12.5 following measurement of phase angle exhibits a larger error than the
limitations specified in [8] measurement of dynamic modulus due to the extremely small
value of the time lag between the two signals.
Let F and U be the phase angle of the force and
displacement signal, respectively. The phase angle of the
complex modulus E * can be computed:
 E*  F  U  
III. RESULTS AND DISCUSSIONS

The investigating temperature was from -10 to 60 °C, and
the frequency was swept from 0.1 to 10 Hz. The sample was
left in the temperature chamber for at least 2 hours before
testing to assure a homogenous temperature field inside the
sample. In addition, the force level was chosen such that the
maximal tensile strain is lower than 50  so that no
significant damage was made to the sample. The obtained
results will be presented in the next Subsections.
C. Dynamic modulus
Figure 3. Dynamic moduli data and their construction of the
The dynamic moduli of the dry process modified SMA master curve using the horizontal shifting technique
12.5 obtained at various temperatures and frequencies are
presented in Figure 3. The master curve of dynamic moduli Figure 6 shows the Black diagram plotting the dynamic
was constructed by simply horizontally shifting adjacent modulus versus the phase angle. The measured data still form
isotherms [11]. The corresponding shift factors used for a single master curve yet is not very smooth due to the large
constructing the master curve of dynamic moduli are errors of the phase angles. Thus, the material can be
summarized in Table 2 and are illustrated in Figure 4. It can considered to fully obey the time-temperature superposition
be observed that the shift factors can be predicted by the principle (TTSP).
exponential law with a very high R 2 value:
 aT  exp 1.353 103 T 2  3.508 101 T  6.378   
264
Table 2. Shift factors with a series of one spring and two parabolic dashpots as
aT log( aT )
illustrated in Figure 7. The complex modulus of the model is
T (°C)
presented in Eq. (4) [12]:
4
-10 2.50×10 4.398
0 5.00×10 2
2.699 E  E0
 *
EHS ( )  E0   
10 2.00×10 1
1.301 1   [i (T )] k  [i (T )] h
20 1.00×100 0.000
-2
30 5.00×10 -1.301
40 4.00×10-3 -2.398
50 5.00×10-4 -3.301
60 5.00×10-5 -4.301
Figure 6. Black diagram at the reference temperature (20°C)
Since the dynamic moduli are more accurate than the

phase angles, only the data of dynamic modulus were used to
determine the HS parameters. According to the physical
causality principle, the phase angle and the dynamic modulus
Figure 4. Shift factors used in the construction of the master curve are interrelated via Kramers-Kronig relations [13,14]:
of dynamic modulus
2  u E (u )
ln E *( )  ln E 
 
0 2  u2
du;
  
2  ln E *(u )
 E ( ) 
 
0 u2  2
du.
Figure 7. Huet-Sayegh rheological model
Table 3. Huet-Sayegh parameters of the modified SMA 12.5

mixture
Parameter Value
E0 ( MPa ) 90
Figure 5. Phase angle data and their construction of the master
curve using the horizontal shifting technique E ( MPa ) 29,500
k 0.40
IV. PREDICTION WITH HUET-SAYEGH MODEL
h 0.45
E. Prediction of dynamic modulus  10.0
The Huet - Sayegh (HS) rheological model is well-known  1.0
in the literature for its excellence in fitting the viscoelastic
Tref (C) 20
behavior of asphalt mixtures while using relatively few
parameters. It comprises one spring that is related in parallel
265
Thus, it is not necessary to use both the dynamic modulus are interrelated with each other via the following equations
and phase angle data in the determination of LVE model [13,14]:
parameters. The less accurate data (e.g., phase angles in this
study) can be omitted in the fitting process and should be 2

E ' ( ) sin(t )
computed from the LVE model [14]. In the case of HS model,
Kramers-Kronig relations are systematically satisfied. The
E (t ) 
 
0

d  , t  0  
phase angle can be determined directly from the HS


parameters as follows: 2 D ' ( ) sin(t )
cos( k / 2) cos( h / 2)
D (t ) 
  
d  , t  0  
A  1 
0

( (T )) k ( (T )) h
 
sin(k / 2) sin( h / 2)
B  
( (T )) k ( (T )) h
E  E0 " E  E0
E ' ( )  E0  A ; E ( )  B 2  
A B
2 2
A  B2
 E ( )    E ( ) 
2 2
E*( ) = ' "
 
 E " ( ) 
 E * ( )  tan  '
1

 E ( ) 
The HS parameters that best fit the material behavior are

Figure 9. Prediction of phase angles using Huet-Sayegh model
presented in Table 3, and the fitting curves are also presented
in Figure 8. The computed phase angles are also presented in
Figure 9, which shows large differences between the
measured and the computed values. Due to the large errors of
the phase angles, the Black diagram predicted by the HS
model does not fit well with the experimental diagram as can
be seen in Figure 10. However, when using the computed
phase angles in place of the measured ones, the predicted
Black diagram agrees well with the hybrid Black diagram
plotting the measured dynamic moduli versus the computed
phase angles as shown in Figure 11.
Thus, in case the phase angles are not accurately measured
like those in this study, they should be computed from the
dynamic modulus data to improve the overall accuracy of the
LVE measurement and prediction.
Figure 10. Black diagram using the measure phase angles
Figure 8. Prediction of dynamic moduli using Huet-Sayegh model

Figure 11. Black diagram when using the converted phase angles
F. Time-frequency interconversion
For analysis in the time domain, the creep compliance where E ' ( ) and D' ( ) are respectively the real part of the
(CC, noted D(t ) ) and (or) relaxation modulus (RM, noted complex modulus and complex compliance, called the
storage modulus and compliance. For the HS model, the
E (t ) ) of the material must be determined. Based on LVE creep compliance and relaxation modulus can be analytically
theory, complex modulus, complex compliance, CC, and RM approximately by using Schapery and Park’s formulas [15]:
266
 arising in the measurement of phase angles do not affect the

 '    (1  n)cos(n )   overall accuracy of the LVE characterization and modeling.
2
E  
' The creep compliance and relaxation modulus were also
 1
E t     computed using the time-frequency interconversion formula
    '   of Schapery and Park. Based on the results obtained in this
study, the LVE behavior of SMA pavement can be
sin(n )   
'
 1 successfully analyzed both in the frequency and time domain,
Dt   . '  
   n E   which is in progress and will be presented in future works.
REFERENCES
where  is Gamma function, and n is the derivative of
[1] H.T.T. Nguyen, “Modelling the mechanical behaviour of asphalt
log(E '()) with respect to log  already presented in detail concrete using the Perzyna viscoplastic theory and Drucker–Prager
in [16]. yield surface”, Road Mater Pavement, vol. 18, pp. 264-280, 2017.
[2] Decision 858-MOT, “Applying current technical standards to
The creep compliance and relaxation modulus of the strengthen the quality of management of the design and
construction of hot mix asphalt pavement of high traffic roadway
studied mixture are respectively presented in Figures 12 and (in Vietnamese)”, Hanoi: Ministry of Transport, 2014.
13. [3] H.T.T. Nguyen, T.N. Tran, “Effects of crumb rubber content and
curing time on the properties of asphalt concrete and stone mastic
asphalt using dry process”, Int. J. Pavement Res. Technol., vol. 13,
pp. 238-244, 2018.
[4] T.N. Tran et al., “Semi-flexible Material: The Sustainable
Alternative for the Use of Conventional Road Materials in Heavy-
Duty Pavement”, in Congrès International de Géotechnique –
Ouvrages – Structures, Singapore: Springer, 2018, pp. 17-24.
[5] H.T.T. Nguyen, T.P. Phan, T.T. Do, D.L. Nguyen, V.T. Tran,
“Analysis of Flexible Pavements Comprised of Conventional and
High Modulus Asphalt Concrete Subjected to Moving Loading
using Linear Viscoelastic Theory”, 2020 5th International
Conference on Green Technology and Sustainable Development
(GTSD), 2020, pp. 302-311, doi:
10.1109/GTSD50082.2020.9303075.
[6] T.T. Do, D.L. Nguyen, V.T. Tran, H.T.T. Nguyen, “Effects of
Forta-fi Fiber on the Resistance to Fatigue of Conventional Asphalt
Mixtures”, 2020 5th International Conference on Green
Technology and Sustainable Development (GTSD), 2020, pp. 312-
Figure 12. Creep compliance computed from Huet-Sayegh model 316, doi: 10.1109/GTSD50082.2020.9303123.
[7] K. Blazejowski, Stone Matrix Asphalt. Theory and Practice, 1st ed.
Boca Raton: CRC Press, 2011.
[8] AASHTO M325, “Standard Specification for Stone Matrix
Asphalt (SMA)”. Washington: American Association of State
Highway and Transportation, 2008.
[9] AASHTO T 342-11, “Standard Method of Test for Determining
Dynamic Modulus of Hot Mix Asphalt (HMA)”. Washington:
American Association of State Highway and Transportation
Officials, 2011.
[10] EN 12697-26, “Bituminous mixtures - Test methods for hot mix
asphalt - Part 26: Stiffness”. Brussel: European Standards, 2004.
[11] J. Ferry, Viscoelastic properties of polymers, 3rd ed. New York:
John Wiley and Sons, 1980.
[12] F. Olard, H. Di Benedetto, “General “2S2P1D” model and relation
between the linear viscoelastic behaviours of bituminous binders
and mixes,” Road Mater Pavement, Vol. 4, pp. 185-224, 2003.
[13] K.S. Cho, Viscoelasticity of Polymers. Theory and Numerical
Algorithms. Dordrecht: Springer, 2016.
Figure 13. Relaxation modulus computed from the Huet-Sayegh [14] T.N. Phan, D.L. Nguyen, T.V. Vu, V.T. Tran, H.T.T. Nguyen,
model with the best-fit parameters M.L. Nguyen, “Determining with high accuracy the relaxation
modulus and creep compliance of asphaltic materials in the form
V. CONCLUSION of sums of exponential functions from mathematical master curves
of dynamic modulus,” J. Mater. Civ. Eng., in press, 2022.
The LVE properties of the dry process modified SMA
[15] R.A. Schapery, S.W. Park, “Methods of interconversion between
mixture have been successfully characterized using the linear viscoelastic material functions. Part II―An approximate
dynamic indirect tensile test method. Due to the analytical method”, Int. J. Solids Struct., Vol. 36, pp. 1677-1699,
nonhomogeneous distribution of the stress and strain in the 1999.
sample, the phase angles are measured as accurately as the [16] H.T.T. Nguyen, T.T. Do, V.T. Tran, T.N. Phan, T.A. Pham, M.L.
dynamic moduli. Thus, only the dynamic modulus data were Nguyen, “Determination of creep compliance of asphalt mixtures
used in the prediction of LVE behavior by Huet-Sayegh at intermediate and high temperature using creep-recovery test,”
Road Mater Pavement, Vol. 22(sup1), pp. S514-S535, 2021.
model. After that, the accurate phase angles can be computed
from the Huet-Sayegh model. With this procedure, the errors
267
Effect of the Limestone Powder Content on the

Properties of Alkali-Activated Slag Mortar
Tai Tran Thanh Chung Pham Duy Tu Nguyen Thanh
taitt@hcmute.edu.vn 1980804@student.hcmute.edu.vn tunt@hcmute.edu.vn
Hyug-Moon Kwon
Civil Engineering Department
Yeungnam University
Gyeongsan, Korea
hmkwon@yu.ac.kr
Abstract: The aim of this study was to investigate the strength in later-age [9][21][22]. Besides, sodium oxide
effect of limestone powder (LP) content on the properties of (Na2O) content and silicate modulus (a SiO2/Na2O ratio) of
alkali-activated slag mortar. Alkali-activated material was alkaline solution also have a strong effect on the setting times,
synthesized by mixing Granulated blast furnace slag (GBFS) workability, microstructure, and mechanical strengths of AAS
and the alkaline solution with a silicate modulus (a SiO2/ based material [17][23][24].
Na2O weight ratio) of 1 and different Na2O dosages of 4, 6, Limestone powder (LP) is one of the most widely used
8%. The used activator was the mixture of sodium silicate mineral admixtures for Portland cement [25][26]. The role of
liquid (Na2SiO3) and sodium hydroxide (NaOH). In the LP in Portland cement matrix was intensively studied in many
mortar mixture, the slag content was replaced by limestone previous publications [27][28]. LP grains were pointed out to
powder up to 20%. The presence of limestone powder was be inert, but acted as nuclei for hydration products [27].
seen to improve the compressive strength and flexural Besides, LP can accelerate the hydration process of C3S, and
strength of mortar after 28 days of ambient curing. In react with C3A [28]. However, up till now, limited studies
addition, the drying shrinkage of mortar samples was also have focused on AAM using LP as an admixture [29][30]. In
assessed in this study. AAM using metakaolin, LP was found to increase the release
of Al and Si ions from metakaolin [31]. While used in AAS,
Keywords: alkali-activated slag, limestone powder, LP mainly exhibited the physical effect: an inert filler in the
mechanical strength, drying shrinkage matrix, acting as nucleation sites [27][28]. Besides, B. Yuan
et al. found that the reaction in AAS is enhanced with LP
I. INTRODUCTION content lower than 5% [32].
Alkali-activated material (AAM) has become an attractive The present research aims to investigate the effect of
topic in the last decades [1]–[4]. AAM can be formed by limestone powder content on the AAS mortar behaviors
mixing an aluminosilicate precursor with the alkaline solution including the flowability, compressive and flexural strength,
[5]. A variety of industrial wastes such as fly ash, and drying shrinkage. Additionally, the properties of AAS
metallurgical slags, or calcined clay (metakaolin) can be used mortar activated with different alkaline solutions are also
as a precursor for AAM synthesis [6]. Granulated blast studied in this research.
furnace slag (GBFS), a by-product of the steel industry, is
used for alkali-activated slag (AAS) formation. It is reported II. MATERIALS AND METHOD
that AAS production is more beneficial for the environment A. Material characterization
than Portland cement manufacture [7][8]. Additionally, AAS
also shows some advantages such as the high early strength The solid materials include blast furnace slag, limestone
[3][9][10], less degradation in the corrosive environment powder and sand. Granulated blast furnace slag (GBFS) from
[3][11][12] or high temperatures condition [13][14]. Hoa Phat steel factory in Vietnam was used in this research.
However, AAS exhibits the higher shrinkage and cracking GBFS has a specific gravity of 2.9 g/cm3 and its surface area
during drying than Portland cement material [15]. from Blaine method is approximately 435 m2/kg. The oxide
composition of GBFS is shown in Table 1. On the other hand,
The behavior of AAS was found to be considerably fine limestone powder (LP) has a mean particle size (D50%)
dependent on the curing condition and alkaline solution of 3 ± 0.3 µm. Table 2 presents the properties of limestone
[9][15]–[17]. High temperature curing can improve the powder (LP). In addition, river sand was used as the fine
strengths of AAS by accelerating the chemical reaction in aggregate in the mortar mixture, and its fineness modulus is
AAS [18][19][20]. In addition, type and concentration of 2.45 .
alkaline solution highly influence AAS properties. Slag
activated by sodium silicate shows the slower strength gain For AAS synthesis, the alkaline solution was made by
than sodium hydroxide activated slag, but obtains the great mixing the pure sodium hydroxide (NaOH) with water glass

(sodium silicate solution). The used water glass has a silicate The flexural strength and compressive strength were
modulus of 3.22 ( a SiO2/ Na2O ratio) and a water amount of measured after 28 days of ambient curing according to
65.4% by mass. standard ASTM C348 and C349 [34][35].
Table 1. Chemical Composition Of GBFS AAS mortar drying shrinkage was accessed by measuring
Oxide composition GBFS (%) the length change of mini-bar (25x25x285 mm) specimens in
accordance with ASTM C596 [36]. The length change was
SiO2 33.81
determined continually in the first week, and at 14 days, 27
CaO 41.24
days, and 28 days. Figure 1 describes the test method for
Al2O3 15.19 mortar flow.
Fe2O3 0.41
MgO 5.54 III. RESULTS AND DISCUSSION
SO3 2.51
D. Flow of mortar
Na2O 0.25
K2O 0.61
The flow of fresh mortar is expressed by the diameter
change of mixture on the flow table before and after dropping.
LOI 0.18
Figure 2 illustrates the flow of all AAS mortar mixtures.
B. Mixture proportion Generally, the flow of all mixtures increased considerably
with using the limestone powder for slag replacement. The
AAS mortar proportions were 1:2.75:0.45 by weight of
highest value of mortar flow was observed in the mixture with
slag, sand, and water, respectively. GBFS was replaced by
the replacement level of 10%. This could be attributed to the
limestone powder content with different percentages by slag
filling effect in the AAS matrix of fine LP particles. As a
weight. Five mixtures with limestone content of 0, 5, 10, 15
result, water in the micro-voids could be released and lubricate
and 20% were named as L0, L5, L10, L15, and L20,
the particles [5]. Nevertheless, the higher content of fine
respectively. The used alkaline solution had three different
particles like LP could result in the higher water demand of
Na2O concentration of 4, 6, and 8 % by slag mass, and the
mortar, causing the reduction in the mortar workability. This
silicate modulus of 1 (a SiO2/ Na2O ratio).
could be demonstrated in AAS mortar with the LP content
more than 10%. On the other hand, AAS mortar mixtures
Table 2. Chemical Analysis of LP
exhibit the increase in mortar flow by raising the Na2O
CaCO3 98.7 % concentration of the alkaline solution.
MgCO3 0.2 %
Fe2O3 0.012 %
SiO2 0.012 %
C. Experiment method
AAS mortar mixtures were made by a 5 L capacity
planetary mixer. The flow of fresh mortar was determined by
the flow table according to ASTM C1437 [33]. Then, fresh
mortar mixtures were casted into prismatic molds with
dimension of 40x40x160 mm for the strength test, and
25x25x285 mm for the drying shrinkage test, respectively.
After demolding, mortar specimens were cured in the
laboratory condition (20  5oC, 60  5% RH) until testing.
Figure 2. Flow of fresh AAS mortar with the different Na2O

dosage of 4, 6, and 8 %.
E. Compressive strength
The compressive strength of AAS mortar specimens was
determined at 28 curing days and the experimental result is
shown in Figure 3.
As is shown by the graph in Figure 3, the replacement of
blast furnace slag by limestone powder led to the compressive
strength gain of specimens. For instance, the maximum
strength increase was 10.8, 17.9, and 19.9% for mixtures with
Na2O dosage of 4, 6, and 8%, respectively. It was noticed that
the acceleration in compressive strength increased with the
limestone content from 5 to 10%, but decreased when the
Figure 1. Drying shrinkage test on the mini-bar (25x25x285 mm) replacement level is higher than 10%. The mortar specimen
AAS mortar attained the highest compressive strength when blast furnace
slag was replaced of 10% by limestone powder. The strength
enhancement could be attributed to the physical effect of fine
Fig. 1. 1 269
limestone powder in AAS matrix, acting as nucleation sites control mortar, the strength gain was found to be the higher in
[32]. In addition, the micro-voids in the AAS matrix could be the mixture with higher Na2O concentration when rising the
filled by fine LP particles [5], which could decrease the total replacement level of limestone powder to 10%. In addition,
porosity, and lead to the strength gain. Nevertheless, this the flexural strength was seen to reduce when the replacement
strength increment was reported to reduce in specimens with level of LP is higher than 10%.
the high LP content of 15 and 20%. This should be resulted
from the reduction of hydration product in AAS matrix due to Figure 7 shows the flexural strength of all AAS mortar
the high LP content. According to previous studies [29][37], using the alkaline solution with three different Na2O content
there is no formation of the hydration products from LP. of 4, 6, and 8%. All mortar mixture specimens showed higher
strength when raising the Na2O content from 4 to 6 and 8% by
slag weight. Nevertheless, increasing the Na2O dosage from 6
to 8% resulted in the strength degradation in mortar with high
limestone content of 15 and 20%.
Figure 3. Compressive strength of AAS mortar with LP different

contents at 28 days
Figure 5 depicts the influence of Na2O dosage on AAS

mortar compressive strength. The strength was observed to Figure 5. Flexural strength of AAS mortar with LP different
increase with a higher dosage of Na2O for all mortar mixtures. contents at 28 days
Sodium oxide (Na2O) was found to accelerate the hydration
significantly, leading to the great strength gain [38][24]. The
graph in Figure 4 indicates that increasing the Na2O dosage
from 4 to 6% resulted in the greater strength gain than. The
strength gain decreased in mortar mixture with the limestone
content of 15 and 20 %.
Figure 6. Relation between the Na2O dosage and flexural strength

of AAS mortar at 28 days
G. Drying shrinkage
Figure 8 displays the results of drying shrinkage for all
Figure 4. Relation between the Na2O concentration and
mixtures with different limestone content from 0 to 20%. In
compressive strength of AAS mortar at 28 days the drying shrinkage test, the measurements have been carried
out for 4 weeks. All AAS mortar specimens showed
F. Flexural strength significant shrinkage in the first week of ambient curing. It is
noted that the more used limestone powder resulted in the
Figure 6 depicts the flexural strength of mortar with the higher drying shrinkage in AAS mortar except for mixture
limestone dosage of 0, 5, 10, 15 and 20%. M5. The highest shrinkage occurred in mixture M15 after 28
As shown in Figure 6, the flexural strength alteration is days of curing. The specimen with 5% limestone powder
similar to the compressive strength. The maximum strength exhibited a shrinkage decrease of approximately 25% in
value was obtained from AAS mortar containing 10% comparison with the control specimen. The high drying
limestone powder by slag weight. In comparison with the shrinkage of this material was pointed out to be attributed to
270
its finer pore structure [39]. LP with fine particles could fill in [6] C. K. Yip, J. L. Provis, G. C. Lukey, and J. S. J. van Deventer,
the pores in AAS structure, resulting in the reduction in drying “Carbonate mineral addition to metakaolin-based
shrinkage. geopolymers”, Cem. Concr. Compos., vol. 30, no. 10, pp. 979–
985, 2008, doi: 10.1016/j.cemconcomp.2008.07.004.
[7] E. Gartner, “Industrially interesting approaches to ‘low-CO2’
cements”, Cem. Concr. Res., vol. 34, no. 9, pp. 1489–1498,
2004, doi: 10.1016/j.cemconres.2004.01.021.
[8] A. R. Sakulich, E. Anderson, C. L. Schauer, and M. W.
Barsoum, “Influence of Si: Al ratio on the microstructural and
mechanical properties of a fine-limestone aggregate alkali-
activated slag concrete”, Mater. Struct. Constr., vol. 43, no. 7,
pp. 1025–1035, 2010, doi: 10.1617/s11527-009-9563-2.
[9] T. Bakharev, J. G. Sanjayan, and Y. B. Cheng, “Alkali
activation of Australian slag cements”, Cem. Concr. Res., vol.
29, no. 1, pp. 113–120, 1999, doi: 10.1016/S0008-
8846(98)00170-7.
[10] A. Fernández-Jiménez, J. G. Palomo, and F. Puertas, “Alkali-
Activated Slag Mortars Mechanical Strength Behavior”, Cem.
Concr. Res., vol. 29, pp. 1313–1321, 1999.
[11] T. Bakharev, J. G. Sanjayan, and Y. B. Cheng, “Sulfate attack
on alkali-activated slag concrete”, Cem. Concr. Res., vol. 32,
Figure 7. Result of drying shrinkage of AAS mortar with different no. 2, pp. 211–216, 2002, doi: 10.1016/S0008-8846(01)00659-
LP content in 4 weeks 7.
[12] T. Bakharev, J. G. Sanjayan, and Y. B. Cheng, “Resistance of
IV. CONCLUSION alkali-activated slag concrete to acid attack”, Cem. Concr. Res.,
vol. 33, no. 10, pp. 1607–1611, 2003, doi: 10.1016/S0008-
From the experimental results and discussion, the
8846(03)00125-X.
conclusions can be summarized as follows:
[13] H. T. Türker, M. Balçikanli, I. H. Durmuş, E. Özbay, and M.
The addition of LP can be seen to enhance the flowability Erdemir, “Microstructural alteration of alkali-activated slag
of fresh AAS mortar. Increasing the Na2O content of the mortars depend on exposed high temperature level”, Constr.
alkaline solution can also increase flow of fresh mortar. Build. Mater., vol. 104, pp. 169–180, 2016, doi:
10.1016/j.conbuildmat.2015.12.070.
The compressive and flexural strength mortar tend to [14] M. Guerrieri, J. Sanjayan, and F. Collins, “Residual
increase when raising LP content from 0 to 10 %, and reduce compressive behavior of alkali-activated concrete exposed to
with LP higher than 10%. elevated temperatures”, Fire Mater., vol. 33, pp. 51–62, 2009.
Raising the Na2O dosage of alkaline solution led to the [15] A. Fernández-Jiménez, J. G. Palomo, and F. Puertas, “Alkali-
mortar mechanical strength gain. The strength enhancement activated slag mortars: Mechanical strength behaviour”, Cem.
tends to reduce in Na2O dosage increment from to 8%, Concr. Res., vol. 29, no. 8, pp. 1313–1321, 1999, doi:
especially in 15, and 20% LP mixtures. 10.1016/S0008-8846(99)00154-4.
[16] M. M. Alonso, S. Gismera, M. T. Blanco, M. Lanzón, and F.
Replacement 5% GBFS by LP resulted in a decrease in Puertas, “Alkali-activated mortars: Workability and
drying shrinkage of AAS mortar. However, drying shrinkage rheological behaviour”, Constr. Build. Mater., vol. 145, pp.
increased with LP content of 10, 15 and 20%. 576–587, 2017, doi: 10.1016/j.conbuildmat.2017.04.020.
[17] S. Aydin and B. Baradan, “Effect of activator type and content
REFERENCES
on properties of alkali-activated slag mortars”, Compos. Part B
[1] Roy D M, “Alkali activated cements, opportunities and Eng., vol. 57, pp. 166–172, 2014, doi:
challenges”, Cem. Concr. Res., vol. 29, pp. 249–254, 1999. 10.1016/j.compositesb.2013.10.001.
[2] F. Puertas and A. Fernández-Jiménez, “Mineralogical and [18] V. Živica, “Effects of type and dosage of alkaline activator and
microstructural characterisation of alkali-activated fly ash/slag temperature on the properties of alkali-activated slag
pastes”, Cem. Concr. Compos., vol. 25, no. 3, pp. 287–292, mixtures”, Constr. Build. Mater., vol. 21, no. 7, pp. 1463–1469,
2003, doi: 10.1016/S0958-9465(02)00059-8. 2007, doi: 10.1016/j.conbuildmat.2006.07.002.
[3] F. Puertas, T. Amat, A. Fernández-Jiménez, and T. Vázquez, [19] S. Aydin and B. Baradan, “Mechanical and microstructural
“Mechanical and durable behaviour of alkaline cement mortars properties of heat cured alkali-activated slag mortars”, Mater.
reinforced with polypropylene fibres”, Cem. Concr. Res., vol. Des., vol. 35, pp. 374–383, 2012, doi:
33, no. 12, pp. 2031–2036, 2003, doi: 10.1016/S0008- 10.1016/j.matdes.2011.10.005.
8846(03)00222-9. [20] E. Altan and S. T. Erdoǧan, “Alkali activation of a slag at
[4] M. Chi, “Effects of dosage of alkali-activated solution and ambient and elevated temperatures”, Cem. Concr. Compos.,
curing conditions on the properties and durability of alkali- vol. 34, no. 2, pp. 131–139, 2012, doi:
activated slag concrete”, Constr. Build. Mater., vol. 35, pp. 10.1016/j.cemconcomp.2011.08.003.
240–245, 2012, doi: 10.1016/j.conbuildmat.2012.04.005. [21] A. R. Brough, M. Holloway, J. Sykes, and A. Atkinson,
[5] N. R. Rakhimova, R. Z. Rakhimov, N. I. Naumkina, A. F. “Sodium silicate-based alkali-activated slag mortars. Part II.
Khuzin, and Y. N. Osin, “Influence of limestone content, The retarding effect of additions of sodium chloride or malic
fineness, and composition on the properties and microstructure acid”, Cem. Concr. Res., vol. 30, no. 9, pp. 1375–1379, 2000,
of alkali-activated slag cement”, Cem. Concr. Compos., vol. doi: 10.1016/S0008-8846(00)00356-2.
72, pp. 268–274, 2016, doi: [22] E. Deir, B. S. Gebregziabiher, and S. Peethamparan, “Influence
10.1016/j.cemconcomp.2016.06.015.
271
of starting material on the early age hydration kinetics, activated metakaolin-based geopolymers”, Constr. Build.
microstructure and composition of binding gel in alkali Mater., vol. 66, pp. 53–62, 2014, doi:
activated binder systems”, Cem. Concr. Compos., vol. 48, pp. 10.1016/j.conbuildmat.2014.05.022.
108–117, 2014, doi: 10.1016/j.cemconcomp.2013.11.010. [32] B. Yuan, Q. L. Yu, and H. J. H. Brouwers, “Assessing the
[23] H. El-Didamony, A. A. Amer, and H. Abd Ela-Ziz, “Properties chemical involvement of limestone powder in sodium
and durability of alkali-activated slag pastes immersed in sea carbonate activated slag”, Mater. Struct. Constr., vol. 50, no. 2,
water”, Ceram. Int., vol. 38, no. 5, pp. 3773–3780, 2012, doi: 2017, doi: 10.1617/s11527-017-1003-0.
10.1016/j.ceramint.2012.01.024. [33] ASTM, “Standard Test Method for Flow of Hydraulic Cement
[24] C. Duran Atiş, C. Bilim, Ö. Çelik, and O. Karahan, “Influence Mortar: C1437-01”, Standard, pp. 7–8, 2001, [Online].
of activator on the strength and drying shrinkage of alkali- Available:
activated slag mortar”, Constr. Build. Mater., vol. 23, no. 1, pp. http://scholar.google.com/scholar?hl=en&btnG=Search&q=in
548–555, 2009, doi: 10.1016/j.conbuildmat.2007.10.011. title:Standard+Test+Method+for+Flow+of+Hydraulic+Ceme
[25] B. Lothenbach, G. Le Saout, E. Gallucci, and K. Scrivener, nt+Mortar#0
“Influence of limestone on the hydration of Portland cements,” [34] ASTM C348, “Flexural strength of hydraulic-cement mortars”,
Cem. Concr. Res., vol. 38, no. 6, pp. 848–860, 2008, doi: Am. Soc. Test. Mater., vol. 04, pp. 1–6, 2002.
10.1016/j.cemconres.2008.01.002. [35] ASTM C349, “Standard test method for compressive strength
[26] P. Hawkins, P. Tennis, and R. Detwiler, The Use of Limestone of hydraulic-cement mortars (Using portions of prisms broken
in Portland Cement : A State-of-the-Art Review, no. June 2003. in flexure)”, ASTM Int., pp. 1–6, 2002.
2019. [36] ASTM Committee C01.31, “ASTM C596-07 Standard Test
[27] J. Péra, S. Husson, and B. Guilhot, “Influence of finely ground Method for Drying Shrinkage of Mortar Containing Hydraulic
limestone on cement hydration”, Cem. Concr. Compos., vol. Cement”, Annu. B. ASTM Stand. Vol. 04.01, vol. 11, no. 5, pp.
21, no. 2, pp. 99–105, 1999, doi: 10.1016/S0958- 1–3, 2007.
9465(98)00020-1. [37] A. R. Sakulich, E. Anderson, C. Schauer, and M. W. Barsoum,
[28] G. Kakali, S. Tsivilis, E. Aggeli, and M. Bati, “Hydration “Mechanical and microstructural characterization of an alkali-
products of C3A, C3S and Portland cement in the presence of activated slag/limestone fine aggregate concrete”, Constr.
CaCO3”, Cem. Concr. Res., vol. 30, no. 7, pp. 1073–1077, Build. Mater., vol. 23, no. 8, pp. 2951–2957, 2009, doi:
2000, doi: 10.1016/S0008-8846(00)00292-1. 10.1016/j.conbuildmat.2009.02.022.
[29] X. Gao, Q. L. Yu, and H. J. H. Brouwers, “Properties of alkali [38] B. S. Gebregziabiher, R. J. Thomas, and S. Peethamparan,
activated slag-fly ash blends with limestone addition”, Cem. “Temperature and activator effect on early-age reaction
Concr. Compos., vol. 59, pp. 119–128, 2015, doi: kinetics of alkali-activated slag binders”, Constr. Build. Mater.,
10.1016/j.cemconcomp.2015.01.007. vol. 113, pp. 783–793, 2016, doi:
[30] A. J. Moseson, D. E. Moseson, and M. W. Barsoum, “High 10.1016/j.conbuildmat.2016.03.098.
volume limestone alkali-activated cement developed by design [39] F. Collins and J. G. Sanjayan, “Effect of pore size distribution
of experiment”, Cem. Concr. Compos., vol. 34, no. 3, pp. 328– on drying shrinkage of alkali-activated slag concrete”, Cem.
336, 2012, doi: 10.1016/j.cemconcomp.2011.11.004. Concr. Res., vol. 30, no. 9, pp. 1401–1406, 2000, doi:
[31] A. Cwirzen, J. L. Provis, V. Penttala, and K. Habermehl- 10.1016/S0008-8846(00)00327-6.
Cwirzen, “The effect of limestone on sodium hydroxide-
272
The Educational Philosophy of Existentialism with

the Development of Personalized Learning for
Learners in Massive Open Online Courses
(MOOCs) – The Case of Ho Chi Minh City
University of Technology and Education
Thi Thao Tran Tran Phuong Thao Hua Thi Chu Tran
Faculty of Political Sciences Faculty of Foreign Languages Department of Education
Ho Chi Minh City University of Ho Chi Minh City University of Chu Van An Continuing Education
Technology and Education Technology and Education Center
thaott@hcmute.edu.vn huatranphuongthao@hcmute.edu.vn tranchudhsp21@gmail.com.vn
Abstract: Massive open online courses (MOOCs) are a result, we propose a framework based on the educational
recent evolution of open educational resources and services philosophy of existentialism to provide a personalized
that can be used to address various higher education needs learning guide to the majority of MOOCs students, in order
in the transition to education 4.0. However, in addition to to satisfy students with different engagement styles and how
their benefits, MOOCs have many drawbacks. Limiting they are taken into account in MOOCs at HCMUTE.
MOOC developers to a one-size-fits-all approach makes it
II. BACKGROUND AND THE CONTEXT OF THE
increasingly difficult to address learners' specific needs. The
STUDY
article clarifies personality-oriented educational
characteristics of existential philosophy in education to pay MOOCs debuted in 2008 with the University of
attention to MOOCs at Ho Chi Minh City University of Manitoba’s Connectivism and Connective Knowledge
Technology and Education. According to the ideals stated in MOOC [2] and have been a hot issue since Laura Pappano
this study, a significant goal in implementing this school of declared 2012 the year of the MOOC in an article for The
New York Times [3]. MOOC models are always evolving,
philosophy's educational philosophy at Ho Chi Minh City
and they can be classified into three basic types based on
University is to create appropriateness in MOOCs for each
their pedagogy. cMOOCs are based on the "connectivism"
learner. philosophy, in which there are no predefined curriculum
requirements and participants decide the focus topics [4]. By
Keywords: educational philosophy, existentialism,
employing a connectivist pedagogical approach, which
personalized learning, MOOCs views knowledge as a networked state and learning as the
process of establishing these networks through the use of
I. INTRODUCTION online and social tools; xMOOCs take a constructivist
Massive Open Online Courses (MOOCs) have benefited pedagogical approach, with well-defined instructions for
millions of students as a result of recent advancements in short-term, structured learning and outcomes [5, 6].
online education (MOOCs). MOOCs offer a new approach xMOOCs duplicate online the traditional model of an expert
to lifelong learning by allowing students to access a wealth teacher and learners as knowledge consumers with video
of high-quality learning resources at their leisure, such as tutorials and graded tasks; QuasiMOOCs pertain to web-
watching video lectures, completing assignments, and based tutorials such as Open Educational Resources
participating in forums. Meanwhile, students with a variety (OER)which are not courses and do not involve social
of learning goals, such as preparing for a future career or interaction.
college life, enroll in those MOOC platforms. Despite their
many benefits, MOOCs have several drawbacks. One of To summarize, MOOCs are distinguished by their vast
these is that MOOC students do not have access to size, openness, and peer assessment. However, there are
personalized support, and a lack of interaction between some disadvantages to MOOCs: because they involve a
instructors and students reduces their learning experience large number of participants (thus the term “massive”),
[1]. One proposed solution is to provide personalized course management is highly hard, and instructor
guidance for MOOCs based on the foundation theory of the engagement is limited to the most crucial activities. This
educational philosophy of existentialism. limits individuality.
The goal of personalized learning in MOOCs is to The advent of individualized education, such as
provide MOOC learners with efficient learning resources MOOCs, is credited to the progress of information and
and feedback tailored to their unique needs. This paper is to communication technologies. Personalized learning
conduct a review of the personality-oriented educational emerged from the broader notion of personalization, which
characteristics of existential philosophy in education. As a is inextricably tied to technology and used in a wide number
of disciplines, including healthcare systems, educational

activities, economic transactions, and political elections... B. Methodology

Following that, individualized learning was established and
advocated since individual differences in learning must also The research methodology combines a philosophical
be recognized in the field of education, and individualized theory with suitable research methods and links a
learning is an effective strategy to increase student philosophical viewpoint into actual and appropriate research
motivation and learning outcomes. The educational mission strategies to increase the validity of the research. According
is to discover the individual so that a better teaching method to the philosophical research paradigm, although there are
can be developed. many types of research methodology applied to identifying
the students' selection behaviors, they are all divided into
Personalized learning is tailoring learning for each two general levels: quantitative and qualitative research.
student’s strengths, needs, and interests - including enabling Given that this study will be descriptive research as well,
students’ voice and choice in what, how, when, and where applying both qualitative and quantitative research methods.
they learn - to provide flexibility and support to ensure The authors will be using quantitative research methods to
mastery of the highest standards possible. [7] find out the personalized learning of MOOCs and student
behavior on surveys. The research findings are statistical
We can see from the definition above that the
numbers, data, percentages, and the number of subjects used
educational process has been customized to the unique
for comparative purposes to find a relationship between
circumstances and needs of each learner, that learners are at
empirical phenomena and quantitative expression; they are
the center of the process, and that learners can be highly
also known as empirical methods derived from
motivated and self-regulated in a personalized learning
philosophical positivism and objectivism.
environment. This approach has shifted the focus of
traditional education away from mass education and toward C. Data Collection and Analysis Method
educating specific individuals or groups, away from general
curriculum and instruction, and toward individualized The primary data collection instrument will be a mixed
curriculum and instruction. questionnaire that includes both open-ended and closed-
ended questions. The open-ended question is similar to a
However, several characteristics of MOOCs were qualitative research question in that if the respondent is
already known from the earlier historical introduction. allowed to answer freely, they can provide more accurate
MOOC is an acronym for Massive Open Online Course. and detailed personal thoughts and feelings about MOOCs’
MOOC students enroll in the programs for several reasons. challenges and future, and the researchers will also gain an
Some students spend the majority of their time viewing expanded result in a short period of time.
videos, while others repeat exercises, and still, others
explore the course in quest of information. According to the IV. THEORETICAL FRAMEWORK
most recent meta-analysis of MOOCs, research and As early as the mid-twentieth century, Van Cleve Morris
development for MOOCs have centered on i) the potential proposed existentialism as a frame of reference for an
and challenges of MOOCs for universities; ii) MOOC educational program geared towards personal development.
platforms; iii) learners and content in MOOCs; and iv) the Existentialism, he believes, is a suitable foundation for a
quality of MOOCs and instructional design issues [8]. teaching method in which education will be more concerned
The main goal of our work is to review and analyze with developing the man’s emotional aspect, the capacity to
studies on the personality-oriented educational love, appreciate, and respond affectionately to the world
characteristics of existential philosophy in education and around him [9].
provide personalized guidance to MOOC students, which According to Scott Webster, existentialism as a
potentially benefits student engagement and learning theoretical framework is essentially an educational
experiences and plays an important role in the further philosophy concerned with people’s “becoming.” This
development of MOOCs at Ho Chi Minh City University of philosophy teaches us how to educate a person to develop
Technology and Education. into an independent individual in the direction of his or her
choice. As a result, existentialism is used as a criterion of
III. RESEARCH METHODOLOGY
reference in educational research and practice. “The
A. Research Philosophy Paradigm emphasis upon the individual allows the framework to be
applicable for addressing specific educational issues
This research is primarily conducted using two involving the learner, such as pedagogy, curriculum, and
paradigms of research philosophy: positivist research and policy” [10].
phenomenological research. The purpose of phenomenology
is to investigate subjective living experiences. The Thus, this paper uses existentialism as a theoretical
phenomenological study is a technique for deciphering an framework to aid in the research and development of
individual's consciousness, perception, emotion, and personalized learning in MOOCs for HCMUTE students.
intentions through the examination of the actuality, Additionally, as required by existentialism, increased
appearance, element, and feel of items. Additionally, this individualization of instruction necessitates the
research attempts to investigate positivism philosophically. identification of students' learning styles. The theoretical
It is based on scientific investigations of data derived from framework also incorporates pertinent research on cognitive
students' sensory knowledge, whereas in positivist research, styles and learning styles.
we first conduct statistical analysis and then examine the
facts of the statistical data and the hypothesis of the result to A. Existentialism and Educational Philosophy of
determine the effect of personalized learning in MOOCs at Existentialism
HCMUTE.
274
Existentialism is primarily a European philosophy that existential terms, despair refers to the emotional response to
emerged before the turn of the twentieth century and gained a breakdown in one or more of one’s “pillars” of self or
popularity following World War II (1939–45). identity.
Existentialism's seeds can be traced back to an earlier 5. Authenticity and Inauthenticity:
period in the history of philosophy. During the eighteenth
century, reason and nature were elevated in importance, In an existentialist’s view, authenticity means remaining
objectivity was emphasized heavily, resulting in industrial true to oneself in one's own freedom or refusing to allow
and technological advancements, and science was elevated others to define the realm of one's agency, and owning who
to a new level of importance. The man was also managed as one really is. Inauthenticity is the condition wherein one
an object from a scientific standpoint. Man became a slave chooses to believe that one possesses an essence, nature, or
to machines as industrial society developed. In response to destiny that is unyielding to human choice.
this situation, existentialism arose as a protest against C. Educational Philosophy of Existentialism
society, asserting the supremacy of man’s individuality.
1) The aim of Education: Existentialists believe that
Existentialist writings are scattered throughout the works
the most important kind of knowledge is about the human
of numerous philosophers, including Friedrich Nietzsche,
condition and the choices that each person has to make and
Soren Kierkegaard, Gabriel Marcel, Martin Heidegger,
that education is a process of gaining consciousness about
Jean-Paul Sartre, Karl Jaspers, Abbagnamo, Bardyaev, and
the freedom to choose and the significance of personal
Albert Camus, to name a few.
responsibility for one's decisions. As a result, group norms,
Maxine Greene, George Kneeler, and Van Cleve Morris authority, and established order – whether social, political,
are well-known existentialists in American education who philosophical, or religious – are disregarded. Existentialism
emphasize individualism and personal self-fulfillment. rejects universal norms, customs, traditions, and everlasting
truths; in this regard, it contrasts with idealism and realism.
B. Fundamental research on Existentialism
2) The Existentialist Curriculum: Curriculum, in
1) Existence precedes Essence1[11]: This philosophy Existentialism, represents a world of information to
begins with man, but not as a thinking entity with a defined discover. This information should give a solid foundation
nature or essence. A man first exists, he meets himself and for unfettered analysis and critique as well as individual
defines himself. Existence precedes value and essence. creative effort. Existentialism emphasizes varied curricula
Because man is nothing at all, he becomes what he produces that meet each individual’s needs, abilities, and aptitude.
of himself. In his subjectivity, man oscillates between Self-realization is part of the curriculum. According to
freedom and existential sorrow. Existentialism is frequently existentialism, the curriculum should be mainly liberal since
associated with fear, death, and liberation. “liberal learning is most likely to lay a foundation for
2) Importance of Subjectivity: Existentialism is a human freedom” [13].
subject-oriented philosophy. By probing into one's Existentialism and Methods of Teaching: According to
subjectivity, one can discover one's true self and authentic existentialism, methods of teaching must develop the
role in life. This is a creative process that leads to new ideas. creative abilities of the learners. Education creativity is most
important for existentialists. They emphasize on application
3) Freedom, choice, and responsibility:
of such instructional techniques which appeal to feeling,
Freedom is the core of existential philosophy. emotion, creativity, and deeper meaning of life. To fully
Existentialism emphasizes human freedom and responds develop the individual, existentialists advocate for
against any determinism, as determinism of any kind limits individualized instruction that develops the learner's
the human personality's ability to flourish and to extend and aesthetic, moral, and emotional selves in addition to his
explain its different abilities, capabilities, and personalities. scientific rational selves. He should develop his capacity to
True freedom requires facing options, making decisions, and love, appreciate, and emotionally respond to the world
accepting them. Existentialism's emphasis on freedom is around him. Existentialists suggest that we should first view
linked to the responsibility that freedom entails. Freedom students as individuals and empower them to determine
and responsibility are intertwined, and defining one reveals their own education and lives. While it is true that each
the other. Man is free in his activities and responsible for student brings to school a unique set of experiences that will
them. Existentialism holds man solely responsible for who influence their personal choices, existentialists argue that
he is, as well as his own destiny. schools and their institutions should be free spaces where
students are encouraged to try new things in order to foster
4) Anxiety, dread, and despair: diversity in education, not just in terms of curriculum, but
Anxiety is one of the moods that man experiences as a also in terms of how things are taught. Certain learners
result of his emotions and intellect. It is a psychological benefit from one method while others benefit from another.
state that sheds light on the human predicament. It is the Numerous educational opportunities should be available to
result of a combination of anguish and rage. Existential them.
anxiety is very subjective, individualistic, and subjective. Existentialism and School: The aim of school should be
When a man is anxious, everything in this world becomes to nurture self-discipline and cultivate self-evaluation thus it
worthless or him. Man has a generalized dread of is humanizing school. The school should provide an
nothingness. Man has an irrational fear of nothingness. atmosphere where the individuals develop in a healthy way.
Kierkegaard equates anxiety with original sin [12]. In The school is preeminent since a place for having
experience with life existentialism. So, schools should try to
1
L’existence précède l’essence: the famous dictum of Sartre immerse their students in life.
275
3) Existentialism and Teacher: According to best approaches for emphasizing personal experience, as
existentialists, the teacher creates an educational existentialism suggests.
environment in which the student can make contact with
himself, become aware of it, and achieve self-realization. 2) Learning styles
This necessitates that the teacher adopts an existential Learning styles are typically defined as the preferred
stance. He should also have had self-realization experiences method of learning or processing information by
in order to be able to guide students through this process. individuals. It represents the learning and information
The teacher's responsibility is to assist students in defining processing strategies that people preferred to use. Both
their own essence by exposing them to a variety of possible cognitive and learning styles are psychological notions that
life routes and providing an environment in which they can characterize how people perceive and process information,
freely choose their preferred way. while learning style is more variable and
4) Existentialism and Teacher-Student Relationship: dependent. Learning styles can be measured in various
Existentialism emphasizes the importance of the teacher- ways. Kolb developed a four-stage experiential learning
student relationship being more informal, intimate, lively, cycle that includes concrete experience, reflective
and personal than it is in the traditional school. All major observation, abstract conceptualization, and active
existentialists decry the meaninglessness of formal experimentation. Kolb defined four learning styles:
responses and relationships. They assert that the teacher- diverging, assimilating, converging, and accommodating.
student relationship must be extremely close, involving [16]. On the other hand, Felder & Silverman classified it as
intensely exploratory intimacy. Both the teacher and the intuitive/sensitive, global/sequential, visual/verbal,
students must always have their own distinct personalities. inductive/deductive, and active/ reflective. The model
The teacher can educate the student if he or she can develop identified 32 learning styles [17]. Stangl’s theory
an authentic dialogue and mutuality with the students, as M. distinguished learners as acting, hearing, reading, or seeing
Buber suggests regarding “I-Thou” relationships [14]. learners. The Felder-Silverman Learning Style Model will
be applied in this study.
Thoroughly, the student develops an awareness of the
meaning of what the teacher as a person and the author of Five different dimensions have been discussed in Felder
the book as a subject present to him, and thus the subject and Silverman's theory.
takes on a personal significance for him. Sensing and intuition are the dimensions that distinguish
The existential educational philosophy emphasizes people in their perception of the world. Sensitive people use
freedom, authority, individual choice, and responsibility. their senses to gather data, whereas intuitive people use their
There is no authority in education without inner freedom. imaginations to examine the world. Sensitive people prefer
There is no freedom without creative work and an orderly evidence, facts, and experiments, whereas intuitive people
environment. Individuals supported by the school will prefer theories and concepts.
choose and go through certain stages of learning. Initially, Visual and auditory learners receive information
the teacher’s authority is derived from experience and differently. Visual learners prefer images, diagrams, and
maturity. Ultimately, authority is the value developed by the demonstrations over texts or other written explanations,
learner. Choice, responsibility, freedom, and discipline go while auditory learners prefer verbal explanations. Besides
hand in hand, and education should train students in each of that, explaining things to others could help auditory learners
these. learn.
D. Cognitive style and Learning styles Inductive and deductive are two different reasoning
progressions conducted by different people. Inductive
1) Cognitive style people work from specific observations to general theories,
inferring principles; deductive people work from border
Cognition is people’s mental processes collection which generalizations to specific hypotheses, deducing
includes their awareness, perception, reasoning, and consequences.
judgment of the external world. Thus, cognitive style refers
to an individual's preferred and habitual methods of Sequential learners process information sequentially and
digesting and organizing external data. It is a manifestation learn in steps. They can think convergently and analyze
of a person's personality and cognitive processes. There are problems without fully understanding them. But global
numerous theories about cognitive styles, including learners need a big picture before they can learn specifics.
reflection impulsivity, field dependence – field They are good at divergent thinking and synthesis.
independence, holist – serialist, and deep-level surface-level The active and reflective dimensions discuss how people
processing. This paper will discuss only the field-dependent put information into practice. Active learners like group
and field-independent theories. This paper has cited one performance, discussion, and application, while reflective
instruction framework Magoulas & Chen, 2006[15]. learners like to think and work on their own.
Because the main purpose of this framework is to provide
instructional decision support learners have more V. RESULTS, DISCUSSIONS, AND CONCLUSION
opportunities to control the learning activities from all This study employs data collection and qualitative
aspects of goals, instructions, assessments, and contents. methods to collect and analyze data on students at Ho Chi
Additionally, this instructional framework is based on Minh City University of Technology and Education in order
the study of instructional science and cognitive science. to identify existing problems with MOOCs courses at UTE.
Besides that, the framework also includes highly This study chose 1004 students at random from a total of
constructivist and prescriptive approaches—which are the 3000 students enrolled in the first 37 MOOCs courses at
HCMUTE.
276
The data collection method included (a) focus group According to surveys, students have a variety of distinct
interviews, and (b) a brief online survey, conducive for two cognitive and learning styles, and therefore require
months in February and March 2022. appropriate instruction with differences in learning goals,
instructional strategies, assessment, content, and
A. Students’ awareness of MOOCs individualized support.
The focus group interviews assessed students' awareness C. Adaptive Learning Goals.
of MOOCs. An email outlining the study's goal and asking
for volunteers was sent to all 1004 students at the Ho Chi The below table (Table 2) discusses the respondents’
Minh City University of Technology and Education (n = feedback on the learning goals for individual courses. Most
1004). Inquiries about participants' experiences with of the participants (161) hold the opinion that the instructors
MOOCs were followed by clarification questions. have given learning goals for the individual courses;
A brief online survey was created using Google Forms however, one-third of them (63) think the goals are general
and published via HCMUTE mailing lists in order to reach ones aimed at the masses, instead of more narrowed groups.
as many students as possible who will be enrolled in In addition, the participants do not have the experience of
available MOOCs during the first term of the school year having “customized” learning goals.
2021–2022. Volunteerism was encouraged. There were 229 Table 2. When you take a course on a MOOC, is there usually a
respondents. set of learning goals that you can choose from?
Only 35 (15.3 percent) of the 229 students who Frequency Percent
responded to the brief online poll revealed any prior Yes, there are. 161 71,6
knowledge of MOOCs. Furthermore, only 15 (6.6%) of the No, there is always a general goal for 63 28,0
229 participants had ever registered for more than three everybody.
No, there are no learning goals at all 1 0,44
MOOCs. Additionally, the 35 students who had prior Total 229 100,0
knowledge of MOOCs mentioned three primary sources for
this knowledge, which comprised the following: D. Adaptive Instructional Approaches
(1) Friends Table 3. Satisfaction of different instructional methods on MOOCs
(2) HCMUTE’s social media
Satisfaction of different instructional methods on MOOCs
(3) Academic Affairs Office and Faculty for High-
Unsatisfactory Completely
Quality Training (students) Unsatisfactory
172 (75%) out of 229 students have a basic (students)
understanding of how MOOCs work before enrolling. 5 Course instruction and 32 1
syllabus
(2%) students out of 229 learn about the style of pedagogy Course videos 15 3
used in HCMUTE’s MOOCs (xMOOCs), while the
Reading resources 25 3
remaining 224 (98%) are uninterested.
B. Individual Differences Presentation on MOOCs Learning notes (logs) 35 7
As mentioned previously, this study employs Magoulas Quiz during the videos 25 7
and Chen’s cognitive style framework and the five- Tasks (e.g. multiple choices) 29 6
dimensional Felder-Silverman Learning Style Model, so in
order to determine whether there are correlations between Discussion boards/forums 27 12
users’ learning styles and their MOOC learning experiences, Links to other social media 30 7
we must first understand the users’ learning styles. Instead learning groups
of having people take a standard learning styles test, two Group problem-solving and 28 6
questions were used for each dimension. collaborative tasks
Table 1. Learning styles Case study 30 5
Styles Semantic groups Frequency Percent

Active Try something out 95 42
Regarding the level of satisfaction with different
Social oriented 156 69 teaching methods on MOOCs, Table 3 shows that many
Reflective Think about material 132 58 students are still dissatisfied with the way MOOCs are
Impersonal oriented 71 32 conducted, with 15 students dissatisfied with course videos
Sensing Existing ways 97 42
concrete materials 80 35
in particular; 25 students are dissatisfied with their reading
careful with details 52 23 materials; 25 students were dissatisfied with the Quiz during
Intuitive new ways 212 93 the videos; 27 students are dissatisfied with discussion
abstract material 15 6,6 boards/forums, while 28 are dissatisfied with group
not careful with details 2 09
problem-solving and collaborative tasks; 29 students were
Visual pictures 22 9,7 dissatisfied with Tasks (for example, multiple-choice);
Verbal spoken words 74 32
written words 74 32
Links to other social media learning groups and Case Study
difficulty with the visual style 59 26 did not satisfy 30 students. Particularly, 32 students are
Sequential Detail-oriented 97 42 dissatisfied with the Course instruction and syllabus, and 35
sequential progress 64 28 students are dissatisfied with the Learning notes (logs).
from parts to the whole 68 30
Global overall picture 120 53 In general, many students are still dissatisfied with the
non-sequential progress 44 20 various methods of instruction on MOOCs.
relations/connections 65
277
E. Adaptive Assessment graphics, audio, video, text, etc., and denoted by hyperlinks)
can ensure additional learning resources for learners."
Table 4. Statements of the course assessments on MOOCs
Statements of the course assessments on MOOCs
Furthermore, 23 students disagreed and 1 student
strongly disagreed with the statement "course content is
Disagree Totally disagree always arranged at different levels and an equivalent rate,"
(students) (students)
There are enough self-assessments (e.g., 22 1
and 23 students disagreed and 3 students strongly disagreed
questions, and exercises, for which you with the statement "course materials always include a
have to check the answers by yourself and variety of resources."
can help you to identify your own
progress). To sum up, there are still some students who do not
There are enough formal assessments (e.g., 20 2 completely agree with the course content of MOOCs.
select answers, quiz, for which the
correctness is clear and can indicate your G. Individualized Support
mastering of some specific contents during
a course). Individualized supports follow the individual learners’
I could always receive timely feedback. 29 4 preferences and differences, one important aspect of
I could receive different forms of feedback. 30 6 individualized support in MOOCs is to help the learners to
learn effectively. So, we used open questions to gather
Table 4 demonstrates that many students still disagree information about the process students have undergone
about the appropriateness of assessment tools and receiving while studying with the MOOC. In order to analyze the
feedback from teachers. answers, we used content analysis. “The goal of some
content analysis work is to discover psychological
There are 22 students who disagreed and 1 student characteristics about the individuals or groups who created
strongly disagreed with the statement that "There are the messages” Krippendorff (1980) presents the advantages
enough self-assessment exercises (e.g., questions, exercises of content analysis: it is unobtrusive, accepts unstructured
where you have to check the answers yourself and can help material, and is context-sensitive. It can also process
you determine your own progress)". symbolic forms and cope with large amounts of data. While
There are 20 students who disagree and 2 students who analyzing the content of the diaries, special emphasis was
strongly disagree with the statement that "There are enough placed on the affective aspects, usability issues, and their
formal assessments (e.g., answer selection, quizzes, where intention to take MOOCs in the future. When data analysis,
correctness is obvious and can show you've mastered some we found that students have undergone difficulties in
specific content in a course)". providing personal support. A few quotations are presented
below:
29 students disagreed and 4 students completely
disagreed with the statement that "I can always get a timely “There is no one to guide, students mainly study on their
response". own.”
30 students disagreed and 6 strongly disagreed with the “Due to a lack of assistance and support from lecturers
statement that "I can get different forms of feedback." and tutors, I am unsure of what to do correctly or
incorrectly.”
In summary, the survey results show that providing “Due to the fact that some exercises lack answer
feedback on MOOCs is still insufficient and timely for instructions, I am unsure whether I'm performing them
learners to create their own personalized learning process. correctly or incorrectly.”
F. Adaptive Content Concerning instructor feedback, there was no unanimity
Table 5. Statements of the course contents on MOOCs among the students. Some students expressed high levels of
satisfaction with the instructor's feedback, while others
Statements of the course contents on MOOCs expressed dissatisfaction. The following statements indicate
Disagree Totally disagree that you are dissatisfied with the instructor’s feedback:
(Students) (Students)
“The teacher’s response is a little sluggish.”
The course materials always 21 1
contain hypermedia forms (contents contain “The most difficult part of taking a MOOC course is
graphics, audio, video, text, hyperlinks, etc. communicating with the lecturer in a timely way so that the
and indicated by
hyperlinks).
lecturer can help with the difficult exercises.”
The course contents are always organized on 23 1 “It’s difficult to communicate directly with the lecturer
different levels (from simply explaining the when you’re experiencing difficulties.”
subject to how to apply them in reality)
The course materials always include multiple 23 3 “When I post homework questions on a forum, I don't
types of resources (questions, exercises, get answers right away. This, I've found, makes me less
examples, and other activities).
curious.”
According to the findings, 21 students disagreed and 1 H. Summary of the results

completely disagreed with the statement that "course The results in this paper proved that almost all the tests
materials always contain hypermedia (content containing of the learners showed no correlation between learners’
278
model (personal data and information) and different and learning styles accessible, individuals should avoid
presentations of the contents and experiences of navigation focusing exclusively on one style and instead incorporate a
on MOOCs. Especially for the learning styles, there was no variety of ways for their personal use. The fact that students'
correlation within even one dimension of the four learning styles vary between courses and over time is a
dimensions. point that should be brought up to them.
The results for adaptive learning goals revealed that The planning of MOOCs concerning learning styles is
MOOC courses only provided general and static goals for complicated, as the individual needs of each learner cannot
all learners. Except for the fact that most participants be taken into account. The planning must be done in such a
expected more types of instruction approaches on MOOCs, way that students with varying learning styles have access to
the majority of findings revealed that personalized features a variety of learning methods. The table below can help
of instruction do not exist on MOOCs. Furthermore, in this teachers recommend learning materials for each learning
test, the course assessment, course content, and individual style.
support on MOOCs did not meet the personalized criteria.
Table 6. Recommendations in MOOCs
The factor of individualized support, while showing a
negative trend, remained ambiguous. Learning
Recommendations in MOOCs
style
To summarize, existentialist research on education has
revealed the diversity of learners. However, the research Active Linear text, general vision maps, chat, forum, and emails are all
findings indicate that not all MOOCs at HCMUTE use available.
personal user modeling, which serves as the foundation for The navigation is based on arrows (back and forward), and
personalized learning. MOOCs may have collected some printings are provided.
Speculating on possible questions and responding to them with
user data, but not all of the necessary data for implementing other students.
personalized learning.
Reflective Lesson objectives, case studies, conceptual maps, and slide shows
I. Limitations of the study that incorporate both linear and non-linear text. Online assistance
and email.
By collecting questionnaire responses, this study Possibility of writing summaries of previously learned material
presented the personalized learning experience of HCMUTE and emphasizing activities that require learners to watch and
listen.
MOOCs users. As a result, some data would be missing due
to the researcher's subjective interpretation of the questions, Sensing Facts, concrete material and data, hands-on activities, and
practical material.
as well as a lack of communication between the researcher Applying theory into practice, relating information to the real
and the participants. world.
Slideshows, hypertext, a response system, a digital library, media
Additionally, the qualitative questions were insufficient clips,
to obtain a trustworthy judgment. graphics, audio objects, case studies, conceptual maps,
multimedia
This is because this section may involve research across slideshows, graphics, digital movies, audio objects, and linear
the disciplines of philosophy, education, information text.
technology, and even psychology. With minimal questions,
Intuitive Theories and concepts.
assessing the user's learning style was challenging (normally Allowing children to explore options, cultivating their creativity
they should take other standard learning style tests). As a and imaginative ability, and eliciting interpretations that connect
result, the number of additional questions had to be reduced data and facts.
to accommodate the user's learning style, which weakened Lesson objectives, conceptual maps, presentations with text and
the results for other factors. multimedia elements, graphics, digital movies, audio objects, and
linear text.
J. Conclusion and future work Visual Graphics, tables, flowcharts, photos, videos, demos, conceptual
maps, color-coded notes with highlighters, and multimedia-
Individualized education has also been a topic of enhanced slides and animations.
discussion in recent years. Educational researchers have Slideshows, a digital library, case studies, and a strong emphasis
conducted numerous studies on pertinent concepts such as on synthesis.
personalized learning and self-directed learning. However, Verbal Text-based material that includes audio objects, lesson objectives,
the majority of research on individualized learning is hypertext, slideshows, a digital library, and conceptual maps.
conducted by technology researchers. MOOCs at Possibility of writing summaries of learning materials, working in
groups, and discussing and lecturing on learning
HCMUTE, a typical synthesis of educational philosophy
and educational sciences, were chosen as the context for Sequential Guidance and a predefined path of learning.
examining whether and how individualized learning has
Global Slideshows, a response system, media objects, and an open
been realized. On the basic principle of existentialism, course structure are all included in this course.
understanding students’ learning styles is necessary in order
to improve the quality of learning on MOOCs and attract
more students to participate in MOOCs. Questionnaires can be distributed to students at the
beginning and end of a course to ascertain their background
It is important to emphasize students' awareness of their information and to solicit feedback. When redesigning a
learning styles. Everyone should be encouraged to take and course, feedback is important. It is also important to solicit
view the results of Felder's Learning Style test (ILS). feedback throughout a course in order to ascertain its
Students can access helpful ideas and appropriate learning shortcomings. Students may respond more enthusiastically
objects for their learning styles on the test's website. to the feedback questionnaire if it is administered during a
Because there are numerous learning style questionnaires course, as they will derive benefit from it.
279
REFERENCES [9] V.C. Moris, “Existentialism and the Education of Twentieth

Century Man”, Educational Theory. Eng., Vol. 11, Iss.1, pp.
[1] S. Nishikant, W. Jack, Ng. Andrew and S. Peter,
52-60, 1961, DOI: 10.1111/j.1741-5446.1961.tb00042.x.
“Roundtable Discussion State-of-the-Field Discussion”,
MOOCs FORUM, Vol.1, Iss. P, pp. 6-9, Sept. 2013, DOI: [10] S. Webster, “Existentialism: Providing an ideal framework
10.1089 / mooc.2013.0006. for educational research in times of uncertainty ®”, in AARE
2002: Problematic futures: educational research in an era of
[2] J. Mackness, S. Mak, and R. Williams, “The ideals and
uncertainty; AARE 2002 conference papers, Australian
reality of participating in a MOOC”, in Proc. 7th
Association for Research in Education, Coldstream, Vic.
International Conference on Networked Learning, University
2002, pp.1-15. Retrieved from http://hdl.h
of Lancaster, 2010, pp. 266-274.
andle.net/10536/DR0/DU:30033908.h
[3] Pappano. L “The Year of the MOOC”, Nov. 2, 2012,
[11] J.P. Sartre. J.P, L'existentialisme est un humanisme. France:
[Online]. Available:
Gallimard, 2015, p.26.
https://www.nytimes.com/2012/11/04/education/edlife/massi
ve-open-online-courses-are-multiplying-at-a-rapid-pace.html [12] A. Hannay, Ed., The Concept of Anxiety: A Simple
[Accessed 3/17/2022]. Psychologically Oriented Deliberation given the Dogmatic
Problem of Hereditary Sin. New York: Liveright, 2014, pp.
[4] Blanco A F, García-Peñalvo F J, and Sein E M, “A
134-186.
methodology proposal for developing adaptive cMOOC”, in
Proc. First International Conference on Technological [13] E.J. Power, Philosophy of education: Studies in philosophies,
Ecosystem for Enhancing Multiculturality, 2013, pp. 553– schooling, and educational policies. New Jersey: Prentice-
558, doi:10.1145/2536536.2536621. Hall, 1982, p.145.
[5] A.F. Blanco, E.M. Sein, and F.J. García-Peñalvo, “From [14] W. Kauffman, trans., I and Thou, New York: Charles
massive access to cooperation: lessons learned and proven Scribner’s Sons, 1972.
results of a hybrid xMOOC/cMOOC pedagogical approach [15] M. Specht, S. Y. Chen and G. D. Magoulas, “Adaptable and
to MOOCs”, International Journal of Educational adaptive hypermedia systems”, Univ Access Inf Soc 5, 325
Technology in Higher Education. En,. Vol. 13, Iss. 1, pp. 1- (2006). https://doi.org/10.1007/s10209-006-0053-9.
13, Dec. 2016, doi: 10.1186/s41239-016-0024-z. [16] D. Kolb, Experiential learning - Experience as the Source of
[6] O. Skrypnyk, S. JoksimovicÌĄ, V. KovanovicÌĄ, D. Learning and Development, USA: Pearson Education, Inc.,
GasÌŇevicÌĄ, and D. Dawson, “Roles of course facilitators, 2015, pp. 97- 151.
learners, and technology in the flow of information of a [17] R. Felder & L. Silverman, “Learning and teaching styles”,
cMOOC”, The International Review of Research in Open Journal of Engineering Education. Eng., Vol. 78, No.7, pp.
and Distributed Learning. Eng., Vol. 16, No.3, pp. 188 – 674 – 681, 1988.
217, June 2015, DOI: 10.19173/irrodl.v16i3.2170. [18] A. K. Neuendorf, The content analysis guidebook. Thousand
[7] S. Patrick, K. Kennedy, and A. Powell, “Mean What You Oaks, CA: Sage, 2002.
Say: Defining and Integrating Personalized, Blended and [19] K. Krippendorf, Content analysis: an introduction to its
Competency Education”, International Association for K-12 methodology. Thousand Oaks, CA: Sage, 2002.
Online Learning (iNACOL), Vienna, Virginia, US, Rep. VA
22182, p.4, Oct. 2013.
[8] O.Z. Richter, A. Bozkurt, U. Alturki, and A. Aldraiweesh,
“What Research Says About MOOCs – An Explorative
Content Analysis”, International Review of Research in
Open and Distributed Learning. Eng., Vol.19, No.1, pp. 242-
259, Feb.2018, DOI: 10.19173/irrodl.v19i1.3356.
280
Transportation Infrastructure Strategy for

Sustainable Development: A Case Study of Vietnam
Mekong Delta
Le Thu Huyen
Faculty of Economics-Transport
University of Transport and
Communications
Hanoi, Vietnam
lethuhuyen@utc.edu.vn
Abstract: Despite the ideal location as the Asian gate to SWOT analysis. It should be emphasized that there have
the ocean, Mekong Delta still cannot make full use of its been only a few quantitative studies on logistics strategy. In
strength (i.e. agricultural and aquaculture production order to obtain precise information for the logistics strategy
capacity) with a rather poor and inappropriate infrastructure development, it is required to conduct both qualitative and
system. The paper aims at proposing a strategy to develop quantitative analyses.
regional multimodal transportation infrastructure in case of
The paper, applying methods of SWOT and MCA (multi-
the Mekong Delta (MKD). SWOT Analysis and expert
criteria assessment), has developed a framework model to
judgment are combined to establish the development policy
analyze the logistic strategy qualitatively and quantitatively,
in the field of logistics and multi-modal infrastructure in an
then finally to reveal the optimized logistic strategy.
explicit manner. The strategic proposal for the Mekong
Delta region is based on offensive strategies which are ideal II. METHODOLOGY
for the region with the target of rapid growth in the planning
period. It should also be emphasized that different regions in SWOT analysis was firstly proposed in Concepts for the
Firm Strategy (Kenneth R. Andrews, 1971). The method has
the country may have different development strategies due
been frequently used in analyzing competition information
to the natural characteristics and social-economics
[5]. SWOT methodology is helpful in proposing strategies
requirements.
to take advantage of opportunities and strengths and to
Keywords: sustainable transportation, sustainable overcome threats as well as weaknesses (Talancón, 2006).
development, transportation strategy, infrastructure Then the most efficient and reasonable option was proposed.
development, Mekong Delta MCA evaluation approaches have been applied
increasingly the last decades for policy evaluations in
I. INTRODUCTION
various decision-making contexts (Grafakos et al. 2010). It
Despite the ideal location as the Asian gate to the ocean, is also applied in analyzing selected opportunities with both
the Mekong Delta still cannot make full use of its strength data and qualitative knowledge to assess the expected
(i.e. agricultural and aquaculture production capacity) with a impacts of implementation and their contribution to green
rather poor and inappropriate transportation infrastructure growth. In the research, the key dimensions of the
system. Recently, Vietnam logistics network has not been Sustainable Development Plan of Vietnam [14] were used to
operated in an integrated manner, resulting in a high drive the assessment, and to guarantee coherence with
logistics cost (around 20% GDP), nearly double that of national priorities.
developed nations (Blancas et al., 2014).
MCA and SWOT methodologies may help clarify
MKD region is even classified as the worst case in complex problems, and the quantitative analysis will be very
logistics field in the whole country. Inefficient air and sea useful to find the best solution to a specific problem.
transportation systems, lacking the land supporting
infrastructure and facilities, including warehousing and The procedure for developing the framework model
depot, are hampering the growth of efficient logistics combining SWOT and MCA is described as follows:
business. Therefore, it has raised the need for a sufficient (1) From the literature review, elements of MKD logistic
infrastructure masterplan to overcome the challenges of and transportation strategy have been determined as
sustainable development implementation. strength (S), weaknesses (W), opportunity (O), and threat
For such a long time, the region has experienced fierce (T). Such factors have been collected and integrated from
domestic and international competition. In this context, (i) existing government’s policies in the multi-modal
researches have been conducted to optimize the regional transport development in the Mekong Delta and other
logistics strategy [2], [8], and some researches are regions in the country, as well as implementation progress;
conducted with SWOT methodology [9], [11]. (ii) experiences in the planning and development of the
Mekong Delta, especially those related to transportation and
However, there is often a lack of quantitative analysis in logistics strategies; (iii) emerging directions in the socio-

economic development of the country and Mekong Delta, Ho Chi Minh city to export. The regional products are
including opportunities for collaboration with the Mekong transported from MKD to Cat Lai Port in the long distance of
upstream countries; and, (iv) latest achievements and 200-500km by road for export. This burdens the cost of
development in planning and implementing logistics and export goods, and reduce the advantage competitiveness [6].
transportation strategies. W2. Traffic congestion and delay. Due to the high cargo
(2) Setting the logistics and transportation strategy as flow, the status of traffic overloading and congestion in some
overall objective, selecting the systematic layer of S-W-O-T key seaports (e.g. the seaport of Cat Lai) happen frequently.
Therefore, the shippers may have opportunities to increase
elements, from which arranging such factors into the
the service prices. At the same time, the time to release cargo
indicator layer, thus establishing a hierarchical model. Then
may be extended, which makes enterprises to waste
SWOT matrix is established with such strategies allowing warehouse cost as well as waiting time. Therefore, recently
for determining the scenarios of offensive, defensive, IWT system have not been improved so far (Hoàng Anh
adaptive, and survival for MKD development. Dũng 2019). Network inconsistency is also the reason to
The priority of the elements in every layer will be limit road supply capacity, then leading to traffic congestion
critically discussed. Finally, logistics and transportation and delay. For example, the segment of National highway
infrastructure development strategies will be recommended 1A from Ho Chi Minh city to Can Tho is 04 lanes (in both
for the MKD region. directions). However, in the road, there are some bridges of 2
lanes, leading to traffic congestion.
III. SWOT ANALYSIS OF MKD INFRASTRUCTURE
W3. Inefficient inland waterway. Yin Yin Lam et al.
A. Current situation of MKD infrastructure (2019) said that the main freight transportation mode in
From a literature review, factors of Strength, Vietnam now is road transport. This is one of the highest-
Weaknesses, Opportunities, and Challenges of MKD ranking reason influencing logistics activities in Vietnam
infrastructure have been summarized in the following part. [16]. This is also correct for freight transportation in MKD.
So far, road transportation in MKD covers about 70% of
The strengths are those of: freight transport.
S1. In the important aspect of the logistics industry in Based on Luis (2014), lack of budget to maintain IWT is
MKD, the region has the great potential for economic also an obstacle of the market development. When we can
development, especially in agriculture and aquaculture. determine and select IWT routes with expected
Investment in logistics infrastructure may create great value capacity/design, it is required to organize maintenance in
and contribute to regional growth, as well as possibility to order to ensure designed capacity.
extend in the national level.
At the time being, IWT maintenance in MKD has to face
S2. The region has the big strength with a condensed the difficulty of budget deficit. The current budget for IWT
river system connecting with major ocean routes in regions maintenance is mainly devoted to dredging and shore
and international [4]. Extended IWT network can be taken protection constructions, which are extracted from the
full use to develop multi-modal transport to reduce transport national budget. The budget is always deficit. Although there
and logistics cost. is no specific estimation of the minimum maintenance
budget required, available budgets can only meet 50-60% of
S3. With its ideal location, MKD also owns the strength
the requirements.
of connecting crossed-border transport by roads and IWT in
order to develop foreign trade with Cambodia and others in W4. Facilities to support freight transportation are
GMS. insufficient.
Nguyễn Thị Tuyết Nga (2019) analyzed challenges and Based on Luis et al. (2014) currently, only about 35% of
difficulties in logistics industry in MKD. The research has container flows in the Mekong Delta are transported by
highlighted issues related to logistics development in the inland waterways. To improve IWT, it is essential to
region, including issues of transport, institutions, industries improve the facilities for handling container at river ports.
and association inter-disciplines, development of logistics Currently, most river ports in the Mekong Delta are not
enterprises in the region, investment budget to develop equipped with infrastructure for container handling. In
logistics industry, human resources [5]. addition, the weak last-mile connectivity has restricted
transportation by the inland waterway network. In order to
The following are weaknesses in the regional logistics
load containers from ships onto trucks and ensure efficient
infrastructure.
cargo transfer, container terminals and container handling
W1. Based on Pham Van Dang (2019), in MKD region, equipment such as cranes and forklifts need to be equipped at
there are lack of seaports, especially ports with high capacity ports.
to serve big container ships for import-export activities. At
W5. Poor connectivity. There is lack of connection
the moment, in the region, the seaat port system is rather
between farmers (agricultural production) with transportation
scattering, with low capacity. They cannot take full
network and ports. In MKD, production or freight
advantages of MKD inland waterway. At the same time,
transportation enterprises are normally at the small scales.
operation and connectivity is inconsistent at the low
Decentralized goods hinder development, such as IWT
efficiency among different transportation modes inside the
routes and sustainable multimodal transport. Farmers do not
region and between the region with the export market.
have the capital to invest in modern technological equipment
Therefore, more than 70% of agriculture and aquaculture
to develop the supply chain, improve productivity and
volumes produced in MKD have to transport by container to
quality. Smaller transportation companies have higher costs.
282
For example, fixed costs per truck are higher than those of Ho Chi Minh city area. Habits and dependence on shippers
companies with large trucks, resulting in low profits. Small- also make logistics in the MKD area difficult to develop.
scale production and transportation lead to higher costs. In T3. Moreover, MKD has big challenges from climate
addition, because the Mekong Delta mainly produces change. Under Han Entzinger (2016), MKD is one of the hot
agricultural products (>70%), the quantity of exported goods points globally in climate change. The region depends on
is higher than that of imports, leading to an imbalance in typical agriculture, but also is vulnerable to impacts of
transport demand. One-way transportation makes freight climate change, including sea-level rising, salinity intrusion,
costs higher. and sudden climate change phenomena such as storms and
W6. Lack of professional human resources. Not only in flooding. From GIZ (2018), to 2100, 38 % of the Mekong
MKD but also in the whole country, human resources in the Delta area may be completely under the sea level. More than
logistics industry are facing many problems such as 50% of the 720 km long coasts is eroded - of which more
insufficient quantity, low professional qualifications... than 10% is eroded at a rate of 20 to 50 meters per year.
Developing human resources in the logistics industry in
Vietnam is an urgent requirement at the time being. B. Establishing SWOT matrix
Currently in Vietnam, up to 53.3% of surveyed enterprises After determining factors of S-W-O-T, the SWOT matrix
lack staff with professional qualifications and knowledge in has been established in order to develop strategies to
logistics, only a very small percentage. 16.7% of enterprises develop logistics infrastructure in MKD. The main goal of a
are satisfied with the professional qualifications of their SWOT matrix may help the decision maker to determine
employees, up to 30% of enterprises have to retrain meet their critical strategies: consolidating the strengths and
employees after recruitment (Pham Van Tai (2019)). minimizing the weaknesses, taking advantage of
Relating to the regional opportunity, it can be described opportunities, and eliminating or reducing threats.
as follows. Table 1. Matrix of Diagnosis SWOT
O1. National policy support and hinterland economic
O1 O2 T1 T2
development open a great opportunity for MKD to develop.
The logistics industry has great potential to be developed in S1 OFFENSIVE DEFFENSIVE
Vietnam. During the last 10 years, the LPI rank of Vietnam S2
has been developed a lot under the evaluation of the World
Bank. In 2018, Vietnam jumped 25 spots in the World W1 ADAPTIVE SURVIVAL
Banks’s Logistics Performance Index 2018, ranking W2
39th among 160 countries.
O2. [12] determines MKD opportunities relating to the Strategies for MKD logistics development have also been
typical geographic locations in Southeast Asia. However, the revealed from literature review and expert interview. They
opportunity can be enhanced just after the implementation of are described in the following section.
Kra channel projects. Currently, the ideal location of MKD is
From GIZ (2018), MKD may seize opportunities and
competed by Cai Mep Thi Vai and ports in Ho Chi Minh
address challenges by (i) Strengthening the coastal area of
city.
the Mekong Delta and enhancing the adaptability of
O3. Being located close to the southern key economic citizens; (ii) Supporting Vietnamese agencies to make
region is a weakness, but also a strength of the region. coastal and inland areas ready to respond to environmental
Traffic congestion, especially around such port areas as Cat
changes; and (iii) Identifying pathways for sustainable
Lai. Currently, limited expansion opportunities in Ho Chi
development and climate change adaptation.
Minh City may create opportunities for logistics industry
development in the Mekong Delta. Besides land availability, Based on reviewing strategies for developing a national
the Mekong Delta has the advantage of developing the logistics system (GIZ 2019), as well as recommendations to
logistics industry to support the southern key economic improve Vietnam logistics system (WB (2019)), the
region. research has determined such logistics infrastructure
Last but not least, the main threat of MKD logistics strategies to apply in the MKD region, aiming at different
system is the competition from neighboring and international targets:
regions. (1) For Integration with Global Markets:
T1. Pham Van Dang (2019) says that MKD agriculture is i. Re-orient transport and spatial planning to
facing more and more competitive pressure in export and the support critical value chains
risk from foreign competitors with advanced equipment, ii. Reconfigure the network of international
technology, and available experience in services and gateways
logistics. That will be huge challenges for agricultural and iii. Create “economic densities” along new corridors
fishery production enterprises in the Mekong Delta that are
(2) For Integration across Domestic Markets
still using outdated production lines, poor logistics activities,
with slow development, leading to big problems in the i. Upgrade connectivity “software” to serve
storing, preserving, transporting, and consuming goods. domestic markets
Therefore, logistics service development solutions will bring ii. Overhaul market infrastructure and logistics in
high efficiency in reducing costs, saving time and costs and cities
improving product quality. (3) For Resilience
T2. Currently, most of the value-added services (e.g. i. Promote multimodal transport as a resilient
warehousing, containers, inspection, etc.) are concentrated in strategy
283
From that, strategies for logistics infrastructure criteria groups. This survey aims at experts on the relative
development in MKD can be extracted and arranged in the importance of the assessment criteria.
matrix as listed below: Dealing with the problem’s complexity in the logistics
S-O strategy: industry and logistics infrastructure, experts with good
knowledge in the industry were chosen to conduct
SO1. Hub development is very essential to develop the
interviews. 30 experts selected have answered research
integrating logistics centers and satellite hubs for MKD, questionnaires. They come from (i) freight transport
connecting with ICD, IWT/seaports (Can Tho, My Thoi), companies and LSPs; (ii) shippers who are
airports, stations, industrial zones, border economic zones agriculture/aquaculture companies; and (iii) logistics
(in Kien Giang, An Giang) in order to take full use of IWT management experts who have been working as
network to increase cargo volumes through transportation transport/logistics authorities for more than five years. These
terminals. This strategy may also encourage enterprises of people also have a wide range of experiences in developing
all transportation modes, gathering cargo and improve logistics/infrastructure system in Vietnam big cities. The
transportation efficiency. questionnaire was designed to obtain experts’ opinions about
SO2. Based on Luis (2014), to make IWT system become the importance of assessment criteria for the logistics
an option with high competitiveness to transport products infrastructure strategies in MKD.
with high value (rather than bulks), it is required to build a Resulting from the analysis, the assessment criterion of
reliable warehouse and logistics network, with high Effectiveness was evaluated by calculating the expected
competitiveness in delivery time. The strategy may be met influences of logistics infrastructure solutions aiming at four
only with frequent maintenance works. objectives. These objectives are the freight transport/logistics
development, enhancing economics efficiency, increasing
SO3. Promoting and mobilizing investment, e.g. the traffic safety, and environmental protection. In a similar
development of a port, IWT, road infrastructure, and manner, the criterion of Applicability is indirectly assessed
logistics facilities in the form of PPP by estimating difficulties in implementing logistics strategies
S-T strategy: in the MKD region. From the analysis, financing for the
ST1. Developing the system of ICD, ports and enhancing measures and public acceptance are determined as the main
logistics facilities difficulties. Financing for the implementation of an
intervention represent the affordability to the state or local
ST2. Enhancing regional connectivity and logistics provinces of implementing measure in practice.
connection, focusing on multimodal transport and terminals,
After collecting opinions and assessments of experts on
such as the IWT network, key seaports. Promoting logistics strategies in the MKD region, results were coded
sustainable and smart technology in agriculture and and encoded using Excel software. The comparison matrix is
aquaculture. developed with the measurement scale of Linkert. The
W-O strategy resulting relative score of importance may then be
WO1. Positioning air logistics for high-value products determined for further analysis.
WO2. Concentrating all resources to complete The estimation results of effectiveness and adaptability
transport/logistics infrastructure connecting to seaport criteria provided the formation of priority classes of logistics
strategies. The following table gives the detailed results of
W-T strategy the qualitative assessment of logistics strategies for the
WT1. Enhancing human resources for professional MKD.
logistics services. Table 2. Qualitative assessment of defined strategies
WT2. The government intervenes with other regions to
help more experience and create competitive advantages for Solutions Effectiveness Applicability Priority
criteria criteria Ranks
MKD infrastructure.
SO1 2.3 2.1 1
IV. PRIORITY RANKING OVER MKD STRATEGIES SO2 1.6 1.7 4
The research applies the project of Multi-Criteria SO3 2.1 1.8 2
Assessment (MCA) to evaluate strategies to develop MKD ST1 1.3 1.1 9
infrastructure. A literature review revealed that MCA allows ST2 2.1 1.7 3
screening policy evaluations under considerations such WO1 1.7 1.5 5
aspects of socio-economics, mobility, traffic safety, and
WO2 1.7 1.2 6
environments, which ultimately provides an overall
comparison among logistics strategies. In addition to this, the WT1 1.1 1.4 8
involvement of multiple stakeholders can increase the WT2 1.5 1.3 7
acceptance of assessment results.
Two groups of criteria have been applied for evaluating From the calculating results, the most likely expected
and ranking the solutions for logistics infrastructure impacts of strategies are SO1 “Establishing hubs”, SO3
development in MKD. They are criteria of Effectiveness and “Promoting and mobilizing investment into infrastructures”
Applicability (or Barrier). Effectiveness represents the and ST1 “Developing ICD, ports and IWT facilities”. The
expected impacts and Applicability represents the main other strategies should follow as they may be effective only
barriers in implementation of these solutions. An expert in the long term or difficult to implement in the next coming
survey has been conducted to evaluate the importance of two time.
284
V. CONCLUSION [4] Luis C. Blancas and M. Baher El-Hifnawi (2014). Facilitating

Trade through Competitive, Low-Carbon Transport: The Case
The paper has taken Mekong Delta as a case study to for Vietnam’s Inland and Coastal Waterways, World Bank.
apply methodologies of SWOT and MCA qualitatively and DOI:10.1596/978-1-4648-0105-1.
quantitatively at the same time. [5] L.J. Peng, X.X. Wang, Y. Deng et al. (2014). “SWOT
analysis method and its application in the case of competitive
Though the region has big potential in production and intelligence research”, Intell. Mag. vol. 7, pp. 15-17, 2014.
export, the ocean transportation and logistics system with a [6] GIZ Report (2018) Strategic Program of comprehensive
lack of supporting infrastructure on the landside (e.g. management of coastal area and salinity forest, Golden Sky
warehousing and depot facilities) are now hampering the Co., Ltd.
growth of efficient logistics practices in the region. [7] GIZ report (2019). Research on orientation strategy to
develop logistics services in period 2021-2030
The study brings a better understanding of the
[8] Grafakos et al. (2010). Multi-criteria analysis weighting
complicated solutions for MKD logistics infrastructure from methodology to incorporate stakeholders’ preferences in
different stakeholders’ perspectives, then finally reveal the energy and climate policy interactions, International Journal
best logistics strategy. The optimal logistical strategy in of Energy Sector Management, September 2010, DOI:
MKD selected shows the fact that MKD disadvantages were 10.1108/17506221011073851,
bigger than its advantages. https://www.researchgate.net/publication/233728578
[9] Nguyễn Thị Tuyết Nga (2019). Solutions for logistics
In summary, the research applied in the paper can be business in Mekong Delta, Journal of transport science and
considered as the initial stage in supporting the decision- technology, No. 33-08/2019.
making process for logistics infrastructure development [10] Phạm Văn Đang (2019). Logistics solutions for aqua-culture
strategy in Vietnam. In such developing countries as industry in Mekong Delta. Journal of transport science and
Vietnam, a better understanding and assessing the impacts of technology, No. 33-08/2019
policymakers’ decisions on logistics systems may greatly [11] Phạm Văn Tài (2019). Developing logistics to support import-
export activities of Mekong River Delta area, Journal of
contribute to sustainable development. transport science and technology, No. 33-08/2019, pp. 65-71
ACKNOWLEDGMENT [12] Talancón, H. P. (2006). La matriz FODA: Una alternativa
para realizar diagnósticos y determinar estrategias de
The authors are grateful to Vietnam Ministry of Science intervención en las organizaciones productivas y sociales.
and Technology and German Federal Ministry of Education Contribuciones a La Economía, (2006-09)
and Research for their generous financial support, under the [13] Institute of South Planning and Construction (2016). Report
Joint Project of For_Mobility (No. NDT/DE/21/30). of Construction Masterplan in MKD region until 2030, vision
to 2050, NXB Ministry of Construction
REFERENCES [14] UN (2017). The National Action Plan for the implementation
of the 2030 sustainable development agenda,
[1] Blancas Mendivil, Luis C.; Isbell, John; Isbell. (2014). https://vietnam.un.org/index.php/en/4123-national-action-
Efficient logistics: a key to Vietnam's competitiveness, plan-implementation-2030-sustainable-development-agenda
Washington, D.C., World Bank Group, DOI: 10.1596/978-1- [15] World Bank (2019). Vietnam Development Report 2019,
4648-0103-7 https://openknowledge.worldbank.org/handle/10986/33139
[2] Han Entzinger và Peter Scholten (2016). Adaptation with [16] Yin Yin Lam, Kaushik Sriram, and Navdha Khera (2019).
climate change by migration: A case study of MKD, Strengthening Vietnam’s Trucking Sector: Towards Lower
International Organization of Migration (IOM), Logistics Costs and Greenhouse Gas Emissions, World Bank.
https://publications.iom.int/system/files/pdf/vietnam_survey_r Publishing and Knowledge Division.
eport_vn_0.pdf
[3] Sustainable Development of Inland Waterways Transport in
Vietnam: Strengthening the Regulatory, Institutional and
Funding Frameworks, WB. Publishing and Knowledge
Division.
285
Study on Behaviour of Short Pile Groups in Soft

Ground with Sand Leveling on a Small Scale Model
Using Schneebeli Analogue Soil
Sy Hung Nguyen Thi Phuong Huyen Tran
Department of Transport Engineering, Faculty of Civil Department of Civil Engineering,
Engineering The University of Danang - University of Technology and
HCMC University of Technology and Education Education
Ho Chi Minh City, Viet Nam Da Nang city, Viet Nam
sihung.nguyen@hcmute.edu.vn huyen.tran158@gmail.com; ttphuyen@ute.udn.vn
https://orcid.org/0000-0002-9164-9109
Abstract: This paper focuses on the behaviour of pile cause the shallow foundation to settle or cause negative
groups on soft ground with sand leveling on a 2D small scale friction to pile foundations [2, 3]. However, if the ground level
model using analogue soil of Schneebeli. The Mekong Delta has been leveled for a long time and the residual consolidated
has many areas where low-rise buildings are built on soft settlement is negligible, the upper-level sand layer will
ground with a sand backfill layer of thickness between 2 - 5m. become a solid layer to settle the foundation.
The option of using small (section smaller than 30×30 cm) Recently, short piles with a diameter of 200 to 250 mm and
and short (length shorter than 5m) piles driven into the sandy a length of 2 to 4 meters plugged into the leveling sand layer
soil layer can be applied to low-rise buildings with the above have been used as a foundation solution in building low-rise
geological conditions. In the small-scale model, we use buildings. However, little work has investigated and
Schneebeli analogue soil to replace the leveling sand and demonstrated clearly this solution. The question is how the
foam instead of the weak clay layer. Short reinforced concrete pile spacing in the pile’s group and the distance from the tip
piles are replaced by steel bars with a square cross-section. of the pile to the soft soil layer affect the soft soil layer.
The vertical loading tests were carried out on 3-piles groups. Group action in piles foundations has resulted in many
Experimental parameters are changed in the test program, recorded failure cases or excessive settlement, even though
including the pile spacing s, and the distance from the tip of loading tests made on a single pile have indicated the
the pile to the boundary of the two soils, H, whereas the satisfactory performance. A typical case of foundation failure
thickness of the soil layers is unchanged. When the pile is the single pile driven to a satisfactory set in a compact or
spacing s is unchanged and the distance from the tip of the stiff soil layer underlain by soft compressible clay. The latter
pile to the boundary of the two soils, H decreases, the formation is not stressed to any significant extent when the
deformed soil area under the pile tip decreases. In the case single pile is loaded, but when the load from the superstructure
that H is fixed and s is increased, the failure of soil under the is applied to the whole group, the stressed zone extends down
pile tips becomes more and more localized. into the soft clay. Excessive settlement or complete general
shear failure of the group can then occur [4].
Keywords: Schneebeli soil analogue, pile group, vertical
Analytical methods on pile group settlement and the
loading test, sand leveling, weak soil
concept of ‘Interaction effect’ in pile groups were introduced
I. INTRODUCTION and pioneered by Poulos [5]. Poulos studied the behaviour of
pile groups, the interaction between two piles, and the
A. Short pile groups in soft ground with sand leveling superposition of the individual effects of adjacent piles in the
Demand for low-rise houses of four floors or less in the groups in consideration of the influence of parameters such as
Mekong River Delta is very high. However, the surface soil type of pile group, L/d ratio, the relative depth h/L, Poisson’s
layer is very weak, and its thickness can be up to 15 m, even ratio, and spacing s/d (where L is the pile length, d is the pile
more than 30m [1]; the construction conditions are diameter, h is the distance from the ground to the hard rock, s
complicated, and the foundation construction becomes costly. is spacing between piles). Some researchers conducted load
In the case of construction sites having filled sand above soft tests on pile groups in the field to study the pile group
soil, the foundation may suffer from the excessive settlement efficiency in clay or sand [6], [7]. Others studied pile groups
due to consolidation, which causes construction deterioration. on small physical models [8-12]. However, the mechanism of
pile-soil interaction under load has not been elucidated,
Using sand to level on soft soil has both advantages and especially in the case that the pile is located in a good soil
disadvantages in construction. The sand layer will act as a layer, but the underneath is a soft soil layer.
form of load and a drainage boundary, making the soft soil
layer below consolidated. The settlement of sand leveling To study the interaction mechanism of the pile group-soil,
areas after the first two years can reach from a few dozen we can study the deformation of the ground under the action
centimeters to more than 100 centimeters, depending on the of the force acting on the pile group. Ekisar and Otani used X-
thickness of the leveling. The consolidation of soft soil may rays CT to study the phenomenon of soil arching on the
reinforced embankment with rigid foundation piles [12].

White and Bolton used the Digital Image Correlation method friction angle is 210, the bi-dimensional Poisson coefficient is
to study displacement and strain paths during plane-strain 0.9, and the deformation modulus is E=107500()0,5.
model pile installation in the sand. They showed how shear
strain developed at the pile tip during pile driving [13]. Dolzhenko [25] performed biaxial compression
Towhata and Aoyama used the visualizing test & image experiments on the same material as in [20]. The author
analysis (PIV) combined with tactile sensors and acoustic obtained an internal friction angle of 240. Dolzhenko [21] also
emission (AE) to study how the ground deformed in the executed experiments to measure the static pressure
subsurface domain where the significant effects of the coefficient K0, which fluctuates in the range of 0.65 to 0.74.
interaction occurred on the load -settlement relationship in the By counting the bars per unit area, the author proposes the
pile group [15]. The above studies were conducted in void ratio of the analogue soil as 0.21 [21].
laboratory experiments and using natural soil. However, these The Schneebeli model is widely used because of the
methods require complex technologies and techniques. following advantages [26]:
Another approach is to use the finite element method. Ju The Schneebeli analogue soil exhibits volume expansion
[16] carried out settlement analysis using PLAXIS 3D finite under deformation, similar to compact sandy soils; The
element package for pile-group located in speech strata. deformation modulus of the soil depends on the stress
Fuchun et al. implemented settlement analysis of pile groups intensity; Analogue soil is heavier than natural soils with an
using the PIGLET finite element package [17]. Gowthaman average density of 6.2. That allows a size reduction of 1/3 with
and Nasvi used PLAXIS to model single and group piles in the stress state remaining in the model; Easy simulation
silty sand layers with different compactions. As the repeatability of experiments is guaranteed; The model does
nonlinearity affects the interaction of the pile group, LE and not need surface support (only a 4-corner bracket is needed)
MC models fail to predict the behaviour of the pile group. The because the bars are self-leveling, unlike when using real sand
soil nonlinearity and the interaction effects are clearly and avoiding the contact edge effect. The model is
captured in the HS model the and incorporation of both straightforward to process the image and measure the strain
nonlinear effects leads to the real prediction of the settlement field by griding squares or coloring the bars.
behaviour of piles group [18]. However, reliable settlement
prediction of pile foundations at typical working loads However, the Schneebeli model still exists some
remains one of the major geotechnical engineering problems disadvantages [26]:
[16], [19]. Besides, the selection of suitable material models The friction angle is relatively small, only from 20 to 250,
and modeling parameters based on the geological conditions while the friction angle of compact sand is 30-400; It cannot
and loading conditions still remains a huge challenge in simulate 3D problems, for example, the problem of the
numerical simulations of settlement behaviour of pile arching effect on the pile head.
foundations [18].
Analogue material Schneebeli is used to simulate
In this study, the authors studied the pile group-soil cohesionless soils. Many authors have used it to understand
interaction, the deformation of the ground under the action of and simplify soil behaviour from 3D to 2D. Kastner used a
the pile group by using the Schneebeli analogue material Schneebeli model for deep excavation in urban [24], Marsouri
instead of sandy soil and foam instead of clayed soil, in a [26] used it to simulate semi-rigid earth retaining wall, Al
small, simple physical 2D model. Abram [28] and Dolzhenko [2] used it to study the stress field
B. Schneebeli analogue material when digging Tunel. Amina (1996) used the Schneebeli
model to study soil reinforcement with geotextiles. Caudron
The Schneebeli analogue material was introduced in 1956 [29] studied the collapse of underground tunnels and the
by Schneebeli [20]. This material is constructed of round impact on the structure above [30]. A. B. Bolt [31] also used
stainless steel bars of 60mm length. These steel bars include the Taylor-Schneebeli model to simulate the collapse of earth
three diameters 3mm, 4mm, and 5mm with volume ratios of retaining walls depending on boundary conditions.
34%, 35%, and 31%, respectively. The density of the material
fluctuates between 6.42 and 6.53 [21]. In the study on soft soil improvement by vertical rigid piles
by two-dimensional physical modeling, Jenck [32] used
Schnebeeli bars can also be made from other materials Schneebelli, foam, and steel elements instead of cohesionless
such as wood, plastic, or aluminum [22] or round steel bars soils, the soft soil, and the rigid piles, respectively [32].
with chamfered edges [23]. The diameter of the bars can vary
from 2mm to 20mm. For different materials, the density of the II. SMALL TWO-DIMENSIONAL MODEL
material and the angle of internal friction will be different
depending on the modelization purpose. It is found that the C. Analogue materials
model using Schneebeli material is very suitable to study This study uses analogue material Schneebeli to simulate
cohesionless soil behaviour in terms of macro-mechanics as the backfill sand layer. The foam material is used to replace
well as in micromechanics aspect. the weak clay layer.
Schneebeli [20] performed biaxial compression tests on In order not to have to redefine the parameters of materials,
aluminum alloy Schneebeli bars with smooth and roughened the authors used the common Schneebeli material as in the
surfaces. The author found that the internal friction angle study of Dolzhenko [21]. In which the Schneebeli bars are
ranges from 240 to 260 with smooth surfaces, whereas with made of stainless-steel material, there are three diameters of
rough surfaces, the friction angle is 34 to 350. 3, 4, and 5 mm. The length of each bar is 6mm. The weight
ratio is 34, 35, and 31% corresponding to each of the above
Kastner [24] tested on stainless steel plated steel
diameters (Fig.1). Parameters of Schneebeli analogue
Schneebeli rods in a large triaxial compressor with high
material, taken from previous studies [21], [25].
chamber pressure (up to 500 kPa). The measured internal
287
To consider the cohesion in the case of cohesive soils, modeling can be found in the literature. Scale models allow
different approaches have been employed. Ghabraie et al. [33] for fast, multiple reproductions of real works, and can be
used a sand-plaster mixture to simulate the bedding planes and driven to the failure point. The choice of material properties is
the strata to study longwall excavation. He et al. [34] used more extensive and it is easy to vary different parameters in
gypsum to model the behaviour of the ground strata to order to study their influence.
simulate the behaviour of mine roadways. Chemenda et al.
[35] used liquid and solid hydrocarbons to study the landslide One of the difficulties is to satisfy the similar conditions
and slope stability in three dimensions. as reality. To be transposable to a full-scale problem, a scale
model must obey the laws of similarity, which guarantee the
similarity of the mechanical behaviour between the reduced
model and its associated prototype. These laws make it
possible to apply the solution obtained on a model to a real-
size problem. The establishment of the laws of similarity is
based on a dimensional analysis by considering the invariance
of the general equations of mechanics with the change of units.
These equations are the general dynamics equations, the mass
conservation equation, and the laws of behaviour of the
Figure 1. Image of Schneebeli analogue material different materials [40].
In 2D models, the cohesive soil in the model can be
simulated by foam [32], rubber foam [36, 37], and peat [38].
In this study, foam material replaced soft soil, which has the
parameters shown in Table 1.
Table 1. Parameters of foam material
Tested
Contents Unit Test Results
according
Density DIN 53420 Kg / m3 40
Compressive stress
DIN 53421 N / mm2 0.24 – 0.29
at 10% strain
Flexural strength DIN 53423 N / mm2 0.46 – 0.65

Figure 2. Schematic view of the test apparatus. 1: Foam
layer; 2: Schneebeli analog material layer; 3: Steel piles; 4:
Shear strength DIN 53427 N / mm2 0.27 – 0.32 Guide bars system; 5: Frame and loads; 6: dial gages; 7:
Model’s frame.
Tensile strength DIN 53430 N / mm2 0.42 – 0.59
James [40] distinguishes two types of physical modeling
Modul of elasticity
in soil mechanics:
DIN 53457 N / mm2 10.1 – 13.5
in compression 1) The scale model is itself a prototype; its behavior is
confronted with analysis methods with a view to improving
D. Dimensions of model these methods. The results obtained in these experiments are
The dimension of the model's frame is 1.0 m long, 70 cm generally not applicable for the design of real structures but
high, and 6 cm wide. The thickness of the foam layer is 30 cm. are of great importance for establishing design principles. The
The thickness of the layer of Schneebeli analogue material is rules of similarity are then not of prime importance (Roscoe,
30 cm. In the model, hollow steel piles with a square cross- 1968).
section of 3x3cm replaced the reinforced concrete short piles.
The pile group includes three piles (Fig. 2). 2) The reduced model satisfies the principles of
similarity, and it makes it possible to predict the behavior of
The piles are connected to the loading beam with sufficient the real prototype structure. Centrifuge modeling is often
rigidity to be considered as not deforming during loading. adopted because it allows the same level of stress to be
Loads are steel weights stacked on the loading beam. On the maintained. The same loading paths of the structure are then
surface of the layers of foam and Schneebeli material, a 5x5cm simulated.
grid is drawn to facilitate the observation of the strain field.
In our study, we are in the first category of models defined
E. Scale of model by James (1972) [40]. The proposed reduced model does not
We use small physical models, which should make it strictly respect the rules of similarity, especially more than
possible to quantitatively predict the actual structure's analogue materials are implemented. The purpose of this
behaviour while being aware of the limits to apply to results. model is not to quantitatively reproduce the behaviour of a real
The recourse to analogical methods, essentially in similarity structure but to constitute an experimental database that will
by reduced models, consists in reproducing the structure and be compared with the dimensioning methods and which will
its boundary conditions, the loading system, and the laws of serve as a reference for modeling approaches. However, the
behaviour of the prototype while respecting the conditions of analogue material made up of steel rolls is heavier than
similarity [39]. The advantages and disadvantages of physical conventional soils; its density is 62kN/m3, limiting the
288
distortion of similarity between model and prototype because with pile spacing of 6D and 9D, the displacement line has
the level of constraint is more or less maintained for a three vertices corresponding to the position of the three pile
downscaling of 3. tips. The displacement below the middle pile top is slightly
larger than that of the two sides. Thus, in the case of 3D pile
F. Testing program spacing, the pile group works as a large diameter pile, with an
The experimental program includes 18 experiments. In evident stress concentration phenomenon. Conversely, the
which the piles spacing s changes in 3 values 3D, 6D, and 9D, piles tend to work independently when the pile spacing s=3D
where D is the pile diameter. The distance H from the tip of and s=6D.
the piles to the junction of two layers of foam material –
Schneebeli is 5cm, 10cm, and 15cm. For each pair of values With the same load and pile length (i.e., distance H), the
(S, H), two experiments were conducted. maximum displacement will decrease as the distance between
piles increases. However, as H increases, the difference in the
Table 2. Parameters of real and model largest displacements narrows significantly. For example, the
maximum displacement difference between the 3D and 9D
Parameter
Real Model pile spacing when the load is 90 kg is 100% when H=5cm and
only 40% when H=15cm.
Thickness of sand layer/
2.5÷4 m 30 cm s= 3D, P=60kg s= 3D, P=75kg s= 3D, P =90kg
Analog material Schneebeli
Thickness of clay layer/
10÷50 m 30 cm
foam layer
Section of piles 20x20÷25x25 cm 3x3cm
Length of piles 2÷3,5 m 15÷25 cm
L/D ratio 8 ÷ 17,5 5÷8 S= 6D, P=60kg S= 6D, P =75kg S= 6D, P =90kg
For testing preparation, the foam layer is first put in place.

Next, the piles are fixed to the original position with (S, H)
determined. Then, Schneebeli bars are stacked by hand on top
of the foam layer.
Loading is carried out by level. The vertical displacement S= 9D, P =60kg S= 9D, P =75kg S= 9D, P =90kg
of the piles is measured by two dial gages, and then the
average value is taken. Both differential displacements of the
material surface outside the boundary piles and at the interval
between the piles a, and the deformation contour of the foam
layer surface are measured with a steel ruler. The error is ± 0.5
mm.
III. EXPERIMENT RESULTS
Figure 3. Images of the model with H= 5cm; s, and P vary
G. Load-settlement curve
s= 3D, P=60kg s= 3D, P=75kg s= 3D, P=90kg
The images of the experiments, when H=5cm and
H=15cm, pile spacing s=3D, 6D, and 9D, at 60, 75, and 90 kg
of load, are shown in Fig. 3, 4. The curves of load-settlement
are shown in Fig. 5.
If s is constant, with the same applied load, the pile
displacement decreases as the pile length increases, i.e., with
H decreasing (Fig. 5). That can be explained by the reason that s= 6D, P=60kg s= 6D, P=75kg s= 6D, P=90kg
as the pile gets longer, the pile bearing capacity increases
because the friction force increases, and the friction force
plays a dominant role in the pile load capacity.
If H is constant, displacement of piles with the same load
will decrease as the distance between piles increases.
However, the difference in displacement between the two
s= 9D, P=60kg s= 9D, P=75kg s= 9D, P=90kg
cases s=6D and s=9D is relatively small. Thus, the pile group
efficiency increases as s increases.
H. Deformation contour of the foam layer surface
The deformation contour of the foam surface
corresponding to different pile spacings and pile lengths at the
90 kg load class is shown in Fig. 6.
When the pile spacing is 3D, the displacement line is a
concave curve with a single vertex corresponding to the
maximum displacement at the central pile position. While Figure 4. Images of the model with H= 15cm; s, and P vary
289
Considering the case of H = 15 and s = 6D or s = 9D, at I. Difference in the surface displacement of the Schneebeli
loads of 60 – 90 kG, the horizontal lines in Schneebeli material layer inside and outside the pile group a
below the pile tips and the surface of the foam layer turn into The difference in the surface displacement of the
convex curves with three vertexes under three pile tips. With
Schneebeli layer inside and outside the pile group a at P=90
s = 3D, the horizontal lines under the pile tip located in the
kG with s an H vary is shown in Fig. 8. If H is fixed, the
Schneebeli material have the same shape above. However, the
foam layer surface is deformed into a convex curve with only displacement a decreases when the pile spacing increases.
a vertex below the center pile. This phenomenon is very With H=5cm, when s changes in 3D, 6D, 9D, then a is 7cm,
similar to the results of Ikuo Towhata and al. [15] and 3cm, and 2cm, respectively. Thus, the difference in a 3D and
Gowthaman S. [18]. Thus, when s = 3D, the material below 6D or 9D is very significant, while this difference is relatively
the pile tip is compacted by the superimposition of action of small between 6D and 9D. a when s=3D decreases very
the piles. quickly as H increases, while for s=6D and s=9D, this
reduction is not large (Fig. 8). Thus, the impact of the pile
5 cm, s = 3D, 6D, 9D
3D, H=5, 10, 15 cm
Axial load (kg) Axial load (kg) group on the soft soil layer will decrease as s and H increase.
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
0 0
20 20 P=0 P = 60 kG P = 90 kG
40 40
60 60
80 80
Settlement (mm)
Settlement (mm)
100 100
3D-5 120
120 9D-5
3D -10 140
140 6D-5
3D-15 160
160 3D-5
180 180
200 200
220 220
240 240
260 260
3D, H=5, 10, 15 cm 10 cm, s = 3D, 6D, 9D

Axial load (kg) Axial load (kg)
0
0 20 40 60 80 100 120 140
0
0 20 40 60 80 100 120 140
Horizontal lines deform under loading with H=15 cm and s=3D.
20 20
40 40
60 60
80 80
Settlement (mm)
Settlement (mm)
100 100
120 120 9D-10

6D-5 140 6D-10
140
6D-10 160 3D-10
160
6D -15 180
180
200 200
220 220
240 240
260 260
9D, H=5, 10, 15 cm 15 cm, s = 3D, 6D, 9D

Axial load (kg) Axial load (kg)
0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140
0 0
20 20
40 40
60 60
80 80
Settlement (mm)
Settlement (mm)
100 100
120 9D-5 120
140 9D-10 140
9D-15
6D-15
Figurea at P=90 kG with s and H vary
160
9D-15 160 3D-15
180 180
200 200 IV. CONCLUSION
220 220
240
260
240
260
A relatively simple small physical model, using Schnibeeli
analogue material to replace sandy soils and foam to replace
soft clayed soils, has helped to study the behaviour and soil-
Figure 5. Load – Settlement curves pile group interaction in the case of short piles located in the
sandy soil layer above a weak clay layer. A loading test
program with three H values and three pile spacings s=3D, 6D,
and 9D shows that the load capacity of the pile group increases
as the pile length increases (H decreases) and s decreases.
By plotting on the surface of Schneebeli and foam material
and observing, then analyzing the shape change of the grid
H = 10 cm X (mm) during the loading process, the study has shown the pile-soil
-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35
0 group interaction. In the cases of s= 6D and 9D, the piles tend
2 to work independently, which is manifested by the local
4 3D-10-90 deformation of the horizontal lines around the piles and under
6 Y (mm) 6D-10-90 the pile tips, including the surface or foam layer. However,
8 9D-10-90 with s=3D, the local deformation under the pile tips only
X (mm)
-35 -30 -25 -20 -15 -10
H = 15 cm
-5 0 5 10 15 20 25 30 35
occurs in the part of the Schneebeli material close to the pile
0 tips. Meanwhile, the entire Schneebeli material under the pile
2 tips is compacted, and the foam surface is deformed into a
4 3D-15-90 concave curve with the top lying below the center pile tip.
6 6D-15-90
8 Y (mm) 9D-15-90 By comparing the maximum displacement of the surface
of the foam layer, it shows that the smaller spacing s and the
shorter distance from the tip of piles to the clay layer H, the
Figure 6. Curves of deformation contour of the foam surface
at P = 90 kG
more significant the impact of the pile group on the clay layer.
290
The results in this study show compatibility with many [20] Dolzhenko Nataliya, “Etude expérimentale et numérique de modèle
research results in the form of analysis, experimental, or réduit bidimensionnel du creusement d'un tunnel”. Diss. Lyon, INSA,
2002.
simulation methods. Therefore, using Schneebeli and foam
[21] R. Gourves and F. Mezghani, “Micromécanique des milieux
materials as analogue materials for cohesionless and clayed granulaires approche expérimentale utilisant le modèle de
soils to simulate foundation geotechnical problems is a Schneebeli”, Revue Française de Géotechnique 42 (1988): 23-34.
potential method. [22] B. Terrade, A. S. Colas, and D. Garnier. "Assessing and repairing
masonry retaining structures by yield design analysis and experimental
REFERENCES modelling." Structural Analysis of Historical Constructions:
Anamnesis, Diagnosis, Therapy, Controls. CRC Press, 2016. 1407-
[1] Nguyen Uyen, Soil reinforcing in civil construction. Construction
1414.
publishing house, Ha Noi, 2009.
[2] V. C. Trinh, “Negative friction on piles and its effect on construction”, [23] R. Kastner, “Excavations profondes en site urbain”. Thèse Génie Civil,
Lyon: INSA de Lyon, 1982; 409p
in The 3rd National Science Conference on Construction Incidents and
Damages, 2005. [24] N. Dolzhenko, P. Mathieu, R. Kastner, “Modélisation expérimentale et
numérique bidimensionelle d’ouvrages souterrains urbains. Réalisation
[3] D. T. Duong, Q. H. Pham, “Testing the neutral surface method in
d’essais biaxiaux et oedométrique sur le sol analogique de Taylor-
calculating settlement of pile groups”, Journal of Science and
Technology in Civil Engineering (STCE) vol. 23(3), pp. 62-68, 2015. Schneebeli”, In : XVIIIème Rencontres de l’AUGC à Lyon,
«Aménagement urbain», 2000, 27-28 mai, p 281-288
[4] M. Tomlinson, J. Woodward, “Pile design and construction practice”,
Fifth edition, Taylor & Francis, 2008. [25] Jenck, Orianne, “Le renforcement des sols compressibles par
inclusions rigides verticales. Modélisation physique et numérique”,
[5] H.G. Poulos, “Analysis of the settlement of pile groups”, Geotechnique Diss. INSA de Lyon, 2005.
4: 449-471. 1968.
[26] F. Masrouri, “Comportement des rideaux de soutènement semi-
[6] A. S. Vesic, “Experiments with Instrumented Pile Groups in Sand, flexibles”, Th: Génie Civil, Lyon: INSA de Lyon: 1986; 247p
Performance of Deep Foundations”, ASTM STP444, 177-222, 1969.
[27] Al Abram, Ismail, “Etude sur modèle réduit bidimensionnel du champ
[7] Yu, Ching-Han, “On Design and Construction of Pile Group de déplacement induit par le creusement d'un tunnel à faible
Foundation of Taipei 101”, Geotechnical Engineering Journal of the profondeur: interaction avec les ouvrages existants”, Diss. Lyon,
SEAGS & AGSSEA, Vol.42 No.2, pp.56-69, 2011. INSA, 1998.
S. H. Nguyen, T. P. H. Tran, and Q. M. Pham, “Experimental Study on [28] A. Benrabah, J. Gielly, “Behavioural Study of a Reinforced Analogical
Pile Group Efficiency in Various Types of Soil Using a Small–Scale Soil Under External Loads”, Geotextiles and Geomembranes 14 (1996)
Physical Model”, CIGOS 2021, Emerging Technologies and 43-56
Applications for Green Infrastructure, Lecture Notes in Civil
Engineering 203, 2022. https://doi.org/10.1007/978-981-16-7160- [29] Caudron, Matthieu, “Etude expérimentale et numérique de l'interaction
9_108 sol-structure lors de l'occurence d'un fontis”, Diss. Lyon, INSA, 2007.
[8] B. V. H. Lan, “Research on the group effect to the capacity and
[30] A. F. Bolt, E. Dembicki, and G. A. Horodecki. "The failure mechanism
of a trench wall caused by additional soil pressure." Geotechnical
settlement of vertical piles group”, Thesis of doctor. Southern institute
aspects of underground construction in soft ground. 1996. P.63-67
of water resources research, 2017.
[31] O. Jenck, D. Dias, R. Kastner, “Soft ground improvement by vertical
[9] Nihar Gogoi, “A Model Study of Micropile Group Efficiency under
rigid piles two-dimensional physical modelling and comparison with
Axial Loading Condition”, International Journal of Civil Engineering
Research. ISSN 2278-3652 Vol. 5, Number 4, pp. 323-332, 2014. current design methods”, Soils and Foundations. Volume 45, Issue 6,
December 2005, Pages 15-30
[10] C. Harish, M. Manjunatha, “Experimental study on load settlement
behaviors of micro pile under vertical loading”, International Research [32] B. Ghabraie, R. Gang, and John V. Smith. "Characterising the multi-
Journal of Engineering and Technology (IRJET), Vol. 03, Issue: 07, seam subsidence due to varying mining configuration, insights from
2016. physical modelling." International Journal of Rock Mechanics and
Mining Sciences 93 (2017): 269-279.
[11] S. Darci, A. Soltani, and M. A. Arjomand, “Experimental
Investigations on the Efficiency Coefficient of Pile Groups”,
[33] He, Man-chao, et al. "Physical modeling of failure process of the
excavation in horizontal strata based on IR thermography." Mining
International Journal of Advanced Biotechnology and Research, Vol-
Science and Technology (China) 19.6 (2009): 689-698.
7, Special Issue3, pp. 2344-2350, 2016.
[12] T. Ekisar, J. Otani and J. Hironaka, “Visualization of soil arching on
[34] A. Chemenda, S. Bouissou, D. Bachmann, Three-dimensional physical
modeling of deep-seated landslides: new technique and first results.
reinforced embankment with rigid pile foundation using X-ray CT”,
Geotextiles and Geomembranes, Vol. 32, 44 – 54, 2012. Journal of Geophysical Research, 110 (4) (2005),
10.1029/2004JF000264
[13] D. J. White and M. D. Bolton, “Displacement and strain paths during
plane – strain model pile installation in sand”, Géotechnique, Vol. 54, [35] Hewlett, W. J., and M. F. Randolph. "Analysis of piled
No. 6, 375 – 397, 2004. embankments." International Journal of Rock Mechanics and Mining
Sciences and Geomechanics Abstracts. Vol. 25. No. 6. Elsevier
[14] Aoyama, Shogo, et al. “Application of advanced procedures to model Science, 1988.
tests on the subsoil behavior under vertical loading of group pile in
sand”, Indian Geotechnical Journal 46.1 (2016): 64-76. [36] van Eekelen, J.M. Suzanne, A. Bezuijen, and O. Oung, "Arching in
piled embankments; experiments and design calculations." BGA
[15] Ju J, “Prediction of the Settlement for the Vertically Loaded Pile Group International Conference on Foundations: Innovations, observations,
Using 3D Finite Element Analyses”, Journal of Marine Georesources design and practice: Proceedings of the international conference
& Geotechnology 33 (3): 264-271, 2013. organised by British Geotechnical Association and held in Dundee,
[16] Fuchun XUE, Jianlin MA and Liping YAN, “Three-dimensional FEM Scotland on 2–5th September 2003. Thomas Telford Publishing, 2003.
analysis of bridge pile group settlement in soft soils”. Geotechnical [37] H. G. Kempfert, D. Zaeske, and D. Alexiew, "Interactions in reinforced
Special Publication ASCE 220: 135-143. 2011. bearing layers over partial supported." Geotechnical Engineering for
[17] S. Gowthaman, and M. C. M. Nasvi, “2D and 3D Numerical Transportation Infrastructure: Theory and Practice, Planning and
Simulation of Load-Settlement Behaviour of Axially Loaded Pile Design, Construction and Maintenance: Proceedings of the Twelfth
Foundations”, American Journal of Civil Engineering and European Conference on Soil Mechanics and Geotechnical
Architecture, vol. 5, no. 5 (2017): 187-195. doi: 10.12691/ajcea-5-5-2. Engineering, Amsterdam, Netherlands, 7-10 June 1999. Taylor &
[18] B. A. McCabe and B. M. Lehane, “Behaviour of axially loaded pile Francis US, 1999.
groups Driven in clayey silt”, Journal of Geotechnical and [38] M. Al Heib, F. Emeriault, and H. L. Nghiem. "On the use of 1g physical
Environmental Engineering: 401-410, 2006. models for ground movements and soil-structure interaction
[19] G. SCHNEEBELI, “Une mécanique pour les terres sans cohésion”, problems." Journal of Rock Mechanics and Geotechnical
Compte rendus des séances de l'Académie des Sciences. Paris:1956, Engineering 12.1 (2020): 197-211.
Tome 243, pp. 2647-2673 [39] R. G. James, “Some aspects of soil mechanics model testing”. In: Parry
R. H. G. Ed. Proc. of Stress-strain behaviour of soils. Roscoe Memorial
Symposium, 1972, Cambridge, Angleterre. pp 417-440
291
Research and Preparation of Thinking Strategies for

Quality of Textile and Garment Products
Manufactured in Vietnam
Nguyen Phuoc Son Nguyen Ngoc Chau Nguyen Thi Tuyet Trinh
Faculty of Fashion and Tourist Faculty of Fashion and Tourist Faculty of Fashion and Tourist
University of Technology and University of Technology and University of Technology and
sonnp@hcmute.edu.vn chaunn@hcmute.edu.vn trinhntt@hcmute.edu.vn
Abstract: As defined by ISO-8402 standard: “Quality is  Circle B is called a set of characteristics requested by
the totality of features and characteristics of a product or customers for that product.
service that bear on its ability to satisfy stated or implied
needs”. Based on this, a system of toolsets, methods, and  The Q overlapping intersection between the two circles
quality objectives is prepared to achieve output standards of is a characteristic area of a textile and garment product
textiles & garments in strict compliance with the that meets the customer's requirements, and is called the
requirements of global buyers for garment manufacturing "quality zone".
and processing CMT (Cut, Made, Trim) or FOB (Free on  M is the area of characteristics of textile products not
Board) at Vietnamese garment companies. This paper needed by customers; N is the area in which customers
studies the proposal for the preparation of quality-defining require characteristics not fulfilled by the textile and
mathematical simulation and five quality thinking solutions: garment product.
1. Quality Check (Inspection); 2. Quality Control; 3. Quality
Assurance; 4. Quality Management; 5. Total Quality  In reality, how can we make the Q area grow and reach
Management. The outcome has contributed to cutting down an ideal level, meeting the requirements of customers.
wastefulness and cost, creating a safe and convenient That is, when Circle A approaches (nearly covers) Circle
working environment, maximizing production efficiency as B (refer to figure 1), conducting market research,
well as business revenue, and building trust with global studying customer preferences, and facilitating
buyers. production to satisfy the strict requirements of global
buyers are of utmost importance [3, 4, 5].
Keywords: quality, quality control, quality assurance,
quality management, total quality management
I. INTRODUCTION M
As defined by ISO-8402 standard: “Quality is the
totality of features and characteristics of a product or service Q
that bear on its ability to satisfy stated or implied needs”. N
Based on this, a system of toolsets, methods, and quality
A B
objectives is prepared to achieve output standards of textiles
& garments in strict compliance with the requirements of
global buyers for garment manufacturing and processing
CMT (Cut, Made, Trim) or FOB (Free on Board) at Figure 1. Describing the definition of quality of textile and
Vietnamese garment companies [1, 2]. Hence, quality is a garment products
value quantity, has no standards, is changed over time, is
subjective, and is not synonymous with "perfection". If II. QUALITY THINKING STRATEGY
quality is too low, it will not be able to meet the set A. 1st Strategy: QC – Quality Check (Inspection).
requirements; Quality too high as compared to requirements
will result in high cost and is not necessary sometimes. The When defects (mistakes) in textile and garment products
following definitions described with the mathematical Venn are found during the inline inspection, refer to figure 2.
diagram method, provide us with a proper understanding of Vietnamese textile and garment companies tend to inspect
quality. at the final stage of production only. This solution fails to
provide corrective and preventive actions and is unable to
 Circle A is called a set of available characteristics of a address the root cause of textile product defects in strict
textile and garment product. compliance with customer requirements. In our survey,
Vietnamese textile-garment companies often come up with

a solution to provide more personnel for the product Quality

inspection department at the end of the production process Control
Five (5) basic conditions
for their production lines. This is a management mistake should be controlled:
and increases cost. - Human resources
- Methods and processes
- Suppliers
- Equipment
Production - Information
inspection
Quality check
Figure 3. Quality Control

Figure 2. Quality Check
All information must be verified and approved by an
B. 2nd Strategy: QC – Quality Control. authorized person. The information must be always up-to-
date and passed on to those who are in need for the earliest
From the Japanese point of view, they demand product
use [6].
quality more rigorously - Zero defects. This shows that, in
the 1st strategy, checking the quality of textile and garment C. 3rd Strategy: QA – Quality Assurance.
products is an action that covers regulations on complex and
Quality Assurance should be based on customer trust in
large-scale production. From there, they would think of a
the suppliers, based on our good understanding of the
"preventive" solution as an alternative to a "detective" one.
organizational structure of personnel, facilities, human
Making good quality a textile and garment product is just
resources, and means of the suppliers. On the other hand,
like building a house, first of all, a solid foundation must be
suppliers must fully have reasonable evidence to prove their
built. Hence, controlling the quality of textile and garment
ability of quality assurance (see Table 1).
products means controlling five (5) basic conditions.
Table 1. Quality assurance demonstration and evidence
 Control of human resource quality:
Everyone, from the top management to workers, should Quality Assurance
be trained to carry out their assigned tasks. They must be
sufficiently experienced to apply methods and procedures,
and know how to use equipment, facilities, etc. Everyone Demonstrating quality control Quality control evidence
thoroughly understands their duties and responsibilities for Quality manuals Test certificates
product quality. They have the necessary guidance Procedures Test reports
documents on duties and enough facilities to carry out their Technical regulations
Regulations on personnel's
work. There are all other necessary conditions to enable capabilities
their work to achieve the desired quality. Reviews by customers on
Regulations on processes
technical ability and organization
 Control of methods and processes: Demonstrating quality control Quality control evidence
Methods and processes must be appropriate, meaning
that textile and garment products and services in support of A reasonable organizational structure for activity
textile and garment production will fulfill the set supervision.
requirements. Quality Assurance
 Control of suppliers:
It is required to know how to select suppliers capable of
- Demonstrating quality control.
meeting all necessary conditions. The contents of an order - Quality assurance evidence
must be clear, accurate, and complete. A purchase order - Designating persons in charge of
must include all technical requirements, such as product quality assurance
characteristics, weight, inspection and test conditions,
certificates, terms of packaging, packing, transportation,
warranty, storage, etc. Purchased materials and accessories
must have a handover record and be stored in appropriate
conditions until they are used.
 Control of equipment quality:
Equipment must be suitable for use and ensures good
working conditions and technical requirements. In addition,
it must be safe, non-polluting, and clean.
 Quality control of information: Figure 4. Quality Assurance
293
D. 4th Strategy: QM - Quality Management.

Total Quality Management
The quality assurance strategy does not cover the price
of products, as we have full discretion to select suppliers for
the lowest price and apply the same quality terms. Hence, - Internal and external supplier
textile and garment companies must calculate economic and customer relationship
efficiency to achieve a low-cost target. Textile and garment
companies must expand their markets, search for suppliers,
increase their profit rates, reform their financial activities
and choose investment options. To help Vietnamese textile
and garment companies to achieve this, technical
specifications of a textile and garment product must fulfill
global buyers' requirements, regulations of the global
market, and these companies must carry out a SWOT
analysis - Strengths, Weaknesses, Opportunities, and
Threats for their textile and garment businesses.
Quality Management
Figure 6. Total Quality Management

- Economic calculation of quality cost.
- Optimization of lowest expenses III. CONCLUSION
- Financial goals Findings of the research, analysis, and simulation of
mathematical Venn diagrams show that Vietnamese textile
and garment companies will be able to win more orders
from global buyers for textile and garment products
manufactured in Vietnam and achieve realistic values
contributed to their companies once they understand the
definition and importance of product quality, building trust,
and reputation with customers. Production cost and product
cost are reduced (by a survey) by 4% based on CMT or
FOB compared to those of other countries in the region such
as China, Bangladesh, India, etc.
Figure 5. Quality Management
ACKNOWLEDGMENT
E. 5th Strategy: TQM – Total Quality Management The research team would like to express its sincere
From the four (4) strategies mentioned above, we can gratitude to:
summarize the following; Quality check of textile and - The Organizing Committee of the 2022 6th
garment products to find mistakes and defects in International Conference on Green Technology and
manufactured products must be carried out; Quality control Sustainable Development who allowed us to take part in the
includes control of five (5) basic conditions to prevent and conference in Nha Trang City.
detect possible mistakes and defects; Quality assurance is - Ho Chi Minh City University of Technology and
aimed at building trust among global buyers, satisfying their Education (HCMUTE) for its financial support.
quality requirements outside the plant (suppliers, services); REFERENCES
Quality management is aimed at optimizing operations to
achieve high quality and economic results. [1] M. D. F. Hanif, H. T. Sadia, M. H. Chaion, M. A. S. Rafi,
M.D. J. Uddin, M. R. Repon, T. Islam, “Quality improvement
Hence, total quality management which is; aimed at in readymade garments industry by traffic light system”,
achieving quality management on an overall scale to fulfill Journal of Textile Engineering & Fashion Technology, June
necessary and sufficient conditions for internal; needs of 2020.
Vietnamese textile and garment companies, and outside [2] Hilletofth P, Hilmola O. P, “Supply chain management in
suppliers, support services, and global buyers buying textile fashion and textile industry”, Int J Services Sciences, pp. 127–
and garment products covers the management of [6, 7, 8]: 147, 2008.
[3] Nguyen Phuoc Son, Nguyen Thi Tuyet Trinh, “Researching for
- Information application of lean production in redesigning sewing workshop
- Training layout in ground apparel company, waste reduction, labour
productivity enhancement”, Journal of Technical Education
- Attitudes, manners, conduct
Science (JTE HCMUTE; ISSN: 1859-1272), pp. 69-77,
- Goals December 2016.
- Organization [4] Nguyen Phuoc Son, Ho Thi Minh Huong, Tran Dai Nguyen,
- Internal relationship between departments and units within “Motion study application to eliminate redundant loss time in
a company production, reduce wastefulness and improve labor
productivity at garment companies”, Vietnam
- Specific means, tools, and conditions to perform the work.
294
Electromechanical Journal (ISSN 2615 - 9910), pp. 193-203, [7] Rahman M. M, Masud A, “Quality improvement in garments
Special issue magazines, December 2021. industry through TQM approach”, International Conference on
[5] Nguyen Phuoc Son, Pham Ngoc Thuy Ngan, “A research on Mechanical Engineering, Dhaka, Bangladesh, 2011.
clo 3D software, the optimal solution for technical designers at [8] Katzell, R. A, Yankelovich, D, Fein M, Ornate, D. A & Nash,
garment companies – following precisely the size specification A, “Work Productivity and Job Satisfaction”, The
of a garment from a 2D pattern. Satisfying past fashion Psychological Corporation, New York, 1975.
requirements from global buying offices”, Vietnam
Electromechanical Journal (ISSN 2615 - 9910), pp. 204-210,
Special issue magazines, December 2021.
[6] Iqbal M. A, Yousuf M. F. H, Salam M. A, “Analysis of an
effective quality control system: a study on woven garments”,
EPRA International Journal of Research and Development
(IJRD), pp. 144–152, 2018.
295
Building Students’ Self-Reliant on Problem-Based

Learning by Embedding Mind of Engineering
Design Concept at HCMC University of
Xuan Tien Vo
School of Mechanical Engineering
HCMC University of Technology and Education
tienvx@hcmute.edu.vn
Abstract: Capabilities of employing engineering Solving this problem, Intel has started the HEEAP, which
principles and concepts from training areas to solve intends to improve instructional activities of faculties and
engineering problems; using knowledge of engineering learning ways of students, in which process of knowledge
science to analyse existing conditions and define transfer would be enhanced by increasing interaction between
requirements and constraints related to solving a problem by them. And the most important innovation act is integrating
developing a complete and accurate problem statement are Problem-based learning in courses, conducted in engineering
essential skills, especially in engineering field, for education programs at HCMUTE. Profiting from the program,
accustoming highly demands of industry 4.0. Self-reliant, PBL has been successfully and wide integrated at HCMUTE
and become the most popular hands-on activities in the
therefore, plays a special role in Career Development. This is
classroom in order to provide students with a more interactive
also reason that Project-based learning (PBL) has officially
higher education experiences. In fact, Intel was not the only
been applied for training student’s skill of hand-on in company which had had difficulty in employment qualified
technical field in HCMUTE since 2012 and becomes even a workforce for its new ATM due to shortage of skilled labor
certain policy at the university. However, a lack of systematic because Mori has figured out difficulties in recruitment
approach in engineering design process, which could be qualified work force at Japanese companies in Vietnam in his
useful to support students effectively hand on his trouble, research: “Many enterprises are still seeking: multi-skilled
causes them usually skip essential steps during design process production-line leaders who can manage and improve the
such as defining design’s problems or controlling/testing entire production process rather than remain limited to a
project’s results to make their design unstable and unreliable. specific one; engineers or technicians who are skilled at high-
It is very important to integrate mind of engineering design precision processing in making metal, plastic parts, molds and
process as strategic approach in PBL in order to support dies; and, middle managers who can lead day-to-day
students build up their design thinking. The article informs improvement of factory operations by identifying and
results extracted from empirical and practice-based research resolving problems from a broad and long-term
as well as provide some teaching experiences which can raise perspective…” He also proposed that “Vietnam needs
instructor’s awareness in TVET training to justify approach industrial human resources who can initiate and manage the
in designing instructional activities in order to encourage improvement of production operations, rather than those who
learners to develop capability of self-reliant skill in learning. can perform only standard tasks.” and figured an “Engineers”
as a person: “who lead the improvement of factory layout, jigs,
Keywords: PBL, engineering design concept, systematic person-hours, and in-house logistics flows in close
approach cooperation with the production and design departments of an
enterprise’s headquarters”. This is also Imagine of
I. INTRODUCTION “Monozukuri”– “making thing” issued by Japan International
Cooperation Agency (JICA) Office in Vietnam, Monozukuri
There is a long history of why Project-based Learning
- high-skill manufacturing, which “implies manufacturing for
(PBL) becomes a policy at HCMC University of Technology
the primary purpose of achieving customer satisfaction
and Education Ho Chi Minh City (HCMUTE). Media’s
through high quality in the spirit of a proud and dedicated
exaggeration of skilled labor’s shortage at the Intel’s
artisan, rather than just making profits” [4], [5].
“recruitment test [9] is indeed most suitable reason for
HEEAP’s accomplishment to solve the issue of inadequate In addition, UTE has also conducted CDIO’s approach and
quality of human resources and the capacity of providing a built all syllabi with CDIO form in which courses are
qualified and capable workforce in Vietnam. PBL has been suggested conducted with embedded PBL in order to fulfil
applied at University of Technology and Education Ho Chi important industry need, which employers expect from their
Minh City since 2011 when the first cohort of faculties engineers who can Conceive, Design, Implement, and Operate
returned from Arizona State University after they had been complex products and systems in a modern, team-based
trained in the frame of the project “Higher Engineering environment. Crawley has advised that building syllabi should
Education Alliance Program (HEEAP), initiated by Intel. be based on concept of CDIO, abbreviation of Conceive –

Design – Implement – Operate and fulfil such standards as (1) An absence of rationale (engineering design concept as
Design-implement experiences, integrated learning guideline for student’s project implementation), and
experiences Integrated learning experiences that lead to the
acquisition of disciplinary knowledge, as well as personal and (2) an absence of detail (lecturer’s observation and
interpersonal skills, and product, process, and system building controlling if students go through design process step by step
skills. According to Crawley et al., learning activities should or not in order to be sure that student’s relevant skills could be
be active and teaching and learning based on active and formed during project proceeding).
experiential learning methods. Moreover, he also suggests that Involving with these issues, Amborse et al. have definitely
teaching activities should focus on students’ cognitive and emphasized that effective teaching must begin with a
affective development that includes analytical reasoning and consideration of how students learn. And Herbert has
problem solving, experimentation, investigation and additionally stated that “Learning results from what the
knowledge discovery, system thinking, creative thinking, student does and thinks and only from what the student does
critical thinking, professional ethics and other responsibilities and thinks. The teacher can advance learning only by
[2]. influencing what the student does to learn”. In fact, “Learning
Finally, engineering education should support students in is not something done to students, but rather something
well-preparing for their career path by practicing essential students themselves do. It is the direct result of how students
skills, attitudes as well as the depth and breadth of knowledge interpret and respond to their experiences — conscious and
in order to implement duty related to their professional field unconscious, past and present”. Learning is actually a
confidently and help them to become capable of adaption in continuous intellectual process in which knowledge should be
continuous changes in technological, economic conditions. consistent and unique when students should bring their prior
This has been discussed by The National Society of knowledge to new course and combine them with fresh
Professional Engineers (NSPE) which initiates Engineering knowledge in order to propose most appropriate solutions for
Body of Knowledge (BOK) – a vision of requirements on given problem or issue which they have to challenge due to
engineers who will take part in labor mart – figures out precise finishing their project. However, their prior knowledge can
categories of capability needed for engineering career: help or hinder their coming learning process. If students’ prior
knowledge is robust and accurate and activated at the
- “Knowledge consists of comprehending theories, appropriate time, it provides a strong foundation for building
principles, and fundamentals; new knowledge. Conversely, their students’ prior knowledge
may impede new learning if it is inert, insufficient for the task,
- Skills are the abilities to perform tasks and apply activated inappropriately [1].
knowledge; and
Students can make connections between pieces of
- Attitudes are the ways in which one thinks and feels knowledge through finishing their projects just when
in response to a fact or situation.” knowledge is structured accurately and intently. Lecturers
The collection of BOK can be explained that “knowledge should organize knowledge transferring process scholarly in
is what one knows in a fundamental sense, skill is what one is order to support students to be better able to retrieve and apply
able to do with what one knows, and attitude is how one their knowledge effectively and efficiently during their project
responds to a variety of situations” generally [8]. implementation. In contrast, when knowledge is connected in
inaccurate or random ways, students can fail to retrieve or
II. CONSIDERATIONS OF PBL AND VITAL ROLE apply it appropriately. This can happen through trying and
OF SELF-RELIANT IN ENGINEERING EDUCATION failing process of students. It results insufficient learning way
Problem-based learning (PBL) is an effective learning and costs so much time of them because they usually rely on
method, based on brain-based educational methods. In PBL, what they can find via Internet (see Fig. 1)
students are encouraged with a real-life issue that they are
required making a decision, or with a real-life problem that
they are asked for presenting a solution which should be
suitable to solve their concerns. Therefore, PBL engages
students and motivates them through giving them enough
space to be creative when they can suggest solutions after their
brainstorming process. When students work in groups that can
stimulate social connection for improving their cooperative
learning skills. PBL requires also students’ creative thinking Ask another person
and makes their fragmented past knowledge to become Finding from books
unique. However, there is a fact that lectures and instructors 11
Internet searching
are teaching a course in professional field because they are an
expert in your field — but they are certainly not an expert in Self-thinking
how to teach others about professional fields. They are mostly
concerned with what should be transferred during the course Figure 1. Student’s responses when they have to deal
rather than detect mechanisms and conditions that promote with a question or problem
student learning may find themselves best ways to accumulate
new knowledge and bridging between them and prior To support students’ development in self-reliant, students
knowledge. Consequently, PBL might disadvantage following must acquire essential knowledge and component skills in
issues: their professional field by practicing and integrating them as
well as know when and how to apply what they have learned.
Moreover, students must develop not only the component
297
skills and knowledge necessary to perform complex tasks, Through these activities, students can understand deeply
they must also practice combining and integrating them to the problem, should be solved. Consequently, they can gather
expand their adaptive capability for preparing to challenge relevant information, which are essential for developing
changes in future working world. From instructors’ conceptual design. It makes communication between team
perspectives, it is important that educator should focus on members and teams more effectively and aids them depict
conscious awareness of these elements of developing self- their thinking more precisely. It will support their engagement
reliant to help students learn more effectively. On the other in getting involved with reality and know exactly what they
hand, student should become self - directed learners through should contribute to social as an engineer. When they
proceeding PBL and must learn to monitor and adjust their persuade their colleagues to believe in their solution for
approaches to learning. Learners may engage in a variety of project, they can develop their skills for negotiation and make
metacognitive processes to monitor and control their learning their feeling more solid so that they can achieve as well as
— assessing the task at hand, evaluating their own strengths anchor new knowledge.
and weaknesses, planning their approach, applying and
monitoring various strategies, and reflecting on the degree to
which their current approach is working. Unfortunately,
students tend not to engage in these processes naturally. When
students develop the skills to engage these processes, they
gain intellectual habits that not only improve their
performance but also their effectiveness as learners.
Finally, organization and preparation of learning process
should harmonize neuroscience of learning and teaching
which the science of the relation of the nervous system to
learning and behavior. Lecturers should understand how
learning occurs effectively, how transfer in learning occurs as
well as how neural connections can be created during learning
process. Student’s cognitive process in learning can be
strengthened via repeated experiences so that neural
connections – called as synapses – can be formed that make
neural firings and transmissions of information more rapid.
When students apply learnt knowledge from their memory and
combine them with new one to give appropriate solutions for
new uses and situations that they have made transfer of their
knowledge because transfer involves using information in Figure 2. The design model of Kosky chosen to embedded in PBL
new ways or in new situations. This does not happen [6]
automatically. Students must learn them through experiences
or determine them on their own through problem solving Embedding design concept in PBL provides students’
exposed during finishing their project. For that reason, project implementation didactical approach in conducting
lectures’ instruction plan must be specified clearly so that their design and leads their project to become successful more
interventions can be tailored to specific needs so that easily. This can be proved by most of students’ projects’ high-
organized teaching activities engage learners and hold their grade of completion. In example of the sea-bin’s project,
attention in order to help them produce better learning. students have focused on their environmental concern of trash
collection by designing model of sea-bin. They had started
III. RESULTS AND DISCUSSIONS with Kosky’s model and applied physical concept of
buoyancy to develop collecting movement principle for their
To build student’s self-reliant, Kosky’s design concept project. Finally, they had finished their project with a real
(see Fig. 2) has been embedded in courses which PBL is model and implement experiments on it.
targeted as a standard teaching method. Students are asked
looking for an issue or problem related to real life or
environmental concern in recommendation their projects.
They are engaged to describe their ideas or observations about
troubles, difficulties with their own words and they have to
persuade their colleagues that these concerns or problem are
from reality and need to remedy. This is, actually, a
brainstorming process because students have to present their
ideas to another group and discuss about alternative solutions,
which they want to choose. After defining their concern, they
should generate alternative concepts in order to be able to
promote most suitable solution for the concern which they are
interested in or problem which they really want to solve. They
should give defense for what they believe with 4W-1H: what
their concern is? Why they choose it? Where they intend to
apply the solution? Who will get benefit from the solution?
And how they solve the problem? (see Fig. 3) Figure 3. Design model derived from Kosky by lecturer’s
consultant
298
In the case of smart-bin’s project, the project had done in students in applying their prior knowledge and in connection
previous semester. Students would like to advance their with new one which they have just learnt in order to build
project to make its completion because they had not done connectivity between related professional subjects. Moreover,
enough experiment on their project therefore the smart bin they are capable of self-reliant when making decision
didn’t work smoothly. To take advantages, they are engaged mentally through solutions’ recommendation for a certain
to plan different cases in which working conditions of the concern or problem which they would to solve or get involved
smart bin are specified such as: with. In fact, knowledge is just applied effectively when they
are grouped together if they have relationships to each other,
- Set the detective range of the first sensor is 380 mm e.g. materials sciences, metallography and metals testing,
at the totally empty condition and 100 mm at the full metal processing technology should be grouped or make
condition. The remaining space is divided into 4 relationship for supporting in applying new knowledge and
levels. teaching methods. There is always a debate of how we can
- Set the maximum detective range of the second organize learning process: Should we divide essential
sensor is 250mm. knowledge into separated courses or make connections
- Set the delay for the users to put the trash in the bin between them to create learning field which students can find
at 3 seconds. essential knowledge and skills through combining their learnt
knowledge?
- The indicate LEDs display 4 different colors to show
the remaining space. REFERENCES
- The alert notification and email will send [1] Ambrose, S. A., Bridges, M. W., DiPietro, M., Lovett, M.C., & Norman,
continuously. M. K... “How Learning works: Seven Research-Based Principles for
Smart Teaching”, 1st Edition. San Francisco, United States of America:
- Set the opening angle of the servo motor at 50°. Jossey-Bass Inc., 2010, p. 3.
[2] Crawley E. F., Östlund J. M. S. Edström D. R. B. K., “Rethinking
For this setting working conditions, the bin could not open Engineering Education: The CDIO Approach”, 2nd Edition, Springer
although it is full (didn’t have enough space for the lid to International Publishing Switzerland, 2014, 36, pp. 50-51.
rotate). In addition, the lid still opening in full condition, it also [3] Eberly Center, “Learning Principles: Theory and Research-based
opened although it was in an accidental case so that setting Principles of Learning”. Pittsburgh: Carnegie Mellon University.
Online: https://www.cmu.edu/teaching/principles/learning.html (retrieve
should be optimized. Students, after that, change the detective d 15.01.2022), (n.d.).
range at the full condition to 140 mm, as well as justified [4] Japan International Cooperation Agency Vietnam Office – JICA, “Policy
operation program to force the lid to close in full condition. Paper Promoting Tripartite Partnerships to Tackle Skills Mismatch:
By increasing the delay time up to 5 seconds, the bin can now Innovative Skills Development Strategies to Accelerate Vietnam’s
work more stable. Finally, students have to update program to Industrialization”, Vietnam Office, Hanoi, 2014, p. 9.
[5] Junichi Mori, Nguyen Thi Xuan Thuy, Pham Truong Hoang, “The final
remove delay time and to keep the lid open until the users go report to Hiroshima University’s COE project entitled “Research on
away. The bin is now working effectively, stable and accurate Cooperation in the Fielded of Skill Development Education and
after changing the opening angle of the servo motor to 60°. Economic Development”: Skill Development for Vietnam’s
Results from empirical survey after research implementation Industrialization: Promotion of Technology Transfer by Partnership
between TVET Institutions and FDI Enterprises”, United Nations
show increasing students’ capacity for independent thought Industrial Development Organization, Vietnam Country Office, Hanoi,
when they can, students can combine essential knowledge and 2009, pp. 4-5.
component know how to apply what they have learned during [6] Kosky, P., Wise, G., Balmer, R., & Keat, W., Exploring Engineering: An
their project finishing (see Fig. 4). Introduction to Engineering and Design, 2nd Edition. Amsterdam:
Academic Press, Elsevier, 2010, p. 352.
[7] Hans-Dieter Hoepfner Hermann Koch, “Self-reliant Learning in
Technical Education and Vocational Training (TEVT)”, Regional
Association for Vocational Teacher Education in Asia (RAVTE) RAVTE
Secretariat, Chiang Mai, Thailand, 2015
[8] National Society of professional engineers (NSPE), “Professional
Engineering Body of Knowledge”, First Edition, Virginia, USA, 2013, p.
4.
[9] The Du, H., Trung Dung, H., Xuan Thanh, Ng, & Anh Tuan, D. Th.,
“Intel products Vietnam 10-year investment: impact study report 2006-
Ask another person 2016”. Ho Chi Minh City: Fulbright University Vietnam.
Finding from books Online: https://fsppm.fulbright.edu.vn/en/policy-papers/policy-
research/intel-products-vietnam-10-year-investment-impact-study-
Internet searching report-2006-2016/ (retrieved 04.02.2022), 2018
Self-thinking [10] HCMC University of Technology and Education Ho Chi Minh City,
“Regular Nr. 117 regarding Innovation of teaching and learning activities
at HCMUTE”, Ho Chi Minh City, 2012
Figure 4. Student’s responses when they have to deal with a [11] Xuan Tien Vo, “Integrating hands-on design project in project-based
question or problem after research is done learning in order to enhance student’s competency of engineering design:
A case study in HCMC university of technology and education”. In:
IV. CONCLUSION Proceeding of 1st Hanoi International Forum on Pedagogical and
Educational Sciences, Vietnam National University, University of
Embedding engineering design concept in PBL can create Education Vietnam National University Press, Hanoi, 2021
more free space and creative energy for students in completing [12] Vo, X. T., “Problem-based Learning (PBL) with embedded design
their project confidently and strengthen their cognitive concept to strengthen students’ cognitive activities”. In: TVET@Asia,
activities when they could know exactly where they should issue 18, pp. 1-16. Online: http://tvet-online.asia/issue/18/problem-
based-learning-pbl-with-embedded-design-concept-to-strengthen-
begin and how they can set up targets which they really want students-cognitive-activities/ (retrieved 31.07.2021), 2022.
to achieve. Learning by doing mentally is best way to engage
299
Modeling and Comprehensive Assessment of

Construction Risks: A Perspective of PPP
Transportation Projects
Ha Duy Khanh Soo-Yong Kim Nguyen Van Khoa
Dept. of Civil Engineering Dept. of Civil Engineering Dept. of Civil Engineering
Ho Chi Minh City University of Pukyong National University Ho Chi Minh City University of
Technology and Education Busan, Korea Technology and Education
Ho Chi Minh City, Vietnam kims@pknu.ac.kr Ho Chi Minh City, Vietnam
khanhhd@hcmute.edu.vn khoanv@hcmute.edu.vn
Abstract: This paper presents a comprehensive Through studying projects in the UK, Scotland, Spain, and
assessment of risks for PPP transportation construction Portugal, Barnier [3] has confirmed that PPP is suitable for
projects. Like other construction projects, PPP-based infrastructure projects. However, the risk is inherent in
transportation projects contain significant risks that affect infrastructure development projects in the form of PPP. They
performance and efficiency. This study has used a structured arise because of uncertainty about future outcomes, which
questionnaire distributed to 200 respondents to collect data. may negatively affect the project's service delivery or
The critical risk factors of PPP transportation projects yielded profitability [4].
in the implementation phase have been analyzed and ranked. Although the PPP methods have been applied worldwide
The results indicated that the critical risks impact the project in construction projects, better performance outcomes have
efficiency in terms of schedule and cost with 9.23%. This yet to be certified [3]. It is especially true in Vietnam due to
value has been justified through a case study project as having differences in technology conditions, average income levels,
a low deviation. In addition, this study has explored the and investment policies [5]. There have been several studies
financial risks using Monte-Carlo Simulation. Three financial on PPP projects in Vietnam, such as Nguyen [2], Huynh [6],
indices have been used, including NPV, PP, and IRR. The Phan [7], Nguyen, and Le [8]. However, these studies only
simulation results show that the probability of obtaining NPV focus on understanding the influencing factors, significant
is 94.8% based on the monthly revenue. Moreover, the safety barriers, and appropriate solutions for successful PPP
coefficient of investment is relatively high compared to the cooperation. Thus, this study investigates considerable risk
expected values. The study's findings encourage investors to factors of PPP transportation projects in the construction stage
join PPP projects many more in the future. and simulates financial risks due to changes of interest and
cash flow in the operation stage through a case study.
Keywords: risk, Public-Private Partnership, PPP, II. LITERATURE REVIEW
transportation project, construction management
In Vietnam, Trinh [9] has studied and provided several
I. INTRODUCTION solutions for risk management in transportation projects with
the perception of risks as negative or adverse impacts. Some
Vietnam has used many state funds to build transportation
typical risks have been found, including business inequality,
projects in recent years. However, the state investment
limited policy, lengthy payment procedure, and profit fraud.
fundings for the construction industry are always limited.
Vo [5] has also added some risk factors of PPP, such as
Therefore, government agencies have tried to find other
unclear contract responsibility and lack of operation
funding from the private sector. This cooperation is called a
experience of project stakeholders. Phan [7] has pointed out
public-private partnership (PPP, 3P, or P3). Typically, it can
some shortcomings in the pilot implementation of PPP
be used to finance, build, and operate construction industry
involving lack of a legal framework, low synchronism,
projects. In general, Wikipedia [1] indicated that the delivery
conflicts of interest, and risk-sharing mechanism among
models of PPP include five types of partnership: operation and
parties. Nguyen and Le [8] have proposed some solutions to
maintenance (O&M), build-finance (BF), build-operate-
promote the role of the government and the private sector in
transfer (BOT), build-own-operate-transfer (BOOT), build-
developing investment in the PPP form. They are: developing
own-operate (BOO), build-lease-transfer (BLT), design-
a policy framework and standardizing regulations for
build-finance-maintain (DBFM), design-build-finance-
activities related to PPP delivery; building a synchronous legal
maintain-operate (DBFMO), design-build-operate-transfer
framework for PPP; building a management apparatus;
(DBOT), design-construct-manage-finance (DCMF).
strengthening investment promotion and communication
Nguyen [2] stated that PPP is essential for optimizing activities to attract the attention of domestic and foreign
investment efficiency by providing high-quality public investors.
services and solving the lack of investment capital. The
Overseas, Shouqing, and Tiong [10] studied to identify a
success of a PPP construction project depends on the type of
list of risk factors in BOT projects in China in the 1990s. This
project, the transparency and consistency in planning and
list includes six types of risks classified based on their origin.
investment, and the level of government participation.
Risk is understood as harming project performance. The

study's main result is to determine the level of risk according Table 1. Respondent’s profile and reliability
to the risk factor by multiplying the probability of occurrence
Reliability
by the level of impact of that risk factor. Finally, they also Profile Percentage
suggest risk reduction measures through discussion and expert Frequency Effect
interviews. Project stakeholders
+ Owners 55 0.893 0.901
Similarly, Mohammed [11] used survey results by + Consultants 25 0.865 0.845
questionnaire technique to allocate risk factors in PPP + Others 20 0.877 0.789
infrastructure development projects in Nigeria. The results Work positions
showed that the government retains seven risks relating to (1) + Managers 54 0.811 0.872
political and policy; (2) changes in the legal framework; (3) + Employee 46 0.834 0.756
delays in project approvals and permits; (4) trustworthiness of Experience year
+ < 10 years 42 0.764 0.834
third parties; (5) the contract risk changes many times; (6) + ≥ 10 years 58 0.815 0.892
economic fluctuations and (7) lack of suitable financial
Total 100 0.802 0.886
instruments. The remaining risks are allocated to the private
partner. C. Analysis and Simulation Approach
Li et al. [12] studied risk allocation in PPP/PFI  For the construction stage:
infrastructure development projects in the UK. Based on a
synthesis of previous studies, they have established a list of The five-point Likert scale has been adopted to qualify the
risk factors according to the source of risk and the concept that probability of occurrence and the level of impact of the risk
risk includes both negative and opportunity. Sachs et al. [13] (see Table 2). The percentage of increase for frequency and
analyzed the political risks and opportunities of PPP in China effect of risk for transportation projects is proposed by three
and 14 selected Asian countries. They have classified risks experts in the pilot test stage.
into two groups based on the source of risk, including
insurable risks and uninsurable risks. Checherita and Gifford Table 2. Risk evaluation scale
[14] found substantial risks affecting PPP projects in the US. Scale
Based on the source of risk, they have divided risks into Meaning Not at Mode- Extre-
typical and unique risks. In addition, Barnier [3] identified risk all
Slightly
rately
Very
mely
factors based on a negative perception of risk and then
discussed issues for the successful development of PPP Value 1 2 3 4 5
projects. Frequency
<10% 10- 30- 50- 70-90%
30% 50% 70%
III. RESEARCH METHODOLOGY < 2% 2% - 5% - 10% - > 20%
Effect
5% 10% 20%
A. Questionnaire Design
Based on the literature review, this study has found
twenty-five risk factors affecting PPP transportation project The risk point (RP) of each factor is then calculated based
efficiency. The questionnaire involves three parts. They are on the formula below:
(1) respondents' agreement on the mentioned risk factors; (2) n
Fi  E i
assessment of the effect of the risk on project cost; and (3) RP =  (1)
respondent's personal information. i 1 25
B. Data Collection and Check Where: F and E are the frequency and effect of risk factors,
Firstly, this study has conducted a pilot test to check the respectively.
suitability of the questionnaire's content and scale with three The predicted performance risk (PPR) of the project in
construction experts. These experts have more than 20 years terms of schedule and cost is then identified based on the
of experience and work in managerial positions in the formula below:
transportation construction field. The test results showed that
they strongly agreed with the questionnaire. Therefore, the PPR = RP×FxE (2)
questionnaire could be used to collect massive data for
Meanwhile, the actual performance risk (APR) is
prospective analyses.
calculated base on the formula below:
Secondly, this study has distributed two hundred
questionnaires to the selected respondents using the non- Actual - Approved
APR = (3)
probability sampling method. Most of the respondents come Approved
from owners with the work position of managers or higher and
experience years of greater than ten years. They mainly work  For the operation stage:
for government agencies and organizations for transportation This study also uses the Monte-Carlo simulation. It is a
investment and management. The detailed personality of the computational technique that helps to calculate and quantify
respondents is mentioned in Table 2. risk and then make decisions. This technique is widely used in
Thirdly, this study has adopted Cronbach's alpha test to finance, management, energy, manufacturing, engineering,
check the scale's reliability. The test results are also described and research and development. The simulation results provide
in Table 2. Compared to the commonly used rule of thumb, the decision-maker with a range and probability of outputs
the value of Cronbach's alpha of 0.802 and 0.886, corresponding to the choices. The procedure of the Monte-
respectively, for frequency and effect is outstanding. Carlo simulation is below:
301
- Step 1: Create a project plan. Rank Content RP

- Step 2: Identify sensitive and uncertain variables. 11 Inappropriate project implementation time 0.518
- Step 3: Define the correlation between variables, 12 Obstacles of service providers 0.516
including direction and strength. 13 Challenging weather and geological conditions 0.513
- Step 4: Develop a simulation model. 14 Unclear specifications and types of materials 0.511
15 Loose legal framework 0.489
- Step 5: Analyze the results, including statistical values
and probability distributions. 16 Changes in the government policies 0.483
17 Unsuitable contract 0.481
As mentioned earlier, the financial efficiency of
transportation projects is assessed using the Net Present Value 18 Political issues 0.480
(NPV) method. Theoretically, the project is worth the 19 Lack of survey on the existing buildings 0.471
investment if NPV > 0. The NPV relies on three parameters
20 Late site clearance 0.461
involving cash flow (CF), profit rate (i), and the number of
timer periods (t). The formula of NPV is expressed as follows: 21 Schedule delays 0.460
n 22 Reworks 0.448
CF
NPV =  t
(4) 23 Outdated technologies and methods 0.444
i1 (1+i)
24 Poor safety implementation 0.423
In (4), CF is calculated by subtracting costs from revenue. 25 Late material supply 0.420
For the transport PPP project, the annual costs are almost fixed
Mean 0.513
costs. Moreover, its revenue primarily comes from tolls of
passing vehicles. Therefore, the change in CF depends only on B. Simulation Results
the change in revenue. In the case of fixed CF and interest,
there will be a value of time that makes NPV = 0. This value This study has adopted a case study to demonstrate the risk
of time is called the payback period (PP). The assessment level of PPP transport projects. The project information is
criterion is PP greater than the expected PP. In the case of shown in Table 4 and Table 5. It is kindly noted that two
fixed CF and time, a value of interest will also make NPV = significant indices are used to assess the project efficiency in
0. This interest value is called the internal rate of return (IRR). the construction stage involving schedule performance and
The assessment criterion is IRR greater than the minimum cost performance. In addition, three financial indices have
attractive rate of return (MARR). been used to assess the investment efficiency in the operation
stage, including NPV, PP, and IRR.
IV. FINDINGS
Table 4. Brief information of the case study project
A. Risk Assessment and Ranking
Property Content
Based on (1), the risk point of each factor of PPP Project name Vinh Loc Bridge
transportation projects is calculated as shown in Table 3.
Investor Hai Den Logistics & Trade Ltd., Co.
Because this study uses the five-point scale, the cut-off risk
Length: 382m, load: HL93 (including road to
value is 3x3/25 = 0.36. The results indicated that all twenty- Scope
bridge: 344m)
five risk factors have a risk point greater than 0.36. Among Type/ Level Transportation project with level III
them, the three highest-ranked factors include “Lack of skilled
workers” (RP = 0.618), “Missing bid quantity in tender” (RP Table 5. Detailed information of the case study project
= 0.599), and “Inconsistent norms of construction works” (RP
= 0.592), and the three lowest-ranked factors include Information Value Unit
“Outdated technologies and methods” (RP = 0.444), “Poor Approved duration in the beginning 24 Month
safety implementation” (RP = 0.423), and “Late material Actual duration at the completion 27 Month
supply” (RP = 0.420). Expected time of toll collection 34 Year
Billion
Approved capital 145
Table 3. Risk and ranking VND
Billion
Actual capital 162
Rank Content RP VND
Monthly cost 130.2
1 Lack of skilled workers 0.618 - Maintenance: 11.2
Million
2 Missing bid quantity in the tender 0.599 - Management & operation: 43.0
VND
- Depreciation: 70.0
3 Inconsistent norms of construction works 0.592 - Others: 6.0
4 Poor qualification of the employee 0.590 Monthly revenue from the toll
- Minimum: 497 Million
5 Delayed payment 0.574 - Average: 710 VND
6 Errors and changes in design 0.573 - Maximum: 923
Minimum attractive rate of return
7 Bad volatility of the market 0.564 0.89 %/ year
(MARR)
Low level of cooperation between the projects Payback Period (PP) 24 year
8 0.539
parties
9 Expenses due to implicit payments 0.538 Based on (2) and (3), the project schedule and cost
performance can be calculated and shown in Table 6. The
10 Low financial ability 0.521
302
results showed that the APR of schedule and cost is 12.5% and Table 8. Probability of gaining NPV and IRR using a uniform
11.72%, respectively. Meanwhile, the PPR for both of them is distribution
9.23%. Therefore, the deviation between them is 2.24% and Probability NPV (million VND) IRR (%/year)
1.46% for schedule and cost. It indicates that the project
performance is not good though the risks have been well 0% -654.80 1.312
understood before construction. 10% 12,536.30 1.867
Table 6. Project performance risk 20% 26,617.94 2.433
Project 30% 40,145.71 2.953

PPR APR Deviation
performance risk 40% 54,806.29 3.495
Schedule delay (%) 12.50 3.27
9.23 50% 68,402.91 3.981
Cost overrun (%) 11.72 2.49 60% 82,120.68 4.457
Selecting the probability density functions is the most vital 70% 96,082.21 4.928
action for the Monte-Carlo simulation because it significantly 80% 109,897.85 5.383
affects the outcomes. There are three popular probability 90% 123,871.39 5.833
distributions: uniform, triangular, and normal. This study uses
the uniform and normal distribution because the 100% 138,202.06 6.285
characteristics of the monthly revenue are minimum, average,
 Simulation using a normal distribution:
and maximum.
 Simulation using a uniform distribution: The simulation results show that the value of NPV ranges
from nearly -104.9 billion VND to 251.9 billion VND. The
The simulation results show that the value of NPV ranges standard error is 467.5 million VND. The results are shown in
from nearly -654.8 million VND to nearly 138.2 billion VND. Fig. 3, Table 9, and Table 10. Thus, the probability that shows
The deviation is 181 million VND with a confidence level of the NPV value less than the initial expected NPV is 50.07%,
95%. The results are shown in Fig. 1, Table 7, and Table 8. and the probability that the IRR is less than the initial IRR is
The results show that the monthly revenue is never enough to 49.93%. The probability of NPV equals the initial investment
return the initial capital. capital is approximately 94.8%.
Figure 1. Chart of NPV using a uniform distribution Figure 2. Chart of NPV using a normal distribution
Table 7. Statistics of NPV and IRR using a uniform distribution Table 9. Statistics of NPV and IRR using a normal distribution
NPV (million Statistics NPV (million VND) IRR (%/year)

Statistics IRR (%/year)
VND)
Trials 10,000 10,000
Trials 10,000 10,000
Base Case 68,777.86 3.994
Base Case 68,777.86 3.994
Mean 69,372.49 3.924
Mean 68,345.72 3.915
Median 69,787.52 4.030
Median 68,409.86 3.981
Mode --- ---
Mode --- --- Standard Deviation 46,748.10 1.699
Standard Deviation 40,155.68 1.430 Variance 2,185,385,223.03 0.029
Variance 1,612,478,366.76 0.020 Skewness -0.0430 -0.4317
Skewness 5.6927E-04 -0.1091 Kurtosis 3.02 3.59
Kurtosis 1.80 1.82 Coeff. of Variability 0.6739 0.4330
Coeff. of Variability 0.5875 0.3652 Minimum -54,949.76 -5.175
Minimum -654.80 1.312 Maximum 197,895.41 9.635
Maximum 138,202.06 6.285 Range Width 156,845.17 14.809
Range Width 138,856.87 4.973 Mean Std. Error 467.48 0.017
Mean Std. Error 401.56 0.014
303
Table 10. Probability of gaining NPV and IRR using a normal  The risk factors of PPP transportation projects should
distribution be carefully investigated before construction to protect
Probability NPV (million VND) IRR (%/year)
the original project efficiency.
 The methods used to respond to PPP transportation
0% -54,949.76 -5.175
projects' risks need to be proposed based on the critical
10% 8,886.78 1.716 risks.
20% 30,405.42 2.580 ACKNOWLEDGMENT
30% 45,084.22 3.138 We would like to thank the respondents who participated
in the survey for their help with the questionnaire. We are also
40% 57,646.34 3.598
grateful to the Management Board of the Vinh Loc Bridge
50% 69,786.18 4.030 Project in An Giang Province for providing the necessary data
for the case study. We especially thank Mr. Vo Phu Hien for
60% 81,486.68 4.435
his help with the data collection.
70% 94,197.00 4.865
REFERENCES
80% 109,186.23 5.360
[1] Wikipedia, Public-Private Partnership, retrieved March 28,
90% 128,899.01 5.992 2022,
https://en.wikipedia.org/wiki/Public%E2%80%93private_part
100% 197,895.41 6.635 nership
[2] T. N. H. Nguyen, “Investment in the forms of PPP in
In addition, the financial indices, including IRR and PP transportation infrastructure projects in Vietnam”, Research
have also been analyzed to explore the investment safety in project No. B2011-06-04, Ministry of Transportation, 2013.
the operation stage of the project (see Table 11). In detail, [3] M. Barnier, Guidelines for successful Public Private
IRR/MARR = 4.55 and PP/Expected PP = 0.63. Overall, these Partnerships, retrieved October 15, 2018
http://europa.eu.int/comm/regional_policy/sources/docgene
indices are quite high compared to the general level of r/guides/PPPguide.htm
construction projects.
[4] P. Burger, J. Tyson, I. Karpowicz and M. D. Coelh, “The
Effects of the Financial Crisis on Public-Private Partnerships”,
Table 11. Financial indices at payback timing IMF, International Monetary Fund, pp. 3-22, 2009.
[5] P. H. Vo, Simulation of risks of PPP construction projects in
Index Unit Value An Giang, Master thesis, Ho Chi Minh City University of
IRR %/year 6.085
Technology and Education, 2019.
[6] T. T. G. Huynh, Public-Private Partnership for the
PP year 21.3 infrastructure development of Vietnamese transportation
projects, Ph.D. dissertation, Ho Chi Minh City University of
IRR/MARR 4.55 Economics, 2010.
PP/Expected PP 0.63 [7] T. B. N. Phan, “PPP: A solution for the problem of the
infrastructure development of urban transportation in Ho Chi
Minh City”, Journal of Development and Integration, vol. 10,
V. CONCLUSION AND RECOMMENDATIONS no. 20, pp. 18-25, 2013.
Compared with other construction projects, a construction [8] T. N. H. Nguyen and H. M. Le, “Development of PPP forms
project in the form of PPP also contains critical risks during in construction and investment for infrastructure of
transportation projects in Vietnam”, Journal of Economics and
and after the construction phase. These risks significantly Development, vol. 168, pp. 3-7, 2011.
affect the project's performance. Using a survey by a [9] T. A. Trinh, “Study on solutions of risk management in
structured questionnaire with two hundred intently selected transportation construction projects in Vietnam”, Ph.D.
personnel, this study showed that the probability of the effect dissertation, Hanoi University of Transportations, 2006.
of the risks on the efficiency of PPP transportation projects is [10] W. Shouqing and R. L. K. Tiong, “Evaluation and management
relatively high, from 42.0% to 61.8%. The top three risk of foreign Exchange and revenue risks in China’s BOT
factors in the construction stage are "Lack of skilled workers", projects”, Construction Management and Economics, vol. 18,
pp. 197-207, 2000.
"Missing bid quantity in the tender", and "Inconsistent norms
[11] I. Y. Mohammed, K. Bala and S. U. Kunya, “Risk allocation
of construction works". preference in Public-Private partnership infrastructure projects
Through a case study PPP project, the impact of risk on in Nigeria,” Journal of Engineering and Applied Science, vol.
4, pp. 34-42, 2012.
the project performance has been simulated and quantified. In
[12] B. Li, A. Akintoye, P. J. Edwards and H. Cliff, “The allocation
detail, the risk level of the project performance is about 9.23% of risk in PPP/PFI construction projects in the UK”,
compared to the planned performance in the construction International Journal of Project Management, vol. 23, pp. 25-
stage. Moreover, at the payback timing, the probability of 35, 2005.
gaining the NPV equal to the initial investment capital is [13] T. Sachs, R. Tiong and W. Shouqing, “Analysis of political
94.8%. In addition, the investment safety coefficient related to risks and opportunities in public private partnerships in China
IRR and PP of this project is relatively high. and selected Asian countries”, China Management Study, vol.
1, no. 2, pp. 126-48, 2007.
Based on the analysis results, this study recommends [14] C. Checherita and J. Gifford, “Risk Sharing in Public- Private
several issues as follows: Partnerships: General Considerations and an Evaluation of the
U.S. Practice in Road Transportation”, 11th World Conference
 Investors should join in PPP projects much more on Transportation Research (WCTR), University of California,
because they have a high rate of return. Berkeley, CA, pp. 24-28, 2007.
304
Fabrication of Graphene Oxide from the Graphite

Rod of a Disposed Battery
Huy-Binh Do, member, IEEE Hoang-Trung Huynh Thien-Trang Nguyen
Department of Materials Technology Department of Materials Technology Department of Materials Technology
binhdh@hcmute.edu.vn
https://orcid.org/0000-0003-3274-5050
Van-Cuong Pham Tien-Luat Nguyen Anh-Vu Phan-Gia

Department of Materials Technology Department of Materials Technology Department of Materials Technology
Maria Merlyne De Souza

EEE Department
University of Sheffield
Sheffield S13JD, UK
Abstract: We report a new electrochemical exfoliation subsequently mixed and stirred for 10 minutes. An anode and
method to produce graphene oxide (GO) on a large scale from a cathode were obtained from disposed graphite rods with the
disposed graphite rod, recycled from a zinc-carbon battery. diameter of 6 mm (Pinaco, the eagle brand). The tip of the
GO is exfoliated with plasma from the tip of the cathode. The cathode was prepared by grinding to a diameter of ~ 200 m.
properties of GO are characterized by Raman, SEM and EDX All electrodes were cleaned in acetone, isopropanol (IPA), and
analysis. A moderate ratio of ID/IG  0.75 in Raman spectra annealed in 24 hours at 130oC.
indicates that the product includes a mix of GO and graphite.
Keywords: graphene oxide, plasma-assisted exfoliation,

battery recycling, disposed graphite
I. INTRODUCTION
Since the discovery of graphene in 2014, two-dimensional
(2D) materials are attracting considerable interest, for
example, as ribbon FETs to be introduced in mainstream
CMOS. Graphene exhibits outstanding mechanical, thermal
and electronic properties [1] and can be synthesized by a
variety of methods such as Chemical vapor deposition (CVD),
epitaxial growth, scotch tape method, and exfoliation of
graphite in solution. CVD methods can be employed at scale
[2], however, mass production of graphene by CVD has still
proved to be a challenge due to the high cost of production in
an inert environment at high temperature. Plasma-assisted
electrochemical exfoliation is a highly efficient method to
fabricate graphene/graphene oxide from graphite [3]. The
purpose of this study is to investigate a continuous production
of GO from recycled graphite, recovered from a zinc-carbon
battery. The safe disposal of batteries is the leading cause of
environmental concern, due to pollution from toxic metals and
graphite rods [4]. The advantage of reutilization of graphite
rods from battery is reduced pollution as well as cheap and
sustainable methods of production of graphene/graphene
oxide (GO).
II. EXPERIMENTAL Figure 1. An electrochemical cell, indicating the placement of
The electrolytic solution in this work was prepared by the cathode and anode. The cathode can be moved freely along
dissolving KOH in 200 ml H2O and (NH4)2SO4 in 40 ml H2O. an axis perpendicular to a bottom of the chamber. Plasma occurs
KOH and (NH4)2SO4 solutions were prepared separately, at a surface when inside the electrolytic solution.

The experiment was conducted in an electrochemical III. RESULTS AND DISCUSSION

chamber illustrated in Fig. 1(a), with a cap designed to have a Fig. 2 shows the field emission scanning electron
freely moving cathode. All electrodes were connected to a DC microscope (FESEM) images of graphite powder, directly
power supply with the bias voltage increased gradually to 60 ground from the graphite rod, and exfoliated graphene oxide
V. The initial temperature of the chamber was recorded at collected after filtration, for comparison. A stacked layer
30oC. To exfoliate graphene, the cathode was slowly moved structure of graphite is observed in graphite powder in Fig.
downwards in steps of 20 m until a plasma occurred around 2(a), whereas the plasma-assisted electrochemical exfoliation
the tip of the cathode (Fig. 1(a)). At this moment, several creates randomly crumpled nano-sheets, closely associated
bubbles were released at the anode and gas was exhausted out with each other due to strong  bonds in Fig. 2(b). The
of the chamber as shown in Fig 1(b). During the experiment, exfoliated product includes several nanometer-scale plates
GO was exfoliated and floated on the surface of the electrolyte (Fig.2(b)). At high magnification, GO sheets are observed to
solution, and the plasma was maintained for more than 1 hour. be scrolled as indicated by red arrows in the inset of Fig. 2(b).
GO was washed in deionized (DI) water and collected using The purification of GO after filtration, is determined in EDX
vacuum filtration through Poly(vinylidene fluoride) (PVDF) spectrum as shown in Fig. 3. It is found that oxygen dominates
membrane with an average pore size of 0.45 m. It was  10% in weight in filtration products, while other metals,
subsequently dried at 50oC for 2 hours. Hitachi FE-SEM S- including Al, Si, K, and Ca, are observed within the marginal
4800 with Horiba EDX system was employed to obtain errors of EDX measurement (less than 2%). There are no toxic
Scanning electron microscope (SEM) and energy dispersive materials observed.
X-ray spectroscopy (EDX). Fourier transform infrared (FTIR)
spectra were recorded by Thermo Nicolet 6700 FT-IR optical
spectrometer, and Raman spectra were measured by Horiba
XploRA ONE system.
(a) (b)
Figure 2. SEM image of (a) graphite and (b) graphene oxide obtained from plasma-assisted exfoliation method. Inset of figure b is
a magnification of this figure. A stacked layer and bulk structure of graphite is observed in figure a, while crumped nano-sheets are
obtained in figure b. Graphene oxide is scrolled illustrated by red arrow in the inset of figure b.
Fig. 4 shows the FTIR spectra of graphite power and

graphene oxide. The highly intense peak at 3432.7 cm1 is due
to the O-H stretching vibration, indicating the present of OH
or COOH functional groups. The symmetry and asymmetry
stretching vibration of C-H bond is clearly observed at 2848.4
and 2917.8 cm1. Very weak stretching vibration is observed
for the carbonyl C=O peak at 1733.7 cm1. The peak at 1635.4
cm1 is assigned to an un-oxidized C=C stretching vibration.
The most intense peak is at 1033.7, which is related to the C-
O stretching vibration of the alkoxy group. The results
obtained from FTIR (Table 1) indicate the present of several
oxygen containing functional groups in agreement with those
reported in [5-9]. The FTIR spectrum of the graphene oxide
Figure 3. EDS spectrum of GO after filtration. There is no powder is similar to that of the graphite powder, indicating
toxic metal observed in the samples. that the graphite rod is oxidized before exfoliation.
306
highest intensity is at 1568.8 cm-1, representing the G band,

due to the first order scattering of the E2g phonon of sp2 C
Graphite powder
atoms [11]. All sp2 carbon materials such as graphite,
Graphene oxide
Transmitance (a.u)
amorphous carbon, carbon nanotube, and graphene exhibit the
G band. Weak bands are related to D band at 1342.3 cm-1 and
2D band at 2682.3 cm-1. The D band is known as a defect band
derived from edge effects, structural defects, and the sp2 bond
that breaks the symmetry. The 2D peak is due to the linear
C-H dispersion of its electronic bands, giving rise to triple or
C=O
1733.7 2848.4 double resonance. The Raman spectra of graphene oxide
C=C C-H depend not only on the phonon properties but also on their
1635.4 2917.8 electronic structures. In Fig. 5, the frequencies of D, G, and
OH
3432.7
2D bands are shifted to higher values, reaching 1354.4,
C-O
1033.7
1581.6, and 2703.2 cm1, respectively, in agreement with
previous reports [5, 12]. It is seen that G and D bands change
750 1500 2250 3000 3750 significantly during the exfoliation of graphite to graphene
oxide, illustrated by a broadening and a right shift of these
Wave number (cm1) bands shown in Fig. 5. The right shift of the G band is
supposed to be the higher resonance frequency of isolated
double bonds present in a single layer GO [13]. An increase
Figure 4. FTIR spectra of graphite powder (obtained from the in ID/IG ratio from 0.18, corresponding to graphite (black line),
disposed graphite rod in battery) and graphene oxide. Many to 0.75, for graphene oxide (red line), in Fig. 5 indicates that
oxygen containing functional groups were observed. the size and crystallinity of graphitic materials decreases,
increasing in isolated graphene domains [11, 12]. The
Table 1. Assignment of characteristics FTIR absorption observed results are in agreement with those reported in [5].
bands of G and GO The results indicate that graphite has been exfoliated to GO. It
No. Wave number (cm1) Functional groups
is also known that the intensity ratio of the D and G bands
(ID/IG) helps to estimate the defects in GO sample, where
1 3432.7 –OH Stretching Vibration higher ratio implies more defects in the graphene sheet.
2917.8 Symmetric and asymmetric
2
2848.4 stretching vibration of –CH2– IV. CONCLUSION
3 1733.7 –C=O stretching vibration In conclusion, the disposed graphite of a battery is used to
4 1635.4 C=C stretching Vibration produce graphene oxide. The process is designed to work for
more than 1 hour. The SEM, EDX, and Raman characteristics
5 1033.7 C–O stretching of alkoxy group verify the obtained GO. There are no toxic elements observed
in the exfoliated GO, so this study opens a new way to
sustainably produce cheap GO and protect the environment
from battery waste.
1564.8 Graphite powder
1581.6
Graphene oxide ACKNOWLEDGMENT
1354.4
1342.3 This work was supported by Ho Chi Minh City University
Intensity (a.b)
of Technology and Education, Vietnam.

2682.3 2703.2
REFERENCES
ID/IG = 0.75
[1] M. Pumera, "Graphene-based nanomaterials for energy
storage", Energy & Environmental Science,
10.1039/C0EE00295J vol. 4, no. 3, pp. 668-674, 2011, doi:
10.1039/C0EE00295J.
ID/IG = 0.18
[2] J. Zhao, S. Pei, W. Ren, L. Gao, and H.-M. Cheng, "Efficient
Preparation of Large-Area Graphene Oxide Sheets for
1000 1500 2000 2500 3000 3500 Transparent Conductive Films", ACS Nano, vol. 4, no. 9, pp.
5245-5252, 2010/09/28 2010, doi: 10.1021/nn1015506.
Raman shift (cm1) [3] D. Van Thanh, L.-J. Li, C.-W. Chu, P.-J. Yen, and K.-H. Wei,
"Plasma-assisted electrochemical exfoliation of graphite for
rapid production of graphene sheets", RSC advances, vol. 4,
no. 14, pp. 6946-6949, 2014.
Figure 5. Raman spectra of graphite powder (black line) and [4] H. Al Mahbub, N. Deb, M. R. Abedin, S. Abedin, and M. S.
graphene oxide (red line). Spectrum of graphene oxide
Khan, "Metal recovery from waste dry cell batteries",
broadens and shifts to the right, indicating the present of isolate
Proceedings of the waste safe, pp. 1-9, 2015.
double bond in single layer graphene oxide.
Figure 5 shows Raman spectra of graphite and exfoliated [5] M. Bera, P. Gupta, and P. K. Maji, "Facile one-pot synthesis
of graphene oxide by sonication assisted mechanochemical
graphene oxide, which reveals structural changes during the
approach and its surface chemistry", Journal of nanoscience
chemical conversion from graphite to graphene oxide. The and nanotechnology, vol. 18, no. 2, pp. 902-912, 2018.
principle is that C=C double bonds give rise to high intensity
in the Raman spectra [10]. For graphite (black line), the
307
[6] D. C. Marcano et al., "Improved Synthesis of Graphene [11] N. A. Kumar, H.-J. Choi, Y. R. Shin, D. W. Chang, L. Dai,
Oxide", ACS Nano, vol. 4, no. 8, pp. 4806-4814, 2010/08/24 and J.-B. Baek, "Polyaniline-grafted reduced graphene oxide
2010, doi: 10.1021/nn1006368. for efficient electrochemical supercapacitors", ACS nano, vol.
[7] D. R. Chowdhury, C. Singh, and A. Paul, "Role of graphite 6, no. 2, pp. 1715-1723, 2012.
precursor and sodium nitrate in graphite oxide synthesis", Rsc [12] C. Fu, G. Zhao, H. Zhang, and S. Li, "Evaluation and
Advances, vol. 4, no. 29, pp. 15138-15145, 2014. characterization of reduced graphene oxide nanosheets as
[8] J. Song, X. Wang, and C.-T. Chang, "Preparation and anode materials for lithium-ion batteries", Int. J. Electrochem.
Characterization of Graphene Oxide", Journal of Sci, vol. 8, no. 5, pp. 6269-6280, 2013.
Nanomaterials, vol. 2014, p. 276143, 2014/03/11 2014, doi: [13] S. Stankovich et al., "Synthesis of graphene-based nanosheets
10.1155/2014/276143. via chemical reduction of exfoliated graphite oxide", carbon,
[9] Y.-M. Long et al., "Shifting and non-shifting fluorescence vol. 45, no. 7, pp. 1558-1565, 2007.
emitted by carbon nanodots", Journal of Materials Chemistry,
vol. 22, no. 13, pp. 5917-5920, 2012.
[10] B. Paulchamy, G. Arthi, and B. Lignesh, "A simple approach
to stepwise synthesis of graphene oxide nanomaterial," J
Nanomed Nanotechnol, vol. 6, no. 1, p. 1, 2015.
308
Design and Optimization of a Compliant

Mechanism for Vibration-Assisted Drilling
Hai-Thanh Nguyen Van-Khien Nguyen Phan-Khanh-Tam Nguyen
Ho Chi Minh City University of Ho Chi Minh City University of Ho Chi Minh City Technical and
Technology and Education Technology and Education Economic College
ORCID: 0000-0002-8987-6185 ORCID: 0000-0001-7274-2810 nguyenphankhanhtam@hotec.edu.vn
Huy-Tuan Pham Quang-Khoa Dang Pham Son Minh

ORCID: 0000-0002-8987-6185 ORCID: 0000-0003-0615-6180
Abstract: Compliant Mechanisms (CMs) are external VADrilling incorporates the piezoelectric actuator into cutting
force-based dynamic and variable mechanisms based on the tools and it operates according to the principle of resonance
elastic deformation of materials with advantages such as [4-7]. This requires a certain degree of sophistication in design
compactness, high accuracy, ease of fabrication, and as well as operation where the natural frequency must match
application in many different fields. In this work, a case study the vibration frequency of the PZT. Some other studies
of using flexible hinges in drilling applications for a non- suggest a more approachable solution in which cutting tools
resonant vibration-assisted drilling application is proposed. are integrated into the PZT. These stay stationary in operation
The geometric design of the mechanism was optimized using in a fashion similar to VATurning [8-11].
an FEA based on an evolutionary algorithm optimization To overcome the disadvantages of the resonance-based
(NSGA-II) procedure, aiming to maximize the natural processing method mentioned above, a new compliant
frequencies. The optimization procedure is coded in mechanism (CM) design that could facilitate the vibration
MATLAB whereas the features in ANSYS ADPL support of the workpiece during a drilling operation is proposed in
and enhance the calculation process. Finite element analysis this paper. This design uses a PZT as a source of vibration
(FEA) was performed to verify the accuracy of the design and operates according to the non-resonance principle in
model. Figure 1. The geometrical design of the CM is optimized
based on the NSGA-II optimization procedure to obtain a
Keywords: compliant mechanism, FEM, cptimization, structure inheriting as large as possible natural frequency.
genetic algorithm, vibration-assisted drilling
I. INTRODUCTION
Vibration assisted machining (VAM) is a hybrid
machining method that combined small amplitude vibration
into the movement of the cutting tool or workpiece during
processing machining [1,2]. The above combination brings
many benefits such as the reduction of cutting force and heat
generated during processing, the minimization of burr
formation. VAM could help to improve the quality of the
processing surface for brittle materials with high hardness as
well as hard-to-cut materials [3]. Among traditional methods
of processing such as turning, milling, drilling, and grinding,
drilling is a seemingly simple process compared to other
methods in which it consists only of rotational and purification
motion. However, the actual process of machining and chip Figure 1. Vibration-assisted machining structure
removal occurs simultaneously in the same space. This leads 1) Cutting tool. 2) Worktable. 3) Frame. 4) Spacer. 5) Toggle Clamp.
to a huge impact on surface accuracy and quality due to 6) Sliding shaft. 7) Flexure hinge. 8) Piezoelectric actuator
unfavorable chip drainage conditions.
II. PROBLEM FORMULATION
Therefore, studies to overcome the above shortcomings
have been carried out, one of which deals with vibration- A. Design Concept
assisted drilling (VADrilling), which proves to have many In this section, a 1-DOF micro-positioning structure is
outstanding advantages. Most of the current research on developed to integrate into a drilling machine that enables

the vibration of the workpiece. Due to the Z-axis movement structure away from destruction due to the frequency
of the structure, a corner-filleted flexible hinge of a matching effect. This is the most essential objective function.
rectangular cross-section [12] is used. Fig. 2 shows the
geometric parameters of a filleted hinge. Next, the stress (𝜎𝑚 ) is considered the second important
objective function of the optimal cycle because the elements
of stability and fatigue failure are the main concern for all
compliant mechanisms.
The optimization problem of the VADrilling platform is
formulated in Table 1 with two objective functions, five
design variable, and four constraint functions.
Table 1. The formulation for the design optimization problem
1. Objectives:
- 𝑓1 = 𝑀𝑎𝑥(𝑓0 ) (1)
- 𝑓2 = 𝑀𝑖𝑛(𝜎𝑚 ) (2)
2. Design variables: L1, L2, T1, T2, R1, R2

3. Constraints:
- Design space: W × H = 200 × 200 (mm)
Figure 2. Design parameters of the filleted hinge i. Design variable limits (mm)
Fig. 3 shows a 1-DOF micro-positioning structure which 12 ≤ L1 , L2, T1, T2 ≤ 20 (3)
includes a frame, a work-table, sliding shafts, a piezoelectric
1 ≤ R1 , R2≤ 4 (4)
actuator, and flexure (filleted) hinge elements. In this concept,
the frame and the worktable connected via 16 flexure hinges ii. Input stiffness:
evenly distributed on each edge to increase rigidity. The PZT
actuator will transmit the vibration directly to the worktable. 𝐾𝑖𝑛 ≤ 0.15𝐾𝑝𝑧𝑡 (𝐾𝑝𝑧𝑡 = 480N/μm)
The optimization of CMs is according to the flow chart

process in Fig. 4. Through genetic algorithms, MATLAB
and ANSYS are used to optimize the size of the structure.
ANSYS is used to simulate the behavior of the structure.
The facial element model is used to check binding
conditions and target functions. SOLID185 element is used
to analyze and simulate the design problem. The material of
the structure is aluminum alloy (Al 7075 - T6) with the
properties of the material as follows: Young’s modulus E =
71.7 GPa, Poisson’s coefficient is 0.33, yield strength is 503
MPa, and density ρ =2810 kg /m3.
The total number of elements in FEM for the design of
the problem is 127.000 elements. During the process of
optimization, the number of generations is set to 100 and the
population of each generation is 20. The number of
variables used in this structure is 6 variables. The order of
the variables is expressed specifically in Table 2.
III. RESULT & DISCUSSIONS
ANSYS Mechanical ADPL is used in this study to
simulate displacement and stress of the structure. The
Figure 3. Design parameters of a 1-DOF for VADrilling Machine
simulation results help us to verify and assess the
performance of the structure.
B. Design Optimization
The optimal design shown in Table 2 is reconstructed in
Solidworks and imported into ANSYS. The purpose of this
The flexure hinge system has been designed and simulated step is to re-examine the results of stress, displacement and
with Ansys APDL combined with NSGA-II to achieve natural frequency for optimal design. The simulation results
optimum performance. The natural frequency should be as are shown in Fig. 5, Fig. 6, and Fig. 7. It can be seen that
large as possible to ensure that the structure is not affected by with the input displacement of the structure being 𝐷𝑖𝑛 =
the resonance phenomenon and works well in operation. The 14 μm, the output of the structure is𝐷𝑜𝑢𝑡 = 14 μm. The
prevention of resonance possibility is to keep the whole
310
maximum Von-mise stress 𝜎𝑚𝑎𝑥 = 77.36 MPa meets the maximum stress σ𝑚𝑎𝑥 = 77.36MPa satisfies the allowable
durability requirements of the material. strength constraints of the material. Fig. 7 shows that the first
natural frequency 𝑓0 = 1326.9 Hz matches the objective of
the problem. Besides, it can be seen that the parasitic motion
in the x,y-direction is completely eliminated. This indicates
satisfactory rigidity in the structure, which is not disturbed by
the input displacement of PZT and facilitates the process of
fabrication.
Figure 5. Von-Mises Stress distribution
Figure 4. The flowchart for multi-objective optimization using GAs

Figure 6. Total displacement of the mechanism with z-input from
PZT
Table 2. Values of the optimum design variables.
Design variables Values
L1 12 mm
L2 16 mm
T1 02 mm
T2 10 mm
R1 03 mm
R2 03 mm
The graph of the relationship between input

displacement (z𝑖𝑛 ) and stress (σ) of the structure is shown in
Fig. 8. According to the graph, the simulation result is a linear
Figure 7. Natural frequency at Mode shape 1
function in the PZT input range from 0 μm to 14 μm and the
311
IV. CONCLUSION REFERENCES

This study presents the 1-DOF micro-positioning structure [1] Gu, Y. and et al., Micromachines, 2018, vol. 9, 499.
using flexure compliants utilized for VADrilling machining. [2] D. E. Brehl and T. A. Dow, "Review of vibration-assisted
The design parameters of the hinges were optimized based on machining", Precision Engineering, vol. 32 (2008), no. 3, pp.
finite element technique combined with NSGA-II. The 153-172.
optimal natural frequency, output displacement and the stress [3] “Optimization Design of a 2-DOF Compliant Parallel
Mechanism Using NSGA-II Algorithm for Vibration-Assisted
were discovered to be 1326.9Hz, 14µm and 77.36MPa, Milling”, Physics and Mechanics of New Materials and Their
respectively. Applications 2020, pp 332-338.
In the next step, the design structure will be fabricated to [4] Todic Rajko, Bartulovic Ante, “Vibration assisted deep hole
micro-drilling: a preliminary experimental study”, Annals of
examine the influence of the vibration effect on the machining DAAAM & Proceedings (2016), pp. 822+.
performance. [5] B. Azarhoushang, J. Akbari, “Ultrasonic-assisted drilling of
Inconel 738-LC”, International Journal of Machine Tools &
Manufacture, vol. 47 (2007), pp. 1027–1033.
[6] Vahid Ostad Ali Akbari, Keivan Ahmadi, “Substructure
analysis of vibration-assisted drilling systems”, The
International Journal of Advanced Manufacturing Technology,
vol. 113 (2021), pp. 2833-2848.
[7] Dang Q.K., Chang P.L., Dang T.N., Weng F., Uan J.Y., Wang
D.A., “Ultrasonic spot welds of gas diffusion layer to proton
exchange membrane of fuel cells”, Journal of Materials
Processing Technology, vol. 266 (2019), 208-216.
[8] Hung Chu Ngoc et al., “A Step-by-Step Design of Vibratory
Apparatus for Ultrasonic-Assisted Drilling”, International
Journal of Advanced Engineering Research and Applications,
Vol. 4-3 (2018), pp. 139-148.
[9] J. Ramkumar, S. K. Malhotra, and R. Krishnamurthy, “Effect
of workpiece vibration on drilling of GFRP laminates”, Journal
of Materials Processing Technology, vol. 152 (2004), pp. 329-
332.
[10] P. Mehbudia, V. Baghlania, J. Akbaria,b*, A.R. Bushroab,
N.A. Mardi, “Applying ultrasonic vibration to decrease
Figure 8. Parasitic motion and stress result by simulation drilling-induced delamination in GFRP laminates”, Procedia
CIRP, vol. 6 (2013), pp. 577 – 582.
ACKNOWLEDGMENT [11] Kien H.T. et al., “Theoretical and Experimental Study of
This work belongs to the project grant No: B2021-SPK-02 Vibration-Assisted Turning”, International Conference on
funded by Ministry of Education and Training, and hosted by Green Technology and Sustainable Development 2020, pp. 37-
44.
Ho Chi Minh City University of Technology and Education,
[12] Nicolae Lobontiu, “Compliant mechanisms: design of flexure
Vietnam. hinges”, CRC Press LLC (2021).
312
Eco-friendly Adsorbent Formulated from Rubber

Shell Biochar to Remove Chromium (Cr(VI)) from
Aqueous Solutions
Linh My Nguyen
Department of Environmental Technology
Ho Chi Minh City
University of Technical and Education (HCMUTE)
linhnm@hcmute.edu.vn
Abstract: This research aims to create a new Many physicochemical methods have been developed
environmentally friendly adsorbent material that can lately to eliminate heavy metal ions from water sources such
effectively remove chromium from aqueous solutions. The as chemical precipitation, membrane, ion exchange,
optimal parameters for operation such as pH, contact time, coagulation/flocculation, and extraction processes. These
adsorbent dose, and temperature were investigated. This methods all proved efficiency; however, the ion exchanging,
study chemically prepared the biosorbent from rubber seed and membrane filtration technologies usually require high
shell (Hevea brasiliensis) biochar coated with the operating cost. Although the chemical precipitation process
glutaraldehyde crosslinked chitosan and modified with has been used widely due to its ability to fulfill all the factors,
succinic acid in order to attach functional groups into the it is not eco-friendly as it still produces a large amount of
material (CRSC). It showed that the optimal pH for the secondary solid waste that needs to be further treated.
Cr(VI) adsorption process is 3,0. The pseudo-first-order Adsorption, a method that could be able to combine the
reaction and pseudo-second-order reaction kinetic models advantage of both using adsorbent and recycling agricultural
were used to correlate the experimental data and determine residue, has notably attracted attention lately [6, 15-18].
the kinetic parameters. The experimental data have been Rubber plants are grown a lot in the Southeast region (46.4%)
analyzed using the Langmuir and Freundlich models. Results of Vietnam, mainly Binh Phuoc, Binh Duong, Tay Ninh,
suggested that a pseudo-second-order equation could Vung Tau. Rubber productivity has continuously increased in
satisfactorily fit the adsorption kinetics and the Freundlich the past 10 years, Vietnam is now one of the leading countries
model described the result better than the Langmuir one. The in the world in terms of rubber yield and export. Rubber seed
adsorption capability value (qmax) for Cr is 41,152 mg/g. The shells are abundant and usually burnt to dispose, which could
thermodynamic parameters, such as Gibbs free energy not utilize the potential nor bring more value out of the ago-
change (∆G), enthalpy change (∆H), and entropy change waste. Recycling this waste to create new environmentally
(∆S), were calculated. The negative value of ∆G indicated the friendly materials should be considered as an appropriate
spontaneous nature of Cr(VI) adsorption onto CRSC method to approach this problem.
biosorbent.
Thus, the objective of this research is to examine the
Keywords: Adsorption, crosslinked chitosan, heavy capacity of the new composite adsorbent, which was
metals removal, rubber shell biochar, thermodynamic chemically formed by combining rubber seed shell-derived
biochar and chitosan, to adsorb Cr(VI) in different conditions
I. INTRODUCTION such as pH, agitation time, and temperature. The efficiency
In current years, contamination of heavy metals in was further analyzed by studying the adsorption isotherm.
aqueous solutions, which mainly comes from escalating
economic development such as mineral mining, metal
plating, photographic industries, and battery manufacturing, II. CHEMICALS AND METHOD
is difficult to remove completely by conventional wastewater A. Chemicals
treatment systems[1-6]. Studies showed that if the
concentration of heavy metals exceeded the permissible limit Rubber shell was collected from the field in Binh Phuoc
of the World Health Organization, it would lead to many province, Vietnam. Raw material was then cleaned with tap
serious diseases, and threaten our health. For instance, water and ground into small pieces and dried at 60oC within 1
exposure to chromium may affect lung and bone, excess lead day in the oven. Chromium standard solution (1000 mg/L)
could result in brain, kidneys, and nervous system damage, was provided by Merck Co. Commercial chitosan with 85%
while drinking water that contains hexavalent chromium is degree of deacetylation and glutaraldehyde (25%) solution in
associated with lung cancer, etc[7-14]. Therefore, the water was provided by Sigma Aldrich Co. All other reagents
wastewater treatment should be conducted strictly and used in this study were analytical grade.
effectively to remove all the trace heavy metals that may be
present in the water.

B. The procedure for material preparation of two materials PSC, CCS and has adsorption capacity.
The procedure to prepare the materials was followed by Especially, the peak 1643.3 cm – 1 shifted to 1652.9 cm – 1.
some previous studies[19-21]. Rubber seed shell was Functional groups on the material include –OH, –COO, –NH2,
calcinated at 500oC within 4 hours in a furnace. After cooling –COOH, –C – H, –CH2. As shown in Figure 1 that CRSC3 (e)
down at room temperature, the collected coal was ground and after desorption, the material tends to recover the same peaks
sieved into 0.015mm. The materials were well mixed with as CRSC1, the metal peaks in intensity have decreased
H3PO4 (1:3) in 24 hours and regenerated in microwave at 450 compared to the measured CRSC2.
W in 10 minutes. A mixture of 10 g -chitosan, 1g -biochar and
240 mL acid acetic (7%) was soaked overnight and dropped (a)
into 200 mL of an alkali coagulating solution
(H2O:methanol:NaOH = 4:5:1, w/w/w) to prepare highly (b)
swollen spherical beads.
(c)
The cross-linking process was carried out by adding wet
beads (200 mL in volume) into a glass beaker containing (d)
glutaraldehyde solution. The final beads (CRSC) were (e)

washed several times with distilled water to remove the
glutaraldehyde residues and dried at 70oC for 24 hours.
C. Batch experiments Figure 1. FTIR spectra of (a) Chitosan (CCS), (b) Modified
Adsorption experiments were performed by batch charcoal at 500oC (PSC), (c) Chitosan combined with rubber seed
shell biochar before adsorption (CRSC1) ), (d) Chitosan combined
equilibration method. Stock solution containing 1000 mg/L with rubber seed shell biochar after adsorption (CRSC2), (e)
was prepared. Batch adsorption experiments in duplicate were Chitosan combined with rubber seed shell biochar after desorption
carried out by mixing 2 g/ L of biosorbent with 100 mL (CRSC3)
solution. The contents were shaken thoroughly using an
orbital shaker incubator at a speed of 200 rpm at 303 K. The D. Effect of pH on the adsorption capacity
solution was then filtered, and residual chromium Figure 2 shows the effect of initial pH on adsorption of
concentration was determined by Atomic Absorption CRSC. The amount of removed metal ion decreased with the
Spectrophotometer. The pH measurements were carried out increasing of pH from 1 to 7. The CRSC's ability to remove
with Horiba pH meter. The effect of agitation time was Cr(VI) was maximum at pH 3.0 with an adsorption capacity of
conducted at 303K and under a stirring rate of 200 rpm with 2 14.1 mg/g at initial Cr(VI) of 50 mg/L. The result is the same
g/ L of biosorbent. The effect of pH on the adsorption of as previous studies in which the optimum pH is low from 1.0-
Cr(VI)was studied in a pH range of 1.0- 7.0 by adding 3.0 for the removal of Cr6+ [9, 11, 22, 23].
appropriate buffers.
The amount of Cr(VI)adsorbed was calculated according
to the following equation:
(𝑪𝟎−𝑪𝒆)𝒙𝑽
𝒒𝒆 = (1)
𝒎
where qe is the amount of Cr(VI)adsorbed onto CRSC (mg/g),
C0 and Ce are the Cr(VI)concentrations in the solution initially
and at equilibrium (mg/ L), respectively, V is the volume of
the solution (L), and m is the mass of biosorbent used (g).
C. Material characteristics
Figure 1 shows the FTIR spectrum of the CCS (a) obtained
with the Chitosan signals as at the peak 3402,3cm-1
(oscillations of the O-H and N-H2 bonds overlap). This peak Figure 2. The effect of initial pH on adsorption capability of
is more pronounced in CRSC1(c) indicating enhanced CRSC
hydrogen bonding. In particular, the presence of strong
E. Effect of agitation time
intensities at 1643,2cm − 1, (CRSC1) and 1658,9 cm − 1, 1378,4
cm − 1 has a combined carboxylate functional group. The peaks The effects of agitation time on the adsorption of Cr(VI)
at 1056,7 cm − 1 and 1028.9 cm − 1 in CRSC1 show C–H on the CRSC as well as chitosan and pristine rubber seed shell
groups. The FTIR spectral absorption band after adsorption of biochar are illustrated in Figure 3. A rapid adsorption occurs
metal ions decreases in intensity indicating a shift to the low within the first 60 min. The adsorption equilibrium is
frequency bands showed that there is linkage between completely obtained after 120 mins. Therefore, 120 mins of
functional groups and metal oxide as shown in Figure 1. contact time is chosen as the optimal contact time.
CRSC2(d) has weak intensity adsorption peak of 894,7 cm – 1.
The changes in spectra after adsorption of metal ions showed
that the successful preparation of materials has the properties
314
The Gibbs energy (∆G) (kJ/mol), the enthalpy change

(∆H) (kJ/mol), and the entropy change (∆S) (J/mol K) of an
adsorption process can be estimated by [31]
𝛥𝐺 = −𝑅𝑇 𝑙𝑛 𝐾𝐿 (5)
where T is the absolute temperature (K) and R is the universal
gas constant (8.314 J mol-1 K-1). The values of ∆H and ∆S can
be obtained from the slope and intercept of the graph of ln KL
against 1/T based on the van’t Hoff equation
𝛥𝑆 𝛥𝐻
𝑙𝑛 𝐾𝐿 = − (6)
𝑅 𝑅𝑇
Table 2 shows the thermodynamic parameters of the

adsorption process of Cr6+ on CRSC. The negative ∆G values
Figure 3. The effect of agitation time on adsorption of CRSC
which indicated the metal ions adsorption on the CSRS was
F. Adsorption isotherm and thermodynamic spontaneous. The values of ∆H and ∆S were -32,91 and -
The following equation described the linear form of the 0,0998, respectively. Both enthalpy change (∆H) and entropy
Freundlich model [2, 24-26]. change value (∆S) reveal the spontaneous at low temperature.
𝑙𝑛 𝑞𝑒 = 𝑙𝑛 𝐾𝐹 + (1⁄𝑛) 𝑙𝑛 𝐶𝑒 (2) Table 2. the thermodynamic parameters of the adsorption of Cr6+ on

CRSC at pH 3.0
where KF is a constant indicating the relative adsorption
capacity of the biosorbent (mg1-(1/n) L1/n / g), and n is a constant Temperature KL ∆𝐺 𝑜 ∆𝐻 𝑜 ∆𝑆 𝑜
indicating the adsorption intensity. (K) (kJ/mol) (kJ/mol.K) (kJ/mol.K)
308 1.951 -2148,38 -32,91 -0,099
The values of KF and n found from the graph of ln qe and 313 1.420 -1758,88
ln Ce at different temperatures are reported in Table 1. The n 18 1.011 -1148,42
value indicates the trend of the adsorption process; if n = 1, a IV. CONCLUSION
linear adsorption occur; if n < 1, it indicates that is a chemical
process; if n > 1, it indicates that is a physical process [27]. In this study, the ability of Cr(VI) adsorption on chemically
The obtained correlation coefficient R2 is larger than 0.991, rubber seed shell biochar modified with crosslinked chitosan
indicating the feasibility of the application of the Freundlich was investigated. The optimal pH value for adsorption was
equation. In this research, the value of n is found between 2,38 3.0. The experiment data followed Freundlich isotherm. The
and 2,65 that is larger than 1.0, demonstrating physical maximum amount of Cr(VI) adsorbed is 41,15 mg/g. The
process of metal ion adsorption on the CRSC. equilibrium contact time is 120 mins. The negative ∆G values
which indicated the metal ions adsorption on the CRSC was
In order to reveal the adsorption behavior on the surface spontaneous. The values of ∆H and ∆S were -32,91, and -
[28-30], the Langmuir isotherm model was applied. The 0,099, respectively. Both of enthalpy change (∆H) and
linearized Langmuir equation is: entropy change value (∆S) reveals the spontaneous at low
𝐶𝑒 1 temperature. This adsorbent is of low cost; its utility can be
= 1 (3)
𝑞𝑒 𝐾𝐿𝑞𝑚𝑎𝑥 ( () )
𝑞𝑚𝑎𝑥 𝑒
economical and can be viewed as a part of a feasible waste
management strategy.
where qmax is the maximum quantity of adsorption (mg/g), and
KL is the Langmuir constant (L/mg). REFERENCES
The values of KL and qmax found from the graph Ce/qe and [1] Wan Ngah, W.S., L.C. Teong, and M.A.K.M. Hanafiah,
Adsorption of dyes and heavy metal ions by chitosan
Ce, listed in Table 1, compare the experimental data with the
composites: A review. Carbohydrate Polymers, 2011. 83(4):
ones fitted by the Langmuir equation (Equation 6) at different pp. 1446-1456.
temperatures.
[2] Montazer-Rahmati, M.M., et al., Kinetics and equilibrium
𝑞𝐿𝑒 𝑚𝑎𝑥 studies on adsorption of cadmium, lead, and nickel ions from
𝑞𝑒 = (4)
1+𝐾𝐿 𝑞𝑚𝑎𝑥 aqueous solutions by intact and chemically modified brown
algae. Journal of Hazardous Materials, 2011. 185(1): pp. 401-
Table 1. The isotherm parameters of the adsorption process of Cr6+ 407.
on the CRSC at pH 3.0 [3] Ahmad, A., et al., Chemically oxidized pineapple fruit peel for
Temperat Langmuir Freundlich the adsorption of heavy metals from aqueous solutions.
ure KL Qmax R2 KF n R2 Desalination and Water Treatment, 2016. 57: pp. 6432-6442.
(K) (L/mg) (mg/g) (mg/g)( [4] Subki, N.S., R. Hashim, and N.Z.M. Muslim, Heavy metals
mg/L)n adsorption by pineapple wastes activated carbon: KOH and
308 0,0193 41,152 0,8820 1,32 1,56 0,99 ZnCl2 activation. AIP Conference Proceedings, 2019.
313 0,0241 32,258 0,9097 1,856 1,61 0,968 2068(1): p. 020035.
6 6
318 0,0269 26,954 0,828 2,167 1,62 0,974 [5] Ali Redha, A., Removal of heavy metals from aqueous media
4 by adsorption. Arab Journal of Basic and Applied Sciences,
2020. 27(1): pp. 183-193.
315
[6] Alalwan, H.A., M.A. Kadhom, and A.H. Alminshid, Removal [19] Foo, K.Y., L.K. Lee, and B.H. Hameed, Preparation of
of heavy metals from wastewater using agricultural activated carbon from sugarcane bagasse by microwave
byproducts. Journal of Water Supply: Research and assisted activation for the remediation of semi-aerobic landfill
Technology-Aqua, 2020. 69(2): pp. 99-112. leachate. Bioresource Technology, 2013. 134: pp. 166-172.
[7] Kumar, R., et al., Adsorption of chromium(VI) from aqueous [20] Suc, N.V. and D. Kim Chi, Removal of rhodamine B from
solution and electroplating wastewater using fungal biomass. aqueous solution via adsorption onto microwave-activated
Chemical Engineering Journal, 2008. 135(3): pp. 202-208. rice husk ash. Journal of Dispersion Science and Technology,
[8] Colica, G., P.C. Mecarozzi, and R. De Philippis, Adsorption 2017. 38(2): pp. 216-222.
and Recovery of Chromium from Industrial Wastewaters By [21] Nguyen, M.L. and R.-S. Juang, Modification of crosslinked
Using Saccharomyces cerevisiae in a Flow-Through System. chitosan beads with histidine and Saccharomyces cerevisiae
Industrial & Engineering Chemistry Research, 2012. 51(11): for enhanced Ni(II) adsorption. Journal of the Taiwan Institute
pp. 4452-4457. of Chemical Engineers, (0).
[9] Thabede, P.M., et al., Adsorption studies of toxic cadmium(II) [22] Agrafioti, E., D. Kalderis, and E. Diamadopoulos, Arsenic and
and chromium(VI) ions from aqueous solution by activated chromium removal from water using biochars derived from
black cumin (Nigella sativa) seeds. Journal of Environmental rice husk, organic solid wastes and sewage sludge. Journal of
Chemical Engineering, 2020. 8(4): pp. 104045. Environmental Management, 2014. 133: pp. 309-314.
[10] Imran, M., et al., Effect of biochar modified with magnetite [23] Shooto, N.D., Removal of lead(II) and chromium(VI) ions
nanoparticles and HNO3 for efficient removal of Cr(VI) from from synthetic wastewater by the roots of harpagophytum
contaminated water: A batch and column scale study. procumbens plant. Journal of Environmental Chemical
Environmental Pollution, 2020. 261: pp. 114231. Engineering, 2020. 8(6): pp. 104541.
[11] Ogata, F., et al., Potential of virgin and calcined wheat bran [24] Pigatto, G., et al., Chitin as biosorbent for phenol removal
biomass for the removal of chromium(VI) ion from a synthetic from aqueous solution: Equilibrium, kinetic and
aqueous solution. Journal of Environmental Chemical thermodynamic studies. Chemical Engineering and
Engineering, 2020. 8(2): pp. 103710. Processing: Process Intensification, (0).
[12] Qu, J., et al., KOH-activated porous biochar with high specific [25] Fereidouni, M., A. Daneshi, and H. Younesi, Adsorption
surface area for adsorptive removal of chromium (VI) and equilibria of binary Cd(II) and Ni(II) systems onto
naphthalene from water: Affecting factors, mechanisms and Saccharomyces cerevisiae and Ralstonia eutropha cells:
reusability exploration. Journal of Hazardous Materials, 2021. Application of response surface methodology. Journal of
401: p. 123292. Hazardous Materials, 2009. 168(2–3): pp. 1437-1448.
[13] Kabir, M.M., et al., Highly effective agro-waste based [26] Farooq, U., et al., Adsorption of heavy metal ions using wheat
functional green adsorbents for toxic chromium(VI) ion based biosorbents – A review of the recent literature.
removal from wastewater. Journal of Molecular Liquids, Bioresource Technology, 2010. 101(14): pp. 5043-5053.
2022. 347: p. 118327. [27] Ngah, W.S.W. and S. Fatinathan, Chitosan flakes and
[14] Chen, X., et al., Nanoscale zero-valent iron particles chitosan–GLA beads for adsorption of p-nitrophenol in
supported on sludge-based biochar for the removal of aqueous solution. Colloids and Surfaces A: Physicochemical
chromium (VI) from aqueous system. Environmental Science and Engineering Aspects, 2006. 277(1–3): pp. 214-222.
and Pollution Research, 2022. 29(3): pp. 3853-3863. [28] Eser, A., et al., Removal of nickel(II) ions by histidine modified
[15] Sud, D., G. Mahajan, and M.P. Kaur, Agricultural waste chitosan beads. Chemical Engineering Journal, 2012. 210(0):
material as potential adsorbent for sequestering heavy metal pp. 590-596.
ions from aqueous solutions – A review. Bioresource [29] Popuri, S.R., et al., Adsorptive removal of copper and nickel
Technology, 2008. 99(14): p.6017-6027. ions from water using chitosan coated PVC beads.
[16] Hegazi, H.A., Removal of heavy metals from wastewater using Bioresource Technology, 2009. 100(1): pp. 194-199.
agricultural and industrial wastes as adsorbents. HBRC [30] Ghaee, A., et al., Adsorption of copper and nickel ions on
Journal, 2013. 9(3): pp. 276-282. macroporous chitosan membrane: Equilibrium study. Applied
[17] Sadeek, S.A., et al., Metal adsorption by agricultural Surface Science, 2012. 258(19): pp. 7732-7743.
biosorbents: Adsorption isotherm, kinetic and biosorbents [31] Sarı, A., et al., Adsorption of Pb(II) and Ni(II) from aqueous
chemical structures. International Journal of Biological solution by lichen (Cladonia furcata) biomass. Biochemical
Macromolecules, 2015. 81: pp. 400-409. Engineering Journal, 2007. 37(2): pp. 151-158.
[18] Falizi, N.J., et al., Evaluation of MBR treated industrial
wastewater quality before and after desalination by NF and
RO processes for agricultural reuse. Journal of Water Process
Engineering, 2018. 22: pp. 103-108.
316
A Study on the Effect of the Shape of the Center

Rib on Thermal Resistance on the Dual-Layer
Micro-channel Heat Sink
Hung-Son Dang Thi-Anh-Tuyet Nguyen
Faculty of Vehicle and Energy Engineering Department of Mechanical Engineering
Ho Chi Minh University of Technology and Education Ho Chi Minh University of Technology and Education
sondh@hcmute.edu.vn ntatuyet@hcmute.edu.vn
Abstract: The energy consumption for electronic ratios. Ansari [6] studied three–dimensional numerical
component rising dramatically, to catch up with the demand, calculation for conjugate heat transfer which was performed
heat sinks is the main role of heat exchangers used for cooling on a full domain of heat sink by solving Navier – Stokes and
down the devices due to the simplicity of fabrication, energy equations for laminar flow using finite–volume solve.
reliability, and, low cost. By changing the shape of the center Huang [7] investigated the new design of a heat sink called a
rib of the Dual-Layer Microchannel Heat Sink (DL-MCHS) Double-layered staggered pin-fin heat sink (DL-SPFHS). The
this study figures out the effect of the shape on thermal result shows that the maximum temperature on the bottom
resistance. The Computational Fluid Dynamics software wall of DL-SPFHS is 13.2K lower than DL-MCHS. Khan [8]
(Ansys Fluent 14) was used to establish the parameters of the evaluated the thermal and hydraulic performance of various
DL-MCHS and simulate the model. The Minitab17 software geometric shapes of a microchannel heatsink using Navier –
was also used to analyze the simulated results and, showed Stokes equations. Khan [9] has been using three–dimensional
the good achievement of changing the shape of the center rib Navier – Stokes analysis and multi-object algorithm to study
on thermal resistance, the validation for the selected model the inverse trapezoidal cross-section of the microchannel heat
was implemented with the different temperatures of the sink. Deng [10] studied the thermal and hydraulic
overall model is 26.5℃, creating the thermal resistance value performance of a double-layered microchannel heat sink
is 0.1325 K/W. (DL-MCHS) by comparing with the single-layered in
different cross-sectional. Zhang [11] studied a novel design
Keywords: micro-channel heat sink, Taguchi method, having dense fins with lesser thickness at the upper layered
optimization, thermal resistance, CFD and compare spare fins with greater thickness in the lower
I. INTRODUCTION layer to further improve the overall thermal performance of
DL-MCHS. Kadam [12] made a review on twenty-first-
Tuckerman and Pease [1] were the first to research micro- century cooling solutions using a microchannel heat sink.
channel applications. Their work was designed to evaluate Kewalramani [13] took a study on thermal and hydrodynamic
the performance of force liquid flows in Microchannel Heat characteristics in the microchannel heat sink to determine the
Sinks (MCHS) and the results revealed that the minimum parameters required in modeling the heat sink as a fluid-
thermal resistance per unit area 𝑅𝑇 (0.1𝐾/𝑊𝑚2 ) was saturated porous medium. Ahmed [14] made a review on
obtained. After the first investigation of MCHS by optimization of thermal design of microchannel heat sink.
Tuckerman and Pease, more than a thousand papers have Zheng [15] investigated the efficiency of heat dissipation for
been reported about the investigations of various a hierarchical microchannel heat sink, the hierarchical heat
configurations of MCHS by using many different theoretical sink shows better heat transfer performance than the
and experimental methods. With the development of traditional one. Brunschwiler [16] reported an experimental
technologies, MCHS is not only used in the industry of investigation of a novel, high-performance ultrathin manifold
electronic and electrical engineering but also used in other microchannel heat sink. Achaya [17] reviewed the
fields, such as Chemical and Biological applications. In the application of a microchannel heat exchanger. Xu [18]
biological industry, many microchannel devices have been studied the flow rate distribution of hydrocarbon fuel in
studied extensively for different applications, and the spiral parallel channels with different cross-section shapes. Vasilev
microchannel device is the most interested one. [19] studied the effect of microchannel heat sink
Hsu [2-3] studied the thermal behavior using the robust configuration on thermal performance and pumping power.
design and Taguchi analysis. Olayinka [4] used the Ahmed [20] investigated the effect of geometrical parameters
constructal design technique to optimize the geometry of a on laminar water flow and forced convection heat transfer
multi-layered microchannel heat sink. The object of the characteristics in the grooved microchannel heat sink. Xie
numerical investigation was to minimize the peak [21] integrated the internal vertical Y-shaped bifurcations to
temperature, and the results showed that as the pressure drop investigate thermal and fluid flow. Gunnasegaran [22]
increased, the peak temperature decreased with the pressure studied the effect of geometrical parameters on heat transfer
drop between 20 and 60kPa. Elbadawy [5] evaluated the heat characteristics. Zhu [23] investigated the fluid flow and heat
transfer characteristics of water flowing through single and transfer characteristics of the microchannel heat sink with
double-stack rectangular microchannels with different aspect different groove shapes.

Through all the above studies, we see that a lot of The detail of the cross-section on the ribs of the channel
scientists have spent time and effort on the research of will be described in Figure 3 below.
microchannel heat exchangers. However, in reality, we have
The robust design (Taguchi) and statistics software
not found an optimal algorithm or solution for this device,
establish and select the parameters factor. The corresponding
although the application and popularity are very large.
test sample then will be printed according to the value fix
For single-layered micro-channel devices, although it is with an orthogonal array table. The result will then calculate
easy to fabricate because of its simple structure, the and analyze through the Taguchi algorithm and then verify
disadvantage of temperature distribution is the inherent by Minitab 17 software. This present study uses L9
problem that this device struggles with, furthermore, the orthogonal arrays to identify the parameter for the classroom
speedily increasing capacity of the electronic chip and as shown in Table 2 below.
devices, the single-channel heat exchangers no longer adapt
to that demand. a)
The alternative solution that seems to be outstanding is
the double-layered and dual-layered channel, with the ability
to dissipate a large amount of heat from the heat source,
minimizing the temperature difference between the input and
output of the device. However, the dual-channel heat
exchanger means that the processing of fabrication will
become more complicated, consume more resources,
increase the cost, and especially consume more power for the
pumping system.
Besides that, there are technical barriers to microchannel
devices with sizes in nanometers scale. b)
From the above problems, this study was purposed with high
expectations will create premise ideas for providing solutions
to create the most optimal microchannel devices.
II. METHODOLOGY
In this paper, the performance of the heat sink of the
device will be term thermal resistance, which is mentioned,
and the formula calculated according to [6].
𝑇𝑠,𝑚𝑎𝑥 −𝑇𝑓,𝑖𝑛
𝑅𝑡ℎ = (1)
𝑞𝐴𝑏
Where, 𝑇𝑠,𝑚𝑎𝑥 is the maximum temperature of heat sink, c)

𝑇𝑓,𝑖𝑛 is the temperature of coolant fluid flow in the inlet of the
device, 𝑞 is the heat flux, and 𝐴𝑏 is the area of base.
The cross-section of the overall channel will be shown in
Figure 1.
Figure 1. Substrate microchannel heat sink Figure 2. Substrate microchannel heat sink a) rectangular b)
trapezoidal c) triangular
The parameters of the channel ribs will be built according
to different profiles according to Figure 2.
Based on the factors in actual parameter collection and
modeling, there were three factors were selected with three
levels for each factor as shown in Table 1 below.
Table 1. Parameters and values levels
Levels
Parameters Factors
1 2 3
P1: Channel width(mm) 0.4 0.6 0.8
P2: Channel height(mm) 0.3 0.5 0.7
P3: Ribs cross-section Rec Trape Tri
Figure 3. cross-section of ribs
318
Table 2. L9 orthogonal arrays Table 3. Thermophysical properties of coolants and substrate

materials of the microchannel
TN P1 P2 P3
1 1 1 1 Materials 𝜌(𝑘𝑔 𝐶𝑝 (𝐽 𝐾𝑓 (𝑊 𝜇(𝑃𝑎𝑠)
/𝑚3 ) /𝑘𝑔𝑜 𝐶) /𝑚𝑜 𝐶)
2 1 2 2
3 1 3 3 Coolants Water 999 4187 0.587 0.001155
4 2 1 2
Substrate Aluminum 2719 871 202.4
5 2 2 3 material
6 2 3 1
7 3 1 3 III. RESULTS
8 3 2 1 After conducting numerical simulations and collecting
9 3 3 2 temperature data, the mean value of the temperature is
obtained according to Table 4.
After establishing the factors and their related levels
Process and calculate data according to Taguchi in
based on the previous study and evaluation of the reference
smaller is better cases shown that response. Setting the
articles, Inventor software was used to build the 3D model
average table of SN shows that the height factor of the
(Figure 4).
channel has the greatest influence on the channel
temperature, followed by the ribs profile factor of the channel
as shown in Table 5 below.
Numerical calculation using the Taguchi approach gives
the same results as validation using Minitab 17 software as
shown in Figure 6.
Table 4. Means value of temperature
Parameters W H S Means
Model 1 400 300 REC 27.4622
Model 2 400 500 TR 25.7160
Model 3 400 700 TRI
24.4692
Model 4 600 300 TR 27.3529
Figure 4. 3D model of the channel
Model 5 600 500 TRI 26.8206
The simulation was carried out using Ansys software. The Model 6 600 700 REC 22.6664
computational domains of the channel are described in Figure Model 7 800 300 TRI 27.2755
5 below.
Model 8 800 500 REC 26.6913
Model 9 800 700 TR 23.6224
From ANOVA theory on hypothesis testing, P-value < α

= 5%. We conclude that factor H affects the product.
Table 5. SN value of temperature
Level W H S
1 -28.2404 -28.7434 -28.1366
2 -28.1390 -28.4336 -28.1369
3 -28.2365 -27.4489 -28.3524
∆ 0.1114 1.2945 0.2158
Rank 3 1 2
Figure 5. 3D computational domains of the channel

The initial condition of computational with ambient
temperature is 25℃, the heat flux is 50 w/cm2 and the inlet
velocity of the coolant is 1m/s.
The thermophysical properties of coolants and substrate
materials of the microchannel are described in Table 3 below.
Figure 6. Main effects plot for SN ratios
319
Based on the Means values analyse by Minitap19 [4] Olayinka Omowunmi Adewumi, Tunde Bello-Ochende, Josua P
software, we choose the best combination for the Meyer, “Geometric Optimisation of Multi-Layered Microchannel Heat
Sink with Different Flow Arrangements”, 15th international heat
temperature: Width P1: W = 600 μm, Height P2: H = 700 μm, transfer Conference, August 2014 DOI: 10.1615/IHTC15.hte.009148.
Section P3: S = Trapezoidal, and the selected model number: [5] Ibrahim Elbadawy and Md Fatouh, “Heat Transfer Characteristics Of
W2H3S2. Water Flowing Through Single And Double Stack Rectangular
Microchannels”, 16th International Conference on Applied Mechanics
We conclude that factor P2: height H has the greatest and Mechanical Engineering, 27-29 May, 2014.
influence on the model’s temperature followed by the cross- [6] Danish Ansari, Kwang-Yong Kim, “Performance analysis of a cross-
section of the ribs trapezoidal. channel double-layer microchannel heat sink for electronic cooling”,
Conference Paper, November, 2015.
The validation for the selected model was implemented [7] Yicang Huang, Hui Li, “Numerical analysis on the thermal
with the different temperatures of the overall model is 26.5℃ performance of a novel double-layered heatsink with staggered pin
(figure 7) creating the thermal resistance value is 0.1325 fins”, Heat Transfer Research, 50(8):757–772 (2019).
K/W. [8] Aatif Ali Khan, Kwang-Yong Kim, “Evaluation of Various Channel
Shapes of a Microchannel Heat Sink”, International Journal of Air-
Conditioning and Refrigeration August 2016.
[9] Aatif Khan, Sun-Min Kim and Kwang-Yong Kim, “Multi-Objective
Optimization of an Inverse Trapezoidal-Shaped Microchannel”, Heat
Transfer Engineering, DOI:10.1080/01457632.2015.1060772.
[10] Daxiang Deng, Guang Pi, Weixun Zhang, Peng Wang and Ting Fu,
“Numerical Study of Double-Layered Microchannel Heat Sinks with
Different Cross-Sectional Shapes”, Entropy 2019, 21, 16;
doi:10.3390/e21010016.
[11] Yong‐Dong Zhang, Miao‐Ru Chen, Jung‐Hsien Wu, Kuo‐Shu Hung
and Chi‐Chuan Wang, “Performance Improvement of a Double‐Layer
Microchannel Heat Sink via Novel Fin Geometry - A Numerical
Study”, Energies 2021, 14, 3585, doi.org/10.3390/en14123585.
[12] Sambhaji T. Kadam, Ritunesh Kumar, “Twenty first century cooling
solution: Microchannel heat sinks”, International Journal of Thermal
Sciences, 85 (2014), 73-92.
[13] Gagan V. Kewalramani, Amit Agrawal, Sandip K. Saha, “Modeling of
Figure 7. Contour temperature of the selected model microchannel heat sinks for electronic cooling applications using
volume averaging approach”, International Journal of Heat and Mass
Compare with the smallest thermal resistance of the Transfer, 115 (2017), 395–409.
following article [4] 50K/W, [6] 0.17K/W, [7] 0.05K/W, [8] [14] Hamdi E. Ahmed, B.H. Salman, A.Sh. Kherbeet, M.I. Ahmed,
0.1K/W with even more complicated geometry. “Optimization of thermal design of heat sinks: A review”, International
Journal of Heat and Mass Transfer, 118 (2018), 129–153.
IV. CONCLUSION [15] Rui Zheng, Yongjin Wu, Yahui Li, Guilian Wang, Guifu Ding, Yunna
Sun, “Development of a hierarchical microchannel heat sink with flow
The results of the present study can provide a huge field reconstruction and low thermal resistance for high heat flux
motivation for future research on optimal thermal resistance dissipation”, International Journal of Heat and Mass Transfer, 182
or even more micro-channel devices. (2022), 121925.
[16] Thomas Brunschwiler, Bruno Michel, “Experimental Investigation of
Based on the Means values analyse by software, we an Ultrathin Manifold Microchannel Heat Sink for Liquid-Cooled
choose the best combination for the temperature, and the Chips”, Journal of Heat Transfer, August 2010 DOI:
selected model number is W2H3S2. Factor P2; the height H 10.1115/1.4001306.
has the greatest influence on the model’s temperature [17] AR Acharya, “Review On Applications of Micro-Channel Heat
Exchanger”, International Research Journal of Engineering and
followed by the cross-section of the ribs trapezoidal. Finally, Technology, Volume: 07 Issue: 03, Mar 2020.
the validation for the selected model was implemented with [18] Yuguang Jiang, Yaxing Xu, Jiang Qin, Silong Zhang, Khaled
the different temperatures of the overall model is 26.5 ℃, Chetehouna, Nicolas Gascoin, Wen Bao, “The flow rate distribution of
creating the thermal resistance value is 0.1325 K/W. hydrocarbon fuel in parallel channels with different cross-section
shapes”, Applied Thermal Engineering 137 (2018) 173–183.
ACKNOWLEDGMENT [19] M.P. Vasilev, R.Sh. Abiev, R. Kumar, “Effect of microchannel heat
sink configuration on the thermal performance and pumping power”,
We acknowledge the support and time provided by Ho International Journal of Heat and Mass Transfer, 141 (2019), 845–854.
Chi Minh City University of Technology and Education [20] Hamdi E. Ahmed, Mirghani I. Ahmed, “Optimum thermal design of
(HCMUTE), as well as the use of their facilities in this study. triangular, trapezoidal and rectangular grooved microchannel heat
sinks”, International Communications in Heat and Mass Transfer,
2015.
REFERENCES
[21] Gongnan Xie, Han Shen, Chi-Chuan Wang, “Parametric study on
[1] Tuckerman DB and Pease RFW (1981), “High-performance heat thermal performance of microchannel heat sinks with internal vertical
sinking for VLSI”. IEEE Electron Device Letter 5:126-129. Y-shaped bifurcations”, International Journal of Heat and Mass
[2] Cheng-Hsing Hsu, Jui-Chin Jiang, Hung-Son Dang, Thi-Anh-Tuyet Transfer, 90 (2015), 948–958.
Nguyen, “Investigating the design parameters on the multi-layer micro- [22] P. Gunnasegaran, H.A. Mohammed, N.H. Shuaib, R. Saidur, “The
channel heat sink by using Quality Function Deployment and Taguchi effect of geometrical parameters on heat transfer characteristics of
method with the enlarged outlet”, IEEE Explore 2017(IE), DOI: microchannels heat sink with different shapes”, International
10.1109/ICASI.2017.7988170. Communications in Heat and Mass Transfer, 37 (2010), 1078–1086.
[3] Cheng-Hsing Hsu, Jui-Chin Jiang, Hung-Son Dang, Thi-Anh-Tuyet [23] Qifeng Zhu, Huixue Xia, Junjie Chen, Xinmin Zhang, Kunpeng Chang,
Nguyen, Investigating the Designed Parameters of Dual-Layer Micro- Hongwei Zhang, Hua Wang, Jianfeng Wan, Yangyang Jin, “Fluid flow
Channel Heat Sink by Design for Six Sigma (DFSS), IEEE Explore and heat transfer characteristics of microchannel heat sinks with
2017 (IE), DOI: 10.1109/ICASI.2017.7988155. different groove shapes”, International Journal of Thermal Sciences,
2020.
320
Experimental Study of the Effect of Heat Input on

Tensile Strength and Microstructure of the Weld
Using the Orbital TIG Welding Process
Thien Tran Ngoc Ngoc-Huy Dinh An-Duong Tra
Faculty of Mechanical Engineering Faculty for High-Quality Training Faculty for High-Quality Training
thientn@hcmute.edu.vn 18143014@student.hcmute.edu.vn 18143005@student.hcmute.edu.vn
Kha-Duy Doan Binh-Minh Ngo Anh-Duc Pham Duc

Faculty for High-Quality Training Faculty for High-Quality Training Faculty for High-Quality Training
18143004@student.hcmute.edu.vn 18143027@student.hcmute.edu.vn 18143008@student.hcmute.edu.vn
Abstract: The automatic orbital TIG welding process is an the welding speed [2-5] and improve the penetration [6-9] in
automatic pipe welding process and is increasingly applied in advanced OTW processes.
fields that use pipes to transport liquids or gases such as Shipilov, A. V. et al. [10] established a physical-
biomedical, food, beverages, oil, and gas,… Welding mathematical model of melting of wire in Orbital non-
parameters are one of the main causes that directly affect consumable electrode welding for the optimum wire feed rate
weld quality, through welding parameters we calculate heat is the ratio between the wire feed rate, the power and the
input results to evaluate weld quality by tensile strength and diameter of the arc jet resulting in stable melting of the wire.
microstructure of the weld. This paper focuses on finding the It has shown that the simplest method of increasing the
affection of heat input and filler wire speed on the tensile stability of the process of orbital welding in relation to the
strength and the microstructure of the pipe weld. The pulsed feed of the filler wire synchronized with the current
experiment is carried out with the material SS304 stainless pulses. In the same way, Chen et al. [11] presented the
steel pipes, 2 mm of thickness and 76 mm of outer diameter, relationship between wire filler speed and energy to reach
and welding parameters such as welding current from 80A to metallic transfer stability with the pulse current added during
120A, arc length from 2mm to 4mm, welding speed from 4 the welding process. Aamlr Sohall et al. [12] investigated the
mm/s to 5.5 mm/s, filler wire speed from 3.8mm/s to effect of TIG welding parameters on the properties of 304L
6.4mm/s. As the result, this study shows that the highest automated girth welded pipes Using an Orbital welding
tensile strength of the weld is 751 MPa with the heat input is machine and filler wire is 0.8mm. It found that the tensile
0,663 kJ/mm and the filler wire speed is 5.3 mm/s. Besides, strength of the joint has increased by applying pulse current
the weld zone of this weld achieved full penetration and finer and reached a maximum value is arc travel speed of
80mm/min and wire feed rate of 330 mm/min. It also showed
microstructure. On the other hand, this experiment shows
that increasing the wire feed rate without an increase in current
some failed welds because of incompatibility between heat
decreases the strength of the weld.
input and filler wire speed.
In all of those results, for each wire filler rate, there is an
Keywords: heat input, orbital TIG welding, filler wire, energy value suitable for melting the wire in a constant way.
tensile strength, microstructure Hence, the target of this study is to consider the effect of heat
input which is made from various welding parameters on the
I. INTRODUCTION profile and tensile strength of welds.
Orbital TIG Welding (OTW) is an automatic Tungsten
Inert Gas (TIG) welding process to join the tubes or pipes. In
this welding process, an Orbital weld-head is responsible for
carrying an electrode and electric arc going aground the weld
joint to create a required weld (Fig.1). OTW is having a
number of applications in different fields. Their applications
are increasing day by day now [1]. Therefore, the requirement
of this process improves productivity and weld quality which
has attracted the attention of producers. To achieve these
problems, using filler metal is an effective solution to increase Figure 1. The principle of Orbital TIG welding

II. EXPERIMENTAL MATERIAL AND METHOD
A. Experimental Material
The material used in this experiment is the SS304 tube
which is widely used in the biomedical, food and beverage,
oil and gas industries,… The outside diameter of the tube is
76 mm and the thickness is 2 mm. The orbital tube welding
process used ER304 wire filler that is homogeneous with the
basic metal and has 1mm in diameter. The chemical
composition of the materials used in the experiment is shown
in Table 1.
Figure 4. Profile for tensile testing according to ISO 6892-1
standard
Table 1. Chemical composition of material SS304 [14]
Chemical composition of basic metal and wire filler (%wt) Heat input (HI) is calculated according to the following
C Si Mn P S Cr Ni formula:
17.5 8 U × I
≤0.07 ≤ 1.0 ≤ 2.0 ≤ 0.045 ≤ 0.03 - - HI = (1)
19.5 10.5 1000 × V𝑠
Where:
B. Experimental Method  HI – Heat Input (J/mm)
A welded is jointed by 2tubes Ø76mm, length of a tube is  U – Voltage (V)
100 mm (Fig. 2) based on the tensile test standards (Fig. 3 &  I – Amperage (A)
4). The type of automatic OTW head used in this experiment  Vs – Welding speed (mm/s)
is an open arc-welding head, with continuous wire
compensation during the welding process. The workpiece is Table 2. Welding parameters and heat input of welds
SS304 with 2mm of thickness, the diameter of the Thorium
electrode is 2.4mm, the electrode angle is 60 and the Code
LA I Vs U Vf HI
shielding gas is used by Argon gas with a gas flow rate of (mm) (A) (mm/s) (V) (mm/s) (J/mm)
12LPM. Five single welds are made with random variations S1 2 100 5.5 22 4.4 400
of welding parameters such as amperage (I), welding speed S2 2.5 110 4.0 24 5.3 660
(Vs), arc length (LA), and welding voltage (U) will create five
S3 3 120 4.4 22 5.9 600
different heat input levels (Table 2). These heat input levels
will provide an amount of heat to melt the filler metal wire at S4 3.5 80 4.9 21 6.4 343
different speeds. Therefore, the mission of this research is to S5 4 90 5.3 22 3.8 374
find the appropriate heat input levels for the size and speed of
wire compensation in the automatic tube welding process. After welding samples are welded, they will be observed
weld profiles to compare with the visual test criteria according
to AWS D18.1 standard (Fig. 5).
(A) (B)
Maximum misalignment 15% of Nonproduct contact surface
tube wall thickness maximum convexity
Figure 2. Sample of tube prepared for welding process
Maximum weld face width (W)
(C) Minimum weld face width

Minimum face width (0.75W)
(D)
Figure 5. The visual test criteria according to AWS D18.1
standard
Next, to observe the profile and microstructure of the weld,

the sample preparation is carried out according to the
procedure shown in Fig. 6. The image of the microstructure of
the welds and the different positions of the weld zones will be
Figure 3. Position of reduced section tensile recorded by the microscope.
322
But, when compared to the tensile strength of base metal, from

1. Cutting 2. Mounting 3. Grinding
500 to 700 Mpa [14], only sample S2 met this criterion, even
higher than the referenced range. Besides, sample S1 has the
width and height of the weld not exceeding the allowable
range, but the tensile strength value is quite high, just lower
than the tensile strength of S2. This may be due to the greater
penetration of S1 than S3, which will be clarified in their
6.Observation 5. Etching 4. Poilishing microstructure. In contrast, the tensile strength of sample S4
has the lowest value due to the intermittent formation of
metallic transfer droplets from filler wire, the cause of them
has been analyzed above.
Figure 6. The sample preparation process for observing the
microstructure
Table 4. The tensile strength of welds
III. RESULTS AND DISCUSSION Code HI (j/mm) Vf (mm/s) Tensile strength (Mpa)
S1 400 4.4 415
C. Visual Testing S2 660 5.3 751
Following Fig. 7, the weld is recorded outside the tube to S3 600 5.9 280
observe the surface of the weld being. It was shown that S4 343 6.4 9
S5 374 3.8 103
samples S1, S2, S3 and S5 formed a continuous weld, but
sample S4 did not form a weld due to the intermittent The graph in Fig. 8 shows the relationship between HI and
formation of metallic transfer droplets. The sample S4 failed wire filler speed, they affect the tensile strength of the samples.
because the welding current and welding voltage are too small It is easy to see that, there is no linear relationship between HI
and the welding speed is quite high, which leads to a small and tensile strength, similar to the relationship between Vf and
heat input (HI) (Table 2) not enough heat to melt the base tensile strength. HI is too low and Vf is too high so non-form
metal and filler wire. On the other hand, the welding speed is a weld, sample S4 is an example. Besides, low HI and low Vf
high, so HI only melts the metal droplets from the filler wire can form a weld, but the tensile strength is also very low as
that stick on the surface of the tube (Fig. 7d). Welds S1, S2, the sample S5 because HI is not enough to melt the base metal.
S3 and S5 are compared to the AWS D18.1 standard visual In addition, the addition of a filler wire takes away the heat of
inspection criteria. The results are shown in Table 3, there the weld pool. For the same reason, sample S3 had rather high
were 2 welds that pass 4 criteria are samples S2 and S3 HI and Vf but did not give required tensile results.
because their HI is 660 J/mm and 600 J/mm respectively,
which is provide suitable heat for the diameter and filler wire Tensile strength (Mpa)
speed, 5.3 mm/s for sample S2 and 5.9 mm/s for sample S3. 800
751
700
600
500
400 415
300 280
200
100 103
0 9
343 374 400 600 660
HI (J/mm)
Tensile strength (MPa)
(a)
Tensile strength (Mpa)
a) S1 b) S2 c) S3 d) S4 e) S5
800
Figure 7. The surface of 5 welds 751
700
600
Table 3. Dimensions of weld’s surface compared to AWS D18.1
500
standard
400 415
Code W (mm) h (mm) (A) (B) (C) (D) 300 280
S1 3.64 0.6 P F F P 200
S2 4.73 0.3 P P P P
100 103 9
S3 5.24 0.2 P P P P
0
S4 2.56 0.9 F because weld was not formed
S5 3.44 0.2 P P F P 3,8 4,4 5,3 5,9 6,4 Vf (mm/s)
Note: “F” is fail, “P” is pass, “W” is width of weld, h is height of weld
Tensile strength (MPa)
D. Tensile Testing
Tensile strength results of 5 samples in Table 4 show that (b)
samples S1 and S2 have the highest tensile strength among 5 Figure 8. The chart of the relationship between HI and tensile
samples with values of 415 Mpa and 751 Mpa, respectively. strength (a), between Vf and Tensile Strength (b)
323
E. Microstructure by the shielding gas blowing the spherical metal drop out of
the position joint or the unstable transmission in the
mechanical of the arc-welding head. The reason for the low
tensile strength of samples S3 and S5 is the small HI and the
center of the weld deviates from the joining line. The cause
may be arc deviation or unstable transmission of the arc-
welding head during the welding process.
Following Fig. 9a) and Fig. 9b) show that samples S1 and
S2 both achieved complete penetration, but welding of the
fusion of sample S2 (Fig. 9b) is better, so the tensile strength
of sample S2 (751 Mpa) is higher than sample S1 (415 MPa),
which shows that, with high HI and suitable Vf, it will create
a) S1 a weld with the good penetration and the good width of fusion
for the base metal in the automatic tube welding process.
Continue to analyze microstructure images of samples S1
and S2 with bigger magnification to see the bonding ability
between the weld zone and the base metal. Comparing Fig.
10a) and Fig. 10c), the dendrite structure of sample S2 (Fig.
10c) is smaller and finer than sample S1 (Fig. 10a), which is
Defects also mentioned in the study of Subodh Kumar et al. [13].
Although the HI of sample S2 is higher than S1. However, the
increase in the wire filler speed also creates a reasonable
amount of heat to melt the base metal and the filler metal wire.
Therefore, the bonding ability between the base metal and the
weld zone of sample S2 is better than S1. In addition, the
b) S2 dendrite structure of the weld zone of sample S2 is also
smaller and finer than S1, which is another proof that the
matching between HI and Vf will play an important role in
creating a weld that meets the required quality.
Gap Dendrites
c) S3
a) Bonding Zone (BZ) b) Weld Zone (WZ)

S1
Dendrites
Gap
d) S4
c) Bonding Zone (BZ) d) Weld Zone (WZ)

S2
Figure 10. The microstructure of S1 and S2 with magnification
Gap X500
e) S5 IV. CONCLUSION
Figure 9. The macrostructure of section of 5 welds: S1 (a), S2 Following the evaluation and analysis results, this study
(b), S3(c), S4(d), S5(e) makes some conclusions:
In Fig. 9d), the penetration is low and is located off the  The establishment of a good relationship between heat
joining position. Their penetration is the lowest because HI is input (HI) and filler wire speed (Vf) is important in
too small and Vf is too fast, which has been analyzed in the forming a quality joining and increasing the
previous parts. Welds that are out of position can be caused productivity of the automated tube welding process.
324
 With the diameter of filler wire being 1mm, this study [4] Opderbecke, T., Guiheux, S., “TOPTIG: Robotic Tig Welding
shows that sample welding S2 has a tensile strength of with Integrated Wire Feeder”, Welding International. Vol. 23,
Issue 7, 2009, pp. 523-529.
751 Mpa when HI is 660 J/mm and Vf is 5.3 mm/s. In
[5] Wu, Q., Lu, J., Liu, C., Shi, X., Ma, Q., Tang, S., Fan, H., Ma
addition, there is a weld that has a lower tensile S., “Obtaining Uniform Deposition with Variable Wire Feeding
strength in sample S1, tensile strength is 415 MPa, HI Direction During Wire-Feed Additive Manufacturing”,
= 400 J/mm, and Vf of 4.4 mm/s. Materials and Manufacturing Processes, Vol. 32, Issue 16, 2017,
pp. 1881-1886.
 When using 1 mm of filler wire in diameter, HI can be [6] S.C Juang, Y.S Tarng, “Process Parameter Selection for
calculated from 400 J/mm to 660 J/mm and Vf can be Optimizing the Weld Pool Geometry in The Tungsten Inert Gas
set from 4.4 mm/s to 5.3 mm/s. Welding of Stainless Steel”, Journal of Materials Processing
Technology, Vol. 122, Issue 1, 2002, pp. 33-37.
 When increasing HI, it is necessary to consider the [7] T.S Chern, K.H Tseng, H.L Tsai, “Study of the Characteristics
oxidation occurring inside the tube, especially when of Duplex Stainless Steel Activated Tungsten Inert Gas Welds”,
welding for tubes in the food, beverage, and Material & Design, Vol. 32, Issue 1, 2011, pp. 255-263.
biomedical industries. [8] CR Heiple, JR Roper, RT Stagner, RJ Aden - Weld. J., “Surface
Active Element Effects on the Shape of GTA, Laser and
 For welding samples with low HI, the option of Electron Beam Welds”, Welding Journal, Vol. 62, Issue 3, 1983,
preheating the filler wire can be used before entering pp. 72-77.
the weld pool. The purpose of this option is to prevent [9] Jun Yan, Ming Gao, Xiaoyan Zeng, “Study on Microstructure
and Mechanical Properties of 304 Stainless Steel Joints by TIG,
the filler wire from taking away the heat of the weld Laser and Laser-TIG Hybrid Welding”, Optics and Lasers in
pool or introducing gas into the tube during the Engineering, Vol. 48, Issue 4, 2010, pp. 512-517.
welding process, which is increasing the weld [10] Chen, S., Zhang, S., Huang, N., Zhang, P., Han, J., “Droplet
penetration and will reduce the damage of oxidation. Transfer in Arcing-Wire GTAW”, Journal of Manufacturing
Processes, Vol. 23, 2016, pp. 149-156.
 Make attention to the angle of the filler wire, the [11] Zhao, D.B., Chen, S.B., Wu, L., Dai, M., Chen, Q., “Intelligent
distance between the filler wire and the electrode Control for the Shape of the Weld Pool in Pulsed GTAW with
because it will affect the metal transition state from the Filler Metal”, Welding Journal, Vol. 80, Issue 11, 2001, pp. 253-
filler wire into the welding pool. 261.
[12] Aamir Sohail, Abdul Aziz, Muhammad Imran, Osama Junaid
ACKNOWLEDGMENT M and Sagheer Ahmed, “Effect of TIG Welding Parameters on
the Properties of 304L Automated Girth Welded Pipes Using
This study is supported by Ho Chi Minh City University Orbital Welding Machine”, Journal of Material Science, Vol. 5,
of Technology and Education (HCMUTE). Issue 4, October 2017, pp. 136-143.
[13] Subodh Kumar, A.S. Shahi, “Effect of heat input on the
REFERENCES microstructure and mechanical properties of gas tungsten arc
[1] Omkar Joshi & Dr. Arunkumar, “Overview of Orbital Welding welded AISI 304 stainless steel joints”, Materials and Design,
Technology”, International Journal of Innovative Science, Vol. 32, 2011, pp. 3617–3623.
Engineering & Technology, Vol. 2, Issue 12, 2015. [14] DIN EN 10088-3, Grade AISI 304/EN 1.4301.
[2] Richardson, M, “Semiautomatic Cold Wire Feeder Systems
Increase GTA Productivity”, Welding Journal, Vol. 71, 1995,
pp. 51-54.
[3] Anantha, P.MR., Baskar, N., Devakumaran, K., “A Study on
Process Characteristics and Performance of Hot Wire Gas
Tungsten Arc Welding Process for High Temperature
Materials”, Materials Research, Vol. 20, 2017, pp. 76-87.
325
A Study of Customer Satisfaction in Online Food

Delivery Service Quality during the Covid-19
Pandemic: Baemin’s Case Study
Hong- Xuyen Thi Ho Ngoc-Anh Ha Thi Ngoc-Tra Tran Thi
Faculty of Economics Faculty for High-Quality Training Faculty for High-Quality Training
Ho Chi Minh 71307, Vietnam Ho Chi Minh 71307, Vietnam Ho Chi Minh 71307, Vietnam
xuyenhth@hcmute.edu.vn 19124002@student.hcmute.edu.vn 19124004@student.hcmute.edu.vn
Abstract: The paper studies the factors affecting customer Veloutsou showed that factors such as price, social influence,
satisfaction when using Baemin food delivery service during habit, trust, convenience, and app quality have a positive
the Covid-19 period in Ho Chi Minh City based on qualitative impact on customer satisfaction before the Covid-19
method combined with the quantitative method. The research pandemic- 19 in which habits have the opposite effect on
results show that the level of customer satisfaction is satisfaction during the Covid-19 pandemic [4]. Prasetyo et al.
influenced by 5 factors: (1) Professional and Reputable, (2) [5] found that hedonic motivation has the most significant
Satisfy customer needs, (3) Pandemic safety, (4) Smart direct influence on intention to use. Furthermore, the price
interface, (5) Discount. All these factors have a positive also has a significant direct effect on actual usage, followed
impact on their online shopping intention. The study also by information quality which has a significant effect on the
pointed out the limitations and provided suggestions to help intention to use and advertising has a significant effect on
improve the quality of food delivery services to help Baemin actual used. Aeny et al. [6] showed that price and service
increase customer satisfaction. quality affects customer satisfaction with Grabbike online
motorbike taxi service in South Cikarang. See-Kwong et al.
Keywords: Covid-19, customer satisfaction, food [7] pointed out that consumer demand is convenient with
delivery, service quality online food delivery. However, they are not willing to
sacrifice food quality. Thus, providing online food delivery
I. INTRODUCTION services is also gradually becoming a tool to survive in the
During the Covid-19 pandemic, the business situation of online food delivery service industry. According to the
foodservice restaurants has been seriously affected by the survey, there are many studies that have conducted research
requirements of social distancing. Therefore, the in the fields of Banking, Hotels, and online transportation
consumption habits of customers also change. More and more services…But there is no research on a specific food delivery
customers are adapting to online shopping. According to data service in the case of Baemin food delivery service. This
collected from the US survey, e-commerce in 2020 has a study helps to find the optimal solutions with the desire to
32.2% increase in revenue compared to 2019 [1], the result help Baemin best meet customers, showing that the factors
indicated fact that consumers have to stay at home and use belonging to Baemin’s service quality affect customer
online services due to social distancing requirements In satisfaction. The results of the study will help Baemin food
Vietnam, according to the 2021 E-Commerce White Paper, delivery service have an overview, thereby improving service
the percentage of people using the internet to shop online has quality to take better care of customers.
increased from 77% in 2019 to 88% in 2020. According to II. LITERATURE REVIEW AND HYPOTHESES
Nielsen statistics, up to 62% of Vietnamese customers want
DEVELOPMENT
to buy food to take home because they don't want to eat in
crowded places [1]. A. Service quality
The epidemic and the requirements of social distancing In this study, food delivery service is understood as the
have affected the business activities of restaurants. Therefore, whole transportation process, the means of transport bring the
online food ordering service is one of the solutions to help utilities provided to customers to meet their food supply
restaurants and food suppliers maintain their business. Pham needs.
Thi Hoang Dung pointed out that the factors including Price When it comes to service quality, we cannot fail to
and Product, have positive impacts on consumers' online mention the very layered contribution of Parasuraman et al.
shopping behavior [2]. The results provide solutions to help [8]. Parasuraman et al. [9] define service quality as “the
food and beverage businesses maintain and develop their degree of difference between consumers’ expectations of
businesses during the Covid-19 period. Zhao and Bacao services and their perceptions of service outcomes”. These
pointed out that a customer's intention to continue using food authors initiated and used qualitative and quantitative
delivery during the Covid-19 pandemic is determined not research to build and test a scale of components of service
only by satisfaction but also by technological suitability, quality (the SERVQUAL scale). The SERVQUAL scale is
trust, longevity brand, and social influence [3]. Gilbert and

adjusted and tested in many different types of services. H6: Pandemic safety has a positive effect on customer
Finally, the SERVQUAL scale includes 22 variables to satisfaction with Baemin food delivery service
measure 5 components of service quality, namely: reliability,
responsiveness, assurance, tangible and intangible means and Convenience of service
H1
empathy. To conclude, due to the characteristics of each type
of service, researchers need to adjust the SERVQUAL scale Contact with customers H2
Level of
to suit each specific study [10] customer
Delivery driver time H3 satisfaction
In summary, Service quality is a broad category and has about
many different definitions depending on the type of service, H4
Food delivery status Baemin’s
but the essence of service quality is generally seen as what
H5 service
customers perceive. Each customer has different perceptions Voucher
and individual needs, so the perception of service quality is quality
H6
also different. Pandemic safety
B. Customer satisfaction
Figure 1. Proposed research model
There are many different perspectives on customer
satisfaction. Customer satisfaction is their response to the
perceived difference between known experience and III. METHODOLOGY
expectations [9], [11]. Thus, the known experience of the The study used qualitative and quantitative methods for
customer when using a service and the outcome after the analysis. Qualitative analysis is applied to consider whether
service is provided. Service quality and satisfaction are the scale in use is suitable for Baemin services or not.
closely related in service research and research results of Quantitative research is used to analyze survey data, for the
Parasuraman et al. [9] indicate that the level of customer purpose of looking at the impact of factors in the model on
satisfaction is higher when the quality of the service is high. dependent factors.
The studies of [12], [13], [14] also demonstrated that service
quality leads to customer satisfaction. In order to achieve a Qualitative research was conducted through focus group
high level of customer satisfaction, most researchers assume discussion, online adjustment on Google meet platform,
that a high level of service quality will be provided by service adding service quality components, and developing a scale
providers because service quality is often considered as an for these factors and a decision scale. Baemin's quality of
antecedent of customer satisfaction [15]. food delivery service leads to customer satisfaction. The
study was conducted by group discussion including people
C. Relationship between service quality and customer who have been using Baemin's food delivery service during
satisfaction the Covid-19 pandemic.
Service quality and satisfaction are closely related in The author first discussed with team members some
service research [9]. The research results indicated that the exploratory open-ended questions to see what factors and
level of customer satisfaction is higher when the quality of under what aspects of food service quality leads to customer
the service is high. In order to achieve a high level of satisfaction. Then, the author introduces the service quality
customer satisfaction, most researchers assume that a high components that lead to customer satisfaction proposed by
level of service quality should be provided by service the author in part II for group members to discuss and give
providers because service quality is often considered as an opinions. Finally, the author summarizes the opinions that are
antecedent of customer satisfaction [15]. The relationship agreed by the majority of members, all components
between service quality and customer satisfaction is positive, mentioned in part II.
service quality has a positive influence on customer
perception. Service quality and customer satisfaction are two Survey object: Customers who have been using
different concepts, while service quality focuses specifically Baemin's food delivery service during the Covid-19
on service components, customer satisfaction is a concept of pandemic.
generality [16]. According to [11] and [17], there is a Sample size: It is expected to choose a common sample
relationship between service quality and customer size for both types of customers who have been using the
satisfaction. service to be surveyed in HCMC.
Based on the named theory, the study has proposed a How much sample size is optimal depends on
research model as shown in Figure 1: expectations about reliability, data analysis method, method
H1: Service convenience has a positive effect on of estimating the parameters to be estimated and the
customer satisfaction with Baemin's food delivery service. distribution law of the set of choices. The researchers believe
that, if using the ML (Maximum Likelihood) estimation
H2: Customer contact has a positive influence on method, the minimum sample size should be from 100 to 150
customer satisfaction with Baemin's food delivery service (Hair et al., 1998) (quoted in Hoang Trong & Chu Nguyen
H3: Delivery driver time has a positive effect on customer Mong Ngoc, 2008). While Hoang Trong & Chu Nguyen
satisfaction with Baemin's food delivery service Mong Ngoc (2008) think that the ratio between the number
of samples and the number of observed variables is 4 or 5. In
H4: Food delivery status has a positive influence on the study, there are all 36 observed variables that need to be
customer satisfaction with Baemin's food delivery service estimated. So, the minimum sample needed is 36 x 5 = 180
observations.
H5: Voucher has a positive effect on customer
satisfaction for Baemin's food delivery service
327
However, to ensure representativeness and redundancy coefficients of total variables greater than 0.3. Thus, all scales
for those who did not respond or responded incompletely, the have enough reliability for EFA exploratory factor analysis.
author chose a sample size of more than 180 people.
Table 2. Summary of the results of testing the reliability of the
Therefore, the author decided to distribute 200 scale
questionnaires. Survey research is conducted mainly for
customers in the city area. Ho Chi Minh City has the highest Factors Number Cronbach’s
customer coverage during the Covid-19 pandemic. of items Alpha
Convenience of service 9 0.841
IV. DATA ANALYSIS AND RESULTS
Contact with customers 5 0.775
D. Descriptive Statistics Delivery driver time 5 0.826
Table 1. Descriptive statistical analysis of the sample 6 0.863
Food delivery status
Std. Voucher 5 0.830
N Minimum Maximum Mean
Deviation
TL1 189 1 5 4.31 .794 Pandemic safety 6 0.849
TL2 189 2 5 4.17 .727
TL3 189 1 5 3.96 .944 F. Results of Exploratory Factor Analysis (EFA)
TL4 189 2 5 4.30 .763
TL5 189 2 5 4.26 .845 The Kaiser-Meyer-Olkin (KMO) measure of sampling
TL6 189 1 5 4.13 .841 adequacy varies between 0 and 1. In this study, the KMO
TL7 189 2 5 4.19 .787 value is 0.922 which points out that factor analysis is relevant
TL8 189 1 5 4.08 .910 and appropriate for this study (Table 2). The Bartlett test
TL9 189 2 5 4.15 .883 (p=0.000<0.05) also determined that the variables all display
TX1 189 1 5 4.24 .826
TX2 189 2 5 4.16 .831
significant levels of correlation
TX3 189 1 5 4.20 .852 Table 3. KMO and Bartlett's test
TX4 189 1 5 4.12 .876
TX5 189 2 5 4.19 .852 Kaiser-Meyer-Olkin Measure of Sampling
TG1 189 2 5 4.17 .827 .922
TG2 189 2 5 4.06 .873
Adequacy.
TG3 189 1 5 4.16 .838 Approx. Chi-Square 2341.863
TG4 189 1 5 4.14 .829 Bartlett's Test of
TG5 189 2 5 4.29 .801 df 276
Sphericity
TR1 189 1 5 4.14 .920
TR2 189 1 5 4.06 .854
Sig. .000
TR3 189 1 5 4.12 .867
TR4 189 1 5 4.17 .829 Table 4. Total varience explained
TR5 189 1 5 4.16 .873 Factor Initial Eigenvalues Extraction Sums of Squared Rotation
TR6 189 1 5 4.20 .813 Loadings Sums of
V1 189 1 5 4.25 .893 Squared
V2 189 1 5 4.15 .837 Loadings
V3 189 2 5 4.14 .807
V4 189 1 5 4.23 .824 Total % of Cumulative Total % of Cumulative Total
V5 189 1 5 4.30 .844 Variance % Variance %
AT1 189 2 5 4.39 .782
1 10.37 43.236 43.236 10.377 43.236 43.236 3.514
AT2 189 1 5 4.20 .831
AT3 189 2 5 4.34 .780 2 1.311 5.461 48.697 1.311 5.461 48.697 3.252
AT4 189 2 5 4.37 .771 3 1.241 5.171 53.868 1.241 5.171 53.868 3.102
AT5 189 1 5 4.13 .886 4 1.049 4.371 58.239 1.049 4.371 58.239 2.980
AT6 189 1 5 4.30 .785 5 1.005 4.188 62.427 1.005 4.188 62.427 2.135
S1 189 2 5 4.24 .807 6 .876 3.648 66.076
S2 189 2 5 4.23 .762 7 .772 3.215 69.291
S3 189 2 5 4.04 .942
8 .735 3.063 72.354
S4 189 1 5 4.11 .850
S5 189 1 5 4.22 .788 9 .644 2.684 75.038
Valid 10 .632 2.632 77.670
N 11 .605 2.521 80.191
189
(listwi 12 .588 2.450 82.641
se) 13 .555 2.313 84.954
(Source: Investigation and analysis team of authors) 14 .495 2.064 87.018
From the descriptive statistics table, we can see that all 15 .448 1.867 88.885
the mean values are greater than 4, which proves that most of 16 .409 1.703 90.587
the research survey subjects agree with the content of the 17 .354 1.476 92.063
18 .349 1.454 93.517
above survey. The standard deviations of all variables are at
19 .336 1.400 94.917
a low level of less than 1. It shows that most of the
20 .320 1.335 96.252
respondents agree with the content of the above survey.
21 .266 1.109 97.360
E. Results of Cronbach’s alpha 22 .239 .997 98.358
23 .213 .888 99.246
The results of Cronbach's Alpha analysis show that all
24 .181 .754 100.000
scales have coefficients greater than 0.7 and correlation
Extraction Method: Principal Component Analysis.
328
G. Correlation analysis results (Pearson coefficient) factors Contact with customers and Food delivered are
combined into one influencing factor, so the dependent
Table 5. Correlation matrix
variable. From the initial 7 factors through the process of
CU NC AT TM V S analysis and evaluation, there are 5 remaining factors: (1)
Pearson ** ** ** Professionalism and prestige, (2) Satisfy customer needs, (3)
CU 1 .713 .688 .655 .546** .725**
Correlation
Pandemic safety, (4) Interface Smart, and (5) Discount both
Pearson
NC .713** 1 .653** .681** .611** .683** have a positive impact on customer satisfaction with Baemin
Correlation
Pearson food delivery service. Below is the research model after
AT .688** .653** 1 .688** .529** .639**
Correlation testing (Figure 3):
Pearson
TM .655** .681** .688** 1 .566** .625**
Correlation
Pearson Professional and
V .546** .611** .529** .566** 1 .567**
Correlation Reputable 0.369
Pearson Level of
S .725** .683** .639** .625** .567** 1
Correlation customer
**. Correlation is significant at the 0.01 level (2-tailed).
Satisfy customer 0.199 satisfaction
needs
about
The results of the correlation analysis in Table 5 show that 0.127 Baemin’s
the dependent variable S is linearly correlated with each Pandemic safety service
independent variable (CU, NC, AT, TM, V), and each 0.126 quality
independent variable is also linearly correlated with each Smart interface.
other. The correlation coefficients between the dependent 0.089
variable S and the independent variables have values from Discount
0.567 (between variable S and variable V) to 0.725 (between
variable S and variable CU). The largest correlation
coefficient between the independent variables is 0.713 and Figure 3. Research model after testing
the smallest is 0.529. Thus, it can be concluded that there is a
When the level of satisfaction increases, Baemin's priority
close correlation between the dependent variable and the
in using services will also be increased, when customers feel
independent variable, but no multicollinearity occurs.
satisfied with the service, they will certainly continue to
Therefore, they satisfy the conditions for regression analysis.
choose to use the service. Besides, they will tend to
H. Results of Regression analysis recommend to friends and relatives to know and use together.
The results of Regression analysis show that the multiple When the level of satisfaction increases, Baemin's priority in
linear regression model has an R square equal to 0.612 and using services will also be increased, when customers feel
an adjusted R square equal to 0.601. This proves that the built satisfied with the service, they will certainly continue to
multiple linear regression model is suitable for the data set at choose to use the service. Besides, they will tend to
60.1%. In other words, with 60.1% change in customer recommend to friends and relatives to know and use together.
satisfaction about Baemin's food delivery service quality On the other hand, in the time of prolonged complicated
during the Covid-19 pandemic is explained by 5 independent epidemic, there are many factors affecting, if not well
variables of the model. controlled the above 5 factors closely will affect the level of
I. Result of variance analysis ANOVA customer satisfaction. Lack of professionalism and
credibility, delay in delivery, service unable to meet customer
Table 6. Analysis of variance ANOVA needs well, or delivery staff not ensuring safety in epidemic
Model Sum of df Mean F Sig.
prevention, Complicated application interface causing
Squares Square confusion for users, or service costs are too high during a
Regression 45.613 5 9.123 57.626 .000b difficult time for everyone due to the epidemic situation,
1 Residual 28.970 183 .158 which will affect the level of customer satisfaction, affecting
Total 74.583 188 the choice of service for the next time their next use.
a. Dependent Variable: S
During the difficult time of the epidemic, the regulations
b. Predictors: (Constant), V, AT, CU, TM, NC
were issued to control the disease well, Baemin needed to
The results of ANOVA analysis of variance (Table 6) take measures to control the above factors to increase the
show that the parameter F has a significance level of 0.000, level of customer satisfaction when the number of customers
showing that the built regression model is consistent with the using Service usage will increase and can gain a foothold in
actual data set, and the included variables are all significant. the food delivery market compared to other mediocre
statistically at the 5% level of significance. In other words, services.
the independent variables in the model are related to the
dependent variable. V. CONCLUSION
Research results show that the scales in the model are
J. Research results
reliable and identify service quality factors affecting
The research results showed that the factor Status of food customer satisfaction using Baemin's food delivery service.
delivered did not affect the dependent variable. Due to the There are a total of 5 factors affecting satisfaction according
epidemic situation, food and food are becoming scarcer than to different levels of impact, which are:(1) Professional and
ever and in compliance with strict government directives, the Reputable, (2) Satisfy customer needs, (3) Pandemic safety,
delivery staff is allowed to go out less so customers no longer (4) Smart interface, (5) Discount.
paying attention to the delivery status of the food. Two good
329
The study also evaluates the influence level of each well as delivery staff, the authors propose a number of
customer group according to gender, age, occupation, and recommendations:
frequency of service use of each customer to each factor in
the research model, from which there is a basis to consider Strengthen propaganda and instruction for all employees,
the impact of each factor on satisfaction among different especially the team of food delivery drivers, about the
customer groups. epidemic prevention manual as well as the knowledge to
know if an employee is unfortunately infected.
Through this research result, the author also makes some
recommendations to help Baemin improve customer There is a support mode for drivers who are infected or
satisfaction when using Baemin's food delivery service isolated due to the nature of their work, moving to many
during the Covid-19 pandemic. places and interacting with many customers.
Professional and reputable: This is the factor that has the Strong communication on the service's information
greatest influence on customer satisfaction. Therefore, the channels about the epidemic situation every day so that
author proposes some recommendations: everyone including employees and customers can catch it
promptly.
In addition to automatically saving the location for the
next order, the author recommends that the application should Smart interface: For online food delivery applications,
have a feature that automatically updates the location flexibly one of the things that enhances customer satisfaction is an
according to the movement of customers in different places, application with a friendly, easy-to-use interface that helps
which helps the booking customers' orders become faster and customers perform the necessary operations. necessary when
easier, thereby improving the professionalism of the service. there is a need to eat without anyone's help. Therefore, the
Equip enough delivery support equipment for staff to ensure author proposes some recommendations:
the quality of food when delivered, strengthen customer Build a professional team to design a more attractive,
feedback updates to get closer, and understand customers' beautiful and cute brand image
tastes better, from which the service will be better. become
much more reputable and professional. Continue to improve the quality of the interface, which is
vivid, clear and easy to operate
Strengthening communication promotion, training a
professional design team to promote Baemin brand closer to Uploading illustrations on a rich, diverse and realistic
customers, especially during the Covid-19 epidemic, interface
especially actively advertising on service Fanpage channel Improve and develop the responsiveness, automation of
Baemin, followed by posting pictures showing the the application interface, chatbox work more efficiently and
professionalism of the service in community Facebook support customers faster
groups specializing in food reviews and similar channels to
reach customers in the best way. Discount: Customers always want to own quality goods
at affordable prices, especially during the pandemic, the
Satisfy customer needs: This is a factor that has a personal economy becomes very limited. Therefore,
significant impact on customer satisfaction when using customers will prefer to order food at the service with the
Baemin's food delivery service during the Covid-19 most discount codes and incentives. Therefore, the author
pandemic. Most customers nowadays tend to have difficulty proposes some recommendations:
choosing which service to use service, so they often lean
towards the service that benefits them and is able to meet the Create more ways to accumulate points to redeem gift and
needs of customers in the best way. One of the main discount codes to facilitate customers to use the service more.
contributors to this is the staff. Therefore, the author proposes There are preferential policies for customers who use the
the following recommendations: service with a large and regular frequency.
Regularly open specialized training courses, training and Apply many promotions for customers who introduce
improving professional knowledge for staff on customer care friends and relatives to know and use Baemin's service
activities.
Limitations and directions for further research:
Employees at the office must regularly update customers'
tastes on many dishes of various restaurants to be able to This research has only stopped in areas within HCMC and
better meet the needs of customers. with the number of surveys just over the minimum of 189
customers. Future research will increase the number of
Promote and improve the sense of responsibility, working respondents and expand further to include all other provinces
style, and attitude of employees, especially delivery drivers, and cities. This study only stops at assessing the level of
in customer service. Increase the opening of monthly customer satisfaction in Ho Chi Minh City in general and
professional training courses on necessary skills for delivery indicates a general direction for all Baemin service staff to
drivers such as the art of communicating with customers, improve in the coming time. the level of customer satisfaction
skills to handle unexpected situations, skills to interact with when using Baemin services for each specific district in the
customers, and so on. city. Ho Chi Minh City because each region will have
Pandemic safety: Facing the increasingly complicated different restrictions on the level of activities affecting
epidemic situation, the government has issued more and more customer satisfaction. Future research will focus more
stringent directives and blocked more areas. Currently, some specifically on each district in Ho Chi Minh City and have a
people have been vaccinated to minimize the spread of the specific comparison of the level of customer satisfaction
disease. For that reason, to ensure the safety of customers as between districts, thereby increasing the rating scale between
quality. Baemin's service quality in each specific district
330
helps the staff operating in the districts see their own Innovation. Journal of Open Innovation: Technology, Market,
limitations so that they can compete to complete their and Complexity, 7(1), 76.
assigned tasks. https://doi.org/10.3390/joitmc7010076
[6] Aeny, N., Ekhsan, M., & Tanjung, A. (2019). The Effect of
This study only assesses the satisfaction level of Service Price And Quality On Customer Satisfaction Online
customers who have been using Baemin food delivery service Transportation Services. Journal of Research in Business,
in particular, not yet assessed the level of customer Economics, and Education, 1(1), 3-10.
satisfaction when using the service. other food deliveries such [7] See-Kwong, G., Soo-Ryue, N. G., Shiun-Yi, W., & Lily, C.
as Grabfood, Gofood, Now, and so on compare with (2017). Outsourcing to online food delivery services:
Baemin's service. Future research can expand these subjects Perspective of F&B business owners. The Journal of Internet
and have a specific comparison between rival food delivery Banking and Commerce, 22(2), 1-18
services with Baemin, and group the strength and weakness [8] Parasuraman, A., L. L. Berry, & V. A. Zeithaml (1991).
of each factor corresponding to each factor. service's brand Refinement and Reassessment of the SERVQUAL Scale,
name, as well as comparing rating scales between services of Journal of Retailing, 67(4), 420-450.
different caliber, thereby helping Baemin to discover its [9] Parasuraman, A., V. A. Zeithaml, & L. L. Berry, (1988).
advantages and disadvantages compared to rival services to SERVQUAL: A Multiple-Item Scale for Measuring Consumer
have a good strategy and attract more customers to use in the Perceptions of Service Quality, Journal of Retailing, 64(1), 12-
future. 40.
[10] Nguyen, T. D. et al. (2003). Measuring service quality of
REFERENCES outdoor amusement parks in Ho Chi Minh City, CS 2003-19,
[1] Ministry of Industry and Trade 2021 Vietnam e-commerce HCMC University of Economics.
report 2021. [11] Spreng, R.A., and Mackoy, R.D (1996). An Empirical
[2] Dung, P. T. (2021). Analysis of factors affecting customers' Examination of a Model of Perceived Service Quality and
online purchase of food in Da Nang city during the Covid-19 Satisfaction, Journal of Retailing, 201-214.
pandemic. Journal of Science and Technology Duy Tan [12] Buttle, F. A. (1998). Word of mouth: understanding and
University, 89. managing referral marketing. Journal of strategic
[3] Yuyang Zhao, Fernando Bacao. (2020). What factors determine marketing, 6(3), 241-254.
customers' continued use of food delivery apps during the 2019 [13] Lee, J. Y., Cole, T. B., Palmiter, R. D., & Koh, J. Y. (2000).
coronavirus pandemic? International Journal of Hospitality Accumulation of zinc in degenerating hippocampal neurons of
Management - Portugal, 91. ZnT3-null mice after seizures: evidence against synaptic vesicle
[4] Jankit Chotigo, Yasuo Kadono. (2021). Comparative analysis of origin. Journal of Neuroscience, 20(11), RC79-RC79.
key factors encouraging food delivery app adoption before and [14] Gilbert, G. R., & Veloutsou, C. (2006). A cross‐industry
during the COVID-19 pandemic in Thailand. Sustainability, 13. comparison of customer satisfaction. Journal of Services
[5] Prasetyo YT, Tanto H, Mariyanto M, Hanjaya C, Young MN, Marketing.
Persada SF, Miraja BA, Redi AANP. (2021). Factors Affecting [15] Mesay, S. (2012). Bank Service Quality. Customer Satisfaction
Customer Satisfaction and Loyalty in Online Food Delivery and Loyalty in Ethiopian
Service during the COVID-19 Pandemic: Its Relation with Open
331
Controllable Green Synthesis and Morphological

Properties of Gold Nanostar
P. Quoc-Duy Huynh Van-Dung Le Chi-Hien Dang
Ton Duc Thang University Institute of Chemical Technology Institute of Chemical Technology
Ho Chi Minh City, Vietnam Vietnam Academy of Science and Vietnam Academy of Science and
duyhuynh2201@gmail.com Technology Technology
tohoahocctb@gmail.com dangchihien@gmail.com
Radek Fajgar The-Ha Stuchlikova Jiri Stuchlik

Institute of Chemical Process Institute of Physics of the ASCR Institute of Physics of the ASCR
Fundamentals of the AS CR Prague, Czech Republic Prague, Czech Republic
Prague, Czech Republic hotheha@fzu.cz stuj@fzu.cz
fajgar@icpf.cas.cz
Thanh-Danh Nguyen*
Institute of Chemical Technology
Vietnam Academy of Science and
Technology
ntdanh@ict.vast.vn
ORCID: 0000-0001-6330-8916
Abstract: Gold nanostar (AuNS) has been given wide mediated approaches which can successfully control the
attention due to the special chemical and physical properties length and the number of spikes by varying the reaction time
for enhanced potential of their applications in sensing, [4], experimental parameters [5] or reagent concentration [6].
catalysis and nanomedicine. Their performance is strongly However, these approaches have required toxic surfactants
dependent on the shape and particle size which particularly such as CTAB (hexadecyltrimethylammonium bromide) [7],
relate to surface plasmonic resonance (SPR). Thus, the Triton X-100 [6] or polymers such as poly-vinylpyrrolidone
controllable synthesis of AuNS without toxic compounds is (PVP) [2], PEG (polyethylene glycol) [8] and also prepared in
crucial to develop sustainable materials. In this study, a novel organic solvents such as N,N-dimethylformamide (DMF) [9]
protocol for the synthesis of AuNS using gold seeds which are harmful to the environment.
combining tween 80 as a novel stabilization agent in the In the present work, we introduce an eco-friendly
presence of hydroquinone is reported. The branch length and approach that could prepare AuNS in water medium. The
size of AuNS (78 – 85 nm) were controlled by a varied method utilizes gold seeds in the presence of hydroquinone
amount of tween 80 (0.38 – 3.86 mM), confirmed by UV-Vis, (HQ) as a reductant and tween 80 as an effective non-toxic
SEM, DLS and zeta potential characterizations. surfactant to control the size and morphology of nanometals.
Keywords: gold nanostar, controllable synthesis, II. EXPERIMENT SECTION

morphology, sustainable material A. Preparation of gold seed
I. INTRODUCTION Seed was prepared by reduction of HAuCl4 using
trisodium citrate as a previous report [2]. 100 μL HAuCl4
Anisotropic gold nanoparticles (AuNPs) have been 25mM was added into a vial-20 mL containing 5 mL distilled
recognized as a useful material due to their excellent optical water under stirring (500 rpm) and heating at 80 ᵒC. Then, 400
properties which can be applied to catalysis, surface-enhanced μL sodium citrate (1%) was quickly added to the mixture and
Raman spectroscopy (SERS) and biological applications [1]. keep heating for 1 hour. As-prepared seed solution was turned
The optical properties strongly depend on the size and into a wine red color.
morphology which lead to interests in anisotropic AuNPs.
AuNS is one of the most interesting forms because the B. Preparation of gold nanostar
branches grow on surfaces of the core exhibit strongly Firstly, 20 μL tween 80 was added into a vial-20 mL
localized surface plasmon resonances (LSPRs) which provide containing distilled water (10 mL). Then, 50 μL of seed
stronger reactivity and optical properties than spherical shape solution was added immediately. The mixture was stirred for
[2]. The core acts as an antenna producing electromagnetic 25 min under 1500 rpm at room temperature. After coating
field enhancement and the morphology of the spikes (length tween 80 on seed surface, 100 μL HAuCl4 25mM were added.
and number) strongly affects plasmon frequency and intensity Then, 300 μL hydroquinone solution (0.1 M) was quickly
[3]. Therefore, methodologies for controlling size and shape added to the reaction and continuously stirred for 30 minutes.
of AuNS has become a noticeable topic in recent years. The solution was turned from slightly pink to greenish-blue.
In previous works, AuNS was usually prepared by seed- The obtained solution was incubated at room temperature in

an additional 30 minutes. Finally, the gold-nanostar solution Table 1. DLS and Zeta Potential of AuNS with Different Tween 80
was centrifuged at 5000 rpm for 20 min, the supernatant Concentrations
removed, and rinsed with distilled water. [Tween 80] Average size PI Zeta potential
(mM) (nm) (mV)
III. RESULT AND DISCUSSION 0.38 82.93 ± 26.91 0.055 -54.7
It has been reported that HQ can act as an efficient 0.77 85.13 ± 20.14 0.08 -54.9
1.54 81.10 ± 6.28 0.112 -56.4
reductant for Au3+ ions to form star-shaped nanoparticles 2.31 81.10 ± 16.16 0.071 -58.4
because HQ can prevent the formation of secondary nuclei 3.09 78.67 ± 15.54 0.051 -54.0
[10]. Furthermore, with the appropriate amount of HQ, the 3.86 83.40 ± 10.21 0.025 -64.1
formation of AuNS quickly reaches a chemical equilibrium.
As shown in Fig. 1, SPR maximum absorption peak at 713 nm
is rapidly formed in 5 min after the injection of HQ into gold
seeds stabilized by tween 80 and kept unchanging until 30
min. Additionally, the absorption intensity at wavelength 520
nm, which is characteristic for the SPR peak of seeds, rapidly
shifts to 713 nm in 5 min, confirming the complete
transformation of spherical seeds to star-shape nanostructure
in only 5 min after the HQ addition process. However, to
assure that all Au3+ ions were completely reduced to gold
atoms, the reaction was carried out for 30 min.
The growth of AuNS was performed in water medium at
room temperature with the support of tween 80 which can
adsorb strongly on gold nanoparticles surface due to hydroxyl
groups and oxygen (C – C – O) in its structure [11]. With an
appropriate amount, the adsorbed tween 80 blocks partially
some of the active sites on the crystalline surfaces of gold
seeds. Therefore, when gold atoms are deposited onto the gold
seeds, the growth of gold crystals is performed at the
unblocked sites whereas the blocked surfaces are suppressed,
thereby the branches of AuNS formed [12].
In this work, gold seeds with a concentration of 1.15x10-3
mM were used to investigate the effect of tween 80
concentrations on the formation of AuNS. As seen in Table 1,
the colloidal solutions of nanogold obtained is stable with
greatly negative zeta potentials in the range from -54.0 to -64
mV, indicating that tween 80 is a suitable stabilizing agent for
the synthesis of nanogold. In particular, there is no change in
particle size with an increase of tween 80 concentration as
seen in DLS data. It confirms the presence of the stabilizer did
not induce the significant changes in particle size of nanogold
in the colloidal solution. Moreover, the polydispersity index
(PI) values are found to be around 0.1 which confirms the
monodisperse distribution of nanogold in the solutions.
0.3
0 min
5 min
Absorbance
10 min
0.2 15 min
20 min
25 min
30 min
0.1
Figure 2. UV-Vis spectra (a) and SEM images of AuNS at different
tween 80 concentrations: 0.38 mM (b), 0.77 mM (c), 1.54 mM (d),
2.31 mM (e), 3.09 mM (f), 3.86 mM (g). The samples synthesized
0.0 from seed concentration of 1.15 x 10-3 mM, [AuCl4]- concentration
400 600 800 of 0.24 mM and HQ concentration of 2.86 mM.
Wavelength (nm) The morphology and shape of AuNS were determined via
the SEM measurement. As increasing the concentration of
Figure 1. UV-Vis spectra for AuNS growth against reaction time tween 80 used in the coating step from 0.38 mM to 3.86 mM,
after the addition of HQ UV-Vis spectra of AuNS samples show a blue shift from 831
nm to 758 nm which indicates morphology transformation of
colloidal solutions (Fig. 2a). At the low tween 80
concentration (0.38 mM) used, the number of gold seed
333
particles was insufficiently covered by tween 80 molecules We also found that there are half-star particles formed at
which led to formation of star-shape but spherical shape was average amount of tween 80 (2.31 mM) which is likely
formed from the uncovered seeds after the growth stage (Fig. because some uncovered facets still exist and locate for
2b). However, the stability of AuNS at this tween 80 branches growth (Fig. 3). However, there is still a large
concentration was still assured because zeta potential amount of tween 80 absorbed onto gold seed surfaces
measurement reached -54.7 mM. When the concentration of reducing Au3+ ion together with HQ. As a consequence, a
tween 80 to 0.77 mM, more gold seeds were covered significant amount of gold atoms are deposited on the core and
efficiently to form star-shaped particles rather than spherical there is no gold atom for the growth of branches, resulting in
particles. Therefore, it suggests that the least concentration of shorter branches. At tween 80 concentration of 1.54 mM,
tween 80 to obtain AuNS should be 0.77 mM. completely star-shape particles are formed as shown in Fig 2d.
Raising the tween 80 amount to 1.54 mM, more facets of Our finding leads to the conclusion that an appropriate amount
gold seed are covered resulting in fewer branches grown on of tween 80 used is crucial for the formation of gold nanostars
seed surfaces (Fig. 2d). Compared to AuNS formed from 0.38 and branches length of AuNS could be controlled via varying
mM and 0.77 mM of tween 80, AuNS formation was more amount of tween 80 in terms of our system.
completely at 1.54 mM tween 80 since some of spherical
particles were formed in Fig 2b and 2c. It can be explained IV. CONCLUSION
that all of seed particles were covered sufficiently to form star- This work successfully synthesized AuNS without using
shaped particles without spherical particles formation. Indeed, toxic compounds. Tween 80, a green capping agent, can be
AuNS formed at lower tween 80 concentration possesses used to control the shape and morphology of AuNS. The
longer wavelengths (831 nm for 0.38 mM) compared with the morphology of AuNS has well-defined SEM images and DLS
high concentrations (809 nm for 0.77 mM and 795 nm for 1.54 techniques. The AuNS can be fabricated efficiently in tween
mM). 80 concentration range between 0.77 and 1.54 mM. The high
As seen in Figs. 2e and 2f, the number of star-shaped amount of tween 80 can produce a great number of gold
particles significantly reduced at a higher concentration of nanoparticles because of the reducing ability of tween 80.
tween 80. In particular, no AuNS is formed at 3.86 mM which ACKNOWLEDGEMENT
is clearly confirmed by the SEM image in Fig. 2g. The UV-
Vis spectra also indicated a blue shift with decreasing amount The authors are thankful to Vietnam Academy of Science
of tween 80. AuNS with fewer branches produced at high and Technology (QTCZ01.01/22-23) and to Czech Academy
tween 80 concentration can contribute to decreasing pattern of of Sciences (Mobility Plus Programme No. VAST-22-02) for
wavelengths. Similar results were found in previous studies funding this work. This research is supported by Operational
when they varied the amount of used surfactant [15, 16]. Programme Research, Development and Education project
SOLID21-CZ.02.1.01/0.0/0.0/16_019/0000760 and the
Moreover, tween 80 can act as an excellent stabilizer as CzechNanolab research infrastructure - LM2018110.
well as a good reducing agent at high concentration [13]. An
increasing number of tween 80 molecules in the reaction can REFERENCES
make gold seeds fully blocked. Therefore, there are no free [1] M. Boronat, D. Combita, P. Concepcion, A. Corma, H. Garcia,
facets to favor the growth of star-shaped particles. Moreover, R. Juarez, S. Laursen and J. de Dios Lopez-Castro, “Making
in addition to reducing the ability of HQ, tween 80 can act as C–C bonds with gold: identification of selective gold sites for
a weaker reductant which speeds up the reduction rate of Au3+ homo-and cross-coupling reactions between iodobenzene and
ions. Fast reduction in growth stage interrupting the kinetics- alkynes”, J. Phys. Chem. C, vol. 116, pp. 24855-24867, 2012.
favored growth for star-shape causes secondary nucleation [2] C.G. Khoury and T. Vo-Dinh, “Gold nanostars for surface-
[14]. To confirm the reducing ability of tween 80, the same enhanced Raman scattering: synthesis, characterization and
procedure was carried out without the presence of HQ. As optimization”, J. Phys. Chem. C, vol. 112, pp. 18849-18859,
seen in Fig. 3, tween 80 can reduce easily Au3+ ions into 2008.
AuNPs with a band of approximately 560 nm which is [3] J. Piella, A. Gónzalez-Febles, J. Patarroyo, J. Arbiol, N.G.
characteristic for SPR peak of AuNPs. Bastús and V. Puntes, “Seeded-growth aqueous synthesis of
colloidal-stable citrate-stabilized Au/CeO2 hybrid
0.20 nanocrystals: heterodimers, core@ shell, and clover-and star-
0 min like structures”, Chem. Mater. Vol. 31, pp. 7922-7932, 2019.
5 min [4] S. Barbosa, A. Agrawal, L. Rodríguez-Lorenzo, I. Pastoriza-
Absorbance
0.15 10 min Santos, R.A. Alvarez-Puebla, A. Kornowski, H. Weller and

15 min L.M. Liz-Marzán, “Tuning size and sensing properties in
20 min colloidal gold nanostars”, Langmuir, vol. 26 pp. 14943-14950,
0.10 25 min
2010.
30 min
[5] G.A. Vinnacombe-Willson, N. Chiang, L. Scarabelli, Y. Hu,
0.05 L.K. Heidenreich, X. Li, Y. Gong, D.T. Inouye, T.S. Fisher,
P.S. Weiss, “In situ shape control of thermoplasmonic gold
nanostars on oxide substrates for hyperthermia-mediated cell
0.00 detachment”, ACS cent. Sci., vol. 6, pp. 2105-2116, 2020.
500 600 700
[6] A.L. Siegel and G.A. Baker, “Bespoke nanostars: synthetic
Wavelength (nm) strategies, tactics, and uses of tailored branched gold
nanoparticles”, Nanoscale Adv., vol. 3, pp. 3980-4004, 2021.
Figure 3. UV-Vis spectral study of gold nanoparticles growth [7] A.M. Fales, H. Yuan and T. Vo-Dinh, “Development of
without Hydroquinone addition hybrid silver-coated gold nanostars for nonaggregated
surface-enhanced Raman scattering”, J Phys. Chem. C, vol.
118, pp. 3708-3715, 2014.
334
[8] X. Xie, J. Liao, X. Shao, Q. Li, Y. Lin, “The effect of shape [13] H.J. Li, A.-Q. Zhang, Y. Hu, L. Sui, D.-J. Qian and M. Chen,
on cellular uptake of gold nanoparticles in the forms of stars, “Large-scale synthesis and self-organization of silver
rods, and triangles”, Sci. Rep. vol. 7, pp. 1-9, 2017. nanoparticles with Tween 80 as a reductant and stabilizer”,
[9] P.S. Kumar, I. Pastoriza-Santos, B. Rodríguez-González, Nanoscale Res. Lett. vol. 7, pp. 1-13, 2012.
F.J.G. De Abajo, L.M. Liz-Marzán, “High-yield synthesis and [14] M.L. Personick and C.A. Mirkin, “Making sense of the
optical response of gold nanostars”, Nanotechnology, vol. 19, mayhem behind shape control in the synthesis of gold
pp. 015606, 2007. nanoparticles”, J. Am. Chem. Soc. vol. 135, pp.18238-18247,
[10] J. Li, J. Wu, X. Zhang, Y. Liu, D. Zhou, H. Sun, H. Zhang, B. 2013.
Yang, “Controllable synthesis of stable urchin-like gold [15] H. Wang, Y. Pu, B. Shan, B., M. Li, “Combining experiments
nanoparticles using hydroquinone to tune the reactivity of gold and theoretical modeling to interrogate the anisotropic growth
chloride”, J. Phys. Chem. C, vol. 115, pp. 3630-3637, 2011. and structure–plasmonic property relationships of gold
[11] P. Suchomel, L. Kvitek, R. Prucek, A. Panacek, A. Halder, S. nanostars”, Inorg. Chem., vol. 58, pp. 12457-12466, 2019.
Vajda and R. Zboril, “Simple size-controlled synthesis of Au [16] A. Casu, E. Cabrini, A. Donà, A. Falqui, Y. Diaz‐Fernandez,
nanoparticles and their size-dependent catalytic activity”, Sci. C. Milanese, A. Taglietti, P. Pallavicini, “Controlled synthesis
Rep., vol. 8, pp. 1-11, 2018. of gold nanostars by using a zwitterionic surfactant”, Chem.
[12] J.E. Ortiz-Castillo, R.C. Gallo-Villanueva, M.J. Madou, and Eur. J., vol. 18, pp. 9381-9390, 2012.
V.H. Perez-Gonzalez, “Anisotropic gold nanoparticles: A [17] J. J. Nordlund, P. E. Grimes, J. P. Ortonne, “The safety of
survey of recent synthetic methodologies,” Coord. Chem. hydroquinone”, J. Eur. Acad. Dermatol. Venereol., vol. 20,
Rev. vol. 425, pp. 213489, 2020. pp. 781-787, 2006.
335
Isolation, Selection and Identification of the

Probiotic Properties of Lactobacillus spp. from
Shrimp Ponds in Nhon Trach, Dong Nai
Doan Thi Tuyet Le Le Thi Thu Huong Phan Pham
No.10 Huynh Van Nghe, Buu Long, No.10 Huynh Van Nghe, Buu Long, No.10 Huynh Van Nghe, Buu Long,
Bien Hoa, Dong Nai, Vietnam Bien Hoa, Dong Nai, Vietnam Bien Hoa, Dong Nai, Vietnam
tuyetledt@lhu.edu.vn lethuhuong@lhu.edu.vn https://orcid.org/0000-0002-6541-3442
Pham Minh Thinh Vo Thi Lan Chi Nguyen Phuoc Trung

No.10 Huynh Van Nghe, Buu Long, No.10 Huynh Van Nghe, Buu Long, No.10 Huynh Van Nghe, Buu Long,
Bien Hoa, Dong Nai, Vietnam Bien Hoa, Dong Nai, Vietnam Bien Hoa, Dong Nai, Vietnam
anhchangnhaque60@gmail.com vothilanchibh@gmail.com nguyenphuoctrung1993@gmail.com
Huynh Minh Hieu Do Minh Anh

No.10 Huynh Van Nghe, Buu Long, No.10 Huynh Van Nghe, Buu Long,
Bien Hoa, Dong Nai, Vietnam Bien Hoa, Dong Nai, Vietnam
hmhieu007@gmail.com minhanhdo@lhu.edu.vn
Abstract: Probiotics are now being used to replace extracellular enzymes that break down excess organic
antibiotics to improve quality in aquaculture, particularly in compounds to improve pond water quality [3, 4]. Recent
shrimp farms. The initiative in seed sources of Lactobacillus studies have also shown that addition of inoculant to shrimp
strains possessing varieties of potential to be used as feed, helped increase resistance for black tiger shrimp and
probiotics is an urgent issue in probiotic production. Isolates maintained good digestion, healthy growth, increased
of Lactobacillus was obtained from this research on pre- survival rate and especially significant increase yield under
selective medium MRS Broth + 50mg/l nystatin and intensive culture conditions [5, 6]. To date, there have been
selective medium MRS agar + 0,5% CaCO3; and then, substantial studies on probiotic strains for shrimp probiotic
production in the world including Vietnam [5, 6, 7, 8]. From
identified by biochemical tests and analyzing their 16S gene
the knowledge of the authors, studies on isolating
sequences. Next, examining the beneficial properties of
Lactobacillus strains with probiotic potential are limited in
Lactobacillus spp. have been identified. The three Litopenaeus vannamei shrimp. Those are the reasons for
Lactobacillus strains identified were L. salivarius conducting this study. The results also have further
LIITA1.2.2, L. reuteri LIVTA2.4.1, and L. plantarum meanings that help to be proactive about seed sources for
LVIITA3.3.9. These strains were alive in the media with pH inoculant as well as contribute to improving the efficiency
range from 1 to 9, salinity from 0‰ to 50‰, and showed of white leg shrimp farming.
resistance to Vibrio parahaemolyticus and E. coli at the
densities of 105 CFU/ml, 106 CFU/ml, and 107 CFU/ml, II. MATERIALS AND METHODS
which are great the potential for probiotic. A. Materials
Keywords: Lactobacillus, Probiotic, L. salivarius, L. Water samples from shrimp ponds and gastrointestinal
tract samples of pacific white shrimp (Litopenaeus
reuteri, L. plantarum, V. parahaemolyticus, E. coli.
vannamei) in Nhon Trach district, Dong Nai province were
I. INTRODUCTION collected. The isolate of V. parahaemolyticus from dead
diseased shrimps was provided by Biotechnology Center of
In recent years, the global demand for seafood is Ho Chi Minh City, and E. coli ATTC25922 strain was
increasing. In the situation that natural seafood is depleting, provided by the Museum of Standard Microbiology -
aquaculture is the source of the future. Probiotics have been Institute of Microbiology Study and Biotechnology,
used to replace antibiotics in aquatic products, helping to Vietnam National University, Hanoi.
increase the survival and development of aquatic animals [1,
2]. Lactobacillus strains have been known as potential B. Methods
probiotic sources playing an important role in the Water and shrimp samples were coded, stored at 4 oC,
gastrointestinal tract of their hosts, helping to stimulate and transferred to the laboratory [9]. Shrimp samples were
hosts’ immunity, competing for adhesion sites. They also washed with alcohol 70o, and the organs were harvested
secrete antibacterial substances such as lactic acid, acetic [10]. Isolates for Lactobacillus strains were placed in PE
acid, bacteriocin that are able to inhibit the growth of bags containing MRS Broth + 50mg/l Nystatin medium,
pathogenic bacteria, and at the same time secrete incubated anaerobically at 37oC [11, 12]. The suspension

was then diluted and spread on MRS agar supplemented fermenting sugars (glucose, fructose, maltose, lactose,
with 0.5% CaCO3 [6] and anaerobically incubate at 37oC for sucrose). According to Bergey's bacterial taxonomy (2011),
48 hours. The Lactobacillus isolates were determined for the above biochemical test results were consistent with the
their lactic acid production by Uffelman reagent [13], their biochemical characteristics of the genus Lactobacillus [16].
antibacterial activity against Vibrio parahaemolyticus and E.
coli by the agar perforation method [14]. Salt tolerance and Molecular identification
Lactobacillus salivarius LIITA1.2.2
pH adaptability of the Lactobacillus isolates were also CTGGCTCAGGACGAACGCTGGCGGCGTGCCTAATACATGCAAGTCGAACGAAACTTTCTTACACCGAA
observed under different NaCl salt concentrations (0‰; TGCTTGCATTCACCGTAAGAAGTTGAGTGGCGGACGGGTGAGTAACACGTGGGTAACCTGCCTAAAAGAA
GGGGATAACACTTGGAAACAGGTGCTAATACCGTATATCTCTAAGGATCGCATGATCCTTAGATGAAAGAT
10‰; 20‰; 30‰; 40‰; 50‰) and pH range (1, 2, 3, 4, 5, GGTTCTGCTATCGCTTTTAGATGGACCCGCGGCGTATTAACTAGTTGGTGGGGTAACGGCCTACCAAGGTG
ATGATACGTAGCCGAACTGAGAGGTTGATCGGCCACATTGGGACTGAGACACGGCCCAAACTCCTACGGG
6, 7, 8, 9) through OD600nm index [11, 15]. AGGCAGCAGTAGGGAATCTTCCACAATGGACGCAAGTCTGATGGAGCAACGCCGCGTGAGTGAAGAAGGT
CTTCGGATCGTAAAACTCTGTTGTTAGAGAAGAACACGAGTGAGAGTAACTGTTCATTCGATGACGGTATC
TAACCAGCAAGTCACGGCTAACTACGTG
III. RESULTS AND DISCUSSION Lactobacillus reuteri LIVTA2.4.1
AGTCGTACGCACTGGCCCAACTGATTGATGGTGCTTGCACCTGATTGACGATGGATCACCAGTGAGTG
C. Isolation and selection GCGGACGGGTGAGTAACACGTAGGTAACCTGCCCCGGAGCGGGGGATAACATTTGGAAACAGATGCTAAT
ACCGCATAACAACAAAAGCCACATGGCTTTTGTTTGAAAGATGGCTTTGGCTATCACTCTGGGATGGACCT
GCGGTGCATTAGCTAGTTGGTAAGGTAACGGCTTACCAAGGCGATGATGCATAGCCGAGTTGAGAGACTG
After proliferation in the MRS Broth medium + 50 mg/l ATCGGCCACAATGGAACTGAGACACGGTCCATACTCCTACGGGAGGCAGCAGTAGGGAATCTTCCACAAT
Nystatin and selection on MRS agar + 0.5% CaCO3 GGGCGCAAGCCTGATGGAGCAACACCGCGTGAGTGAAGAAGGGTTTCGGCTCGTAAAGCTCTGTTGTTGG
AGAAGAACGTGCGTGAGAGTAACTGTTCACGCAGTGACGGTATCCAACCAGAAAGTCACGGCTAACTACG
medium, 24 isolates of rod-shaped bacteria, G+, with TGCC
CaCO3 resolution ring (Fig. 1) and Uffelman color change Lactobacillus plantarum LVIITA3.3.9
GAGAGTTTGATCCTGGCTCAGGACGAACGCTGGCGGCGTGCCTAATACATGCAAGTCGAACGAACTCT
were selected. GGTATTGATTGGTGCTTGCATCATGATTTACATTTGAGTGAGTGGCGAACTGGTGAGTAACACGTGGGAAA
CCTGCCCAGAAGCGGGGGATAACACCTGGAAACAGATGCTAATACCGCATAACAACTTGGACCGCATGGT
CCGAGTTTGAAAGATGGCTTCGGCTATCACTTCTGGATGGTCCCGCGGCGTATTAGCTAGATGGTGAGGTA
ACGGCTCACCATGGCAATGATACGTAGCCGACCTGAGAGGGTAATCGGCCACATTGGGACTGAGACACGG
CCCAAACTCCTACGGGAGGCAGCAGTAGGGAATCTTCCACAATGGACGAAAGTCTGATGGAGCAACGCCG
CGTGAGTGAAGAAGGGTTTCGGCTCGTAAAACTCTGTTGTTAAAGAAGAACATATCTGAGAGTAACTGTTC
AGGTATTGACGGTATTTAACCAGAAAGCCACGGCTAACTACGTGCCAGCA
GCC
Figure 3. Sequence of the 16s rRNA gene of selected bacterial

strains
Among the five suspected Lactobacillus strains, two
Figure 1. Colonies of suspected Lactobacillus sp. on MRS strains were L. salivarius, the other two strains were L.
medium supplemented with 0.5% CaCO3 reuteri, the remaining was L. plantarum. Based on the above
bacterial strain selection process and identification results,
The 24 isolates of suspected Lactobacillus were further the Lactobacillus strains L. salivarius LIITA1.2.2, L. reuteri
selected for their antibacterial activity, out of which, 5 LIVTA2.4.1 and L. plantarum LVIITA3.3.9 (Fig. 3) were
isolates showed. The strongest resistance to V. selected to conduct further investigations of some beneficial
parahaemolyticus (∆D: 5mm to 10mm; Fig. 2). This result biological properties.
was also consistent with previous studies on resistance to V.
parahaemolyticus of Lactobacillus strains [6, 10]. Nystatin E. Some beneficial biological properties
was isolated from Lactobacillus by Ishola and Adebayo-
Antibacterial ability over time
Tayo (2012) [12]. However, there was an improvement in
the isolation process when adding Nystatin to the The selected Lactobacillus strains, L. salivarius
preselective medium, initially inhibiting the growth of fungi. LIITA1.2.2, L. reuteri LIVTA2.4.1 and L. plantarum
Lactobacillus strains were characterized by their ability to LVIITA3.3.9, were further tested for their resistance against
produce acid, so they should be selected on MRS agar typical pathogens in shrimp ponds such as V.
medium supplemented with CaCO3 and rapidly quantified parahaemolyticus and E. coli. The antibacterial activity of
lactic acid by Uffelman reagent. the three Lactobacillus strainswere investigated by
inoculation of the either V. parahaemolyticus or E. coli at
the concentrations of 105 CFU/ml, 106 CFU/ml and 107
CFU/ml. Observations and measurement of the antibacterial
rings were performed at 24 hours, 36 hours and 48 hours
after inoculation (Table 1).
Table 1. Resistance to v. Parahaemolyticus
V.paraha Time (hours)

emolyticu
s (cfu/ml) 24 36 48
L. salivarius 105 15.5±0.1 9.5±0.1 9±0.1

LIITA1.2.2 106 12±0.1 8.7±0.5 7±0.1
(D-d, mm) 107 11.7±0.5 7.2±0.3 7.3±0.5
Figure 2. Diameter of antibacterial ring of bacteria strains
LIITA1.2.2 (9.67mm), LIITA2.1.2(5mm), LIVTA2.4.1 (8mm), L. reuteri 105 10.8±0.6 5.3±0.2 4.7±0.5
LIVTA2.4.3 (6.17mm), LVIITA3.3.9 (10mm) LIVTA2.4.1 106 7.2±0.3 4.2±0.1 4±0.1
(D-d, mm) 107 6.3±0.5 4±0.1 3±0.1
D. Identification 105 15.7±0.5 12.3±0.5 7.3±0.5
L. plantarum
Biochemical identification LVIITA3.3.9 106 12.2±0.3 10.3±0.5 7.7±0.5
Characteristics of the obtained isolates identified through (D-d, mm)
107 11.8±0.6 8.7±0.1 7.3±0.5
biochemical tests were catalase, mobility, and capable of
337
At 24 hours after inoculation, at the density 10 5 -107

CFU/ml of V. parahaemolyticus, the ring diameters of L.
salivarius LIITA1.2.2 ranged from 11.7±0.5mm -
15.5±0.1mm and those of L. plantarum LVIITA3.3.9 were
from 11.8±0.6mm - 15.7±0.5mm. In contrast, the the
antibacterial ring diameter of L. reuteri LIVTA2.4.1 was
only from 1.3±0.5mm - 6±0.7mm. After 36 hours and 48
hours, the diameters of antibacterial zone of the 3
Lactobacillus strains challenged with V. parahaemolyticus
at the concentration of 105 - 106 CFU/ml showed
insignificant change. However, during the period from 24
hours to 36 hours when the concentration of pathogenic
bacteria reached 107 CFU/ml, the antibacterial expression of
the three strains changed signigicantly. After that period, the
antibacterial zone gradually narrowed, and after 48 hours it
only ranged from 3±0.1mm to 7.3±0.5mm (Table 2). Figure 4. OD600nm value of bacterial strains at different sites with
different salinity
Table 2. Resistance to e. Coli
In summary, the salt tolerance threshold of the
E. coli Time (hours) investigated strains is consistent with the physiological
(cfu/ml) 24 36 48 characteristics of Litopenaeus vannamei, and meets the
requirements of bacterial source for probiotic production for
L. salivarius 105 10±1.2 7.3±1.2 7.7±0.5 Litopenaeus vannamei shrimp.
LIITA1.2.2 106 10.7±0.5 7.3±1.2 7.0±0.1
(D-d, mm) pH adaptability
107 6.7±0.5 6.3±0.5 5.3±0.5
L. reuteri 105 6.0±0.7 4.3±1.2 3.7±0.5 The OD600nm measurement results of L. salivarius
LIVTA2.4.1 106 6.0±0.7 4.0±0.7 3.0±0.7 LIITA1.2.2, L. reuteri LIVTA2.4.1 and L. plantarum
(D-d, mm) LVIITA3.3.9 after 24 hours were adjusted to pH 0, 1, 2, 3,
107 5.3±0.5 1.67±0.5 1.3±0.5
4, 5, 6, 7, 8, 9 were respectively shown in fig. 5.
L. plantarum 105 13±0.7 10±0.7 8.0±0.1
LVIITA3.3.9 106 11.7±0.5 9.3±0.5 8.3±0.5
(D-d, mm) 107 7.3±0.5 6.0±0.7 6.0±0.7
After 24 h, the ring diameters of E. coli at 105 CFU/ml -

7
10 CFU/ml of L. salivarius LIITA1.2.2, L. plantarum
LVIITA3.3.9 and L. reuteri LIVTA2.4.1 ranged from
5.3±1.2 mm - 13±0.7 mm. After 36 hours and 48 hours of
culturing, there was a change in the antibacterial ring
diameter of 3 strains of Lactobacillus with E. coli at the
concentration of 105 CFU/ml - 106 CFU/ml. However, in the
period from 24 hours to 36 hours at concentration of
pathogenic bacteria 107 CFU/ml, Table 2 showed that the
antibacterial ability of the three strains gradually descended,
colonies spread within the antibacterial ring and after 48 Figure 5. Obtical density (OD600nm) value of Lactobacillus
hours, it only fluctuated from 1.3±0.5 mm - 5.3±0.5 mm strains at different pH intervals
(Table 2). These antibacterial test results were also
consistent with previously published studies [6, 17]. The result from Fig. 5 showed that all of the isolates had
the ability to grow with the surveyed pH ranges. However,
Salt tolerance
the optimal pH range for growth of each strain was varied
The results of measuring OD600nm after 24 hours of which was 4 – 8 for L. reuteri LIVTA2.4.1 , 5 - 8 for L.
incubation at 37oC of L. salivarius LIITA1.2.2, L. reuteri plantarum LVIITA3.3.9 and 4 – 8 for L. salivarius
LIVTA2.4.1, L. plantarum LVIITA3.3.9 were shown in Fig. LIITA1.2.2 . Of the three isolated Lactobacillus strains, L.
4. It showed that strain L. salivarius LIITA1.2.2, L. reuteri plantarum LVIITA3.3.9 showed superior growth over the
LIVTA2.4.1 and L. plantarum LVIITA3.3.9 had the ability investigated pH ranges. This result is consistent with
to grow in the salinity conditions from 0‰ to 50‰ and previous publications [5, 11]. With suitable pH conditions
optimally at 10‰. As the salinity increased, the OD600nm for Litopenaeus vannamei in Vietnam, which ranges from
value gradually decreased, to the lowest OD600nm as at 50‰ 7.7 to 8.3 [19], the isolates can live in the pH conditions of
salinity. This was consistent with previous studies [5, 11, the white shrimp ponds in Vietnam and can be used to
18]. Moreover, the physiological characteristics of produce probiotic for Litopenaeus vannamei shrimp.
Litopenaeus vannamei can be adapted to pond conditions
with salinity ranging from 5‰ to 50‰ and optimally in IV. CONCLUSION
pond environment with salinity from 10‰ to 25‰ [19]. The research has improved the process of isolating
Lactobacillus strains by increasing the sample on
338
preselective medium (MRS Broth + 50mg/l Nystatin), [8] J.L. Balcazar, Roas-Luna T, “Inhibitory Activity of Probiotic
selecting on MRS agar supplemented with 0.5% CaCO3 and Bacillus subtilis UTM 126 against Vibrio species confers
determing lactic acid production by Uffelman reagent. From protection against vibriosis in Juvenile Shrimp (Litopenaeus
24 pond water samples and 24 gastrointestinal tract samples vannamei)”, Current Microbiology, vol. 55, pp. 409 - 412,
of Litopenaeus vannamei in Nhon Trach district, Dong Nai 2007.
province, 3 strains of L. salivarius LIITA1.2.2, L. reuteri [9] Trần Linh Thước (2007), Phương pháp phân tích vi sinh vật
LIVTA2.4.1 and L. plantarum LVIITA3.3.9 were isolated trong nước, thực phẩm và mĩ phẩm, Nhà xuất bản Giáo dục,
Hà Nội.
and identified. They resisted V. parahaemolyticus and E.coli
[10] K. Kongnum, T. Hongpattarakere, “Effect of Lactobacillus
at pathogenic densities from 105 CFU/ml to 107 CFU/ml.
plantarum isolated from digestive tract of wild shrimp pond
Isolated Lactobacillus strains have the ability to live in growth and survival of white shrimp (Litopenaeus vannamei)
conditions with pH ranging from 1 to 9 and salinity from challenged with Vibrio harveyi”, Fish & Shellfish
0‰ to 50‰, which is a good fit for use in shrimp ponds in Immunology, vol. 32, pp. 170 - 177, 2012.
Vietnam, and meets the requirements of the bacterial strain [11] P.M. Nguyen, “Isolation, identification and characterization
for probiotic production. of Lactobacillus on black tiger shrimp”, International Journal
of Multidisciplinary Research and Development, vol. 1(6),
REFERENCES pp. 153 - 158, 2014.
[1] M. Reyes-Becerril, F. Ascencio, V. Gracia-Lopez, M. E. [12] R.O. Ishola, B.C. AdebayoTayo, “Screening lactic acid
Macias, M. C. Roa, M. A. Esteban, “Single or combined bacteria isolated from Fermented food for Bio-molecules
effects of Lactobacillus sakei and inulin on growth, production”, Aust. J. Tech. 15, vol. 4, pp. 205 - 217, 2012.
nonspecific immunity and IgM expression in leopard grouper [13] Nguyễn Lân Dũng, Đoàn Xuân Mượn, Nguyễn Phùng Tiến,
(Mycteroperca rosacea)”, Fish Physiol. Biochem, vol. 40, Đặng Đức Trạch, Phạm Văn Ty (1976), Một số phương pháp
pp.1169 - 1180, 2014. nghiên cứu vi sinh vật học, tập 2, NXB Khoa học Kỹ thuật,
[2] S. M. Swain, C. Singh, V. Arul, “Inhibitory activity of Hà Nội.
probiotics Streptococcus phocae PI80 and Enterococcus [14] U. Schillinger, F. K. Lücke, “Antibacterial activity of
faecium MC13 against Vibriosis in shrimp Penaeus Lactobacillus sakei isolated from meat”, Applied and
monodon”, J. Microbiol. Biotechnol, vol. 25, pp. 697-703, Environmental Microbiology, vol. 55, pp. 1901 - 1906, 1989.
2009. [15] Nguyễn Văn Phúc, Phan Thị Phượng Trang (2014), “Phân
[3] B.A. Ige, “Probiotics use in intensive fish farming”, J. lập, định danh và xác định các đặc tính có lợi của chủng
Microbiol. Res., vol. 7, pp. 2701-2711, 2013 Bacillus spp. Từ ao nuôi tôm ở tỉnh Bến Tre”, Tạp chí Khoa
[4] M. Maeda, A. Shibata, G. Biswas, H. Korenaga, T. Kono, T. học Đại học sư phạm Thành phố Hồ Chí Minh, số 64, Trang
Itami, M. Sakai, “Isolation of lactic acid bacteria from 94 – 102.
kuruma shrimp (Marsupenaeus japonicus) intestine [16] D, V. Paul, M. G. George, J. Dorothy, R. K. Noel, L.
andassessment of immunomodulatory role of a selected Wolfgang, A. F. Rainey, S. Karl-Heinz and B. W. William,
strain as probiotic”, Mar. Biotechnol, vol. 16, pp. 181-192, Bergey’s Manual Of Systematic Bacteriology, Second
2014. Edition, vol. 3, The Firmicutes, 2011.
[5] Võ Thị Thứ (2006) “Dự án hoàn thiện và triển khai công [17] L. Jayasree, P. Janakiram, R. Madhavi, “Characterization of
nghệ sản xuất chế phẩm sinh học phục vụ xử lý môi trường Vibrio str. associated with diseased shrimp from culture
nuôi trồng thủy sản”. Báo cáo tổng kết khoa học và kỹ thuật, ponds of andhra pradesh (India)”, Journal of the World
Hà Nội. Aquaculture Society 37, vol. 4, p. 523, 2006.
[6] Khuất Hữu Thanh, Nguyễn Đăng, Phúc Hải, Bùi Văn Đạt, [18] C.H. Liu, C.S. Chiu, P.L. Ho, S.W. Wang, “Improvement in
Võ Văn Nha (2009), “Phân lập và tuyển chọn một số chủng the growth performance of white shrimp, Litopenaeus
vi khuẩn có đặc tính probiotic trong tạo chế phẩm nuôi tôm vannamei, by a protease - producing probiotic, Bacillus
sú”, Tạp chí khoa học & công nghệ các trường đại học kỹ subtilis E20, from natto”, Journal of Applied Microbiology,
thuật, số 47, tr. 113-116. số 107, tr. 1031 - 1041, 2009.
[7] Chiu-Hsia Chiu, Yuan-Kuang Guu, Chun-Hung Liu, Tzu- [19] Thái Bá Hồ và Ngô Trọng Lư (2003), Kỹ thuật nuôi tôm thẻ
Ming Pan, “Immune responses and gene expression in white chân trắng. Nhà xuất bản Nông nghiệp, Hà Nội.
shrimp, Litopenaeus vannamei, induced by Lactobacillus
plantarum”, Fish & Shellfish Immunology, vol. 23, pp. 364 -
377, 2007.
339
Production of Cheese from Reconstituted Milk and

Soy Milk with the Addition of Microbial
Transglutaminase
Pham Thi Hoan Trinh Khanh Son
Food Technology department Food Technology department
Education Education
hoanpt@hcmute.edu.vn sontk@hcmute.edu.vn
Abstract: This study was carried out to evaluate the cow's milk). The results showed that the mixture cheese had
possibility of producing fresh cheese by combining an average fat content, but the protein content was enhanced
reconstituted milk with soy milk using microbial compared to other samples. The organoleptic qualities of
transglutaminase (MTGase) as a structural stabilizer. cheese samples were similar during 15 days of storage.
Analytical results showed that the fresh cheese sample made Moreover, mixture cheese was cheaper - an important issue
from 80% (v/v) reconstituted milk and 20% (v/v) soy milk in cheese production, especially, in developing countries.
had a recovery efficiency of 25.02%. Fresh cheese that was
The casein can be coagulated by either an acid or an
made with the above ingredients supplemented with 2 IU
enzyme with different mechanisms. In our study, citric acid
MTGase/g protein at 37oC for 2 hours had the highest dry
was used as a coagulant – a common and inexpensive acid
matter content and cheese solid yield of 39.43 and 45.81%,
that gives good casein coagulation efficiency. Dmytrów et
respectively. Besides, this cheese had the lowest whey
al. [3] used organic acids such as citric acid, malic acid, and
separation (0.18%) compared to other samples. There was a
tartaric acid as coagulation agents in the production of acid
difference in colour (ΔE > 2) between investigated fresh
curd cheese (tvarog). Interestingly, citric and tartaric acids
cheese and traditional cheese; however, the difference did
had higher overall acceptance scores than that of malic
not negatively affect the sensory evaluation score. Fresh
acids. The authors suggested using these acids at a
cheese samples had a protein content of 16.40% and a lipid
concentration of 2% to obtain a product with better
content of 15.65%. In addition, Listeria monocytogenes, and
performance and desired taste.
Staphylococci were not found in this product. The product
had a more stable and uniform texture than the reference In recent years, the application of the enzyme microbial
one. The products were also evaluated for the changes in transglutaminase (MTGase) in industries, in particular in the
acidity, whey separation, and colour during storage (28 dairy industry, is growing fast. Transglutaminase (EC
days). The results showed that the research product meets 2.3.2.13, protein-glutamine γ-glutamyl transferase) catalyzes
the standards (CXS 221-2001) for fresh cheese. the transfer reactions of acyl group (C-CO-R) between the
γ-carboxyamide (C-CO-RNH) groups of peptides. This
Keywords: fresh cheese, transglutaminase, soy milk, reaction results in the creation of cross-linking in protein
reconstituted milk, cross-linking components such as whey proteins, soy proteins, wheat
proteins, beef myosin, and casein …[4]. Due to the cross-
I. INTRODUCTION
linking formation, most products are improved in the
Cheese is a protein-rich food that provides calcium- recovery efficiency [5], [6]. In order to improve the yield
binding structures and causes almost no digestive problems. production of fresh cheese, increase the texture quality, and
However, the current cheese production in Vietnam still improve the water retention capacity, MTGase was chosen
faces many problems such as low efficiency and high cost. to conduct the experiments in this study.
Furthermore, according to nutritionists and FAO, people
should consume both animal-based protein and plant-based Unfortunately, the amount of fresh milk in our country
protein simultaneously. Actually, there are a few studies that currently only meets about 35-40% of the needs of
have approached this issue. Nazim et al., 2013 [1] conducted businesses in the market. The main source of raw materials
a study to compare and evaluate the nutritional parameters is imported milk powder [7]. Therefore, we propose
of two types of fresh cheese made entirely from soy milk research on fresh cheese production using reconstituted milk
(soy cheese) and a mixture of soy and cow's milk. The and soy milk as raw materials in order to evaluate the
results showed an increase in nutritional composition, combination ability of two protein sources, and at the same
especially the vitamin B group in cheese products with soy time can diversify products, improve the yield production
milk. According to a research of Elmahi AKH et al. [2], and reduce the costs of cheese products.
authors compared 3 types of cheese including tofu cheese This study was conducted to (a) investigate the mixing
(100% soy milk), Sudanese white cheese (100% cow's ratio between reconstituted milk and soy milk on the quality
milk), and mixture cheese (50% soybeans milk and 50%

of fresh cheese, (b) determine the cheese processing Temperature: RS2 milk mixture was treated with
parameters during treatment with MTGase (concentration, enzyme
temperature, and time), (c) to evaluate the quality of fresh (2 IU/g protein) at different temperatures (from 37 to 85°C)
cheese through some of the quality criteria, (d) assess the for 2 hrs.
quality properties of fresh cheese samples during storage.
Time: RS2 milk mixture was treated with enzyme (2
IU/g protein) at 37 °C for different reaction times (from 1.0
to 3.0 hrs).
In all of the above experiments, the reference sample
A. Materials (Ref.) was a fresh cheese product made from RS2 milk
Whole milk powder made in New Zealand was mixture without MTGase treatment. In each experiment, the
purchased at Dai Tan Viet Joint Stock Company (New Viet yield production (%H), solid cheese yield (%CSY), total
Dairy), Ho Chi Minh City, Vietnam. Soybeans were solid in cheese (%TScheese), whey separation, and texture
purchased at Xuan Hong Trading and Service Co., Ltd., Ho characteristics of fresh cheese samples were measured.
Chi Minh City, Vietnam. MTGase made in Malaysia was
purchased at the US-Australian science and technology 3) Changes in quality parameters of the fresh soy cheese
development joint-stock company, Ho Chi Minh City, during storage. Fresh cheese samples were produced and
Vietnam. stored in sealed containers at 4±2°C for 28 days. The
organoleptic properties, acidity, and whey separation of the
B. Fresh cheese production products were monitored and measured.
The manufacturing process is shown in Fig. 1. Soybeans C. Analytical methods
(500 g) were washed and soaked with water at room
temperature (30oC) in a ratio of 1:2 (w:w) for 12 hours, then The yield of fresh cheese production. The yield (H, %)
drained. Then, soybeans were grilled with water at a ratio of was determined by a formula [3]:
m (1)
1:1 (w:w) by a Comet CM9989 blender to obtain soy milk H = 1 100
with a dry matter content of 15%. The suspension after m0
grinding was filtered through a filter bag (amalife, 24×36 Where, m1 was the weight of fresh cheese (g); m0 was the
cm). To collect the soy milk, the filtrate was recovered, then weight of reconstituted milk solution (g).
boiled (100°C) with constant stirring for 5 minutes to Cheese solid yield (CSY) is understood as the ratio of the
destroy microorganisms, inactivate enzymes, and anti- recovered solids in the cheese product to the solids of milk
nutritional factors. At the same time, the powdered milk was mixture [8]. CSY was calculated according to the following
reconstituted at 50°C for 30 min. It was then mixed with the formula:
soy milk in various proportions (total volume was 300 mL).
Then, MTGase was added and incubated. The mixtures were H  TScheese (2)
CSY =
then pasteurized (85°C, 20 min) to remove microorganisms TSmilk
and inactivate MTGase. Citric acid (1.5%, w/v) was added
for protein coagulation (85°C, pH 4.6). The curds were Where, H (%) was the yield production, TScheese (%) and
separated from whey and put in a plastic mold (Ø = 5.6 cm). TSmilk (%) was the total solid contents in cheese and in milk
After 15 min of draining, cheese samples were slightly mixture, respectively.
pressed using a weight of 1 kg to achieve a final height of Whey separation (Wh, %) was determined according to
2.0 cm. Finally, products were stored in a sealed container at a method of Dmytrów et al., 2010 [3]. Cheese samples
4±2°C for further analytical measurement. (25 g) were weighed and placed into zip-lock packages.
1) Determination of mixing ratio of soy milk and After 20 hours, the whey leached out of samples at 25°C
reconstituted milk. The reconstituted milk (total solid of was weighted. Wh was calculated by the formula:
15.43%) was replaced and mixed with the soy milk (total M1
Wh = 100 (3)
solid of 14.97%) various ratios (v:v): 10:0 (R); 9:1 (RS1); M0
8:2 (RS2); 7:3 (RS3); 6:4 (RS4). In the other words, the
replacement (%, v/v) of soy milk in the final mixed milk Where, M1 (g) was the weight of separated whey from the
was 0, 10, 20, 30, and 40%, respectively. These mixtures sample; M2 (g) was the initial weight of the sample.
were treated with MTGase (37°C). These samples were then Determination of cheese texture characteristics. Cheese
subjected to the procedure described in Fig. 1 for fresh texture properties were measured by a CT3 Texture
cheese production. Analyzer (Ametek Brookfield, America). Parameters for
2) Determination of MTGase treatment conditions: measurement were: (a) a cylinder force (TA-AACC36) with
diameter of 3.6 cm; (b) test speed of 3.0 mm/s; (c) pretest
Enzyme concentration: In order to investigate the effects speed of 2.0 mm/s; recovery time of 5.0 s; trigger load of
of enzyme concentrations on the fresh cheese properties, 5.0 g and target distance of 3.0 mm with fixture base [9],
MTGase was added to the RS2 milk mixture at different [10].
concentrations (from 1 to 5 IU/g protein). The enzymatic
reaction was conducted at 37°C for 2 hrs. Colour space measurements
The colour space for the fresh cheese samples was
determined by a CR-400 chroma meter (Minolta, Japan)
341
according to Mokrzycki and Tatol, 2011 [11]. The colour notice the difference in colour); 1<ΔE<2 (only experienced
parameters of fresh cheese samples were compared with observers were able to notice the difference in colour);
those of the non-soy milk cheese (R sample). The L*, a*, 2<ΔE<3.5 (inexperienced observers might notice colour
and b* values represent white/black, red/green and differences); and ΔE >3.5 (there was a clear colour
yellow/blue, respectively. The difference in colour (△E) difference between the two samples).
was calculated using the formula (4):
E = a2 + b2 + L2 (4)

Based on the ΔE value, the difference in colour between
the samples was expressed as: 0<ΔE<1 (the observer did not
Whole milk
Soy beans powder
12hrs, 30 oC;
Washing & Soaking Reconstitution 50 oC, 30 min.
Soy beans:water =1:2 (w:w)
*nguyênngnguyên
MTGase
Draining Mixing
2 min.,
Soy beans:water =1:1 Enzyme treatment
Grinding
(w:w)
Filtering Pasteurization 85°C, 20 min.

residue
Boiling 100°C, 5 min. Coagulation Acid citric 1.5%
Whey separation
Whey
Soy milk
Curd cutting
Fresh cheese
Preservation 4±2 oC
Figure 1. A flow chart of fresh cheese production from reconstituted milk and soy milk
Microstructure observation. The experiment was carried evaluated by a panel of 5 assessors trained through a pre-
out to evaluate the effect of MTGase on the microstructure established sensory profile of fresh cheese products.
of fresh cheese by a scanning electron micrograph (SEM) Evaluated attributes were appearance, texture, odour and
[12]. Firstly, fresh cheese samples were frozen at -50ºC for taste. The appearance was judged by the colour and
24h, then they were freeze-dried by equipment Yamato smoothness of the cheeses. The texture was evaluated
DC-401, Japan (30±2 oC, 10-20 Pa, 24h). The cheese through firmness (defined as the amount of resistance to
samples were mounted on a stub and were taken compression offered by a 1.0 cm thick slice of cheese when
photographs with TM4000PLUS (Japan). Observation pushed between the thumb and the index finger until fingers
conditions were selected: Accelerating Voltage (Standard 15 touch each other) [13]. The term "odour" was defined as the
kV Mode 2), Vacuum level (Standard (H)), Detector (BSE). "impressions perceived from the volatile chemical
compounds enter the nasal passage". And "taste" was
Sensory evaluation. The sensory test was taken
determined as "impressions perceived via the chemical
according to the ISO 22935-3:2009 with a scale of 0 to 5
senses in the mouth" [13]. Sensory evaluation sessions were
points. The sensory properties of cheese samples were
342
conducted in the individual booth under fluorescent light. Table 1. Chemical composition of materials
The samples were coded with three random digit numbers.
The testing room was cleaned without a strange odour. Chemical Whole milk Reconstituted Soybean
composition, % powder milk milk
Chemical compounds. Protein, moisture, ash, and
reducing sugar contents were estimated according to
Kjeldahl method (ISO 8968 1:2014), ISO 5537:2004, Protein 25.09±0.51c 3.87±0.00a 7.22±0.00b
AOAC 923.03, and DNS (3,5-Dinitrosalicylic acid) method, Lipid 26.27±0.03 b
4.22±0.01 a
3.89±0.00a
c b
respectively. Lipid content in whole milk powder and fresh Carbohydrate 40.36±0.15 7.45±0.00 2.98±0.00a
cheese was determined according to Mojonnier method Ash 5.82±0.00b 0.99±0.01a 0.88±0.00a
(ISO 1736:2008/ IDF 9:2008) and ISO 1735:2004, Moisture 3.10±0.01a 84.73±0.72b 84.57±0.56b
respectively. a-c
In each row, values with different letters are significantly different (p<0,05).
Titratable acidity (TA, mM NaOH per 100 g of product)

of cheese samples was determined in accordance with the
ISO 11869:2012. Acidity is the amount, in milliliters, of 0.1
mol/l sodium hydroxide solution required to titrate 10 grams
of product to pH 8.3±0.1 and is calculated by the formula:
V  10
TA = (5)
m
Where, V (mL) was the volume of NaOH solution used for
titration; m (g) was the weight of the titrated test portion.
Statistical analysis. Each experiment was done in
triplicate. Data were expressed as means ± standard
deviation. IBM SPSS Statistics Version 26.0 was used to
analyze the variance, and Duncan’s test was performed to
determine the statistical significance between samples at a Figure 2. Changes of yield production (H, %), total solid content
level of 0.05. (TScheese, %), cheese solid yield (CSY, %), and whey separation
(Wh, %) in fresh cheese samples with different proportions of
III. RESULTS AND DISCUSSION reconstituted milk and soy milk 10:0(R), 9:1(RS1), 8:2 (RS2), 7:3
(RS3), 6:4 (RS4). The different letters (a-d) showed significant
D. Chemical composition of materials differences (p < 0.05)
The chemical composition of whole milk powder,
reconstituted milk, and soybean milk were presented in The analytical results (Fig. 2) showed that the yield of
Table 1. Samples of reconstituted milk and soy milk were fresh cheese production (%H) gradually increased from
undergone through experiments and prepared with dry 22.31% (R sample) to 25.02% (RS2 sample) with the
matter content of approximately 15%. Analytical results increase in replacement rate of soy milk from 0 up to 20%.
(Table 1) showed that these two materials had similar levels When this replacement got a raise from 30% (RS3 sample)
of ash, fat, and moisture contents. The carbohydrate content to 40% (RS4 sample), the recovery of fresh cheese tended to
of reconstituted milk was higher than that of soy milk, while decrease. Besides, the quality properties such as total solid
the protein content was in the opposite trend. Therefore, in in cheese (%TScheese) and cheese solid yield (%CSY) also
the hope that the combination of whole milk powder and soy increased from sample R to sample RS2 but decreased
milk will result in (a) the creation of a product that is more slightly from sample RS3 to sample RS4. Besides, whey
enriched in the protein content; (b) the yield of fresh cheese separation did not exceed 0.35% and reached the lowest
production could be increased, (c) reduction of the final level in RS2 samples (0.27%). Thus, at the reconstituted
product costs because of the lower cost of soybeans milk:soy milk ratio = 8:2 it showed a higher product
material, and (d) creation of diversity for fresh cheese recovery and a lower whey separation. Our results are also
products on the market. similar to those of Metwalli et al. [14] or Aworh et al. [15],
E. Determination of mixing ratio of soy milk and in which the ratio of soy milk and whole milk was 2:8 also
reconstituted milk gave a better cheese recovery efficiency (18.17 and 17.4%).
However, these values were lower than that of the cheese
Changes in yield production (H, %), total solid content sample made from whole milk only (18.36 and 18.6%). The
(TScheese, %), cheese solid yield (CSY, %) and whey cheese recovery efficiency in our study (25.02%) was higher
separation (Wh, %) in fresh cheese samples with different than the results of that previous studies due to the different
proportions (v:v) of reconstituted milk and soy milk dry matter content of the raw materials and the different
10:0(R), 9:1(RS1), 8:2 (RS2), 7:3 (RS3), 6:4 (RS4) were coagulation mechanism.
presented in Fig. 2.
The effect of the mixing ratio between reconstituted milk
and soy milk on cheese structure also was analyzed. The
results were shown in Table 2.
343
Table 2 showed that the hardness of cheese samples separation (Wh, %) in fresh cheese samples with different
containing soy milk (RS1-RS4) was harder than that of enzyme concentrations were shown in fig. 3.
cheese without this ingredient (R). The hardness increased
The results showed that the MTGase supplemented
when replacing reconstituted milk with soy milk from
treated samples showed a significant increase in yield
0-20%. However, at the replacement higher than 20%, the
compared to the cheese samples without MTGase addition.
hardness of products was not much different. Meanwhile,
The highest yield (24.98%) was achieved when the enzyme
springiness in cheese samples was quite similar. The
concentration was 2 IU/g protein. The total solid content
difference in hardness of cheese samples could be attributed
(TScheese) and cheese solid yield (CSY) values also gradually
to the higher moisture retained by the soy proteins, which
increased with the enzyme concentration from
plays an important role in the cheese's structure because of
0 - 2 IU/g protein, then tended to decrease. The whey
its continuous solid phase presentation of the product [16].
separation (Wh) of the cheese samples gradually decreased
Theoretically, any change in the nature or quantity of the
from 0.62 to 0.05% as the MTGase concentration increased
protein will also change the structure of cheeses. The higher
from 0 to 5 IU/g protein. These results demonstrated the
the protein content is, the higher the force required to
effectiveness of MTGase in the production of this soy milk
deform a sample of cheese is [8]. Cheese samples with a
substitute cheese.
partial soy milk substitute had a higher protein content.
Colour space measurements of cheese samples were
shown in Table 3. Fresh cheese samples were different in
colour (which was recognizable, ΔE > 2). Fortunately, these
differences did not negatively affect the sensory properties
of the product. Therefore, the partial replacement (20%) of
soy milk not only contributed to the protein diversity in the
final cheese product but also increased the yield production
and dry matter content as well as the structure of the fresh
cheese product. The mixing ratio 8:2 (RS2) was chosen for
further experiments.
Table 2. Effect of mixing ratio (% soy milk replacement of

reconstituted milk) on cheese texture
Sample
Hardness (g) Springiness (mm)
(% soy milk) Figure 3. Changes in yield production (H, %), total solid content
R (0) 123.00±11.76a 2.77±0.01a (TScheese, %), cheese solid yield (CSY, %), and whey separation
RS1 (10) 129.27±3.66a 2.82±0.03a (Wh, %) in fresh cheese samples with different enzyme
concentrations. The different letters (a-d) showed significant
RS2 (20) 202.67±1.16bc 2.79±0.05a
b
differences (p < 0.05)
RS3 (30) 194.10±3.66 2.74±0.16a
c
RS4 (40) 210.10±4.95 2.78±0.05a Table 4. Effect of mixing ratio (% soy milk replacement of
a-c
In each column, values with different letters are significantly different (p<0,05). reconstituted milk) on cheese texture
Enzyme
concentration (IU/ Hardness (g) Springiness (mm)
Table 3. Effect of mixing ratio (% soy milk replacement of g protein)
reconstituted milk) on the colour of cheese
0 (Ref.) 149.01±2.98a 2.67±0.02a
1 160.53±5.80b 2.81±0.02c
Sample
(% soy L* a* b* ΔE 2 234.77±4.15f 2.86±0.02d
milk, v/v) 3 222.17±5.63e 2.83±0.04cd
4 208.43±1.21d 2.82±0.02cd
R (0) 92.04±0.47e 0.93±0.02a 15.77±0.15b 0.00
5 193.53±5.75c 2.75±0.03b
d b
RS1 (10) 88.06±0.34 0.88±0.03 15.69±0.39b 3.98 a-f
RS2 (20) 86.83±0.11c 0.93±0.16c 15.08±0.13a 5.25
RS3 (30) 84.70±0.33a 0.93±0.10d 15.24±0.18a 7.36 The increase in yield production of fresh cheese samples
b c
RS4 (40) 85.60±0.12 0.92±0.02 15.04±0.06a 6.48 using MTGase could be explained by the action of this
a-d
In each column, values with different letters are significantly different (p<0,05). enzyme. These cross-links make the protein network tighter,
and then the dry matter was retained leading to an increase
F. Determination of MTGase treatment conditions in cheese recovery. Based on the tighter and denser network
a) Effect of MTGase concentration on fresh cheese of protein, water was not able to pass through it, so the
quality: Milk prepared from reconstituted milk and soy milk cheese got more water-retaining and avoided whey
with a ratio of 8:2 was treated with MTGase in different separation. Furthermore, the amino acid composition of soy
concentrations: 0.0; 1.0; 2.0; 3.0; 4.0; 5.0 IU/g protein. milk and reconstituted milk was rich in glutamine and lysine
Changes in yield production (H, %), total solid content [14]. This means that soy milk was rich in substrates for
(TScheese, %), cheese solid yield (CSY, %), and whey MTGase to carry out the crosslinking reactions. The results
of our study were similar to published papers [17], [18].
344
Actually, MTGase increases product recovery by 3-10%. On Fig. 4 showed that temperature had a significant
the contrary, the yield production tended to decrease slightly influence on the quality of fresh cheese. The cheese
as the concentration of the treated enzyme continued to recovery yield was similar at temperatures ranging from 37
increase. This could be explained by enzyme competition to 50°C. However, at 37oC, the total solid in cheese and
and/or breaking of the cross-links created during treatment cheese solid yield of the enzyme-treated sample were
[19]. highest. At 85°C, the enzyme was inactivated, so the cheese
production yield on treated and non-treated (Ref.) samples
The structure of cheese samples at different treatment
were insignificantly different. Whey separation (%) of the
concentrations was shown in Table 4. The results showed
Ref. sample (0.78%) was the highest. According to Chen et
that MTGase treated samples were harder and more elastic
al. [21] MTGase was stable over a wide range of pH values
than non-enzyme treated samples. The hardness increased
(from 5 to 8) and temperature (from 10 to 70°C).
(from 149.52 to 234.77 g) gradually due to the increase of
Furthermore, the optimal reaction temperature of MTGase is
the enzyme concentration (0- 2 IU/g protein). At the enzyme
37°C. According to the research results of Brink &
dosage higher than 2 IU/g protein, the hardness decreased
Leskauskaitė [22] the highest yield of cottage cheese was
slightly (from 234.77 g to 193.53g). Besides, springiness
obtained at 40°C for 60 minutes. At a higher temperature
also increased gradually due to the increase in enzyme
(45 to 50°C), the yield production was reduced and the
dosage (0-2 IU/g protein). Higher enzyme dosage did not
product structure was brittle and weaker due to excessive
result in the continuous enhancement of the springiness.
protein polymerization. These changes resulted in a
These results could be explained by MTGase catalyzing the
structure with a higher stress value and prone to fracture.
reaction of glutamine and lysine to induce protein
Our findings on texture properties of cheese samples (Table
crosslinking reactions in the milk gel system. However, if
5) also showed the similarity. The hardness of products
the enzyme dosage is too high, the gel strength gradually
decreased with increasing enzymatic reaction temperature.
decreased. This change could explain by the possible protein
The lower the hardness is, the higher the adhesiveness of the
substrate depletion and enzyme competition causing gel
cheese is [22], and the lower the springiness is.
network breakage [19], [20].
c) Effect of treatment time with MTGase on the quality
b) Effect of treatment temperature on quality of fresh
of fresh cheese: Changes of yield production (H, %), total
cheese: Changes of yield production (H, %), total solid
solid content (TScheese, %), cheese solid yield (CSY, %) and
content (TScheese, %), cheese solid yield (CSY, %) and whey
whey separation (Wh, %) in fresh cheese samples with
separation (Wh, %) in fresh cheese samples with different
different treatment time were shown on Fig. 5.
treatment temperatures were shown in Fig. 4.
60
50 f
e d
40 b c
%(w/w)
30 f e d
c
b
20 a
10 a
0
0 0.5 1 1.5 2 2.5 3
Time (hrs)
%H %TScheese %CSY %Wh
Figure 5. Changes in yield production (H, %), total solid content

Figure 4. Changes of yield production (H, %), total solid content
(TScheese, %), cheese solid yield (CSY, %), and whey separation
(TScheese, %), cheese solid yield (CSY, %), and whey separation
(Wh, %) in fresh cheese samples with different treatment times. The
(Wh, %) in fresh cheese samples with different treatment different letters (a-d) showed significant differences (p < 0.05)
temperatures. The different letters (a-d) showed significant
differences (p < 0.05)
Table 6. Effect of enzymic treatment time on cheese texture
properties
Table 5. Effect of enzymatic reaction temperature on cheese
texture Treatment time
Hardness (g) Springiness (mm)
(hours)
Temperature (°C) Hardness (g) Springiness (mm) 0 (Ref.) 138.20 33.16a 2.77 0.14a
Ref. 127.37±35.51a 2.55±0.08a 1 179.30 15.62ab 2.79 0.02a
37 243.93±12.95d 2.81±0.02b 1.5 210.60 32.84bc 2.77 0.03a
45 231.53±36.65c 2.77±0.04b 2 251.00 21.70c 2.75 0.03a
50 213.93±28.28b 2.72±0.03b 2.5 240.57 19.49c 2.78 0.09a
85 105.70±10.41a 2.58±0.05a 3 211.70 2.16bc 2.77 0.05a
a-d
In each row, values with different letters are significantly different (p<0,05). a-f
345
Fig. 5 showed that the yield of cheese production Fresh cheese treated with MTGase (RS sample) was
increased from 14.74 to 24.91% based on the increase in manufactured and assessed. RS fresh cheese had protein
treatment time (from 0 to 2 hrs). At higher treatment time, content, lipid content, and total dry matter content of 16.40,
the yield significantly decreased. Cheese, obtained at 2 hrs 15.65, and 39.43 %, respectively. Listeria monocytogenes
of treatment, had the higher values of cheese solid yield and coagulase-positive Staphylococci were not detected in
(CSY of 45.81%) and total solid in cheese (TScheese of the final product. It indicated the final product meets the
39.43%) than that of the other samples. The whey separation standards (CXS 221-2001) of fresh cheese.
(Wh) of the non-enzyme treated sample was the highest
The sensory evaluation (Fig. 6) of the RS cheese and the
(0.72%). Thus, there exists a reaction time so that the
control cheese (C samples were made from reconstituted
balance between enzyme and substrate is optimal.
milk without MTGase treatment) showed that these products
According to the research of Darnay et al. [23] the gel
had no difference in appearance and taste. RS cheese was
properties have an increase in viscosity value after 100 to
yellowish and was not creamy white compared to C
140 minutes of incubation. Moreover, the high MTGase
samples. All the products had a slightly sour taste of citric
concentration has a high affinity to form isopeptide bonds
acid (a coagulant). The hardness of the RS sample was
with β-casein in the outer layer of micellar casein resulting
higher than that of the C sample because the RS cheese
in the enhancement of structural stability of micelles. The
sample had a more stable structure caused by MTGase
authors stated the increase in yield production was due to an
treatment [1]. The smell of these two products was also easy
increase in the protein and moisture content of the cheese. In
to distinguish from each other because of the presence of
fact, MTGase enhancement of structural stability by
volatile and non-volatile compounds in the soy milk that
providing cross-linking between hydrophilic parts of
affects the flavor of the cheese. These flavors, if they go
κ-casein during milk protein coagulation and covalent bonds
through a long fermentation process by microorganisms, can
between milk protein molecules [24]. However, if the
cause discomfort to the consumers. However, in our study,
enzymic reaction is too long, the recovery efficiency slightly
the acid coagulation process took place quickly (about 10
decreased. This could be explained that, in too long
minutes), so the smell of soy milk did not negatively affect
treatment time, the substrate for the enzyme reaction was
the organoleptic properties.
gradually depleted, meanwhile, water can react as an acyl
acceptor (acyl donor), leading to the deamination of
glutamine residues to form glutamic acid and ammonia, then
this bond will become weaker [25]. Furthermore, excessive
protein polymerization, which severs the cross-links formed
outside and on the micelle casein surface, leads to reduced
yield and dry matter content in the product [17], [22].
The cheese texture properties of samples at different
enzyme treatment times were presented in Table 6. The
results showed that the hardness of cheese gradually
increases based on the increase of treatment time (from 0 to
2 hrs), then slightly decreased. This could be explained by
the fact that protein plays an important role in the formation
of cheese structure. The higher the protein content was, the
higher the force required to deform the protein was [8].
According to the study of Jaros et al. [25] the hardness of
acid casein gel was a function of protein polymerization; the
Figure 6. Sensory evaluation score of fresh cheese samples
longer the incubation time was, the lower the gel hardness
was. The presence of macromolecule compounds in cheese Scanning electron micrograph (SEM) images of cheese
somewhat reduced the gelation process and prevent the samples were shown in Fig. 7. The images showed a
formation of the homogenous network [26]. rougher, sharper, and more voids surface structure of the C
samples compared to that of the RS sample. It could be
Thus, the fresh cheese production was carried out with
speculated that MTGase was also effective in forming cross-
the following specifications: (a) the mixing ratio of
links between proteins from different sources.
reconstituted milk:soy milk was 8:2; (b) enzymic treatment
conditions: enzyme concentration of 2 IU/g protein at 37°C
for 2 hrs; (c) coagulation by citric acid (1.5 %) at 85°C.
346
C RS
Figure 7. Scanning electron micrograph images of cheese samples

G. Changes in quality parameters of the fresh cheese
during storage IV. CONCLUSION
Changes in titratable acidity (TA) and whey separation The replacement of reconstituted milk with a part of soy
(Wh) of fresh cheese products, during 28 days of storage at milk (20%, v/v) in addition to MTGase (2 IU MTGase/g
4±2°C, were presented in Table 7. protein at 37oC for 2 hours) practically improved the yield
Table 7 showed that the cheese samples had an increase production (39.43%), cheese solid yield (45.81%) and
in titratable acidity and whey separation during storage. texture of fresh cheese. At the same time, the results may
However, the TA and Wh values of the RS cheese were help to solve the problem of shortage of raw milk materials.
always less than that of the C sample. Actually, the TA of The investigated results can be a premise for the next
RS samples decreased in the first week (from 9.5 to 8.73 researches of replacing cow's milk ingredients with the other
mM/100 g). This could be explained by the formation of plant-based milk to create products with a combination of
cross-links between glutamine and lysine resulting in the different protein sources.
release of free ammonia (NH3) which reduces the acidity of
ACKNOWLEDGMENT
the products [3]. During storage, the whey separation of
cheese is the result of dense aggregates of casein molecules, We are grateful to Ho Chi Minh City University of
because casein molecules are extremely flexible, even in the Technology and Education for providing the facilities
presence of degradation. It tends to form more compact needed to complete this study. We sincerely thank the
micelle structures leading to whey separation. These dense following students for their contributions in conducting the
aggregates are spontaneous processes and are explained as a experiments and for their honesty in obtaining these
contraction of the gel system without any external force, valuable results: Phuong The Son (Student ID 17116119),
resulting in the rearrangement of the gel network and whey Pham Trung Thanh (Student ID 17116125).
separation [27]. However, because of MTGase treatment, a
REFERENCES
stable protein network was created and reduced whey
separation. Furthermore, these isopeptide bonds increase the [1] S. Nazim, M. U., Mitra, K., Rahman, M. M., Abdullah, A. T.
M., & Parveen, “Evaluation of the nutritional quality and
volume within the curd matrix, which gives cheese microbiological analysis of newly developed soya cheese”,
structure, thereby increasing whey retention [28]. Moreover, Int. Food Res. J., vol. 20, no. 6, p. 3373, 2013.
the colour of storage cheeses was not significantly different [2] I. E. Elmahi, A. K. H., & ElZubeir, “Physicochemical,
(ΔE < 2). sensory characteristics and cost of production of soymilk
cheese, Sudanese white cheese and their mixture during
storage”, J. Protein Res. Bioinforma., vol. 2, p. 007, 2020.
Table 7. Change of titratable acid and whey separation of fresh
cheese during storage [3] Dmytrów and M. Jasinska and K. Dmitrów, “Effect of
microbiological transglutaminase on selected physicochemical
Storage properties of tvarog”, Ital. J. Food Sci., pp. 449–460, 2010.
Titratable acidity Whey separation (Wh, [4] L. Duarte, C. R. Matte, C. V. Bizarro, and M. A. Z. Ayub,
time
(TA, mM/100g) %) “Review transglutaminases: part II—industrial applications in
(days)
food, biotechnology, textiles and leather products”, World J.
C RS C RS Microbiol. Biotechnol., vol. 36, no. 1, p. 11, Jan. 2020, doi:
10.1007/s11274-019-2792-9.
1 9.82±0.16a 9.50±0.15b 0.84±0.03a 0.17±0.01a [5] W. A Mahmood, “Effect of microbial transglutaminase
treatment on soft cheese properties”, Mesopotamia J. Agric.,
a a b
7 10.23±0.47 8.73±0.28 2.62±0.02 0.91±0.01b vol. 37, no. 4, pp. 19–27, 2009.
14 11.40±0.10b 9.75±0.09bc 4.74±0.04c 1.43±0.01c [6] V. Jooyandeh, H., Mortazavi, S. A., Farhang, P., & Samavati,
“Physicochemical properties of set-style yoghurt as effected
21 12.12±0.21c 10.03±0.08b 5.83±0.03d 2.66±0.02d by microbial transglutaminase and milk solids contents”, J.
Appl. Environ. Biol. Sci, vol. 4, no. 11S, pp. 59–67, 2015.
28 13.36±0.28d 11.10±0.28c 6.31±0.00e 3.18±0.12e
a-d
347
[7] InvestVietnam. See also: [18] F. Bittencourt Luciano and S. Arntfield, “Use of
https://investvietnam.gov.vn/en/nghanh.nghd/27/dairy- transglutaminases in foods and potential utilization of plants
industry.html as a transglutaminase source – Review”, Biotemas, vol. 25,
[8] A. Salinas-Valdés, J. De la Rosa Millán, S. O. Serna-Saldívar, no. 4, Nov. 2012, doi: 10.5007/2175-7925.2012v25n4p1.
and C. Chuck-Hernández, “Yield and Textural Characteristics [19] C. Gauche, J. T. C. Vieira, P. J. Ogliari, and M. T. Bordignon-
of Panela Cheeses Produced with Dairy-Vegetable Protein Luiz, “Crosslinking of milk whey proteins by
(Soybean or Peanut) Blends Supplemented with transglutaminase”, Process Biochem., vol. 43, no. 7, pp. 788–
Transglutaminase”, J. Food Sci., vol. 80, no. 12, pp. S2950– 794, Jul. 2008, doi: 10.1016/j.procbio.2008.04.004.
S2956, Dec. 2015, doi: 10.1111/1750-3841.13126. [20] W. Qi et al., “Whey protein isolate modified by
[9] B. A. Abd El-, “Effect of Milk Fat Replacement with transglutaminase aggregation and emulsion gel properties”,
Vegetable Oil and/or Whey Protein Concentrate on IOP Conf. Ser. Mater. Sci. Eng., vol. 87, p. 012034, Jul. 2015,
Microstructure, Texture and Sensory Characteristics of Fresh doi: 10.1088/1757-899X/87/1/012034.
Soft Cheese”, Int. J. Dairy Sci., vol. 10, no. 3, pp. 117–125, [21] R.-N. Chen, H.-O. Ho, and M.-T. Sheu, “Characterization of
Apr. 2015, doi: 10.3923/ijds.2015.117.125. collagen matrices crosslinked using microbial
[10] O. Benjamin, M. Davidovich-Pinhas, A. Shpigelman, and G. transglutaminase”, Biomaterials, vol. 26, no. 20, pp. 4229–
Rytwo, “Utilization of polysaccharides to modify salt release 4235, Jul. 2005, doi: 10.1016/j.biomaterials.2004.11.012.
and texture of a fresh semi hard model cheese”, Food [22] D. Brink, I., & Leskauskaitė, “Influence of transglutaminase
Hydrocoll., vol. 75, pp. 95–106, Feb. 2018, doi: on the physical and chemical properties of acid milk gel and
10.1016/j.foodhyd.2017.08.033. cottage type cheese”, J. Food, Agric. Environ., vol. 11, no.
[11] T. M. Mokrzycki WS, “Colour difference ΔE - A survey”, 3&4, pp. 119–124, 2013.
Mach Graph Vis., vol. 20, no. 4, pp. 383–411, 2011. [23] L. Darnay, “Textural properties of low-fat set-type yoghurt
[12] G. K. Deshwal, A. K. Singh, D. Kumar, and H. Sharma, depending on mTG addition”, Mljekarstvo, vol. 66, no. 3, pp.
“Effect of spray and freeze drying on physico-chemical, 225–230, Jul. 2016, doi: 10.15567/mljekarstvo.2016.0307.
functional, moisture sorption and morphological [24] F. Razeghi and S. Yazdanpanah, “Effects of free and
characteristics of camel milk powder,” LWT, vol. 134, p. encapsulated transglutaminase on the physicochemical,
110117, Dec. 2020, doi: 10.1016/j.lwt.2020.110117. textural, microbial, sensorial, and microstructural properties
[13] H. T. Lawless and H. Heymann, Sensory Evaluation of Food. of white cheese”, Food Sci. Nutr., vol. 8, no. 7, pp. 3750–
New York, NY: Springer New York, 2010. 3758, Jul. 2020, doi: 10.1002/fsn3.1663.
[14] N. H. Metwalli, S. I. Shalabi, A. S. Zahran, and O. El- [25] D. Jaros, C. Partschefeld, T. Henle, and H. Rohm,
demerdash, “The use of soybean milk in soft-cheese making”, “Transglutaminase in dairy products: Chemistry, physics,
Int. J. Food Sci. Technol., vol. 17, no. 1, pp. 71–77, Jun. 2007, applications”, J. Texture Stud., vol. 37, no. 2, pp. 113–155,
doi: 10.1111/j.1365-2621.1982.tb00161.x. Apr. 2006, doi: 10.1111/j.1745-4603.2006.00042.x.
[15] O. C. Aworh, B. A. Adedeji, and E. C. Nwanekezi, “Effect of [26] D. Jaros, U. Schwarzenbolz, N. Raak, J. Löbner, T. Henle,
partial substitution of cow’s milk with soymilk on yield and and H. Rohm, “Cross-linking with microbial
quality attributes of West African soft cheese”, Int. J. Food transglutaminase: Relationship between polymerisation
Sci. Technol., vol. 22, no. 2, pp. 135–138, Jun. 2007, doi: degree and stiffness of acid casein gels”, Int. Dairy J., vol. 38,
10.1111/j.1365-2621.1987.tb00467.x. no. 2, pp. 174–178, Oct. 2014, doi:
[16] J. Adda, J. C. Gripon, and L. Vassal, “The chemistry of 10.1016/j.idairyj.2013.10.011.
flavour and texture generation in cheese,” Food Chem., vol. 9, [27] L. R. Han XQ, Pfeifer JK, “US patent: Process for making a
no. 1–2, pp. 115–129, Jul. 1982, doi: 10.1016/0308- whey less cream cheese using transglutaminase”, US
8146(82)90073-5. 6,416,797 B1., 2002.
[17] S. M. Taghi Gharibzahedi, M. Koubaa, F. J. Barba, R. [28] P. Di Pierro, L. Mariniello, A. Sorrentino, C. V. L. Giosafatto,
Greiner, S. George, and S. Roohinejad, “Recent advances in L. Chianese, and R. Porta, “Transglutaminase-Induced
the application of microbial transglutaminase crosslinking in Chemical and Rheological Properties of Cheese”, Food
cheese and ice cream products: A review”, Int. J. Biol. Biotechnol., vol. 24, no. 2, pp. 107–120, May 2010, doi:
Macromol., vol. 107, pp. 2364–2374, Feb. 2018, doi: 10.1080/08905431003784465.
10.1016/j.ijbiomac.2017.10.115.
348
Studying the Effect of Biodiesel Blend Ratios

Derived from Rubber Seed Oil on the Technical
Characteristics of Diesel Engines without
Modifying the Structure by Simulation and
Experiment
Nguyen Manh Cuong Huynh Phuoc Son

Faculty of Vehicle and Energy Engineering Faculty of Vehicle and Energy Engineering
HCMC University of Technology and Education HCMC University of Technology and Education
cuongnm@hcmute.edu.vn sonhp@hcmute.edu.vn
Abstract: Biodiesel is derived from rubber seed oil, which used in a diesel engine with little or no change in engine
is one of the environmentally friendly, renewable and structure [1÷5]. Strict emission laws, the exhaustion of fossil
promising biofuels that promise to be used in diesel engines fuels and the relationship of fuel with politics have forced the
without modifying the structure. Vietnam is a country located world to find an alternative to fossil fuels. Many vegetable oil
in the tropical monsoon region, with large areas of rubber esters (biodiesel) have been studied for use in internal
trees in Asia. Therefore, the study of biodiesel fuel from combustion engines and have been shown to have higher
rubber seed oil used for diesel is very necessary and suitable potential to reduce CO2 emissions [6], [7]. The effects of
to our country's current condition. In this study, the model different fuel types on engine performance and specifications
have been widely reported. Common engine parameters on
Mazda WL diesel engine, 4-cylinder, 4 straight-line cylinders
which quantitative effects are included include: Fuel
with biodiesel blends ratio of 0%, 15%, 20%, 25%, 30%
consumption, utility capacity and thermal efficiency. Many
derived from rubber seed oil and DO has been simulated on studies have compared the performance of engines using
ANSYS FLUENT software to analyze fuel consumption biodiesel and its mix with conventional diesel fuel engines.
characteristics. Experimental engine has been using fuel with The results showed that using biodiesel for fuel efficiency and
diesel (DO) and biodiesel blends (B15, B20, B25, B30) higher combustion efficiency than using traditional diesel fuel
derived from rubber seed oil for a Mazda WL diesel engine in diesel [8 ÷ 11].
installed on Mazda 2500 and Ford Ranger cars. Data values
are measured at 75% high load mode with 1500rpm and However, the use of different biodiesel for the same engine
2250rpm. The main parameters of the combustion process will have variable results in engine efficiency, engine
have been analyzed and evaluated from the pressure data performance and engine exhaust due to changes in physical
properties and chemistry of biodiesel [12]. The effects of
recorded in the cylinder. Simulation results show that the
physicochemical properties of fuel on fuel supply systems
average decrease of engine power for all modes is -1.34%, - such as high-pressure pumps, fuel filters and cylinder mixing
2.33%, -3.16% and -3.94% respectively for B15, B20, B25 rates between gas-fuel mixtures have been reported [13]. To
and B30 compared to DO fuel. Experimental results show improve the performance and emissions of biodiesel fuel
that the average decrease of engine power for all modes is - engines and understand the effects of fuel physics on engine
1.49%, -2.62%, -3.58% and -4.35% respectively for B15, efficiency and emissions, a number of detailed studies of
B20, B25 and B30 compared to DO fuel. The study also transmission properties of heat has been made on traditional
shows that: It is possible to use biodiesel fuel with a diesel engines. The rate of heat release, combustion pressure
maximum 20% (B20) blending ratio for Mazda WL diesel and combustion temperature affects the characteristics of
engine without modifying the structure. engine performance and emissions, but most researchers have
so far linked the relationship between efficiency and
Keywords: biodiesel, diesel engine, engine power, engine emissions. Characteristic of biodiesel with experimental
torque, rubber seed oil parameters such as biodiesel mixture ratio, engine speed,
engine load, injection time, injection pressure and
I. INTRODUCTION compression ratio of the engine. However, there are very few
Combustion process in combustion engine determines the published works on combustion characteristics, combustion
power, efficiency, economic level, pollution emission level law and heat generation phenomena corresponding to
and engine features when operating. A large number of studies different biodiesel rates and their mixes [14 ÷ 15].
have shown that biodiesel is one of the renewable, promising Simulation with mathematical models is necessary to
alternative and environmentally friendly biofuel that can be develop a new engine construction or modernize an existing

one. Modeling of the processes inside the cylinder allows a of conservation element; Equation for turbulence kinetic
first approximation to evaluate engine performance, choose energy k; Equation for Dissolution Dissipation Rate :
the rational value of adjustment or constructive parameter, to
reduce material, labor and time required to conduct The main combustion model of J. Abraham: Consider
experimental research. One of the most difficult processes for the variability of substances such as: O2, N2, CO2, CO, H2, H20
simulation is the combustion process in diesel engines. In the duration of time combustion J. Abraham [20] calculates
Diffusion combustion model and turbulence combustion both low temperature and high temperature. In particular, the
model are the base model to study combustion process in rate of time-varying mass of a substance m in the combustion
diesel engines using diesel fuel mixture biodiesel phase. chamber due to chemical changes from one substance to
Theoretical basis diffusion combustion and turbulence model another is determined by the equation:
is the foundation for the integrated simulation and simulation dm  − *m (2)
software from ANSYS 14.5. Ansys 14.5 has the ability to =− m
dt c
optimize geometry, and can automatically adjust the structural
parameters of special designs until they reach the desired Where: ρm: partial density of substance m; ρ*m:
optimal design goals. Based on Ansys 14.5 software, it is instantaneous and local thermodynamic equilibrium value of
possible to set up a combustion model of the diesel-biodiesel partial density; τc: characteristic time to reach equilibrium.
fuel mixture to study engine performance [16, 18]. Simulation
results in this paper are shown through the variation of Ignition Delay model of Kong and Reitz: Ignition Delay
pressure, temperature, fuel level, speed of the mixture of model of Kong and Reitz [20] computes the ignition delayed
different fuels in the same boundary conditions. time of eight reactions between the five. Determine the
coefficients of formation substance’s velocity.
Therefore, the objective of this study is to investigate and
study the characteristics of combustion of biodiesel fuel in Building combustion chamber geometric and discrete
model: The experimental engine is a Mazda WL-Turbo 4-
separated combustion chamber diesel engines such as: the
stroke engine, 4 in-line cylinders are mounted on Mazda 2500
technical characteristics, fuel consumption, with different and Ford Ranger cars with pre-combustion chamber. The
loads condition, biodiesel ratios, thereby contributing to
combustion chamber is divided into 2 parts: the main
research to perfect diesel engines when using biodiesel fuel
combustion chamber and the auxiliary combustion chamber,
derived from rubber seed oil with different blend ratios as well
forming 2 different vortex lines as shown in Figure 1.
as contributions the orientation to promote the use of this fuel
in practice. This paper presents the results of the ANSYS 14.5
software application to find out technical characteristics of
biodiesel (B15, B20, B25, B30) and conventional diesel (DO)
in Mazda WL diesel engine This paper also will present
experimental results to study the combustion characteristics of
biodiesel fuels with different proportions (B15, B20, B25,
B30) derived from rubber seed oil and conventional diesel
(DO) in the pre-combustion chamber diesel engine is
implemented at: Internal Combustion Engine Laboratory -
Department of Transportation Engineering - University of
Technology - Danang University.
Figure 1. Combustion chamber geometric
II. SIMULATION MODEL SETUP
Meshing model: Meshing of the combustion chamber
A. Simulation modeling separating WL-turbo engines is done on Ansys Fluent 14.5
Simulation results to predict the combustion, economy, software. With the automatic meshing feature (Automatic
technical and emissions characteristics of diesel engine using Method), Quad/Tri grid type, the mesh model of the 2D
B15, B20, B25 and B30 biodiesel fuel mixtures. Be the basis combustion chamber includes 9917 nodes and 9677 elements
for evaluating experimental results a real engine. as shown in Figure 2.
Turbulence modeling: In the turbulence modeling k −  ,
the equations are constructed as follows: Under the
assumption of Boussinesq's viscosity [19], to have:
 u u j  2 
−  ui' u 'j =  t  i +  −  k +  i ui  ij (1)
 x  3 
 j xi   xi 
The above equation shows the relationship between

Reynolds stress −  ui u j and mean velocity variation.
' '
The general equation represents the diffusion spray: In Figure 2. Mesh of the Mazda engine combustion chamber
order to model the supply of fuel for diesel engine, simulation
software demonstrates diffusion spray by continuous equation Simulation mode will be as follows: Step 1: Input the
as follows [19]; Continuous equation; Equation of momentum corresponding parameters when the engine is running at 1500
conservation; Full Enthalpy conservation equation; Equation rpm and 2250 rpm with advance angle injection of 10 deg
before TDC, pressure is 114 (bar). Step 2: Enter the fuel
350
models corresponding to DO, B15, B20, B25, B30. For each cylinder, the lift of the injector etc. Indiset 620 device is used
fuel, change the load corresponding to the 25%, 50% and 75%. in the experiment of measuring the indicator parameters of the
Step 3: Run the calculation model and record the results of the engine such as the combustion chamber pressure. The basic
combustion process, power, fuel consumption and emissions. specifications of the Mazda WL engine are shown in Table 2.
B. Characterization of biodiesel fuel used for Simulation Table 2. The specifications of MAZDA WL engine
and experiment
Specifications Symbol Unit Value
The fuel used for this study includes 4 types: biodiesel Max power at 3500 (rpm) Ne kW 85
blends B15, B20, B25 and B30 (eg. B25 is the ratio of 25% Compression ratio ε 19,8:1
biodiesel and 75% conventional diesel) derived from rubber Maximum Torque at 2000(rpm) Me Nm 266
seed oil and traditional diesel fuel (DO fuel): The chemical Cylinder diameter D mm 93
composition of rubber seed oil is mainly triglycerides, so they Length of connecting rod L mm 152
have the full characteristics of a typical ester. Biodiesel Piston stroke S mm 92
synthesized from rubber seed oil contains mainly unsaturated displacement V cc 2499
Number of cylinders i 4
fatty acid roots with a total content of up to 76.69%. This
Number of strokes τ 4
component makes biodiesel flexible at low temperatures. The Advance injection timing TDC  deg 10
total amount of methyl ester present in biodiesel is 100%, IVO 1 deg 10
indicating the very high purity of synthetic fuels. In the IVC 2 deg 44
chemical formula of biodiesel, the weight ratio is 77% C, 12% EVO 3 deg 51
H, 11% O, the ratio between air and standard fuel = 12.5; for EVC 4 deg 9
diesel, the mass density is 87% C, 13% H and the ratio Fuel pressure at start of injection bar 114÷121
between air and standard fuel = 14.5. Proceeding to mix DO
fuel samples with 15%, 20%, 25% and 30% biodiesel (from
rubber seed oil) in volume, to make the mixture stable for a III. THE EXPERIMENT LAYOUT AND
period of 5 days with no impact on mechanics. Then take the INSTALLATION
sample to analyze at the laboratory of the General Department The experiment engine is mounted behind the APA test
of Standards, Metrology and Quality - Technical Center for strip, which is connected to the motor by the connecting shaft.
Quality Measurement No.2 and compare control with DO fuel The schematic diagram of the experiment setup is shown in
samples of the same type with diesel phase into the mix is Fig. 3 and the installation; the experimental engine in Fig. 4.
shown in Table 1.
Table 1. Comparison main characteristics of B15, B20, B25 and

B30 with diesel
Characteristic,
Experiment Diesel B15 B20 B25 B30
units
Sulfur content, ASTM
500 295 238 238 217
mg/kg D5453
Heating value, ASTM D240
43738 42217,3 41910,4 38730 37500
kJ/kg - 06
Lightning flash
ASTM D93 55 68 72,0 74,0 74,0
burn, oC
Kinematic
viscosity at 40oC, ASTM D445 3,070 3,310 3,990 4,693 4,693
cSt
Water and
Figure 3. Schematic diagram of experiment setup
mechanical ASTM
<0,005 <0,005 <0,005 <0,005
impurities, % D2709 – 06
volume
Particulate matter, ASTM
10 <1 <1 <1 <1
mg/l D2276 –06
Density at 15oC, ASTM
840 845,2 848,0 853,4 853,4
kg/m3 D1298
ASTM D 46,9÷49, 46,9÷49, 46,9÷49, 46,9÷49,
Cetane index 45
6751 9 min 9 min 9 min 9 min
C. Experiment Specifications used for Simulation and

experiment
The experimental engine is a Mazda WL-Turbo 4-stroke
engine, 4 in-line cylinders are mounted on Mazda 2500 and
Ford Ranger cars. With pre-combustion chamber: Capacity:
2499 (cc); Maximum torque: 266 Nm@ 2000 rpm; Maximum
power: 85 kW@3500 rpm; The type of combustion chamber Figure 4. Engine experimental installation
separates. The engine is fully mounted with sensors to record
the engine's parameters when working such as fuel pressure Experimental conditions: In order to experiment with
before high-pressure pump, fuel pressure on the high-pressure satisfactory results, the author chooses the initial conditions
pipe, intake air pressure, exhaust pressure, pressure in the for the experimental process:
351
• Do not change temperature conditions such as intake Table 5. Engine power value when using DO and biodiesel by
air temperature (Tintake = 32 ± 0.50C), fuel temperature experiment
(Tfuel = 33 ± 0.50C) and lubrication oil temperature (Toil Biodiesel blend ratio and DO
= 35 ± 10C). speed load (%) DO B15 change B20 change B25 change B30 change
25 12,8 12,5 -2,34 12,3 -3,91 12,19 -4,77 12,03 -6,02
• Flow of cooling water: 15m3/h; Cooling water 1500 (rpm) 50 19,3 18,9 -2,07 18,7 -3,11 18,41 -4,61 18,39 -4,72
temperature in: 40 ÷ 850C; Lubricating oil pressure: 3 75 22,3 22,07 -1,03 21,86 -1,97 21,61 -3,09 21,35 -4,26
÷ 5 bar. average -1,82 -3,00 -4,16 -5,00
Content of experimental modes: In this study, when 25 5,37 5,29 -1,49 5,18 -3,54 5,15 -4,10 5,08 -5,40
2250 (rpm) 50 31,4 30,9 -1,59 30,78 -1,97 30,46 -2,99 30,38 -3,25
using DO fuel, biodiesel blends B15, B20, B25 and B30, the
75 48,3 48,1 -0,41 47,7 -1,24 47,36 -1,95 47,11 -2,46
operation mode of the test engine is constructed such that the
average -1,17 -2,25 -3,01 -3,70
speed of rotation and load of the engine are equivalent to those -1,49 -2,62 -3,58 -4,35
The average value
of common values of vehicles. The test modes according to
the ECE R49 test cycle are shown in Table 3. From the experimental data table 5 shows, the average
reduction of power for all modes with B15, B20, B25 and B30
Table 3. Contents of experimental modes compared to DO fuel are -1.49%, -2.62%, -3.58% and -4.35%,
Experimental fuels DO, B15, B20, B25 và B30 respectively. Thus, if increasing the biodiesel mixing ratio to
Engine speed 1000 rpm÷3000 rpm step: 250 rpm
B30, the engine power will be reduced 3 times (%) compared
to when using the low mixing ratio B15. This can be seen that
Load conditions 25%, 50% and 75% the kinematic viscosity, specific gravity, compression stress of
Repeatability 1 measurement 3 times, the average pressure data 100 the fuel, volatility and calorific value of biodiesel have a great
point cycles influence on the combustion process of the fuel as well as the
engine power.
Limited condition of experiment:
The results show that, at the same engine load mode, the
• Do not change the advance injection timing of high- engine power when using DO gives the largest value and
pressure pump ( = 10deg before TDC) and the gradually decreases when increasing the biodiesel blend
injection pressure of MAZDA WL test engine is 114 ratios, the rule of engine power change when using biodiesel
bar. between simulation and experiment are similar. Considering
the same engine load mode means that the amount of fuel
• Limit the maximum engine speed to 2500 rpm in supplied for a cycle is the same for all fuels, the reduced
25% throttle position and 3000 rpm in 50%, 75% engine power is due to the lower calorific value of biodiesel.
modes. On the other hand, due to reduced ignition delay time, the
• Do not use additives when mixing DO fuel with phenomenon of both combustion and compression occurs for
biodiesel from rubber seed oil into B15, B20, B25, B30 biodiesel so that decrease engine power. However, the
blends. tendency to reduce engine power is different when operating
at low engine speed (1500 rpm) with a larger deviation
IV. RESULT AND DISCUSSION compared to high engine speed rpm (2250 rpm) for the same
Tables 4 and 5 present the results of comparison of engine load mode. This can be seen when increasing the number of
power when using DO with biodiesel according to simulation engine revolutions, leading to an increase in the intake
and experiment at 2 revolutions of 1500 rpm and 2250 rpm coefficient, the vortex of the fuel mixture in the cylinder
with 25%, 50%, 75% load modes. increases, the fuel evaporates faster and more evenly, so the
combustion process occurs better. The engine power deviant
Table 4. Engine power value when using DO and biodiesel by at high engine speed is smaller than at low engine speed. The
simulation trend of engine power change according to simulation and
experiment are shown in the graph of Figures 5, 6.
Load Biodiesel blend ratio and DO
speed
(%) DO B15 change B20 change B25 change B30 change Engine power change trend according to simulation
25 12,13 11,94 -1,57 11,9 -1,90 11,85 -2,31 11,81 -2,64 0
1500 (rpm) 19,02 18,46 -2,94 18,13 -4,68 17,95 -5,63 17,89 -5,94 B15 B20 B25 B30
Engine power change Ne (%)
50
75 21,85 21,63 -1,01 21,28 -2,61 20,94 -4,16 20,47 -6,32

-2
average -1,84 -3,06 -4,03 -4,96
25 5,14 5,07 -1,36 5,01 -2,53 4,95 -3,70 4,89 5,14

-4 y = -0,6869x - 0,1924
2250 (rpm) 50 30,07 29,95 -0,40 29,83 -0,80 29,71 -1,20 29,66 30,07 R² = 0,9979
75 46,97 46,61 -0,77 46,27 -1,49 46,05 -1,96 45,8 46,97 y = -1,0349x - 0,8874
-6
R² = 0,9957
average -0,84 -1,61 -2,28 -2,91
The average value -1,34 -2,33 -3,16 -3,94

Biodiesel/diesel ratio
-8
According to the simulation results in Table 4, the average
engine power reduction for all modes with B15, B20, B25 and
B30 compared to DO fuel are -1.34%, -2.33%, -3.16% and - Figure 5. The trend of engine power change according to
3.94% respectively. simulation
352
Engine power change trend according to experiment The Rule of Trends in Ne change between Si. and Exp.
0
0 DO B15 B20 B25 B30
B15 B20 B25 B30

Engine power change Ne (%)
-1
Value change Ne (%)

-2 y = -0,9689x + 0,7529
-2
R² = 0,9865
y = -1,0795x + 0,8286
-4 -3 R² = 0,9839
y = -1,100x - 0,758
R² = 0,988
-4
y = -1,070x - 0,815 The trend change Ne by simulation
-6
R² = 0,994 The trend change Ne by experiment
-5
-8
Figure 7. Comparison of the rule of trend between simulation and

Figure 6. The trend of engine power change according to the experiment of the engine power when using biodiesel
experiment
Also according to the experimental results, the change in
Compare the power results of the engine between engine power when using DO fuel compared with biodiesel at
simulation and experiment: Because the simulation model is different engine speeds is also shown through the engine
calibrated by economic and technical criteria of Mazda WL power curve in Figure 8. When the engine is running at high
engine measured from experiment. Therefore, in terms of load, the maximum power shifts towards high engine speed
absolute value, we will compare criteria such as engine power, and then decreases rapidly. From the characteristic curve, the
fuel consumption rate. The value of Mazda WL engine power capacity of DO is always larger than biodiesel when the
when using biodiesel fuel changes compared to DO fuel number of revolutions changes at the same load and the
according to simulation and experimental results shown in increase is almost the same for each type of fuel, this is also
Table 6. The comparison results show that the biggest shown in Figure 8.
deviation in engine power between simulation and experiment
is -5.23% at 1500 rpm and 25% engine load mode. The power Experimental engine power characteristic; Ne = f(n)
70
comparison results between simulation and experiment from
Engine load (75%)
Engine power Ne (kW)
Table 6 also show that the trend of changing engine power 60

SOI: 10 deg BTDC
when using biodiesel in the simulation has a close rule with 50 Injection pressure 114 bar
experiment. This is shown in the graph of Figure 7 comparing
40
the trend of changing technical parameters of the Mazda WL
engine when using biodiesel fuel derived from rubber seed oil 30
compared to DO fuel.
20
Table 6. Compare the engine power between simulation and DO B15 B20
10
experiment B25 B30
0
Mazda WL engine power, Ne (kW) 1000 1250 1500 1750 2000 2250 2500 2750 3000
Engine speed n (rpm)
At engine speed n = 1500 rpm At engine speed n = 2250 rpm
Load (%) Figure 8. Characteristic of engine power vs. engine speed by
DO B15 B20 B25 B30 DO B15 B20 B25 B30 experiment
Si. 12,13 11,94 11,9 11,85 11,81 5,14 5,07 5,01 4,95 4,89 Comment: Although biodiesel fuel has an oxygen
component in the fuel, it improves combustion process.
25 Exp. 12,8 12,5 12,3 12,19 12,03 5,37 5,29 5,18 5,15 5,08 However, due to high cetane number, high kinematic
viscosity, high density, high elasticity and poor evaporation,
S-E (%) -5,23 -4,48 -3,25 -2,79 -1,83 -4,28 -4,16 -3,28 -3,88 -3,74 the technical parameters of Mazda WL engine are all reduced.
Based on experimental data, it can be seen that B15 has a small
Si. 19,02 18,46 18,13 17,95 17,89 30,07 29,95 29,83 29,71 29,66
decrease compared to DO and does not decrease significantly
19,3 18,9 18,7 18,41 18,39 31,4 30,9 30,78 30,46 30,38
when at 75% load. Therefore, considering the technical
50 Exp. specifications of the engine, it can be recommended to use
B15 fuel for the Mazda WL diesel engine when in high load
S-E (%) -1,45 -2,33 -3,05 -2,50 -2,72 -4,24 -3,07 -3,09 -2,46 -2,37
mode without modifying the structure. For diesel engines, the
Si. 21,85 21,63 21,28 20,94 20,47 46,97 46,61 46,27 46,05 45,8 high load mode (75% load) is also the mode in the most
frequent operating area.
75 Exp. 22,3 22,07 21,86 21,61 21,35 48,3 48,1 47,7 47,36 47,11
V. CONCLUSION
S-E (%) -2,02 -1,99 -2,65 -3,10 -4,12 -2,75 -3,10 -3,00 -2,77 -2,78 The model Mazda WL diesel engine, 4-cylinder, 4
straight-line cylinder with biodiesel blends ratio of 0%, 15%,
20%, 25%, 30% derived from rubber seed oil and DO has been
353
simulated on ANSYS FLUENT software to analyze [6] M. P. Dorado, E. Ballesteros, J. M. Arnal, J. Gˇmez, and F. J.
combustion parameters, compare economic and technical Lˇpez, “Exhaust emissions from a Diesel engine fueled with
transesterified waste olive oil[small star, filled]”, Fuel, vol. 82,
characteristics as well as emission targets, to ensure reliability no. 11, pp. 1311-1315, Jul. 2003.
and suitability to the reality. [7] Z. Utlu and M. S. Koˇak, “The effect of biodiesel fuel obtained
Through the developed model, it is possible to analyze and from waste frying oil on direct injection diesel engine
performance and exhaust emissions”, Renewable Energy, vol.
evaluate the combustion characteristics of biodiesel in internal 33, no. 8, pp. 1936-1941, Aug. 2008.
combustion engine as well as to explain and evaluate the [8] M. Lapuerta, O. Armas, R. Ballesteros, and J. Fernández,
experimental results. The simulation results show that the “Diesel emissions from biofuels derived from Spanish potential
average decrease of engine power for all modes are -1.34%, - vegetable oils”, Fuel, vol. 84, no. 6, pp. 773-780, Apr. 2005.
2.33%, -3.16% and -3.94% respectively for B15, B20, B25 [9] M. Lapuerta, O. Armas, and J. Rodríguez-Fernández, “Effect
and B30 compared to DO fuel. of biodiesel fuels on diesel engine emissions”, Progress in
Energy and Combustion Science, vol. 34, no.2, pp. 198-223,
The results of the experimental study have also identified Apr. 2008.
factors that can be used to analyze the effects of biodiesel fuel [10] D. Altiparmak, A. Keskin, A. Koca, and M. Guru, “Alternative
derived from rubber seed oil to substitute on economic fuel properties of tall oil fatty acid methyl ester-diesel fuel
indicators, operational parameters of diesel engine. blends”, Bioresource Technology, vol.98, no. 2, pp. 241-246,
Jan. 2007.
Experimental results show that the average reduction of power
[11] A. Monyem and J. H. Van Gerpen, “The effect of biodiesel
for all modes with B15, B20, B25 and B30 compared to DO oxidation on engine performance and emissions”, Biomass and
fuel are -1.49%, -2.62%, -3.58% and -4.35%, respectively. Bioenergy, vol. 20, no. 4, pp. 317-325, Apr. 2001.
Based on experimental data, it can be seen that B15 has a [12] Y. Gao et al., “Experimental study of the spray characteristics
of biodiesel based on inedible oil,” Biotechnology Advances,
small decrease compared to DO and does not decrease vol. 27, no. 5, pp. 616-624, Sep.
significantly when at 75% load. Therefore, considering the [13] B. Tesfa, R. Mishra, F. Gu, and N. Powles, “Prediction models
technical specifications of the engine, it can be recommended for density and viscosity of biodiesel and their effects on fuel
to use B15 fuel for the Mazda WL diesel engine when in high supply system in CI engines”, Renewable Energy, vol. 35, no.
load mode without modifying the structure. 12, pp. 2752-2760, Dec. 2010.
[14] M. Gumus, “A comprehensive experimental investigation of
ACKNOWLEDGMENT combustion and heat release characteristics of a biodiesel
(hazelnut kernel oil methyl ester) fueled direct injection
This work belongs to the project in 2022 funded by Ho Chi compression ignition engine”, Fuel.
Minh City University of Technology and Education, Vietnam. [15] Duong Viet Dung, Tran Van Nam, Nguyen Manh Cuong,
“Comparision of the exhaust components of biodiesel from fish
REFERENCES fat and diesel in internal combustion engine”, 2nd International
[1] G. Hammond, S. Kallu, and M. McManus, “Development of Conference on Green Technology and Sustainable
biofuels for the UK automotive market”, Applied Energy, vol. Development, October 30th - 31st, 2014.
85, no. 6, pp. 506-515, Jun. 2008. [16] Kim, H., and N.Sung, 2004 . "Combustion and Emission
[2] M. Lapuerta, O. Armas, R. Ballesteros, and J. Fernˇndez, modelling for a Direct Injection Diesel Engine". SAE paper
“Diesel emissions from biofuels derived from Spanish potential No.2004-01-0104.
vegetable oils”, Fuel, vol. 84, no. 6, pp. 773-780, Apr. 2005. [17] G.Stiesch, 2003, “Modeling Engine Spray and Combustion
[3] T. Durbin, J. Collins, J. Norbeck, and M. Smith, “Effects of Processes,” Springer-Verlag Berlin Heidelberg. 2003.
Biodiesel, Biodiesel Blends, and a Synthetic Diesel on [18] Kärrholm, F.P., “Numerical Modelling of Diesel Spray
Emissions from Light Heavy-Duty Diesel Vehicles,” Injection, turbulence Interaction and Combustion”. Thesis for
Environmental Science & Technology, vol. 34, no. 3, pp. 349- Doctor of Philosophy, 2008.
355, Feb. 2000. [19] Han, Z.; Reitz, R.D., “Turbulence Modeling of Internal
[4] Daniel Puppan, “Environmental Evaluation of Biofuels”, 16- Combustion Engines Using RNG k-e Models”, Combust. Sci.
Jun-2008. Online]. Available: and Tech. 106, 4-6, p. 267, 1995.
http://www.pp.bme.hu/so/2002_1/pdf/so2002_1_08.pdf. [20] J. Abraham, F. V. Bracco, and R. D. Reitz (1985),
[Accessed: 16-Jun-2008]. “Comparison of Computed and Measured Premixed Charged
[5] A. Ramadhas, C. Muraleedharan, and S. Jayaraj, “Performance Engine Combustion”, Combustion and Flame, vol. 60, pp. 309-
and emission evaluation of a diesel engine fueled with methyl 322.
esters of rubber seed oil”, Renewable Energy, vol. 30, no. 12,
pp. 1789-1800, Oct. 2005.
354
Work from Home during the Covid-19 Pandemic: A

Qualitative Research with Women Taking Care of
Young Children
Hien Phan Thi Thanh Thuy Nguyen Thi Thanh Tram Nguyen Thi Mai
Faculty of Economics Faculty of Economics Faculty of Economics
hienptt@hcmute.edu.vn thuynguyen@hcmute.edu.vn tramntm@hcmute.edu.vn
Abstract: This study aims to interpret the consequences of research of Sullivan that telework creates the flexibility in
work from home (WFH) towards women who have young working arrangements, and increases work-life balance [13].
children (primary school pupils or younger) during the social Employees who are responsible for taking care of
distancing time because of the Covid-19 pandemic. Personal children, elderly parents or relatives might appreciate remote
interview was adopted to collect the data. 15 mothers with working [2], a significant relationship between the number of
young children participated in the interviews which were dependents at home and better work-life balance thanks to
performed online via Google Meet and recorded. The telework. The research [9] also supports this with the
interviews were then transcribed and coded into themes using suggestion that employees are possible to take care of family
Nvivo software. The research shows the advantages and members while working remotely.
disadvantages of WFH which are classified into three
Reference [10] found that remote working results in job
categories, including: work-related, family-related, and
satisfaction. The organizations benefit from remote working
individual-related benefits and drawbacks. With the results of
with increased organizational commitment and work efforts,
this study, WFH has both positive and negative sides. And in as well as employees are able to improve job satisfaction and
the specific situation of Covid-19 pandemic, the well-being [5].
disadvantages overweigh the advantages. But nearly a half of
the interviewees want to continue with WFH after social The disadvantages of WFH were also proved by other
distancing. This poses a problem of job redesigning so that studies. Fonner and Stache stated that WFH creates greater
female employees have opportunities to enhance their life contamination between private life and work, which leads to
satisfaction in general and particularly at work for sustainable life-to-work conflicts [7]. Besides, this causes stress and time
development of organizations. allocation challenges [15]. Work-to-life conflicts are also
caused by the encroaching of work-related worries into daily
Keywords: work from home, Covid-19, qualitative life [11].
research, women, stress According to reference [4], as a result of professional
isolation and diminished organizational commitment, the
I. INTRODUCTION benefits of telework are diminished in the public sector. This
The fourth wave of Covid-19 Pandemic from April 2021 finding was supported by the study [3], with WFH,
has severely affected Vietnam. As a result of social distancing workers are disconnected from their office-based
policy of the Government, enterprises had to adopt WFH for counterparts. Similarly, if not carefully planned, remote
most of the staff. Employees faced a vast change compared working reduces collaborations amongst colleagues
with before daily routines. Especially with women who have and makes development and maintenance social
to take care of young children. In Vietnamese society, women relationships more challenging [6, 14]. Flores has found that
are supposed to do most of the housework and take care of there’s less perceptibility on the fairness in performance
children. Women spend 14 hours more per week than men on evaluation of employees who work remotely [6].
housework, child care and the elderly [3]. It’s clearly far more
It’s reported that remote working negatively impacts
difficult for women to perform family role and work role at
personal well-being, and these effects differ according to
the same time during the pandemic. This study aims at
gender and parental status [12]. With WFH, the more children
clarifying the consequences of WFH toward women who have
parents have the more difficult to manage the work and there
young children (primary school or younger) during the social
can be a lot of distractions that make it difficult to concentrate
distancing time.
on work [9, 14]. Remote employees may find it difficult to
Conventionally, WFH is believed to enhance work maintain motivation [14]
flexibility, reduce distraction, improve productivity, increase
As reported by Felstead and Henseke, remote working can
morale and loyalty, lower stress and lessen monthly expenses
result in increased workloads [5]. This finding
[3]. The similar advantages of WFH were also found by a
was affirmed by Flores in a study which suggested that remote

workers find it difficult to keep balance between work and spending time commuting to work or preparing (makeup,
personal demands. It’s also demanding to make technology dress), they can start their work earlier, for instance, “I can
work and find timely information in a remote environment [6]. save the time commuting from my house to workplace, it’s a
Additionally, remote workers face possibility of insufficient bit far. It takes about an hour, in total, I can save about two to
equipment and resources for remote working [14]. three hours a day. I spend that time working.”(PV10). Saving
time travelling not only from home to the workplaces but also
II. METHOD from the company to other offices as a claim: “Working at the
A. Participants company, it takes a lot of time to move from home to the
company and then from the company to the relevant offices.”
There are three criteria to recruit participants for this study: (PV12).
women who have young child(ren), have a full-time job and
were working from home because of the Covid-19 pandemic. Better performance: there were 4 participants who
With the relationships of the researchers, 18 women who met mentioned this advantage of WFH with 5 references. With the
the requirements were contacted, after articulating about the convenient information technologies, it’s easy for employees
research, there were 15 agreed to participate in the interviews. to contact their co-workers without moving, for example,
“WFH makes it easier for us to communicate with colleagues.
They had different jobs: 3 Lecturers, 1 Screenwriter, 5 Well, I can work from home and talk without having to move.”
Office Workers, 1 Civil Engineer, 1 Sale Person, 3 (PV12). Or another opinion: “it is clear that my productivity
Department Managers, 1 English Teacher. They are taking is better because I can spend time being able to handle the
care of young children: 1 child (6 participants), 2 children (7 backlog of work. Because with my job at the bank, when the
participants), 3 children (2 participants). working hour is over I can’t stay to continue working. But
At the time of the interviews (October and November, when working from home, I can work any hour, and I can
2021), they had been remotely working for 3 to 10 months. spend time reading a lot of documents and instructions from
the bank.” (PV14).
B. Procedure
Comfortable while working: 6 participants mentioned this
The semi-structured, in-depth interviews were conducted
advantage of WFH with 7 references. WFH helps reduce
via video calls with Google Meet. Participants were asked to
pressure from the supervisor like the experience of a
share their own experience of WFH during the pandemic with
participant: “Working from home makes it more mentally
the main questions:
comfortable, reducing a lot of pressure. Especially when I go
- What are the advantages of WFH during the social to work, I often have to deal with urgent requests, it’s really
distancing time? stressful, but now when I work online, everything is handled
in a much lighter way.” (PV1).
- What are the disadvantages of WFH during the social
distancing time? Besides, working in their own homes makes employees
feel snugger and they also can do some other things that help
Besides, to be aware of the aspirations of employees about them feel relaxed, as a statement: “WFH helps me have a more
WFH, the last question was: comfortable working spirit, I can sit and work in a familiar
- After social distancing time, if have rights, will you space of my own family, I can listen to music while working
choose WFH? without affecting anyone and when I’m too stressed, I can take
a break for example, do some other household chores, or play
The video call interviews were recorded under the with the kids, eat, or drink something it will relieve my stress”.
permission of participants and transcribed into documents for (PV12).
analysis. In this research, Nvivo was used to interpret the data
into themes. Furthermore, working without makeup, formal dress and
others’ observations helps employees feel much easier as a
III. RESULTS claim “Yeah, I feel more comfortable, for example, when I go
After interpreting the data with NVivo, the specific to work, I have to put on makeup and then formal dress. I have
advantages and disadvantages of WFH were identified. to pay attention to the way I walk, the way I act, because they
affect my image at work. When I work at home, it is more
A. The advantages of WFH comfortable to wear pyjamas and sit more comfortably, I am
• Work-related advantages: not constrained.” (PV9).
Flexible working time: 9 participants mentioned this • Family-related advantages:

benefit of WFH. When employees work remotely, they do not Family members are more connected: 15/15 mentioned
have to follow the established working hours. With the same these pros with 24 references. By working from home,
workload, they can flexibly arrange time to accomplish employees have more chances to take care of their families,
depending on personal schedules. For example, a participant especially their children. In the pandemic situation, all the
showed her experience: “Well, WFH helps me to arrange my members of families stay at home, this is a vacation for them
time more flexibly, not necessarily what time I have to start to share with each other, like a statement “WFH has a positive
my workday and if there are some tasks that I cannot schedule impact on my family, that is, we have family meals and
during the day I can arrange to do it in the evening.” (PV10). everyone can cook together.” (PV11), or “Family members
Or another participant stated: “I am not constrained in are more connected, people have the opportunity to
working hours.” (PV15). understand each other better” (PV1).
Save the time of traveling and preparing: There were 5
participants showing this advantage of WFH. Instead of
356
Save the living expense: 3/15 mentioned this benefit of shown is the distractions: “I myself don’t work with discipline
WFH. When employees go to work, they spend money on like in the past, I don’t guarantee that it is working
gasoline, lunch, cosmetics, dress, hanging out with co- continuously during office hours, I don’t focus on working like
workers or friends. All these expenses are cut down in the office.” (PV15).
dramatically in the social distancing time with WFH. This
advantage was described by a participant: “For nearly a year Insufficient facilities: 5/15 mentioned these cons with 6
now, I have also saved money on cosmetics and gasoline” references. Inadequate equipment is one of the things that
(PV11), or another: “I also save money on shopping for hinders work. A participant revealed “The first thing I can see
clothes and cosmetics, for example.” (PV14). is the system sync errors. There are documents I need, but due
to the sync error, I can’t find them. And the internet at home
• Individual-related advantages: is sometimes slow. For example, with the same task, it only
takes about fifteen minutes at work, but at home, because of
Better health condition: 8/15 mentioned this with 10 the slow internet, it can take up to 30 minutes and even longer
references. Motorbikes are the most popular vehicle in and sometimes the network is even broken.” (PV15). It’s not
Vietnam with 55 million in 2018 [1]. 11/15 participants in this only about technical errors but also about equipment: “The
study travel to work by motorbike; most of them are afraid of workspace is not as professional as the office, so I can’t
traffic accidents, air and noise pollution. These things are concentrate as well as I am in the office.” (PV3).
shared: “Physically, WFH helps me not get stuck in traffic, I
don’t have to be exposed to dust and noise pollution, it helps Expansion of work: 4/15 women mentioned this
my health better.” (PV12). Besides, staying at home and drawback. Because the employees do not have the clear
working creates “More time for me to exercise” (PV11) and boundaries of working time and personal time, their work can
get better health conditions. be expanded “Normally, if I go to work, I work eight hours a
day, but when I work at home, sometimes I still have to write
B. The disadvantages of WFH scripts at night.” (PV1) With the online meetings, because of
• Work-related disadvantages: low effectiveness of communication, they can be lengthened
“WFH makes online meetings happen more regularly and
Lower income: 5/15 participants informed that WFH there are meetings that last very long through my lunch and
caused a reduction in their incomes because of fewer even dinner. Some days, I have to join meetings from 11.00am
workloads and declined companies’ turnover “Especially, until 06.00pm.” (PV12).
WFH affects my income a lot” (PV4) or a teacher explained
“The tuition can’t be as high as before.” (PV11). • Family-related disadvantages
Increased workload: 2/15 women said that their workloads Family conflicts: 3/15 participants talked about this cons
increased because they had to deal with technology more than of WFH. During the social distancing time, both husbands and
before: “my workload has increased because I have to wives have to work from home, women expect their husbands
prepare the digital learning materials and send them to my to share the housework, which often leads to conflicts. For
students”. (PV11). example: “Sometimes there are conflicts and quarrels in my
family because of trying to arrange this person to help others”
Ineffective communication: 10/15 mentioned this
(PV6) or other experience: “The most obvious negative impact
drawback of WFH with 13 references. To receive feedback
on our family is that everyone has to arrange time between
from others, it took quite longer than the time when they went
work and family, so it often leads to not being able to arrange
to work, like a claim: “When I ask for a professional opinion,
time, leading to the fact that I and my husband argue.” (PV7).
I have to wait until my boss or co-workers is online to receive
the answer.” (PV1). Family’s routines are unexpectedly changed: 5/15
participants mentioned this consequence of WFH. Because the
The effectiveness of communication is also affected by the
boundary between work and family life is not as clear as
misunderstanding each other like: “Because we don’t meet in
before, this results in some changes in arrangements of
person, sometimes we misunderstood each other.” (PV12).
families as an explanation from a woman “Due to working
Especially with the job that need discussions among partners,
from home, the division of time and space between work and
working online makes the process lengthened as an
family life is not as the same as before. For example, family
experience: “With my job, I have to discuss a lot with co-
time such as mealtime or bedtime, wake up time are all later
workers to be able to come up with a plan and policy on
than usual.” (PV15). Or even another situation: “Because it’s
human resource for the company. WFH makes the working
very difficult to arrange time to take care of two children, I
process longer and more complicated and I get more
have to send one child to my parents to take care of.” (PV8).
complaints.” (PV12). Or in the job that they need to consult
for the customers about persuade them to buy products, it • Individual-related disadvantages
seems more difficult to get the deal: “With WFH I have to
consult the customers over the phone, the ability to finalize a More stress: 11/15 women supposed that WFH made them
deal is even more difficult, so my job is affected a lot.” (PV5). feel more stressed, the reasons can come from the lack of
social communication “mentally, there is no communication
Family to work conflict: 9/15 mentioned this cons. At the between colleagues, only communication between family
same time, women have to finish their work role and do members, so when problems happen at work, I get angry more
housework and take care of their children. This fact makes easily. And with the same task, it makes me more stressed than
them cannot concentrate on their work continuously like at before.” (PV12). Or the cause could come from taking care of
workplace. The explanation is: “Because I am a person with children by themselves all the time or having to do much
young children studying online, for example, when I am housework as these three references: “In fact, being at home
working, but my child is studying, I also have to support him. with children too much is also stressful” (PV2); “WFH affects
That makes my work interrupted.” (PV1). Another reason was
357
my personal emotions a lot, I am more stressed because the exchange and approval by many parties.” (PV1); “If I
having to monitor my children’s studies and activities” (PV6); have a choice, I will choose to return to normal work, because
“I often feel stressed because I have to take on many things at with the specificity of my job I have to go to the office.
the same time, job tasks and family duties.” (PV7). Furthermore, at the bank’s headquarters, I can support
customers better.” (PV14). Or the reason is about the
Worse health condition: 4/15 participants believed that availability of equipment: “I will not choose WFH, I prefer to
their health was affected by WFH. This consequence is a result work at the company. At the company, everything is available,
of the expansion of work: “Some days I have to join meetings so it’s more convenient for me and I can concentrate on work
from 11.00pm to 6.00pm, it happens continuously, it really more.” (PV2). And the last reason is about social interaction:
affects my physical condition because during the meeting, I “In my opinion, I probably won’t choose WFH and I also want
can’t stand up, sit down or go out.” (PV12). Or “When to go out to interact with people.” (PV4).
working from home, I do not have time to take a nap. When
going to work normally, after having lunch I usually spend 4/15 chose WFH, because of the job nature: “If my
about 30 minutes taking a nap, but when working from home, children go back to school, I will choose to work from home
I do not have time to sleep. At noon, my children are free so because I think that working from home is very effective for
they are very noisy, by the time they return to online class or my job characteristics and I love working online.” (PV7).
no longer make noise, it’s time to work .” (PV14). Others chose this alternative because WFH helps them save
the life expense and have safety and more time for their
Table 1. Number of references of WFH advantages and children: “After a period of separation and having to work
disadvantages from home, I have started to adapt to the current conditions
Number of References and I have discovered that working from home is a very good
Participants way of working, saving time and money and travel expenses.
Disadvantages Advantages At the same time, it is safer because the distance from my
PV1 8 5 house to work is very long. On the way, there is a lot of traffic.
Less travel will ensure my life safer, help save money on gas
PV 2 4 3 and do not get smoke and dust. At home, I can take care of my
children. Everything can be flexible because the technology is
PV 3 4 4
modern now, so as long as there is a good internet connection,
PV 4 4 3 it’s fine.” (PV11).
PV 5 4 3 3/15 wanted to combine WFH and work at the company.

“If I have rights to choose, I want to adopt WFH partly, 50-
PV 6 7 3 50 for example, then I can go to workplace to handle needed
tasks in the morning and in the afternoon, I can stay at home
PV 7 3 6 and work with online data.” (PV15). Or another supposed
PV 8 3 5 “With my job characteristics, I can change to WFH about 20-
40% .” (PV12).
PV 9 3 3
IV. DISCUSSION
PV 10 2 3
A. WFH advantages and disadvantages
PV 11 6 7 Most of the advantages are compliant with previous
studies like: Flexible working time, better performance, Save
PV 12 2 3 the living expense, Family members are more connected.
PV 13 7 6 These benefits were mentioned by references [2, 9, 13]. But
there are some advantages that have not been pointed out such
PV 14 6 6 as: Saving the time of traveling and preparing, Comfortable
while working, Better health conditions. All these new
PV 15 3 5 findings can be explained that normally, makeup before going
Total 66 65 to work and moving from home to companies take a lot of time
and face air and noise pollution. Besides, wearing formal
Table 1 compares the number of references when the clothes during working hours, and having to keep decent
participants mentioned the advantages and disadvantages of behaviour make them sometimes feel uncomfortable at work.
WFH. It is clear that the weights of them are almost the same The disadvantages of WFH during the pandemic were
(66 and 65 times the participants mentioned the drawbacks shown: Ineffective communication, Family to work conflict,
and the benefits). Insufficient facilities, Expansion of work, Increased
C. Aspirations about WFH workload. All these drawbacks of WFH are compliant with
studies [5, 6, 7, 14, 15]. Besides, there are some other findings
To find out which one overweighs, disadvantages or
have been pointed, such as: Lower income, Family conflicts;
advantages of WFH, the participants were asked when they
Family’s routines are unexpectedly changed. Due to the
have chances to choose, they would pick WFH or not, and the
unique circumstance of the pandemic, all these cons are
results as below:
understandable. The decline in turnover of businesses has
8/15 participants would not choose WFH, the job resulted in lower income for employees. As a result of the
characteristic was shown as the most important reason for pandemic, all the members of families stayed at home working
their choice: “I will not choose WFH because my job needs and studying all the time, this could cause conflicts if one of
358
them did not spend time helping others. And the expansion of Depending on the characteristics of jobs, companies can
work made the routines of families change unexpectedly. make the decision about the extent to which employees can
work from home, such as: 20%, 30%, 40%,… as a statement
B. The disadvantages overweigh the advantages from interviewees “With my job characteristics, I can change
About work – related pros and cons, there are four to WFH about 20-40% .” (PV12). Employees also should
advantages were pointed out: Flexible working time, Save the have the rights to choose which level of WFH they will adopt.
time of traveling and preparing, Better performance, and Because each individual has a different ability to keep the
Comfortable while working; and six disadvantages are balance between work and personal life and different family
mentioned: Lower income, Increased workload, Ineffective situations. We can see this as some interviewees supposed that
communication, Family to work conflict, Insufficient WFH made their health better meanwhile some others said it
facilities, Expansion of work. made their health worse.
With family – related benefits and drawbacks, there are Another factor should be taken into account when
two advantages: Family members are more connected, Save designing WFH is the seasonality of industries. For example,
the living expense; and two disadvantages: Family conflicts; in an interview – who is a manager of Human Resource
Family’s routines are unexpectedly changed. Department stated that “I would like to design one or two
Mentioning individual – related consequences of WFH, weeks for WFH in June and July so employees could work at
there is an advantage: Better health condition; and two home to create better family relationship.” (PV12).
disadvantages: More stress; Worse health condition. When a part of the work is done distantly, performance
Looking at the number of pros and cons, there are 7 appraisal needs to be re-designed as well. As an interviewee –
positive consequences of WFH while 10 negatives ones; a Sales and Marketing Manager concerned “Well, the thing I
looking at the number of references, 66 times were the cons can see most clearly is that I cannot control the employees’
mentioned by the interviewees and 65 times for pros; The performance as well as before. As a manager, I have to
aspirations show exactly how interviewees evaluate how manage my subordinates directly, it ensures more
WFH affects their life. More than a half of interviewees (8/15) effectiveness. Working from home like this, I only meet them
don’t want to continue with WFH, 3/15 want to have a when we have online meetings. The rest of working time, I
combination of WFH and working at the workplaces. 4/15 cannot control the things they are doing.” (PV10). Instead of
want to work from home totally. All the evidence shows the managing by process, managers could change to managing by
fact that, the negative side of WFH overweighs the positive objective to create a better condition for adopting WFH.
one. One of the work – related disadvantages is Increased
During the pandemic, WFH was compulsory, employees workload because the employees had to deal with
did not have other choices. They had to stay at home and technologies and problem of accessing the needed companies’
worked distantly via the Internet. This was really difficult for data. Therefore, when companies supply WFH as an option
women who had to take care of young children. Primary for employees, they need to invest in digitalization so that the
school pupils who studied online needed supervision from related data can be accessed at anytime and anywhere. And
their parents; kindergarten pupils staying at home needed official training courses should be supplied as well.
taking care of; housework needed to be done like cooking, V. CONCLUSION
cleaning up while the work of the companies needed to be
finished. The challenge of arranging everything caused much WFH has two sides: positive and negative. With
stress as in Vietnam, most of above things are believed Vietnamese women who have to take care of young children,
women’s responsibilities. WFH during the pandemic caused more cons than pros. The
challenge of arranging time for work and housework makes
C. The necessary of job re-designing after the them stressed. But the advantages of WFH cannot be denied.
pandemic After the pandemic, some drawbacks will be eliminated or
WFH during the pandemic is a practical test that helps all reduced, to take advantage of the benefits of WFH, as a part
the company have the chance to redesign jobs. Although the of work-life balance policy for women, WFH can become an
cons overweigh the pros, nearly a half (7/15) of interviewees optional part of work.
wanted to have a combination between WFH and working at
the workplaces or WFH only. Furthermore, after the ACKNOWLEDGMENT
pandemic, there are some drawbacks will be eliminated or Thanks for funding from Ho Chi Minh City University of
reduced like: Lower income; Family to work conflict; Family Technology and Education, Vietnam (HCMUTE). This work
conflicts; Family’s routines are unexpectedly changed; More belongs to the project grant No: T2021-09GVT. The
stress; and Worse health condition. Because the economic researchers also want to thank 15 participants for their
conditions will be recovered, all the social activities will be precious time taking part in our interviews to provide data for
back to normal, children will be back to schools and the study.
kindergartens. Demand of WFH now becomes clearer. To take
advantages of positive effects of WFH like: Flexible working REFERENCES
time, Save the time of traveling and preparing, Better [1] ANTD.VN. Cả nước đã có 55 triệu xe máy. An ninh Thủ Đô.
performance, and Comfortable while working; Family 2018 Mar 30; Available at: https://anninhthudo.vn/ca-nuoc-
members are more connected, Save the living expense; Better da-co-55-trieu-xe-may-post348806.antd.
health condition companies could deliberate on redesigning [2] Baard N, Thomas A. Teleworking in South Africa and
jobs. challenges. SA Journal of Human Resource Management.
2010, 8(1), 1–10.
359
[3] Collins AM, Hislop D, Cartwright S. Social support in the [9] Nakrošienė A, Bučiūnienė I, Goštautaitė B. Working from
workplace between teleworkers, office‐based colleagues and home: characteristics and outcomes of telework. International
supervisors. New Technology, Work and Employment, 2016, Journal of Manpower, 2019.
31(2), 161–75. [10] Reshma P, Aithal P, Acharya S. An empirical study on
[4] De Vries H, Tummers L, Bekkers V. The benefits of Working from Home: A popular e-business model.
teleworking in the public sector: Reality or rhetoric? Review International Journal of Advance and Innovative Research.
of Public Personnel Administration, 2019, 39(4), 570–93. 2015, 2(2).
[5] Felstead A, Henseke G. Assessing the growth of remote [11] Sarbu M. The role of telecommuting for work-family conflict
working and its consequences for effort, well‐being and work‐ among German employees. Research in Transportation
life balance. New Technology, Work and Employment, 2017, Economics, 2018, 70, 37–51.
32(3), 195–212. [12] Song Y, Gao J. Does telework stress employees out? A study
[6] Flores MF. Understanding the challenges of remote working on working at home and subjective well-being for wage/salary
and its impact to workers. International Journal of Business workers. Journal of Happiness Studies. 2020, 21(7), 2649–68.
Marketing and Management (IJBMM), 2019, 4(11), 40–4. [13] Sullivan C. Remote working and work-life balance. In: Work
[7] Fonner KL, Stache LC. All in a day’s work, at home: and quality of life. Springer, 2012, 275–90.
Teleworkers’ management of micro role transitions and the [14] O’Kane P, Walton S, Ruwhiu D, University of Otago, Work
work–home boundary. New Technology, Work and Futures Otago group. Remote working during COVID-19:
Employment, 2012, 27(3), 242–57. New Zealand national survey. [Internet]. 2020 [cited 2022 Jun
[8] Nguyễn L. Bình đẳng giới để hướng đến phát triển bền vững. 17]. Available at:
Báo điện tử Chính phủ nước Cộng hoà Xã hội chủ nghĩa Việt https://ourarchive.otago.ac.nz/handle/10523/10211.
Nam. 2021 Nov 19; Available at: http://baochinhphu.vn/Phat- [15] Wheatley D. Work‐life balance, travel‐to‐work, and the dual
trien-ben-vung/Binh-dang-gioi-de-huong-den-phat-trien-ben- career household. Personnel Review. 2012.
vung/453472.vgp
360
Study on Determining the Freezing Mode of Frozen

Fillet Bigeye Tuna (Thunnus obesus)
Dzung Tan Nguyen Linh Khanh Thuy Do
Chemical and Food Technology Faculty Chemical and Food Technology Faculty
Hochiminh City University of Technology and Education Hochiminh City University of Technology and Education
Hochiminh City, Vietnam Hochiminh City, Vietnam
tandzung072@hcmute.edu.vn dtklinh@hcmute.edu.vn
Tel: +84918801670
Chuyen Van Hoang Tuan Thanh Chau

Chemical and Food Technology Faculty Soctrang Vocational College
Hochiminh City University of Technology and Education Soctrang Province, Vietnam
Hochiminh City, Vietnam chauthanhtuan@gmail.com
chuyenhv@hcmute.edu.vn Tel: +84989298529
Abstract: Bigeye tuna (Thunnus obesus), a kind of Water (wet)

Protein (g) Lipid (g) Ashes (g) Energy (kcal)
(g)
delicious seafood, can be processed to be several valuable
products. To maintain the product quality, harvested tunas 72.6 ± 0.61 23.4 ± 0.56 1.8 ± 0.24 2.2 ± 0.45 105 ± 0.5
had been strictly persevered, transported and frozen at low
temperature. This study was carried out to determine the Table 2. Chemical ingredients and energy in 100g fillet bigeyes
technological mode of the freezing process of the fillet tuna tuna
to find the optimum temperature and freezing time to reduce Mineral Vitamin
mass loss and keep its quality. The combining two-level Calcium Phosphor Ferrous Sodium Potassium A B1 B2 PP C
(mg) (mg) (mg) (mg) (mg) (g) (mg) (mg) (mg) (mg)
orthogonal arrays was used to build the relationship between
objective functions and income variables. Results found the 68 482 2.4 7.4 6.8 162 0.32 0.53 20.4 0.12
optimized freezing mode of the fillet tuna is that the freezing
0and 0showed that the bigeye tunas are rich of nutrients,
environment temperature is –42.5oC, and the freezing time is and are the suitable environment for the growing and
2.12h. Carried out the experiment with found freezing mode developing of bacteria. Thus, during post-harvesting, if it is
showed that the temperature of fillet tuna at the end of the not having suitable methods for preservation, the quality of the
freezing process reaches –22.5oC and the yield of mass loss tuna will be rapidly reduced by the effect of internal enzymes,
is 3.1% meaning all amount of product internal water is domestic or contaminated bacteria. These effects lead to the
completely crystallized and the quality loss is negligible. The debasement of using and economical values of tunas. To
freezing mode can be applied to manufacturing to process the preserve fillet bigeye tunas for domestic consumption and
frozen fillet tuna. exporting purpose, the freezing method is commonly used to
preserve the tunas in seafood processing factories, because
Keywords: freezing, bigeye tuna, thunnus obesus, optimal only this method can lower the rate of quality decreasing of
freezing process, fillet tuna the products [3, 4].
I. INTRODUCTION However, when using the freezing method for tuna
Bigeye tuna (Thunnus obesus), commonly known as Bò preservation, some of the following problems should be
Gù (Vietnamese local name), is a large sea fish belonging to noted:
the Scombridae (mackerel) family [1], living in the warm sea, Firstly, the products at the end of the freezing process have
185 km far from the coastline. In Vietnam, the tuna lives to reach an optimum temperature [3, 4]. At that time, all liquid
mainly in Binh Dinh, Phu Yen and Khanh Hoa provinces [2]. water inside the products has been completely crystallized
The tuna, a kind of delicious seafood, having big and then the shelf-life has been extended. At this temperature, the
nutritious eyes, can be processed into various delicious dishes operating freezing systems can be stopped to save energy and
and made to be several valuable products for domestic product costs.
consumption or for exporting to oversea markets. The
harvested tunas had been strictly persevered and transported Secondly, if the freezing rate is low, the size of crystalized
to factories [2], in which the tunas were filleted to be plate water inside the tissue of products will be large lead to
pieces in size of 300×150×20 mm (length×width×thickness), destroying the texture of cells, when the frozen products are
the tuna chemical ingredients were shown in 0and 0 defrosted, the fluid in cells will be leaked, and it causes the
large mass loss of products. Then, their quality will be
Table 1. Chemical ingredients and energy in 100g fillet Bigeye decreased because nutrition in the cell fluid will be loosed. If
tuna the freezing rate is quick or super quick, the water inside the

products is crystallized at its position, and the size of the 1) Temperature measuring
crystallized water is small in micro or nano size, it does not The temperature of the freezing environment was
destroy the texture of cells. When the products are defrosted, measured by temperature sensors, and calculated
their quality are kept as that of initial. But quick frozen or automatically by a computer [7, 8]. The temperature of the
super quick frozen have to be carried out in the deep low surface and center of the frozen fillet tuna fish were measured
temperature, lower than –40oC, since the freezers will spend a by the Dual Digital Thermometer. To measure the center
large amount of energy. These are the unexpected problems temperature of the fillet tuna products at the end of freezing,
occurring in manufacturing [3, 4]. the lenght of the PT100 sensor was placed paralelled to the
Thus, the problem that needs to be solved is how to build length of the frozen piece so that the sensor was inside of the
the optimum freezing temperature and how to minimize. The center position of the piece.
freezing time making products – the frozen fillet bigeye tunas
in size of 300 mm × 150 mm × 20 mm – reach the optimum
temperature, making all water inside the products is
crystallized, and making the mass loss of post-defrosted
frozen products is lowest.
According to some authors [3, 5, 6], crystallizing
temperature of fillet tuna internal water is –1.24oC, and the
average temperature of crystallized water inside the final
frozen fillet tunas are –22.5oC – the Eutectic point of internal
water of products. There are the fundamentals to build the
optimized mathematical problem to find the freezing
technology of the frozen fillet bigeye tuna.
Figure 2. Freezing system DL-3
A. Materials
Whole harvested bigeye tunas caught from Binh Dinh to Average final temperature of the frozen fillet tuna pieces
was determined by the following formula [3, 4]:
Khanh Hoa, were stored at –45 to –40oC environment, then
they were transported to factories. At the factories, they were Ts + Tc
filleted and cut to be plat pieces in size of 300mm length, Y1 = Ta = (1)
2
150mm width, and 20mm thick (see 0), each piece was put
into a PE bag to avoid loss of water and cold burning on the Where: Ta, Ts, Tc (oC) average, surface and center
surface, put into trays and frozen [7, 8, 9]. temperatures of the fillet tuna products at the end of freezing,
respectively.
B. Equipment
Scale Sartorius, Basic Type BA310S ((Sartorius, 2) Defrosting of frozen fillet tuna
Germany): accuracy class: F1 (Guided by International Frozen fillet tuna pieces were defrosted by 3.5 MHz
Organization Of Legal Metrology), Dual Digital ultrasonic at 28oC cooling chamber for 45 minutes. The
Thermometer – range −50  70oC, readability ± 0.05oC – ultrasonic was regularly turned on and off every minute.
(Omron, Japan), the freezer DL-3 – build by the Faculty of 3) Mass loss determining
Chemical and Food Technology, Ho Chi Minh city University Mass of the pre-frozen fillet tuna products, and that of
of Technology and Education, Vietnam (0) – controllable post-defrosted frozen fillet tunas were determined by the Scale
freezing temperature: lowest −50oC, the cooling system (Sartorius), and the yield of mass loss of post-defrosted frozen
controlled by computer [3]. fillet tuna products was calculated by formula [3, 4]:
m1 − m 2 m
Y2 = .100% = .100% (2)
m1 m1
Where: m1, m2 (g) mass of the pre-frozen and post-
defrosted frozen fillet tuna products, respectively.
Y2 (%) Yield of mass loss of post-defrosted frozen fillet
tuna products.
4) Building the freezing mode.
Using the combining two-level orthogonal arrays to build
the relationship between objective functions (Y1, Y2)
including average final temperature of frozen fillet tuna
products (Y1, oC), and the yield of mass loss of post-defrosted
frozen fillet tuna products (Y2, %); and technological factors
Figure 1. Frozen fillet bigeye tuna pieces (X1, X2) directly affecting to the process such as: temperature
of freezing environment (X1, oC), time of freezing process (X2,
C. Methods h), and the thickness of the fillet tuna plat (X3, mm). Because
In this study, some methods had been used: the thickness of the bigeye tuna is standardized as X3 = 20mm,
362
X3 had been omitted to the variable sheet. Thus, the equation was tested by Fisher test [10]. Results built
relationship between objective functions (Y1, Y2) and mathematical modeling of experiments describing the
technological factors (X1, X2) is described by the following freezing process of the fillet tuna pieces.
formula:
Table 3. Levels of technological factors affecting freezing process
b x +  b x x +  ( )
k k k
of fillet tuna pieces
Y = b0 + j j ji j i b jj x 2j − (3)
j=1 j i; j=1 j=1 Levels of experiment
Variance
Where x1, x2 are variables coded from the real variables Factors Lower Medium Upper range
- +
X1, X2 as following formula: (-1)
level, level, level,
(+1) Zi
(-1) (0) (+1)
Xi – X i 0
xi = ; Xi = xi.Xi + Xi0 (4) Temperature of freezing
-45 -45 -40 -35 -35 5
Xi environment, X1 (oC)
Freezing time, X2 (h) 0.4 0.4 2.2 4.0 4.0 1.8
Where: Xi0 = (Ximax + Ximin)/2; Xi = (Ximax – Ximin)/2;
Ximin  Xi  Ximax ; i = 1÷2 Table 4. The combining two-level orthogonal arrays and the
experiment results finding objective function (Y1, Y2) of the
Number of experiment was identified by: freezing process of fillet tuna pieces
N = nk + n* + n0 = 2k + 2k + n0 = 9 (5) Objective
Real
Where: k = 2; Coded variable function
variable
Number of values
nk = 2k = 22 = 4; experiments
X1, X2,
oC
x0 x1 x2 x1x2 x12 – 2/3 x22 – 2/3 Y1 Y2
n* = 2k = 2×2 = 4; n0 = 1 h
Swing arm of orthogonal matrix was identified: 1 –35 4.5 +1 +1 +1 +1 0.333 0.333 –27.77 5.15
= N.2
( k −2) − 2( k −1) = ( 2−2) − 2( 2−1) = 1
9.2 (6) 2k
2 –45 4.5 +1 –1 +1 –1 0.333 0.333 –36.38 4.74
3 –35 0.5 +1 +1 –1 –1 0.333 0.333 –2.19 1.52

Conditions of empirical orthogonal matrix:
4 –45 0.5 +1 –1 –1 +1 0.333 0.333 –2.83 0.99
1 k
( 2 1 2 2
) (
 = N 2 + 2 = 9 2 + 2.1 = 2/3 ) (7) 5 –35 2.5 +1 +1 0 0 0.333 –0.667 –20.45 3.61
6 –45 2.5 +1 –1 0 0 0.333 –0.667 –24.34 3.15

5) Other measurements. 2k
In addition, mathematical and information technical tools 7 –40 4.5 +1 0 +1 0 –0.667 0.333 –31.24 4.65
had been used to solve the optimized problems describing the
8 –40 0.5 +1 0 –1 0 –0.667 0.333 –2.77 1.03
freezing process, and to identify the technical models of the
freezing process of fillet tuna pieces. n0 9 –40 2.5 +1 0 0 0 –0.667 –0.667 –21.93 3.19
III. RESULTS AND DISCUSSIONS

• Temperature of fillet tuna products at the end of freezing
D. Building the mathematical models of the freezing process process
of fillet tuna
The average final temperature of fillet tuna products at the 𝑌1 = 𝑓1 (𝑥1 , 𝑥2 ) = −18.88 + 2.2𝑥1 − 14.59𝑥2 +
2 2
end of the freezing process (Y1, oC), and the yield of mass loss 2.01𝑥1 𝑥2 − 0.34 (𝑥12 − ) + 5.04(𝑥22 − ) ()
3 3
of post-defrosted frozen fillet tuna products are dependent on
the freezing process, and directly affected by technological
factors such as: the temperature of freezing environment (X1, • Yield of mass loss of post-defrosted frozen fillet tuna
o
C), time of the freezing process (X2, h). products
Technological factors X1 and X2 were individually 𝑌2 = 𝑓2 (𝑥1 , 𝑥2 ) = 3.11 + 0.231𝑥1 − 1.829𝑥2 −

surveyed to find the extreme domain of Y1, Y2. The vital 2 2
0.03𝑥1 𝑥2 + 0.24 (𝑥12 − ) − 0.31(𝑥22 − ) ()
surveying domain was found and shown in 0 3 3
The experiments, with levels of technological factors X1, After testing the compatibility by Fisher test, the results
X2 in Table 3, were carried out with the combining two-level confirmed that the mathematical models (8), (9) are totally
orthogonal arrays shown in 0to build the mathematical model compatible with the experimental data shown in 0 Thus, they
Y1 = f1(X1, X2) = f1(x1, x2) và Y2 = f2(X1, X2) = f2(x1, x2) can describe the freezing process of the fillet tuna.
describing the freezing process. Results identified Y1, Y2
depended on x1, x2 shown in the column “Objective function Simulating the mathematical models (8) and (9), the
values” of 0 results received responsive or curved surfaces Y1 = f1(x1, x2)
and Y2 = f2(x1, x2) shown in 0 and 0.
MS Excel 2019 (Microsoft, US) was used to calculate the
experimental data in 0, to calculate coefficients, bj and bji. The The results showed that the relationship between objective
significant coefficients of regression equations (3) were tested functions (Y1 (oC) and Y2 (%)) and the technological factors
by Student’s test, and the compatibility of the regression (x1 and x2) directly affecting the freezing process, is the
363
quadratic nonlinear. It is more complicated than that of first- Based on technical conditions answering the requirements
order linear. Normally, the quadratic nonlinear function of economy and engineering, the freezing process of the fillet
consists of two value ranges on two branches, one branch is tuna (in size of 300mm×150mm×20mm) had been carried out
covariant and the other is inverse depending on a quadratic as quick as possible, and not only whether all internal water of
coefficient. If the coefficient is negative, the parabola curve is the tuna is completely crystallized or temperature of the
convex. In contrast, the parabola curve is concave. The product reaches the optimum temperature TFopt = −22.5oC but
boundary point between these two branches is the extreme also the quality of post-frozen products have to be paid
pitch of the parabola. attention [9, 11]. To solve this problem, the optimized
problem had been established as follow: Let's determine the
optimized problem xopt = (x1opt, x2opt)  x = {-1  x1, x2  1}
to:
 Y = f (x opt , x opt ) = T
 1 1 1 2 Fopt = −22.5

 Y2 min = f 2 (x1 , x 2 ) = Min f 2 (x1 , x 2 )
opt opt

x = (x1 , x 2 )   x = {−1  x1, x 2  1}

(10)
2) Solving the optimized problems describing freezing
process of fillet bigeye tuna.
Solving the optimized problem (10) by Lagrange method
Figure 3. Relationship between frozen fillet tuna pieces and as follow:
technological factors: temperature of freezing environment, and
freezing time L ( x1, x 2 , h ) = f 2 ( x1, x 2 ) + h. ( f1 ( x1, x 2 ) + 22.5 ) (11)
Find extremes by Lagrange method (11) as follow:

L ( x1, x 2 , h )
 =0
 x1
 L ( x , x , h )
 1 2
=0
 x 2 (12)

 L ( x1, x 2 , h ) = 0
 h

 − 1  x1, x 2  1

0.231 − 0.03x + 0.048x
 2 1
Figure 4. Relationship between yield of mass loss of post-  + h  ( 2.2 + 2.01x 2 − 0.68x1 ) = 0
defrosted of the frozen fillet tuna products and technological 
factors: temperature of freezing environment, and freezing time  – 1.829 − 0.03x1 − 0.62x 2

 + h  ( −14.59 + 2.01x1 + 10.08x 2 ) = 0 (13)
The formula (8) showed the coefficient 5.04 linking to x22, −18.88 + 2.2x –14.59x + 2.01x x
thus it is misunderstood that if freezing time (x2) is prolonging  1 2 1 2
- time increase, the temperature of fillet tuna products at the   2   2
 0.34   x12 –  + 5.04   x 22 −  = 0
end of freezing process will be increasing; It is the wrong  3  3

situation. In fact, the freezing process, assessed in the negative −1  x1, x 2  1
temperatures (x1), is belonging to the inverse brand. It means
that when the freezing time is increasing, the product With meshing algorithm under the supporting of Visual
temperature during the freezing process will be decreasing. Basic 6.0 to solve equations (13), the results found that roots
See 0. Similarly, the model (9) showed the quadratic of the equations (10) together with Lagrange multiplier
coefficient −0.31 linking to x22, thus the freezing process is coefficient, h = 0.109, are:
assessed in the covariant brand of parabola, which means that
when the time is increasing, the Yield of mass loss of post- x1opt = −0.502; x2opt = −0.044; (14)
defrosted frozen fillet tuna products will be increasing, see 0. Converting to real variable by equation (4) received:
E. Building and solving the optimized problems describing X1 = −42.5oC; X2 = 2.12h (15)
freezing process of fillet bigeye tuna Replacing the optimal roots x1opt and x2opt into equations
1) Building the optimized problems describing freezing (8) and (9), found:
process of fillet bigeye tuna. Y1 = −22.5oC; Y2min = 3.1% (16)
364
F. Verification of the found optimized problems described water is crystallized. When the temperature of the product
the freezing process of fillet tuna reduces to −1.24oC, some amount of water is crystallized, and
Carrying out the Freezing of the fillet tuna (in size of the crystallized water will separate from the solution. It
300mm×150mm×20mm) on the freezing system DL-3 at the increases the solution concentration, and leads to a decrease in
optimized freezing modes: the environment temperature is the crystallized temperature. Thus, when the temperature of
−42.5oC, and kept constantly during freezing process, and the products decreases to −22.5oC, all internal water in the
freezing time is 2.12h. Temperature of products and yield of products is crystallized (This point is called optimal freezing
mass loss were measured as described in II. Materials AND temperature or the Eutectic point). However, crystallized
METHODS. At the end of the freezing process, the results water is only free and physio-chemically bonded water,
are followed: the average temperature of the center products chemically bonded water is nearly not crystallized.
is −22.57oC, Yield of mass loss of post-defrosted frozen fillet
tuna products is 3.19%. It can be seen that the optimized
freezing process model of fillet tuna is completely consistent
with reality. This technical condition, thus, can be applied to
real manufacturing to preserve the fillet tuna pieces to serve
for domestic consumption and export. Besides, in this study,
the fillet tuna had been frozen in another mode that differs
from the optimized mode: the temperature of freezing
environment −38.5oC to make the temperature of the products
at the end of the freezing process reach −22.5oC then all
internal water of products is crystallized, and the freezing time
is 3.95h. At that time, yield of mass loss of post-defrosted
frozen fillet tuna products is 6.79%.
Figure 6. Dynamics graph of the freezing process fillet tuna pieces
built by experiment
IV. CONCLUSION
This research found the optimized freezing mode of the
fillet bigeye tuna (in size of 300mm length, 150mm width,
and 20mm thick): environment temperature is −42.5oC and
the freezing time is 2.12h. When this mode was applied for
freezing, the temperature of fillet tuna at the end of the
freezing process reaches −22.5oC and the yield of mass loss
Figure 5. The SEM image of ice crystal structure in frozen fillet is 3.1%. It means that all internal water of the product is
tuna in size of 300mm×150mm×20mm. A. The freezing mode: crystallized and the quality loss is negligible. This technology
temperature of freezing environment, -42.5oC, and freezing time, mode, thus, can be applied to manufacturing to serve
2.12h; B. The freezing mode: temperature of freezing environment, domestic trade and exporting purposes.
-38.55oC, and freezing time, 3.95h. The images A and B were
scanned and captured at the same magnification (1000×) REFERENCES
[1] J. B. Graham and K. A. Dickson, "Tuna comparative
What is the reason why when the freezing environment physiology", The Journal of Experimental Biology., vol. 207,
temperature is −38.5oC, and the freezing time is 3.95h, the no. 23, pp. 4015–4024, 2004.
yield of mass loss of post-defrosted frozen fillet tuna products [2] A. D. Lewis, "The tuna fisheries of Vietnam - an overview of
increases to 6.79%. The reason is that in the low rate freezing available information", in 1st Meeting of the Scientific
process, ice crystals in the products are formed in a large size Committee of the Western and Central Pacific Fisheries
Commission WCPFC–SC1, Noumea, New Caledonia, August
(see 0B), then they do destroy the texture of cell membranes 2005.
and tore the tissue membranes [11, 12]. When the products are
[3] N. T. Dzung, The method to determine the rate of freezing
defrosted to consume, the melting process of ices inside the water inside product, 1st ed., Chisinau: Lambert Academic
products will attract cell fluid, lead to mass and nutrition loss Publishing, 2015.
will increase. When freezing at the optimized mode, the mass [4] N. T. Dzung, T. V. Dzung and T. D. Ba, "Building The Method
loss is very low and considered negligible. The reason is that To Determine The Rate of Freezing Water of Penaeus
ices inside the products are formed with a small size (see 0A), Monodon", Carpathian Journal of Food Science and
they cause tiny injury to cell membranes. Then, when the Technology, vol. 4, no. 2, pp. 28-35, 2012.
products are defrosted, nutrients are maintained as initial, and [5] A. C. Cleland and R. L. Earle, "A new method for prediction
of surface heat transfer coefficients in freezing," Bulletin.
the frozen products are high quality. Annex-International Institute of Refrigeration (IIR).
Fig. 6. showed that the crystallizing temperature of water Refrigeration Science and Technology, vols. 1976-1, pp. 361-
368, 1976.
in the fillet tuna products is −1.24oC, not 0oC, the reason is [6] J. Holman, Heat Transfer, 10th, Ed., New York: McGraw –
that product internal water is not pure water, it is solution Hill, 2009.
form. The crystallizing temperature of the pure solvent is not [7] Nguyen Tan Dzung, "Building the Method and the
been changing during the freezing process, but that of solution Mathematical Model to Determine the Rate of Freezing Water
is being decreased continuously during the process [11, 12] to inside Royal Jelly in the Freezing Process", Research Journal
reach the Eutectic point where all amount of product internal of Applied Sciences, Engineering and Technology, vol. 7, no.
2, pp. 403-412, 2014.
365
[8] S. E. Charm and J. Slavin, "A method for calculating freezing [11] A. Banin and A. M. Duwayne, "Effects of salt concentration
time of rectangular packages of food", Annex Bull. Inst. Int. changes during freezing on the unfrozen water content of
Froid, p. 567 – 568, 1962. porous materials", Water Resources Research, vol. 10, no. 1,
[9] D. R. Heldman, "Food properties during freezing", Food pp. 124-128, 1974.
Technol, vol. 36, pp. 92-96, 1982. [12] G. H. Haugvalstad, D. Skipnes and M. Sivertsvik, "Food free
[10] R. Fisher, "Tests of Significance in Harmonic Analysis", from preservative", Journal of Food Engineering, vol. 30, pp.
Proceedings of the Royal Society of London. Series A, 124-142, 2005.
Containing Papers of a Mathematical and Physical Character,
vol. 125, no. 796, pp. 54-59, 1929.
366
Numerical Investigation of the Optimum Operating

Condition in Magnetically Confined Plasma with
Sheared Slab Ion-Temperature-Gradient Model
Thanh Tinh Tran
Thermal Engineering Department, Faculty of Vehicle and Energy Engineering,
Ho Chi Minh City University of Technology and Education,
tinhtt@hcmute.edu.vn
Abstract: In magnetically confined plasma, heat flux velocity shear is introduced to represent one of major
reduction, a desirable characteristic to sustain high- mechanisms to enhance plasma temperature, such as neutral
temperature condition for fusion reaction in tokamak, was beam injection. Due to simplicity of slab geometry,
observed experimentally in literature. However, underline simulations in this paper are much more less expensive than
mechanism of this advantage is still obscure. In this work, we the above works which base on complex global gyrogluid or
performed a computational study of a sheared slab ion- gyrokinetic simulations.
temperature-gradient turbulence to investigate the impact of Numerical results clearly show that increasing parallel
parallel flow fluctuation on heat flux in sheared magnetic ′
mean velocity shear 𝑢∥0 leads to heat flux reduction. The
field. It is found that heat flux decreases with increasing ′
minimum value of heat flux 𝑄 obtains at 𝑢∥0 = 0.6, then it
parallel flow fluctuation up to the onset of parallel-velocity- ′
switches to increasing trend as 𝑢∥0 surpasses the value of 0.6.
gradient turbulence. However, as parallel flow fluctuation This threshold value corresponds to the onset of parallel-
surpasses that threshold value, heat flux switches to an velocity-gradient turbulence which can be found from linear
increasing trend. That implies an optimum operating analysis. By using energetic analysis technique which firstly
condition for confinement enhancement in tokamak. Stronger introduced in our previous work [4], it is found that zonal flow
zonal flow formation is also observed for the entire range of enhances during the period of heat flow reduction. Here, zonal
parallel flow fluctuation. Interestingly, careful energetic flow (ZF), a turbulence-generated mean flow in perpendicular
analyses show distinct behavior of a portion of energy source direction, has significant role in turbulence regulation [4-11].
used for zonal flow generation between before and after the The rest of this paper is organized as follows. In section II,
onset of parallel-velocity-gradient turbulence. That sheds we describe the model adopted in this work, along with
light on the heat flux reduction in tokamak. numerical schemes and conditions. Sections III devotes to
linear and nonlinear simulations for the onset of parallel
Keywords: sheared slab ion-temperature-gradient, zonal velocity gradient instability, heat flux reduction, and
flow, plasma turbulence, parallel velocity gradient, energetics analysis. We conclude the paper in section IV with
fluctuation energy a brief summary of the main results.
I. INTRODUCTION II. NUMERICAL MODEL
Fusion reaction in tokamak, a promising way to provide To start, we consider a ITG model [12] including a parallel
seemly unlimited energy for human demand, has been ′
flow shear 𝑢∥0 :
attracting numerous studies and internationally collaborated
works, e.g. the International Thermolnuclear Experimental
𝜕𝑡 𝑞 = (1 + 𝐾𝛻⊥2 )𝛻𝑦 𝜙 − [𝜙, 𝑞] + (𝛻𝑧 + 𝑠𝑥𝛻𝑦 )𝑢∥ + 𝐷𝑞  ()
Reactor project. In order to maintain a high temperature level
for self-sustain fusion reaction, confinement is a crucial point.
Better plasma confinement means there is less thermal flux
′
𝜕𝑡 𝑢∥ = 𝑢∥0 𝛻𝑦 𝜙 − [𝜙, 𝑢∥ ] − (𝛻𝑧 + 𝑠𝑥𝛻𝑦 )(𝜙 + 𝑝) + 𝐷𝑢  ()
which escapes from the fusion plasma in tokamak sytem.
𝜕𝑡 𝑝 = −𝐾𝛻𝑦 𝜙 − [𝜙, 𝑝] − Γ(𝛻𝑧 + 𝑠𝑥𝛻𝑦 )𝑢∥ + 𝐷𝑝  ()
Heat flux reduction from 30% up to 60% has been reported
experimentally in [1] by investigating the JET tokamak.
Gyrokinetic simulations of Barnes [2] also showed a big drop The above equations base on slab geometry whose coordinates
in heat flux. Recently, Ko et al. [3] presented a reduction of (x, y, z) denote the radial, poloidal, and parallel direction,
40% in ion heat conductivity from global gyrofluid respectively. We use magnetic field 𝐵 ⃗ = 𝐵0 (𝑧̂ + 𝜌𝑠 𝑠𝑥𝑦̂)
𝐿𝑛
simulations of ion-temperature-gradient (ITG) turbulence with shear parameter 𝑠. Here, potential vorticity is defined by
with parallel mean velocity shear in a tokamak. All these 𝑞 = 𝛻⊥2 𝜙 − (𝜙 − ⟨𝜙⟩), where the zonal part of potential 𝜙 is
works indicate clear heat flux reduction phenomenon.
1 𝐿𝑦 𝐿𝑧
However, underlined mechanism of this beneficial ⟨𝜙⟩ = ∫ ∫0 𝜙 𝑑𝑦𝑑𝑧. (4)
𝐿𝑦 𝐿𝑧 0
characteristic is still obscured.
In frame of this work, we perform numerical simulations
a slab ITG turbulence with parallel flow shear to shed light on
the heat flux reduction in tokamak. Here, parallel mean

𝑑 ln 𝑛0 −1 nonlinear simulations are performed to show heat flux

And 𝐾 = (1 + 𝜂𝑖 )𝑇𝑖 /𝑇𝑒 , 𝜂𝑖 = 𝐿𝑛 /𝐿 𝑇 , 𝐿𝑛 = − ( ) ,
𝑑𝑥 reduction for a wide range of parallel flow shear. The
𝑑 ln 𝑇 −1
𝐿𝑇 = − ( 𝑖
) , Γ = 𝛾𝑇𝑖 /𝑇𝑒 , with 𝑇𝑖 and 𝑇𝑒 are ion underline mechanism of this reduction is revealed from
𝑑𝑥 energetics analysis.
temperature and electron temperature, respectively. Equations
𝐿 𝑒𝜙 𝐿 𝑢 A. PVG onset
(1-3) were normalized as follow 𝑛 → 𝜙 ; 𝑛 ∥ → 𝑢∥ ;
𝜌𝑠 𝑇𝑒 𝜌𝑠 𝑐𝑠
𝑝 𝐿𝑛 𝑇𝑖 𝜌𝑠 𝑥 𝑦 𝑧 To study the onset condition of the PVG instability, one
→ 𝑝; Ω𝑐𝑖 𝑡 → 𝑡 ; → 𝑥; → 𝑦; → 𝑧 . And for can perform simple linear simulations. All nonlinear terms,
𝑝0 𝜌𝑠 𝑇𝑒 𝐿𝑛 𝜌𝑠 𝜌𝑠 𝐿𝑛
sake of simplicity, tilde has been removed from fluctuated i.e. Poisson bracket terms in righ hand side of (1-3), are
variables 𝑢̃∥ and 𝑝̃ in (1-3). The last term on righ hand side of neglected in linear simulations. Profile of growth rate 𝛾 as a
′
(1-3) are dissipation terms which simulate damping in plasma: function of parallel flow shear |𝑢∥0 | is plotted in Fig. 1.
𝐷𝑞 = 𝜇⊥ 𝛻⊥2 𝑞 − 𝜇ℎ 𝛻⊥4 𝑞 + 𝜇∥ 𝛻∥2 𝑞 − 𝜇𝑍𝐹 𝑞𝑍𝐹 , 𝐷𝑢 = 𝜇⊥ 𝛻⊥2 𝑢∥ − Growth rate values are almost similar when |𝑢∥0 ′
| are small.
𝜇ℎ 𝛻⊥4 𝑢∥ + 𝜇∥ 𝛻∥2 𝑢∥ , 𝐷𝑝 = 𝜇⊥ 𝛻⊥2 𝑝 − 𝜇ℎ 𝛻⊥4 𝑝 + 𝜇∥ 𝛻∥2 𝑝 ; with However, different behavior of growth rate 𝛾 are observed
𝜕 𝜕 𝜕 𝜕 ′
𝛻⊥ ≃ 𝑥̂ + 𝑦̂ ; 𝛻∥ = + 𝑠𝑥
. The Poisson bracket since |𝑢∥0 | surpasses the value of 0.6. Hence we can confirm
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑦 ′
operator is defined as [𝜙, 𝑓] = −∇𝜙 × 𝑏̂ ∙ ∇𝑓 where 𝑏̂ is a |𝑢∥0 | = 0.6 is the threshold of instability driven by parallel-
unit vector along the magnetic field. flow-gradient turbulence.
′
In equation (2), 𝑢∥0 = 𝑑𝑢∥0 /𝑑𝑥 represents an equilibrium B. Heat flux reduction
parallel flow shear where 𝑢∥0 is the background parallel flow. We carry out nonlinear simulations to investigate effects
′ of parallel flow shear on heat flux level, i.e. confinement, of
Here we assume that 𝑢∥0 is constant in this work for simplicity
purpose. We perform nonlinear simulations using a wide fusion plasma. Each simulation is carefully checked in order
range of |𝑢∥0 ′
| = 0 ~ 1.5 which includes the onset and to ennsure a steady solution. Fig. 2a, b, c show spatio-temporal
′
instability of parallel-velocity-gradient turbulence. Detailed contours of total potential 𝜙 for parallel flow shear |𝑢∥0 | = 0,
description will be made in section III. 0.6 , 1.5 , respectively. After a short initial period, almost
steady zonal flow structures are clearly observed in each case.
The diffusion coefficients 𝜇⊥ in perpendicular direction ′
Fig. 2 also indicates strongly effect of parallel flow shear |𝑢∥0 |
and 𝜇∥ in parallel direction are adopted as 𝜇⊥ = 10−3 and on zonal flow structures for both amplitude and pattern.
𝜇∥ = 0.05 . These values are in the estimated ranges for Higher |𝑢∥0′
| provides higher amplitude of ZF. This implies
perpendicular collisional process and Landau damping, whose that the more PVG energy is provided, the stronger ZF is
detailed description can be found in our previous work [4]. In generated. While number of patterns does not monotonely
order to mimic real tokamak plasmas, we also employ ′
neoclassical poloidal flow damping for zonal flow dissipation, change as |𝑢∥0 | increases. For example, if we consider the
𝜇𝑍𝐹 = 0.01 is selected following Garbet et al [13]. For number of peaks of potential in radial direction in Fig. 2, there
′
numerical stability, the hyper-diffusion coefficient 𝜇ℎ = are 5, 3, and 4 peaks for the case of |𝑢∥0 | = 0, 0.6, and 1.5,
10−6 is used. respectively.
We perform numerical simulations (1-3) by implementing
them in the BOUT++ framework [14]. Numerical schemes are
chosen as the same as in [4] with the fourth order central
difference method in derivetives and the second Arakawa
scheme in calculating the Poisson bracket. The numerical
grids in x, y, z directions are 𝑛𝑥 = 300 , 𝑛𝑦 = 256 , 𝑛𝑧 =
Growth rate 𝛾
120, with the resolution of 𝑑𝑥 = 𝑑𝑦 = 𝑑𝑧 = 0.2. Periodic

boundary conditions are imposed in all directions. The
cyclone base case parameters [15] with 𝜂𝑖 = 3.144 , 𝑠 =
′
0.25, and Γ = 5/3 are used for all simulations in this work. |𝑢∥0 |
III. RESULTS AND DISCUSSION Figure 1. Growth rate from linear simulations including
This section firstly figures out the onset of parallel velocity dissipation terms in (1-3)
gradient (PVG) instability through linear simulations. Then
(a) (b) (c)

′ ′ ′
Figure 2. Spatio-temporal contours of total potential 𝜙 when parallel flow shear |𝑢∥0 | = 0 (a); |𝑢∥0 | = 0.6 (b); |𝑢∥0 | = 1.5 (c)
368
Fig. 3 shows profiles of heat flux 𝑄 = ⟨𝑢̃𝑥 𝑝̃⟩ and parallel −⟨𝜙𝛻∥ 𝑢∥ ⟩ represents a transfer rate between perpendicular
Reynolds stress Π𝑥∥ = ⟨𝑢̃𝑥 𝑢̃∥ ⟩ as functions of parallel flow and parallel energies. The parallel dynamics enhances
′ (reduces) the perpendicular fluctuation energy if 𝑆𝑐𝑜𝑢𝑝 is
shear |𝑢∥0 |. To produce this figure, we take temporal average
during a saturated state. Parallel Reynolds stress increases positive (negative). In the other hand, the compressional
monotonely as parallel flow shear increases. It is obviously source which show a transfer rate between ion temperature
since stronger PVG source, i.e. parallel flow shear, is dynamics and parallel energies is defined by 𝑆𝑐𝑜𝑚𝑝 =
introduced. The same result is also remarked in [4]. However, ⟨𝑝𝛻∥ 𝑢∥ ⟩ = −⟨𝑢∥ 𝛻∥ 𝑝⟩ . Again, a positive 𝑆𝑐𝑜𝑚𝑝 (negative
heat flux has different behavior. Firstly, heat flux decreases as 𝑆𝑐𝑜𝑚𝑝 ) means that ion temperature dynamics enhances
parallel flow shear increases. The reduction of heat flux (reduces) parallel fluctuation energy. The source of PVG and
′
around 11% is observed as |𝑢∥0 | increases from 0 to 0.6 . ITG are defined by 𝑆𝑃𝑉𝐺 = 𝑢∥0 ′
Π𝑥∥ and 𝑆𝐼𝑇𝐺 = −𝐾⟨𝑝𝛻𝑦 𝜙⟩ =
′
When parallel flow shear surpasses the value |𝑢∥0 | = 0.6 𝐾⟨𝑢𝑥 𝑝⟩, respectively.
which closes to the onset of PVG instability, the heat flux then
switches to an increased trend. The behavior of heat flux The perpendicular energy in (5) may in turn separate into
shown in fig. 3 is called heat flux reduction. That implies an ̃⊥ and a zonal part 𝑊𝑍𝐹 as follows
a fluctuated part 𝑊
optimum operating condition in tokamak that corresponds to
the minimum heat flux. ̃⊥ = ∫ 𝑆𝑐𝑜𝑢𝑝 𝑑𝑉 − ∫ 𝑆𝑍𝐹 𝑑𝑉 − 𝐷𝑊
𝜕𝑡 𝑊 ̃ ⊥, (8)
Similar behaviors of heat flux reduction were also reported
in literature for both experimentally and numerically. 𝜕𝑡 𝑊𝑍𝐹 = ∫ 𝑆𝑍𝐹 𝑑𝑉 − 𝐷𝑍𝐹 , (9)
Bonanomi et al [1] observed a heat flux reduction from 30%
to 60% in ABC tokamak. Gyrokinetic simulations [2] where 𝑊 ̃⊥ = 𝑊⊥ − 𝑊𝑍𝐹 , 𝑆𝑍𝐹 = −𝑢𝑍𝐹 Γ𝑃𝑉 represents ZF
presented a big drop in heat flux. Recently, Ko et al. [3] with source, Γ𝑃𝑉 = ⟨𝑞𝑢𝑥 ⟩ , and 𝑢𝑥 = −𝜕𝜙/𝜕𝑦 . The process of
global gyrofluid simulation of ITG turbulence presented a deducing (8, 9) is analogous to the way mentioned in [4].
reduction of 40% in ion heat conductivity. Heat flux reduction Equations (5-9) present the whole picture of energetics which
in our preliminary results is less than the above studies, clearly show how free energy PVG and ITG are introduced
however, only a simple slab ITG model is used in our work. into parallel then to perpendicular dynamic, as well as to zonal
flow formation.
Fig. 4 plots fluctuated, zonal, and total energies as
functions of parallel flow shear. Larger free energy of PVG
obviously excites more fluctuated energy that makes stronger
Fluxes
zonal flow structures, i.e. zonal energy. Increasing trend of

zonal energy in fig. 4 is consistent with larger zonal amplitude
observed in Fig. 2. It should be noted that unlike the results in
our previous study [4] which bases on 3D coupled drift wave-
ion parallel flow, profile of fluctuated energy in fig. 4 shows
unclear onset of PVG instability.
′
|𝑢∥0 | ′
As parallel flow shear |𝑢∥0 | increases, both ZF source 𝑆𝑍𝐹
and coupling energy 𝑆𝑐𝑜𝑢𝑝 are found to increase but at
different rates as mentioned in fig. 5. The above energetics
analysis show that the perpendicular fluctuation gains energy
Figure 3. Profiles of heat flux and momentum flux from 𝑆𝑐𝑜𝑢𝑝 source, then transfers an amount of energy to 𝑆𝑍𝐹
for ZF generation as appearing in (9). The efficiency of ZF
C. Energetics analysis production can be systematically defined by
This subsection aims at energetics analysis of fluctuations
to elucidate a distribution mechanism of the PVG and ITG ∫ 𝑆𝑍𝐹 𝑑𝑉
driven free energy into parallel and perpendicular dynamics, 𝛽= , (10)
∫ 𝑆𝑐𝑜𝑢𝑝 𝑑𝑉
as well as a portion energy goes to zonal flow formation. By
multiplying (1) with – 𝜙 , (2) with 𝑢∥ , and (3) with 𝑝 then which represents how large amount of energy is used for the
taking a volume integral, one can obtain the energy balance ZF generation from a given total fluctuation energy received
equations: from coupling. In other word, 𝛽 could be considered as a
𝜕𝑡 𝑊⊥ = ∫ 𝑆𝑐𝑜𝑢𝑝 𝑑𝑉 − ∫ 𝜙𝐷𝑞 𝑑𝑉, (5) measure of the turbulence regulation efficiency. Fig. 5 also
′
plots 𝛽 as a function of |𝑢∥0 |. On can observe 𝛽 has different
′
𝜕𝑡 𝑊∥ = ∫ 𝑆𝑃𝑉𝐺 𝑑𝑉 − ∫ 𝑆𝑐𝑜𝑢𝑝 𝑑𝑉 + ∫ 𝑆𝑐𝑜𝑚𝑝 𝑑𝑉 + trend as |𝑢∥0 | goes through the onset threshold, and maximum
′
∫ ∥ ∥ 𝜙𝐷𝑞 𝑑𝑉,
𝑢 𝐷 (6) value of 𝛽 is obtained at |𝑢∥0 | = 0.6. That is consistent with
the heat flux reduction results in Fig. 3.
𝜕𝑡 𝑊𝑝 = ∫ 𝑆𝐼𝑇𝐺 𝑑𝑉 − ∫ 𝑆𝑐𝑜𝑚𝑝 𝑑𝑉 + (1/Γ) ∫ 𝑝𝐷𝑝 𝑑𝑉. (7) IV. CONCLUSION
1 1 In this work, we perform linear and nonlinear simulations
Here, 𝑊⊥ = ∫⟨𝜙 2 + |𝛻⊥ 𝜙|2 ⟩ 𝑑𝑉 , 𝑊∥ = ∫⟨𝑢∥2 ⟩ 𝑑𝑉 , and of the slab ITG turbulence model with parallel flow shear to
2 2
𝑊𝑝 = (0.5/Γ) ∫⟨𝑝2 ⟩ 𝑑𝑉 are the perpendicular, parallel, and investigate the underline mechanism of heat flux reduction. A
ion thermal energies, respectively. 𝑆𝑐𝑜𝑢𝑝 = ⟨𝑢∥ 𝛻∥ 𝜙⟩ = wide range of parallel flow shears are employed.
369
multi-scale interactions in the JET tokamak”, Nucl. Fusion,

vol. 58, pp. 124003, 2018.
[2] M. Barnes, F. I. Parra, E. G. Highcock, A. A. Schekochihin, S.
Energies
C. Cowley, and C. M. Roach, “Turbulent Transport in

Tokamak Plasmas with Rotational Shear”, Phys. Rev. Lett.,
vol. 106, pp. 175004, 2011.
[3] S. Ko, S. S. Kim, H. Jhang, and J. Kim, “Global gyrofluid
simulation of ITG turbulence by parallel mean velocity shear
in tokamak”, 4th Asia-Pacific Conference on Plasma Physics,
26-31 Oct., pp. MF1-I39, 2020, remote econference.
[4] T. T. Tran, S. S. Kim, H. Jhang, and J. Kim, “Turbulence
′
|𝑢∥0 | characteristics, energy equipartition, and zonal flow generation
in coupled drift wave-parallel velocity gradient driven
turbulence”, Plasma Phys. Control. Fusion, vol. 61, p. 065002,
2019.
Figure 4. Profiles of fluctuated, zonal, and total energy [5] Y. Kaneda, M. Yokokawa, K. Itakura, and A. Uno, “Energy
dissipation rate and energy spectrum in high resolution direct
numerical simulations of turbulence in a periodic box”, Phys.
Fluids, vol. 15, pp. L21, 2003.
[6] W. Horton, “Drift waves and transport”, Rev. Mod. Phys., vol.
𝑆𝑍𝐹 , 𝑆𝑐𝑜𝑢𝑝 /20, 𝛽
71, pp. 735-778, 1999.

[7] C. Connaughton, S. Nazarrenko, and B. Quinn, “Rossby and
drift wave turbulence and zonal flows: the Charney-Hasegawa-
Mima model and its extensions”, Phys. Rep., vol. 604, pp. 1-
71, 2015.
[8] A. Hasegawa and K. Mima, “Pseudo-three-dimensional
turbulence in magnetized nonuniform plasma”, Phys. Fluids,
vol. 21, p. 87, 1978.
[9] A. I. Smolyakov, P. H. Diamond, and V. I. Shevchenko, “Zonal
′
|𝑢∥0 | flow generation by parametric instability in magnetized
plasmas and geostrophic fluids”, Phys. Plasmas, vol. 7, p. 1349,
2000.
Figure 5. Profiles of ZF source 𝑆𝑍𝐹 , coupling energy 𝑆𝑐𝑜𝑢𝑝 , [10] G. Manfredi, C. M. Roach, and R. O. Dendy, “Zonal flow and
and 𝛽 streamer generation in drift turbulence”, Plasma Phys.
Controlled Fusion, vol. 43, p. 825, 2001.
An important new finding in this paper is that the [11] T. T. Tran, H. Jhang, J. Kim, and S. S. Kim, “A characterization
efficiency of ZF generation is maximum at the PVG instability of the inertial range in forced-damped Hasegawa-Mima
turbulence”, Phys. Plasmas, vol. 24, p. 102318, 2017.
onset threshold. 𝛽 increases at low parallel flow shear, i.e.
′ [12] S. Hamaguchi and W. Horton, “Fluctuation spectrum and
0 < |𝑢∥0 | ≤ 0.6, and decreases at high parallel flow shear transport from ion temperature gradient driven modes in
′
|𝑢∥0 | > 0.6 . Due to important role of ZF in turbulence sheared magnetic fields”, Physics of Fluids B: Plasma Physics,
regulation, the above finding reveals underline mechanism of vol. 2, pp. 1833, 1990.
heat flux reduction. That, therefore, implies an optimum [13] X. Garbet, C. Bourdellel, G. T. Hoang, P. Magnet, S.
Benkadda, P. Beyer, C. Figarella, I. Voitsekovitch, O. Agullo,
operating condition in tokamak. and N. Bian, “Global simulations of ion turbulence with
magnetic shear reversal”, Phys. Plasmas, vol. 8, p. 2793, 2001.
ACKNOWLEDGMENT [14] B. Dudson, M. Umansky, X. Xu, P. Snyder, and H. Wilson,
We acknowledge Drs. SungSik Kim, Hogun Jhang, “BOUT++: A framework for parallel plasma fluid
Juhyung Kim for their great supports. This research used simulations”, Comput. Phys. Commun., vol. 180, p. 1467,
2009.
resources of the KAIROS computing system from Korean
Institute of Fusion Energy. [15] A. M. Dimits, G. Bateman, M. A. Beer, B. I. Cohen, W.
Dorland, G. W. Hammett, C. Kim, J. E. Kinsey, M.
Kotschenreuther, A. H. Kritz, et al., “Comparisons and physics
REFERENCES basis of tokamak transport models and turbulence simulations”,
[1] N. Bonanomi, P. Mantica, J. Citrin, T. Gorler, B. Teaca, and Phys. Plasmas, vol. 7, pp. 969–983, 2000.
JET Contributors, “Impact of electron-scale turbulence and
370
Using Brake and Engine Torque to Control Traction

on Either Side of the Drive Wheel
Tan Tai Phan Van Nhu Tran
Dept. of Mechanical and Dynamic Dept. of Automobile Engineering
School of Engineering and Technology Faculty of Mechanical Engineering
Tra Vinh University University of Transport and Communications
Tra Vinh, Vietnam Hanoi, Vietnam
phantantaickd@tvu.edu.vn vannhu.tran@utc.edu.vn
Abstract: Traction created at the drive wheels is a result [5] and applied on cars with active 4-wheel drive (4WD)
of the dynamic transformation from the engine through the according to the requirements of the wheel torque differential
powertrain to the wheels and acting on the road surface. The system [6]. A traction control unit that only focuses on
traction is limited by the grip-ability between the wheel and controlling engine torque cannot maintain vehicle stability in
the road. The value of traction also depends on the value of complex road conditions. Therefore, the brake pressure
the engine torque output. When the engine torque is too high, controller should be combined with the engine torque
the wheels rotate too quickly while the grip-ability of the road controller [8,9]. Liqiang Jin et al. [10] presented a new traction
is limited, resulting in the wheels turning too fast. If the wheel control system by adjusting engine torque and wheel brake
pressure. The motor torque regulation is based on the
slips, the traction will be down by its decreased grip. Another
optimization of the parameters of the proportional–integral–
hand, the traction on both sides of the drive wheels complies
derivative (PID) controller and the wheel brake pressure
with the law distribution of the active differential. This
regulator is based on the optimized proportional-integral (PI)
distribution rule also depends on the grip-ability tire with the controller. Control in complex conditions such as low slippery
road on either side of the drive wheel, the lower grip-ability pavement and uneven pavement on either side.
of the side wheel, and the lower differential that distributes
traction torque for that side wheel. In this study, the authors However, all of these approaches apply input-output
present a method of controlling the traction at the drive linearization to the widely used 2-state model, which describes
wheels by using the braking torque impacts on the more the longitudinal speed of the vehicle and only the wheel
slipped wheel with the other and reducing traction engine rotational speed. In this paper, the authors apply the LQR
torque as needed. It is a linear-quadratic-regulator (LQR) control method to control traction on both sides of the drive
controller with two outputs such as making the braking torque wheel in the condition that the traction force on both sides of
the wheel is different to improve the engine torque and
acting on the wheel that slips and controlling the engine
traction quality at the drive wheel, helping the vehicle to move
torque by an act on the throttle mechanism. Simulation results stability in a straight line.
under specific conditions show high control accuracy, stable
wheel traction control, and improved engine torque. II. BUILDING A VEHICLE DYNAMIC MODEL
Keywords: traction, differential, LQR controller, braking A. The engine model

torque, traction control The engine is a very complex system. Many aspects affect
an engine's torque output, but the most important are throttle
I. INTRODUCTION position (𝜃𝑡ℎ ) and crankshaft speed (𝑛𝑒 ) . When vehicle
Traction at the drive wheel is limited by the force of grip movement, there is a delay between throttle input and torque
between the wheel and the road. If the grip force on the two output; This time delay issue should be treated. Therefore, the
sides of the wheel is different, the traction force between them engine model is used the inertial medium to establish. In this
will be unequal according to the law of torque distribution of paper, the dynamic model of the engine can be described
the differential. The differential distributes more traction to according to (1). Where: 𝑀𝑐 is the resistance moment of the
the low-grip side of the wheel, increasing that wheel's slip and powertrain, 𝑀𝑒 is the torque of the engine, 𝑡𝑒1 is the impact
loss of traction. In addition, when the wheels slip on the road, force time, 𝑡𝑒2 is the coefficient of adaptation to the kinematic
engine power is lost due to the friction between the road and parameter, s is the coefficient of Laplace transforms, 𝜃𝑡ℎ is the
the tire. To control the traction generated on either side of the throttle opening angle, 𝑛𝑒 is the engine crankshaft rotational
wheels to match their grip with the road and ensure equal speed. The output torque of the engine depends on the throttle
wheel speeds for stable straight-line vehicle movements, opening angle and rotational speed.
improving drivability and performance of the vehicle, it is 𝑒 −𝑠.𝑡𝑒1
necessary to have a traction control unit on both sides of the 𝑀𝑐 = 𝑀𝑒 . (1)
𝑡𝑒2 .𝑠+1
drive wheel. This controller controls the frictional forces
between the wheels and the road on either side to distribute 𝑀𝑒 = 𝑓(𝜃𝑡ℎ , 𝑛𝑒 ) (2)
traction to suit the road conditions [1,2]. Design the engine- B. The tire model
based traction controller according to the input-output
linearization method, including the torsional dynamics of the The mathematical model proposed and selected is the tire
powertrain on the automotive design model [3,4,7]. A traction model by Pacejka [11]. It is based on the physical mock-up of
control system has also recently been applied on electric cars the tire which is used to simulate real-time wheel dynamics.

Pacejka found a mathematical expression describing the tire's semi-axes. 𝑀0 is the torque on the pinion gear of the main
grip on the road surface according to (3). transmission; 𝑀𝑝1 , 𝑀𝑝2 is the braking torque acting on the
𝐶.ar𝑐𝑡𝑔 left and right wheels; 𝑇3 , 𝑇4 is the elastic and resistance
𝐹𝑥 (𝜆) = 𝐷. 𝑆𝑖𝑛 ( 𝐵. 𝜆 ) (3) moments of the left and right semi-axes; 𝑇5 , 𝑇6 is the traction
(𝐵. 𝜆 − 𝐸. ( ))
−ar𝑐𝑡𝑔(𝐵. 𝜆) torque on the left and right wheels.
In which: B, C, D, and E are characteristic quantities of the Then the pinion gear torque of the main
grip force curve. These quantities depend on the vertical transmission 𝑀0 = 𝑀𝑒 𝑖ℎ𝑖 𝜂𝑡𝑙 , which 𝑖ℎ𝑖 is the gear
reaction of the road acting on the wheel. Parameters B, C, D,
transmission ratio; 𝜂𝑡𝑙 is the transmission efficiency of the
and E are affected by: the longitudinal stiffness of the tire(𝐶𝑥 ),
the value of longitudinal grip force at full and maximum powertrain. Assuming the wheel model does not contact the
sliding state, curved shape of the grip characteristic curve. road surface, the component of grip torque is replaced by
torque 𝑇5 = 𝑘1 𝜑̇ 5 ; 𝑇6 = 𝑘2 𝜑̇ 6 , where the coefficient
The coefficient of longitudinal slip is determined by (4) as
follows: 𝑘1 , 𝑘2 represents the different grip characteristics on both
sides of the drive wheel. The powertrain dynamics equation
|𝜔.𝑟𝑏𝑥 −𝑉𝑥 |
𝜆= (4) takes into account the differential effect as (5).
𝑚𝑎𝑥{𝜔.𝑟𝑏𝑥 ,𝑉𝑥 }
𝑎 𝑎
Where: 𝜆 is the coefficient of longitudinal slip; 𝜔 is the (𝐴.(𝑀0 . 21 −𝑇3 )−𝐵.(𝑀0 . 21 −𝑇4 ))
angular speed of the wheel; 𝑉𝑥 is the linear velocity at the 𝜑̈ 3 = (𝐴2 −𝐵2 )
wheel center; 𝑟𝑏𝑥 is the average working radius of the wheel. 𝑎 𝑎
(𝐵.(𝑀0 . 21 −𝑇3 )−𝐴.(𝑀0 . 21 −𝑇4 ))
C. The dynamic model of the body 𝜑̈ 4 = (𝐵2 −𝐴2 ) (5)
The vehicle model nonlinear fifteen degrees-of-freedom (𝑇3 −𝑇5 −𝑀𝑝1 )
𝜑̈ 5 =
𝐽5
(DOF) is used to anticipate vehicle response under road
(𝑇4 −𝑇6 −𝑀𝑝2 )
conditions and different driving maneuvers. The traction 𝜑̈ 6 =
{ 𝐽6
control system controls the longitudinal dynamics of the
vehicle but ignores some factors [10]. In this object, the 𝐽0 𝑎12 𝐽1
𝐴= + + 𝐽3 + 2𝐽2 𝑎22 ;
kinematic model of the vehicle body only considers the 4 4
𝐽0 𝑎12 𝐽1
factors affecting the longitudinal and rolling motion of the With: 𝐵= + − 2𝐽2 𝑎22 (6)
4 4
wheel. It is hypothesized that considering the elements of the 𝑇3 = 𝑐3 (𝜔3 − 𝜔5 ) + 𝑘3 (𝜑3 − 𝜑5 );
transmission system from the engine to the pinion shaft to be 𝑇4 = 𝑐4 (𝜔4 − 𝜔6 ) + 𝑘3 (𝜑4 − 𝜑6 )
rigid, the mass moment of inertia of the driving elements is
reduced to the pinion gear of the main transmission. The Where: 𝑎1 is the transmission ratio of the main
powertrain model that takes into account the influence of the transmission; 𝑎2 is the ratio between the semi-shaft gear and
differential is shown in Fig. 1 [2]. the planetary gear. 𝜑3 ÷ 𝜑6 are respectively the rotation
angle of the left half shaft gear, right half shaft gear, and the
semi-shaft left and right.
III. TRACTION CONTROL LQR CONTROLLER

DESIGN
The active wheels’ torque exceeds the maximum grip
force between the tire and the road surface by a full-throttle
acceleration of the driver, causing both active wheels to slip.
Then we need to reduce the engine torque so that the traction
force at the wheel does not exceed the grip force on the road.
On the other hand, when the vehicle moves on complicated
Figure 1. The dynamic model of the powertrain includes roads, the roughing of the surface road on both sides of the
differential wheel is not the same, leading to different grip forces on both
sides, making the vehicle move unstable; At this time, the
Where: 𝐽0 ÷ 𝐽6 the roster is the equivalent mass moment wheel brake control lever turns faster to keep the traction on
of inertia of the pinion gear, the passive gear, the planetary both sides balanced. Using high-precision control algorithms
gears, the left half-axle gear, the right half-axle gear, the left to reduce wheel rotation by reducing engine torque and acting
side of the semi-axle, and the drive wheel, the right side of brake torque at the wheel is a necessary effect.
the semi-axle and the drive wheel; 𝜔0 ÷ 𝜔6 the roster is the The LQR controller is built to control the angular speed
angular speed of the pinion gear, the passive gear, the deviation of the left (𝜔5 ) and right (𝜔6 ) drive wheels, and
planetary gears, the left half-axle gear, the right half-axle also control the engine torque during slippage on either side
gear, the left side of the semi-axle and the drive wheel, the of the drive wheel exceeding the allowable limit. The control
right side of the semi-axle and the drive wheel; 𝑘3 , 𝑘4 is the system diagram is shown in Fig. 2, in which: the signal output
equivalent stiffness coefficient of the left semi-axle and right (𝑢1 ) acts on the ABS brake actuator (Anti-lock Brake System)
to independently control the braking torque on both sides of
semi-axle; 𝑐3 , 𝑐4 is the drag coefficient of the left and right the drive wheel movement to ensure that when the automobile
372
is moving in a straight line, the angular speed of both wheels matrices are determined 𝐴𝑒 , 𝐵1𝑒 , and 𝐵2𝑢 according to (12,
is equal [12]; The output signal (𝑢2 ) acts on the throttle 13, 14).
control mechanism of the engine to reduce the torque
transmitted to the two drive wheels when the vehicle is −3242.86 3242.86 𝑇
𝐴𝑒 = [ ] (12)
moving on the road with a low coefficient of grip. In this 1622.97 −1622.97
simulation excluding the ABS brake actuator model, the 𝐵1u = [0 0.25]𝑇 (13)
signal 𝑢1 is the braking torque acting on one side of the wheel.
When the grip between the two sides of the wheel is different, 1.04 0 𝑇
𝐵2𝑢 = [ ] (14)
the wheel with a low coefficient of grip will rotate faster, it is 1.04 −0.25
necessary to apply a braking force to that wheel so that its The Q and R matrices are also selected, transformed as
angular speed reduces to balance the other one. When the slip- above, and using Matlab's tools to determine the K matrix as
on on both sides of the wheel is equal and is outside the limit follows:
of slip (𝜆 = 15% ÷ 30%), it is necessary to reduce the engine 𝑇
9
torque transmitted to the two drive wheels to properly 𝑄 = [2.10 0 ] (15)
distribute the traction force to both sides of the drive wheel, 0 2.109
helping the car motion is stable. 1 0𝑇
𝑅=[ ] (16)
0 1
13917.28 −21188.27 𝑇
⇒𝐾=[ ] (17)
32578.91 30247.30
IV. SURVEY RESULTS
Consider the case of the left wheel traveling on the road
with a low coefficient of grip, which is characterized by the
𝑘1,2 coefficient in the model in Fig. 1. The values 𝑘1,2 in the
survey cases are shown in Table 1.
Figure 2. LQR control diagram for automobile traction control Table 1. Case studies of the system
Assuming the left wheel moves on the road with a low Coefficients Cases
coefficient of grip, the system applies the brakes to the left 𝑘1 0 20 40 60 80
wheel, then the right wheel brake torque is 𝑀𝑝2 = 0. Set 𝑢1 = 𝑘2 100 100 100 100 100
𝑀𝑝1 the controller output, 𝑒 = [𝑒1 𝑒2 ]𝑇 where, 𝑒1 , 𝑒2
respectively, is the speed difference between the left semi-axle The parameters of the survey model are as follows: 𝑀𝑒 =
gear speed with the left wheel, and the right semi-axle gear 180 𝑁𝑚 , 𝑖ℎ1 = 4.25 ; 𝜂𝑡𝑙 = 0.96 ; a1 = 4.85; 𝑎2 =
speed with the right wheel, shown in (7, 8). On the other hand, 0.87 ; 𝐽0 = 0.002 𝑘𝑔𝑚2 ; 𝐽1 = 0.6 𝑘𝑔𝑚2 ; 𝐽2 =
when the vehicle is traveling on a road with a low coefficient 0.0003 𝑘𝑔𝑚2 ; 𝐽3 = 2.0035 𝑘𝑔𝑚2 ; 𝐽5 = 𝐽6 =
of grip on both sides of the wheel slipping, then the second 2
4.005 𝑘𝑔𝑚 ; 𝑘3 = 𝑘4 = 32,600 𝑁𝑚/𝑟𝑎𝑑 ; 𝑐3 = 𝑐4 =
output of the control unit 𝑢2 acts on the mechanism to reduce 6,500 𝑁𝑚𝑠/𝑟𝑎𝑑 [2].
the throttle opening angle, causing engine torque drops. With
the assumption that the elements in the powertrain from the D. In case of not using the LQR controller
engine to the pinion shaft are rigid, the torque 𝑀0 depends
When the controller is not used, the angular speed
on 𝑀𝑒 or 𝑀0 = 𝑓(𝜃𝑡ℎ , 𝑛𝑒 ).
difference between the two sides of the wheel is shown in Fig.
𝑒1 = 𝜑̇ 3 − 𝜑̇ 4 = 𝜔3 − 𝜔4 (7) 3. In case 𝑘1 = 0, the left wheel is completely slipped, and its
angular speed increases rapidly to 587.2 rad/s. At that time,
𝑒2 = 𝜑̇ 5 − 𝜑̇ 6 = 𝜔5 − 𝜔6 (8)
the right wheel cannot move because the engine torque is
Transform (5) into the form of the error vector's equation transmitted to the differential and consumed in the left wheel,
of state. resulting in the car not moving. The larger the 𝑘1,2 coefficient
𝑒̇ = 𝐴𝑒 𝑒 + 𝐵1𝑒 𝑀0 + 𝐵2𝑢 𝑢 (9) difference, the higher the angular speed difference between
the two sides of the wheel, resulting in different traction
The control objective to give 𝑒 → 0, according to the LQR torque on the two sides of the wheel, the left wheel receives
control method, the state feedback control law is defined as more torque from the engine. This causes a loss of engine
(10). In which the matrix 𝐾 is found to ensure that the power and instability when the vehicle is going straight. In
objective function 𝐽 according to (11) is minimized. this case, the braking torque needs to be applied to the left
𝑢(𝑡) = −𝐾𝑒(𝑡) (10) wheel to reduce its speed to the same as that of the right
∞ wheel. So a controller must be used and take care of this.
J= ∫0 (𝑒 𝑇 𝑄𝑒 + 𝑢 𝑇 𝑅𝑢) 𝑑𝑡 (11)
E. Using a single-parameter controller to control the
Where: 𝑄 ∈ ℝ𝑛×𝑛 and 𝑅 ∈ ℝ𝑚×𝑚 are the controller design braking torque at the wheel
parameter matrices, where𝑛 is the number of state variables In this case, the LQR controller is designed to control the
(𝑛 = 2), 𝑚 is the number of controller outputs (𝑚 = 2). The braking torque at the wheel with a higher angular speed when
matrix 𝑄 is a positive semi-deterministic symmetric matrix, the vehicle is traveling on a road with different coefficients of
which 𝑅 is a positive definite matrix. Based on the values of grip on either side of the wheel. The survey results shown in
the parameters of the model and the transformation, the Fig. 4 show that when the ability to grip the left wheel is lower,
373
the brake controller will send a signal 𝑢1 = 𝑀𝑝1 to the left In addition, using the brake controller caused the angular
wheel brake to keep 𝜔5 = 𝜔6 . Specifically 𝑘1 = 0 , the speed difference between the two sides of the wheel to
controller sends a signal to the brake actuator to brake the left decrease significantly as the grip increased, the lower the
wheel with a torque of 3,891 Nm. angular speed difference between them (Fig. 5a). Compared
to not using the brake controller, in this case 𝑘1 = 40 the
deviation is very small, only 0.037 rad/s, in the same time this
error reaches 56.44 rad/s when the brake controller is not used
(Fig. 5b).
F. Using a two-parameter controller to control brake
torque and engine torque
The vehicle moves on a road with a complex road surface,
both slippery and the grip on both sides is not uniform, then
the brake torque controller does not respond in time, so it is
necessary to coordinate the engine torque controller to help
Figure 3. The angular speed difference on both sides of the wheel the vehicle move more stable. Use the controller to coordinate
when 𝑘1,2 changing in the case of no LQR control brake and engine torque control to perform control in the case
of different grip on either side of the wheel. The survey results
in Fig. 6 show that the controller outputs a signal 𝑢2 to control
the engine torque to reduce the slip between the wheel and the
road surface. The signal 𝑢2 decreases more as the wheel slip
increases. When 𝑘1 = 40 the controller acts on the throttle
mechanism, and the torque transmitted to the pinion gear on
the drive shaft is reduced to 122.9 N.m in 0.033 s. The time
𝑢2 to act on the throttle mechanism is reduced as the
coefficient of the grip of the wheel on the road decreases (Fig.
6). This causes the engine torque to drop rapidly to avoid
power loss due to wheel slip.
Figure 4. Brake torque acting on the left driving wheel when

𝑘1 changing
Figure 6. Motor torque acting on the drive shaft when 𝑘1 changing

(a). Angular speed deviation on either side of the wheel when using the
brake controller
(b). Angular speed deviation with and without brake control

Figure 5. Angular speed deviation on either side of the driving Figure 7. Brake torque acting on the left driving wheel when
wheel 𝑘1 changing
374
of the two wheels is stabilized at 0.026 s (Fig. 8b). From the

above results, it is shown that the wheel angular speed
controlled by the two-parameter controller is not only much
smaller but also faster and more stable than the single-
parameter controller. This helps to prevent the wheel from
slipping and turning for a long time, causing a loss of power
from the engine to the wheel. At the same time, the two-
parameter controller also controls the wheel's angular speed
about 13 times smaller than the single-parameter controller,
and the angular speed deviation on both sides of the wheel is
also lower. Therefore, it can be said that the two-parameter
controller has high responsiveness, and improves engine
torque and traction quality at the drive wheel, helping the
vehicle to move in a straight line.
V. CONCLUSION
This paper presents a method of controlling the traction on
(a). Control 2 parameters 𝑢1 , 𝑢2 at 𝑘1 = 40 both sides of the wheels based on their ability to grip the road
surface. Thereby, the authors have built dynamic models from
the engine to the drive wheel, taking into account the influence
of the mechanical differential. The design of the LQR
controller combines the control of two parameters such as
wheel brake torque control and engine torque control. In
addition, the simulation results are compared with the results
of the single-parameter controller. Simulation results in
different cases have shown that the two-parameter controller
controls the engine torque to reduce the slip between the wheel
and the road surface, the less time it takes to act on the throttle
mechanism. when the coefficient of the grip of the wheel on
the road is lower. At the same time, the brake torque controller
also acts on the wheel. The larger the speed difference on
either side of the wheel, the higher the braking torque acting
on the wheel. The value of torque acting on the brake
mechanism and wheel angular velocity when using a two-
(b) One parameter control 𝑢1 at 𝑘1 = 40
parameter controller will be much smaller than that of a
single-parameter controller. The two-parameter controller is
Figure 8. Angle speed on both sides of the wheel when controlling highly responsive, improving engine torque and traction
2 parameters and one output parameter quality at the drive wheel, helping the vehicle to move in a
At the same time, the brake torque controller also acts on straight line. However, the new results stop at the simulation,
the wheel when there is a difference in angular speed between the next research towards experimentation on the model.
the two sides of the wheel. The larger the speed difference on
REFERENCES
both sides of the wheel, the higher the braking torque acting
on the wheel (Fig. 7). Like engine torque, the time 𝑢1 to act [1] V. Ivanov, D. Savitski and B. Shyrokau, “A survey of traction
on the brake actuator decreases as the wheel's coefficient of control and antilock braking systems of full-electric vehicles
adhesion to the road decreases. Furthermore, the value of with individually controlled electric motors”, IEEE
Transactions on Vehicular Technology, 64(9), pp. 3878–3896,
torque applied to the brake actuator when controlled by a two- 2015.
parameter controller is much smaller than that of a single-
[2] P.T. Tai, T.V. Nhu, and T.Q. Dung, “Using the Brake Torque
parameter controller. In the case of 𝑘1 = 40, the wheel-acting to Redistribute the Engine Power Transmitting to the Left and
brake torque is 173.7 Nm at 0.029 s time, much smaller than Right Drive Wheels”, Lecture Notes in Mechanical
using a single parameter controller is 2,317 Nm at 0.5 s time Engineering, Springer, Cham, 2021.
(Fig. 4). This proves that thanks to the combination of engine [3] A. Zech, T. Eberl, E. Reichensdörfer, D. Odenthal, and S.
torque control, the braking torque value is significantly Müller, “Method for tire developing slip controllers regarding
reduced and the control time is faster, meeting the a new cascaded controller structure”, Print 14th International
coordination requirements of the designed controller, Symposium on Advanced Vehicle Control (AVEC), 302–307,
simultaneously. It also helps the vehicle to be more stable 2018.
when traveling on slippery and uneven roads. [4] E. Reichensdörfer, D. Odenthal, and D. Wollherr, “On the
stability of nonlinear wheel-slip zero dynamics in traction
Under the same conditions 𝑘1 = 40 , the angular speed control systems”. IEEE Transactions on Control Systems
values of the two sides of the wheel are very different when Technology, 28(2), 489–504, 2020.
using a single parameter and two-parameter controller. In the [5] A. Zech, T. Eberl, C. Marx, and S. Müller, “Analysis of the
single-parameter controller, 𝜔5 = 39.34 𝑟𝑎𝑑/𝑠 , 𝜔6 = potential of a new control approach for traction control
39.30 𝑟𝑎𝑑/𝑠 , the angular speeds of the two wheels are consider a P2-hybrid drivetrain”. In 9th International Munich
stabilized at 0.5 s (Fig. 8a). In the two-parameter controller, Chassis Symposium 2018, 285–303, Springer, 2019.
𝜔5 = 2.980 𝑟𝑎𝑑/𝑠, 𝜔6 = 2.957 𝑟𝑎𝑑/𝑠, the angular speed [6] E. Reichensdörfer, W. Degel, D. Odenthal, and D. Wollherr,
“Nonlinear traction control design, stability analysis and
375
experiments for vehicles with on-demand 4WD torque bias [16] F. Cheli, M. Pedrinelli, F. Resta, G. Travaglio, M. Zanchetta,
systems”, In 2019 IEEE 58th Conference on Decision and and A. Zorzutti, “Development of a new control strategy for a
Control (CDC), 6669–6674, IEEE, 2019. semi-active differential for a high-performance vehicle”,
[7] E. Reichensdörfer, D. Odenthal and D. Wollherr, “Engine- Vehicle System Dynamics, vol. 44, no. sup1, pp. 202–215, Jan.
Based Input-Output Linearization for Traction Control 2006.
Systems”, IFAC PapersOnLine 53-2, 14055–14060, Elsevier,
2020. [17] H. Chou and B. D’andréa-Novel, “Global vehicle control
[8] G. Liu and L.Q. Jin, “A study of coordinated vehicle traction using differential braking torques and active suspension
control system based on optimal slip ratio algorithm”. Math forces”, Vehicle System Dynamics, vol. 43, pp. 261–284, Apr.
Probl Eng. 2016: 3413624, 2016. 2005.
[9] X. Ran, X. Zhao, J. Chen and C. Yang, “Novel coordinated [18] H.-Z. LI, L. LI, L. HE, M.-X. KANG, J. SONG, L.-Y. YU,
algorithm for traction control system on split friction and slope and C. W.U, “PID plus fuzzy logic method for torque control
road”, International Journal of Automotive Technology 17, in a traction control system”, International Journal of
817–827, 2016.
Automotive Technology, vol. 13, no. 3, pp. 441–450, Apr.
[10] L.Q. Jin, M. Ling and J. Li, “Development of a new traction 2012.
control system using ant colony optimization”, Advances in
Mechanical Engineering, Vol. 10(8), 1–12, SAGE, 2018. [19] M. Canale, L. Fagiano, M. Milanese, and P. Borodani,
[11] H.B. Pacejka, Tire and vehicle dynamics. 3rd ed. Oxford, “Robust vehicle yaw control using active differential and IMC
Butterworth Heinemann, 2012. techniques”, Control Engineering Practice, vol. 15, no. 8, pp.
[12] P.T. Tai and T.V. Nhu, “Study on traction control of cars when 923–941, Aug. 2007.
moving on the road with different coefficients of traction on [20] Y. Park and B. Kwak, “Slip controller design for a traction
both sides of the wheel”, Vietnam Mechanical Magazine, control system”, International Journal of Automotive
Special issue, Dec 2021, p.304-309, 2021. Technology, vol. 1, no. 1, pp. 48–55, 2000.
[13] S. Saha, H. P. Ikkurti, and S. Saha, “A robust slip-based [21] D. Rubin and S. Arogeti, “Vehicle yaw stability control using
traction control of an electric vehicle under different road rear active differential via Sliding mode control methods”, In
conditions”, paper ID: 66, DOI: 10.1049/cp.2015.1618, 2015. 21st, Mediterranean Conference on Control and Automation,
[14] H. Kim, S. Lee, and J. K. Hedrick, “Active yaw control for Platanias, Chania - Crete, Greece, pp. 317–322, Jun. 2013.
handling performance improvement by using traction force”, [22] K. Nam, Y. Hori, and C. Lee, “Wheel Slip Control for
International Journal of Automotive Technology, vol. 16, no. Improving Traction-Ability and Energy Efficiency of a
3, pp. 457–464, DOI: 10.1007/s12239-015-0047-9, Jun. 2015. Personal Electric Vehicle”, Energies, vol. 8, no. 7, pp. 6820–
[15] C. Annicchiarico, M. Ranchi, S. Pellari, and R. Capitani, 6840, Jul. 2015.
“Design of a Semi-Active Differential to Improve Vehicle
Dynamics”, in Volume 1: Applied Mechanics; Automotive
Systems; Biomedical Biotechnology Engineering;
Computational Mechanics; Design; Digital Manufacturing;
Education; Marine and Aerospace Applications, Copenhagen,
Denmark, p. V001T02A006, DOI: 10.1115/ESDA2014-
20157, Jul. 2014.
376
Fatigue Life of Accelerated Corroded Steel Plate

Dao Duy Kien Nguyen Thanh Hung Nguyen Thi Thu Hao
Ho Chi Minh City of Technology Ho Chi Minh City of Technology Ho Chi Minh City of Technology
kiendd@hcmute.edu.vn nthung@hcmute.edu.vn hao.nguyen@tcxd.edu.vn
Nguyen Van Hung Hai Dang Phan

VNU University of Engineering and VNU University of Engineering and
Hanoi, Vietnam Hanoi, Vietnam
vanhung.nguyen@vnu.edu.vn haidangphan.vn@vnu.edu.vn
Abstract: In this study, the qualitative investigation on the and provide the local aggressive environment conducive to
effect of corrosion, produced by cyclic corrosion test, on crack initiation. A sectioning method in a laboratory setting
fatigue behavior of corroded steel plate was investigated by has been successfully used for pit examination, shape
fatigue test and FEM analysis. The specimen model was classification, and fatigue life prediction.
exposed to a corrosive environment chamber following the The relationship between fractal dimension and fatigue
ASTM D 6899-03. The relationship between residual life was also studied by some experts and engineers [1-2].
measured thickness and fatigue life of corroded steel plate, Then, the effect of corrosion on fatigue behavior of corroded
the reduction of fatigue life due to corrosion for designation steel was also investigated by [3-8]. The present study was
of fatigue, and the trend of stress concentration factor Kt undertaken as a first step toward developing an integrated
versus time of corroded steel plate in contact with concrete experiment to qualitatively account for the effects of corrosion
was also studied. The stress concentration effect of corrosion on the fatigue behavior of steel plate in contact with concrete.
pits with various pit depth and diameter has been studied. In this study, the qualitative investigation on the effect of
corrosion, produced by cyclic corrosion test, on fatigue
Keywords: accelerated cyclic corrosion test, corroded behavior of corroded steel plate was investigated by fatigue
steel plate, fatigue life test and FEM analysis. The relationship between residual
measured thickness and fatigue life of corroded steel plate, the
I. INTRODUCTION reduction of fatigue life due to corrosion for designation of
Corrosion and corrosion-related problems are fatigue, and the trend of stress concentration factor Kt versus
considered to be the predominate factors that lead to age- time of corroded steel plate in contact with concrete was also
related structural degradation. Structural integrity can be studied.
dramatically decreased by corrosion, particularly the pitting, II. FATIGUE TEST PROGRAM
because fatigue cracks would be much easier to nucleate at
corrosion pits and would propagate rapidly under dynamic A. Fatigue test specimens
load. The 18 corroded specimens among 48 specimens of four
Although safety factor in initial design may delay the types of specimen at 200, 400, and 600 cycles and 3 new
occurrence of fatigue problems, the priority negligence of specimens (un-corroded specimen, were denoted by N) were
corrosion pits in structural integrity analysis may lead to chosen to perform fatigue tests, the model of specimen is
shown in Fig. 1. The summary of characteristics of all
significant overestimation of the damage tolerance ability of
specimens shows in Table 1. All the specimens were carefully
existing steel structures. Corrosion is also an economic cleaned through a blasting treatment to remove rusty paint
burden because some artificial allowance for future corrosion clinging on the specimen surfaces after the concrete block
in initial design and repair operations undertaken to mitigate removal (Fig. 2 and Fig. 3)
the effects of corrosion are usually conservative owing to the
lack of a reliable methodology to predict future corrosion
effects.
Over the past few years, fatigue performance of
corroded aluminum alloy material has been studied
extensively. But for structural steel, it is less than satisfactory
with the existing information. Only a few pieces of literature
can be found on the aspects of this problem. Most studies
concluded that corrosion could dramatically reduce the
fatigue failure resistance of steel material. Zahrai concluded
that fatigue failure of corroded steel plate can be explained by
irregularities that may act as concentrators of stress and strain
Figure 1. Coordinate of the specimen

Table 1. Material properties and chemical composition

(wt.%)
Base metal C Si Mn P S
KS SM490 0.174 0.195 1.519 0.013 0.001
Figure 2. Inside of chamber of the instrument and the arrangement

of specimens
a) Specimen type SC-0
b) Specimen type SC-20
c) Specimen type SC-70 Figure 4. Test setup of fatigue test

C. The reduction of the characteristics of fatigue behavior
due to corrosion
d) Specimen type S As we can see that the fatigue life reduced as the
corrosion level increased. From all that is stated above, in
Figure 3. Corroded specimens after removing the concrete block, general, the principal effect of corrosion is that of a stress
rust, and paint coating
raiser. Then, irregular surface by corrosion on the surface
could raise the localized stresses to a value greater than that
B. Fatigue test condition for crack initiation and cause failure. Pit produced by
All specimens were tested at room temperature in corrosion on the surface can accelerate the speed of crack
laboratory air; tests were performed in accordance with the growth, then, worsen the fracture toughness. Corrosion on
KS B0802 (2008). Fatigue testing was carried out in steel will cause the plasticity and ductility of steel to
Universal Fatigue Testing Machine (UTM) operating at 8 Hz. deteriorate rapidly. Then, the specimens were prone to failure
The vibration loading amplitude was controlled during the under dynamic load. It is important to note that the surface
test, which is governed by sine function; the maximal load is appearance affects the fatigue life greatly, and corrosion
-10kN, and minimum load is -90kN. Fig. 4 shows the caused the surface to become rougher, thereby rapidly
experimental fatigue testing set up of corroded specimens declining fatigue life. Accordingly, the reduction of fatigue
conducted in this study. life S-N curves of specimen type S was reduced faster than
378
Figure 5. Reduction of fatigue life
the other and next was the specimen type SC-70, SC20 and plate in contact with concrete in terms of effective
SC-0, respectively. Those phenomena are related to corrosion thickness, as shown in Fig. 5 and Fig. 6.
progress of each type of specimen which was analyzed in
detail.
D. Fatigue Failure mode of corroded steel structure in
contact with concrete
Once established, the confidence level formulation of
fatigue strength provides the designer with properties that
can be used, with confidence, in design. However, the
authors would be remiss if they did not remind the reader
that these strength properties do not account for such
design details as joints, size effects and environmental
degradation. These design details must be handled outside
of the confidence level formulation with additional safety
This relatively simple set of procedures produces a
reasonable estimation of the tolerance limit for fatigue data
used in the evaluation of damage for steel members in
contact with concrete applications. It should be noted that
Figure 6. Confident curve
of the issues related to estimating a fatigue life curve at a
given confidence level, the most likely to produce large IV. CONCLUSION
differences in estimated lifetime is the choice of a linear or
A total of 21 specimens of four steel plate types in contact
logarithmic axis for stress or strain. the confidence curve with concrete were chosen to perform the fatigue test. The
was proposed to design the fatigue life of corroded steel fatigue life of corroded steel plate in contact with concrete
was evaluated quantitatively in terms of S-N curves, typical
379
of crack, and stress concentration factor using the residual corroded aluminum alloy specimens”, International Journal of
Fatigue, Vol. 59, pp. 282-291. (2014)
thickness. The following conclusions can be drawn from this
[2] Andrea Carpinteri, Andrea Spagnoli, Sabrina Vantadori.:“Size
study: effect in S–N curves: A fractal approach to finite-life fatigue
1. Corrosion will cause a significant decrease in fatigue strength”, International Journal of Fatigue, Vol. 31, pp. 927-
933. (2009)
life, with the corrosion increased the fatigue life decreased. [3] M.E. Biancolini, C. Brutti, G. Paparo, A. Zanini: Fatigue cracks
2. There was only one or two critical crack in each nucleation on steel, acoustic emission and fractal analysis,
International Journal of Fatigue, Vol. 28, pp. 1820-1825.
specimen, although several cracks were observed to initiate (2006).
within the corrosion sections. The results proved that stress [4] Y. Garbatov, C. Guedes Soares, J. Parunov.: “Fatigue strength
concentration factor Kt is inverse ratio with residual experiments of corroded small scale steel specimens”,
measured thickness. The Kt of corroded steel plate in contact International Journal of Fatigue, Vol. 59, pp. 137-144. (2014)
with concrete can be totally estimated by residual measured [5] Shan-hua Xu, You-de Wang.: “Estimating the effects of
corrosion pits on the fatigue life of steel plate based on the 3D
thickness profile”, International Journal of Fatigue, Vol. 72, pp. 27-41.
(2015)
ACKNOWLEDGMENTS
[6] Shigenobu Kainuma, Young-Soo Jeong, Jin-Hee Ahn.:
This work belongs to the project grant No: B2021-SPK- “Investigation on the stress concentration effect at the corroded
surface achieved by atmospheric exposure test”, Materials
07. funded by Ministry of Education and Training, and hosted Science & Engineering A, Vol. 602, pp. 89– 97. (2014)
by Ho Chi Minh City University of Technology and [7] M. Cerit, K. Genel, S. Eksi.: Numerical investigation on stress
Education, Vietnam. concentration of corrosion pit”, Engineering Failure Analysis,
Vol. 16, pp. 2467–2472. (2009)
REFERENCES [8] Muhammet Cerit.: “Numerical investigation on torsional stress
concentration factor at the semielliptical corrosion pit”,
[1] Zhang Chuan, Chen Yanhui, Yao Weixing.: “The use of fractal Corrosion Science, Vol. 67, pp. 225–232. (2013)
dimensions in the prediction of residual fatigue life of pre-
380
A Study on Ultrasonic Shear Horizontal Waves in

Composite Structures
Duy Kien Dao Hoang Ngoc Quy Truong Giang Nguyen
Faculty of Civil Engineering Institute of Mechanic Institute of Mechanic
Ho Chi Minh City of Technology and Vietnam Academy of Science and Vietnam Academy of Science and
Education Technology (VAST) Technology (VAST)
Ho Chi Minh City, Vietnam Hanoi, Vietnam Hanoi, Vietnam
kiendd@hcmute.edu.vn
Duc Tho Le Hoai Nguyen Hai Dang Phan1,2

1
Faculty of Mathematics, Mechanics Institute of Physics VNU University of Engineering and
and Informatics Vietnam Academy of Science and Technology, Hanoi, Vietnam
2
VNU University of Science Technology (VAST) Graduate University of Science and
Hanoi, Vietnam Hanoi, Vietnam Technology, VAST, Hanoi, Vietnam
haidangphan.vn@vnu.edu.vn
Abstract: Composite materials, which are generally that have been considered by numerous scholars and engineers
lighter, stronger and more durable than conventional in the field.
materials, have been increasingly used in aerospace and Research on ultrasonic guided waves in composite
automotive industries in recent decades. Ultrasonic guided materials is comprehensively presented in several textbooks,
waves have been proved to be an advanced technique for e.g., [1, 4-6]. The transfer matrix method (TMM) for modeling
characterizing composite materials and structures since they guided waves in multilayered structures was first provided in
can cover a large inspection area with a low attenuation [7] and then corrected in [8]. However, TMM must deal with
effect. This paper discusses the motions of shear horizontal numerical instability when considering high frequency and
(SH) guided waves generated by a time-harmonic source in thickness. To avoid this numerical problem, the global matrix
multilayered composite structures. We propose explicit method is introduced in [9]. The newest matrix technique
expressions of SH waves in composite structures based on the called the stiffness matrix method (SMM) is proposed in [10].
transfer matrix method. The amplitudes of SH wave motions The SMM has the speed of TMM while maintaining stability
due to the time-harmonic loading are then theoretically like GMM. Although only SH waves are concerned, the TMM
derived using reciprocity theorems. Afterward, the phase and is the most suitable because it does not contain the exponential
group velocity dispersion curves obtained by solving the terms in the diagonal line of the matrix thus the numerical
characteristic equa-tion are superimposed by quantitative instability does not exist. Moreover, the TMM is
amplitude spectra. The improved dispersion curves can straightforward to implement and write the explicit solution.
enhance the wave mode and frequency selection in the To obtain the full solution of guided waves, several analytical
ultrasonic eval-uation of composite structures. approaches have been developed, such as residue theorem
[11], normal mode expansion [12], or reciprocity theorem
Keywords: shear-horizontal waves, laminated [13]. However, to the best of the authors' knowledge, only the
reciprocity theorem holds for closed-form solutions and has a
composites, reciprocity theorem, dispersion curves.
simple implementation for multilayered anisotropic systems,
I. INTRODUCTION see [14-20].
The majority of materials used in the aerospace industry This paper first presents the explicit expressions of SH
nowadays are dominated by composites due to their superior guided waves in composite plates. The dispersion curves are
properties such as good stiffness to weight ratio, high obtained using a root-finding algorithm. Based on the explicit
durability, resistance to corrosion, mildew, noise, or reducing expressions, closed-form solutions of SH waves generated by
electromagnetic responses. These characteristics make a time-harmonics source are derived. The amplitude spectra
composites suitable for commercial aircraft, spacecraft, or are then integrated into the dispersion curves to highlight the
even military jets. Specifically, Boeing 787 Dreamliner, the magnitude of wave signals.
company's newest product line, is 80% composite by volume
and 50% by weight [1]. Composite structures, however, are II. SHEAR-HORIZONTAL WAVES IN A
facing the issue of degradation due to micro-crack or delam- LAMINATED COMPOSITE PLATE
ination. That problem raises a high demand for inspecting and We first consider the propagation of SH guided waves in
monitoring such structures. Ultrasonic guided waves have an orthotropic layer of thickness ℎ, see Figure 1, which has the
proved their efficiency in the large-area inspection of stiffness matrix 𝐶 . The equation of motions of the guided
aerospace composites, see [1-3]. Nondestructive evaluation waves in the plate can be written as
and structural health monitoring are the two mainstream topics

𝜎𝑖𝑗,𝑗 = 𝜌𝑢̈ 𝑖 (𝑖, 𝑗 = 𝑥, 𝑦, 𝑧) (1) [𝑇] (8

1 1 𝑖𝑘𝛽ℎ
0 ][ 1 1 ) −1
where 𝜎𝑖𝑗 denote stress components, 𝑢𝑖 represents =[ ] [𝑒 ]
𝑖𝑘𝛽𝐶44 −𝑖𝑘𝛽𝐶44 0 𝑒 −𝑖𝑘𝛽ℎ 𝑖𝑘𝛽𝐶44 −𝑖𝑘𝛽𝐶44
displacement components, and 𝜌 is the mass density. In pure
shear-horizontal waves, only the motion 𝑢𝑦 is considered, If we consider an 𝑁-layer composite plate, see Figure 2,
and it should also be independent of 𝑦. Thus, the equation of consisting of several orthotropic laminae in the same direction
SH guided waves can be expressed as or the cross-ply direction, the boundary at interfaces should
indicate the continuity of stress and displacement, i.e.
𝐶66 𝑢𝑦,𝑥𝑥 + 𝐶 44 𝑢𝑦,𝑧𝑧 = 𝜌𝑢̈ 𝑦 (2)
𝑢𝑦 (𝑛) 𝑢𝑦 (𝑛−1) (9)
where 𝐶44 and 𝐶66 are components of the stiffness matrix 𝐶. [𝜎 ] = [𝜎 ]
𝑦𝑧 𝑧=0 𝑦𝑧 𝑧=ℎ
The general solution of wave motion propagating in the 𝑥-
direction can be written as
𝑢𝑦 = 𝑈𝑒 𝑖𝑘(𝑥+𝛼𝑧−𝑐𝑡) (3)
where 𝑈 is the unknown amplitude, 𝑘 stands for the

wavenumber, 𝑐 indicates phase velocity, and 𝛼 is the ratio of
wavenumber between 𝑧 and 𝑥 direction.
Figure 2. Laminated composite plate
Therefore, we may derive the relation between the bottom

and top layers as
𝑢𝑦 (𝑁) 𝑢𝑦 (1) (10)

[𝜎 ] = [𝑇𝑇] [𝜎 ]
𝑦𝑧 𝑧=ℎ 𝑦𝑧 𝑧=0
where 𝑇𝑇 is the total transfer matrix which can be computed

as
𝑁
Figure 1. SH guided waves in an orthotropic layer 𝑇𝑇 𝑇𝑇12 (11)
[ 11 ] = ∏[𝑇](𝑖)
𝑇𝑇21 𝑇𝑇22
Substituting Eq. (3) into Eq. (2) yields 𝑖=1
By invoking the traction-free boundary conditions at the

(4) lower and upper surfaces of the plate, we may write
𝜌𝑐 2 − 𝐶66
𝛼1 = −𝛼2 = √
𝐶44
𝑇𝑇21 = 0 (12)
where 𝛼1 and 𝛼2 are two solutions of 𝛼. Equation (12) is called the characteristic equation which
describes the dispersion of SH waves. Using a root-searching
Then the solution of SH waves can be rewritten in the form
algorithm, we then obtain the dispersion curves for both phase
of two partial waves as
and group velocity. The dispersion figures will be shown in
the result section.
𝑢𝑦 = (𝑈1 𝑒 𝑖𝑘𝛽𝑧 + 𝑈2 𝑒 −𝑖𝑘𝛽𝑧 )𝑒 𝑖𝑘(𝑥−𝑐𝑡) (5)
Invoking the boundary condition at 𝑧 = 0, we may write
where 𝛽 = 𝛼1 . Using Hooke's law, the stress expression is
described as (1) (1) (13)
𝑈1 − 𝑈2 =0
𝜎𝑦𝑧 = 𝑖𝑘𝐶44 𝛽(𝑈1 𝑒 𝑖𝑘𝛽𝑧 − 𝑈2 𝑒 −𝑖𝑘𝛽𝑧 )𝑒 𝑖𝑘(𝑥−𝑐𝑡) (6) or
The relation of stress and displacement between the lower 𝑈1 (1) 𝑑
(1) (14)
surface (𝑧 = 0) and upper surface (𝑧 = ℎ) of the orthotropic [ ] = 𝐴 [ 1]
𝑈2 𝑑2
layer can be expressed as
where
𝑢𝑦 𝑢𝑦 (7)
[𝜎 ] = [𝑇] [𝜎 ] 𝑑1
(1) (15)
𝑦𝑧 𝑧=ℎ 𝑦𝑧 𝑧=0 1
[ ] =[ ]
𝑑2 1
where [𝑇] is the transfer matrix given by
and 𝐴 is an arbitrary constant.
Thus, we have
382
𝑛 (𝑝) (𝑝) (𝑥+𝑐 (𝑝) 𝑡)

𝑑
(𝑛)
𝑑
(1) (16) ℬ
𝜎𝑦𝑧 = 𝐵 (𝑝) Σ𝑦𝑧 (𝑧)𝑒 −𝑖𝑘 (25)
[ 1] = ∏[𝑇] (𝑖)
[ 1]
𝑑2 𝑑2
𝑖=1
(𝑝) (𝑝) (𝑥+𝑐 (𝑝) 𝑡)
ℬ
𝜎𝑦𝑥 = 𝐴(𝑝) Σ𝑦𝑥 (𝑧)𝑒 −𝑖𝑘 (26)
The expressions of displacement and stress fields for an
arbitrary layer now can be rewritten as where (𝑚), (𝑝) denote mode orders, and
(𝑗) (𝑗) 𝛽𝑧 (𝑗) 𝛽𝑧
𝑢𝑦 = 𝐴(𝑑1 𝑒 𝑖𝑘𝛽𝑧 + 𝑑2 𝑒 −𝑖𝑘𝛽𝑧 )𝑒 𝑖𝑘(𝑥−𝑐𝑡) Σ𝑦𝑧 = 𝑖𝑘 (𝑗) 𝐶44 (𝛽𝑑1 𝑒 𝑖𝑘 − 𝛽𝑑2 𝑒 −𝑖𝑘 ) (27)
(17)
= 𝐴𝑈(𝑧)𝑒 𝑖𝑘(𝑥−𝑐𝑡)
(𝑗) (𝑗) 𝛽𝑧 (𝑗) 𝛽𝑧
Σ𝑦𝑥 = 𝑖𝑘 (𝑗) 𝐶66 (𝑑1 𝑒 𝑖𝑘 + 𝑑2 𝑒 −𝑖𝑘 ) (28)
𝜎𝑦𝑧 = 𝑖𝑘𝐶44 𝐴(𝛽𝑑1 𝑒 𝑖𝑘𝛽𝑧 − 𝛽𝑑2 𝑒 −𝑖𝑘𝛽𝑧 )𝑒 𝑖𝑘(𝑥−𝑐𝑡)
(18)
= 𝑖𝑘𝐶44 𝐴Σ(𝑧)𝑒 𝑖𝑘(𝑥−𝑐𝑡) It should be noted that, in state 𝒜 , the propagation
direction is positive 𝑥 direction while the waves in state ℬ are
These equations have only one unknown relative traveling in the opposite direction. The indicator 𝑛 has been
amplitude 𝐴. The solutions of SH waves generated by a time- neglected for simplicity.
harmonics source will be obtained using the reciprocity
theorem. This approach has been shown to be an efficient Substituting Eqs. (20)-(26) into Eq. (19), after some
analytical tool for wave propagation problems [15, 17, 21]. manipulations, we have
Basically, a reciprocity theorem describes a relation
between two elastodynamics states. One state can be 𝑃𝑈𝑦 (𝑧0 )𝑒 −𝑖𝑘𝑝 𝑥0
𝐴𝑃 = (29)
considered as actual state 𝒜 which describes the SH guided 2𝐼
waves generated by a time-harmonics load. The other one is
the propagation of free SH waves, called virtual state ℬ. The where
formula of the reciprocity theorem for a multi-body system 𝑁
𝑧 (𝑛+1)
can be written as, see [13], (𝑛) (𝑛)
𝐼 = ∑ 𝑖𝑘𝑝 𝐶66 ∫ 𝛴𝑦𝑥 𝑈𝑦 𝑑𝑧 (30)
𝑛=1 𝑧 (𝑛)
𝑁
∑ ∫ (𝑓𝑗𝒜 𝑢𝑗ℬ − 𝑓𝑗ℬ 𝑢𝑗𝒜 )𝑑𝑉𝑛 It is worth mentioning that the solution in Eq. (29) stands
𝑛=1 𝑉𝑛 (19) for the forward field. For the backward field, the result
remains the same in magnitude but opposite sign in the term
=∫ ℬ 𝒜
(𝜏𝑖𝑗 𝑢𝑗 − 𝜏𝑖𝑗𝒜 𝑢𝑗ℬ )𝑛𝑖 𝑑𝑆𝑛̅ 𝑖𝑘𝑝 𝑥0 .
̅
𝑆𝑛
Where 𝑓 is body force, 𝑛 is the normal unit vector, III. RESULTS AND DISCUSSIONS
𝑆𝑛̅ (𝑛 = 1,2, … , 𝑁) are external boundaries, which are In this section, we present the results of dispersion curves
contours around 𝑉𝑛 without the interface with other bodies. and reciprocity calculation for an eight-layer cross-ply
laminate which is commonly used in aerospace structures
The body load applied at (𝑥0 , 𝑧0 ) can be expressed [22]. The stacking sequence of fiber orientations is (0/90/
in forms of Delta function as 90/0/0/90/90/0) degree. The thickness of a single lamina
is 0.2 𝑚𝑚 , thus the total thickness is 1.6𝑚𝑚 . Material
𝑓𝑦 = 𝑃. 𝛿(𝑥 − 𝑥0 )𝛿(𝑧 − 𝑧0 )𝑒 −𝑖𝑘𝑐𝑡 (20)
properties of the composite plate are listed in Table 1.
Considering layer 𝑛, the expressions of state 𝒜 can be
proposed as a combination of all modes as Table 1. Material properties of the lamina.
Mass
𝑚 Material Stiffness matrix (𝐺𝑃𝑎) density
(𝑚)
(𝑥−𝑐 (𝑚) 𝑡) (𝑘𝑔/𝑚3 )
𝑢𝑦𝒜 = ∑ 𝐴(𝑚) 𝑈 (𝑚) (𝑧)𝑒 𝑖𝑘 (21)
𝑖=1
143.8 6.2 6.2 0 0 0
6.2 13.3 6.5 0 0 0
𝑚
(𝑚) (𝑥−𝑐 (𝑚) 𝑡)  6.2 6.5 13.3 0 0 0 1560
(𝑚)
𝒜
𝜎𝑦𝑧 = ∑ 𝐴(𝑚) Σ𝑦𝑧 (𝑧)𝑒 𝑖𝑘 (22) 0 0 0 3.4 0 0
0 0 0 0 5.7 0
𝑖=1 [ 0 0 0 0 0 5.7]
𝑚
(𝑚) (𝑚) (𝑥−𝑐 (𝑚) 𝑡)
𝒜
𝜎𝑦𝑥 = ∑ 𝐴(𝑚) Σ𝑦𝑥 (𝑧)𝑒 𝑖𝑘 (23) Based on the closed-formed solution obtained in Section
𝑖=1 2, the amplitude spectra of SH guided wave modes are
immersed in the original dispersion curves. The improved
whereas state ℬ can be chosen as mode 𝑝 of free waves, i.e., dispersion curves are called the dispersion curves
(𝑝) (𝑥+𝑐 (𝑝) 𝑡) superimposed by amplitude spectra. These dispersion curves
𝑢𝑦ℬ = 𝐵 (𝑝) 𝑈 (𝑝) (𝑧)𝑒 −𝑖𝑘 (24) of phase velocity and group velocity are shown in Figure 3.
383
As illustrated in the dispersion diagrams, the lowest mode [2] Matt H, Bartoli I, Lanza di Scalea F: Ultrasonic guided wave
(𝑆𝐻0 ) is nondispersive and has the velocity of transverse monitoring of composite wing skin-to-spar bonded joints in
aerospace structures. The Journal of the Acoustical Society of
waves. This mode is usually used in experiments or practical America 118(4), 2240-2252 (2005).
inspections due to the simplicity in observing and processing [3] Gao H, Rose JL: Ice detection and classification on an aircraft
signals. Based on the amplitude spectra, the domains colored wing with ultrasonic shear horizontal guided waves. IEEE
red have a big displacement magnitude and vice versa for the Transactions on Ultrasonics, Ferroelectrics, and Frequency
blue domains. According to this, if the 𝑆𝐻0 is excited in the Control 56(2), 334-344 (2009).
low-frequency zone (< 0.1𝑀𝐻𝑧) the mode signals will get [4] Datta SK, Shah AH: Elastic waves in composite media and
the highest magnitude. Using the proposed dispersion structures: With applications to ultrasonic nondestructive
evaluation. CRC Press, (2008).
diagram, the practitioners can instantly choose a suitable mode
[5] Nayfeh AH: Wave propagation in layered anisotropic media:
and excitation frequency to optimize the calibration and post- With applications to composites. Elsevier, Amsterdam (1995).
processing processes. Therefore, the presented diagram may [6] Rokhlin SI, Chimenti DE, Nagy PB: Physical ultrasonics of
serve as a reliable reference for nondestructive evaluation composites. Oxford University Press, New York (2011).
applications. [7] Thomson WT: Transmission of elastic waves through a
stratified solid medium. Journal of Applied Physics 21(2), 89-
93 (1950).
[8] Haskell NA: The dispersion of surface waves on multilayered
media. Bulletin of the Seismological Society of America 43(1),
17-34 (1953).
[9] Knopoff L: A matrix method for elastic wave problems.
Bulletin of the Seismological Society of America 54(1), 431-
438 (1964).
[10] Rokhlin SI, Wang L: Stable recursive algorithm for elastic
wave propagation in layered anisotropic media: Stiffness
matrix method. The Journal of the Acoustical Society of
America 112(3), 822-834 (2002).
[11] Graff KF: Wave motion in elastic solids. Dover Publications,
New York (1991).
[12] Auld BA: Acoustic fields and waves in solids. Krieger
Publishing Company, Malabar, FL, USA (1990).
[13] Achenbach JD: Reciprocity in elastodynamics. Cambridge
University Press, Cambridge (2003).
[14] Phan H, Cho Y, Achenbach JD: Validity of the reciprocity
approach for determination of surface wave motion.
Ultrasonics 53(3), 665-671 (2013).
[15] Phan H, Cho Y, Le QH, Pham CV, Nguyen HTL, Nguyen PT,
et al.: A closed-form solution to propagation of guided waves
in a layered half-space under a time-harmonic load: An
application of elastodynamic reciprocity. Ultrasonics 96(40-47
Figure 3. Dispersion curves (2019).
[16] Le D, Lee J, Cho Y, Dao DK, Nguyen TG, Phan H: Ultrasonic
IV. CONCLUSION guided waves in unidirectional fiber-reinforced composite
plates. Advances in Condition Monitoring and Structural
This article discusses the motion of SH guided waves Health Monitoring. Springer Singapore (2021).
generated in multilayered composite structures under time- [17] Dao DK, Ngo V, Phan H, Pham CV, Lee J, Bui TQ: Rayleigh
harmonic ultrasonic sources. The explicit expressions of wave motions in an orthotropic half-space under time-
guided waves have been introduced based on the transfer harmonic loadings: A theoretical study. Applied Mathematical
matrix method. Using the elastodynamic reciprocity theorems, Modelling 87(171-179 (2020).
we have calculated the amplitudes of SH guided waves [18] Nguyen P-T, Phan H: A theoretical study on propagation of
generated by the time-harmonic load. The results of the phase guided waves in a fluid layer overlying a solid half-space.
Vietnam Journal of Mechanics 41(1), 51-62 (2019).
and group velocity dispersion curves superimposed by
[19] Nguyen H, Le D, Plan E, Dang ST, Phan H: Theoretical model
amplitude spectra have been presented and discussed. These of guided waves in a bone-mimicking plate coupled with soft-
modified dispersion curves are critical to the selection of tissue layers. Vietnam Journal of Mechanics 43(1), 91-104
suitable wave modes and frequencies in the ultrasonic (2021).
inspection by nondestructive evaluation techniques. [20] Nguyen PT, Nguyen H, Le D, Phan H: A model for ultrasonic
guided waves in a cortical bone plate coupled with a soft-tissue
ACKNOWLEDGEMENTS layer. AIP Conference Proceedings 2102(1), 050007 (2019).
[21] Phan H, Cho Y, Achenbach JD: Application of the reciprocity
This research is funded by Graduate University of Science theorem to scattering of surface waves by a cavity.
and Technology under grant number GUST.STS.ÐT2020- International Journal of Solids and Structures 50(24), 4080-
CH01. 4088 (2013).
[22] Rose JL: Ultrasonic guided waves in solid media. Cambridge
REFERENCES University Press, Cambridge (2014).
[1] Giurgiutiu V: Structural health monitoring of aerospace
composites. Academic Press, (2015).
384
Designing and Modeling Pipe Welding Machine

Tuong Phuoc Tho Phan Phuc Khang
Department of Mechatronics, Ho Chi Minh City Department of Mechatronics, Ho Chi Minh City
University of Technology and Education University of Technology and Education
Thu Duc City, Ho Chi Minh City, Viet Nam Thu Duc City, Ho Chi Minh City, Viet Nam
thotp@hcmute.edu.vn 18146139@student.hcmute.edu.vn
Tran Thanh Nhon Pham Phi Long

Department of Mechatronics, Ho Chi Minh City Department of Mechatronics, Ho Chi Minh City
Thu Duc City, Ho Chi Minh City, Viet Nam Thu Duc City, Ho Chi Minh City, Viet Nam
18145196@student.hcmute.edu.vn 18146166@student.hcmute.edu.vn
Abstract: Pipe circumference welding is always a

challenging operation due to the complex geometry of the
weld. This job requires highly skilled personnel and some
welding errors can occur, especially when it is difficult to
weld cross-shaped pipe joints. This paper presents the design
and development of manipulators for welding various types
of circumferential type pipe joints. This welding technique
consists of a rotating workbench to hold the workpiece and a
TIG welding head mounted on a combined manipulator. The
mechanism will come with a flexible fixture that can
accommodate different pipe angles as well as a variety of
materials. Control software is designed to control the
mechanisms according to the given trajectories. Parameters
affecting the welding process such as arc length, welding
speed, worktable rotation, etc. can be adjusted with the help
of software. Complex pipe welding trajectories are tested for
simulation, thereby evaluating the feasibility of this design.
Figure 1. Pipe welding machine
Keywords: pipe circumference welding, kinematics,
trajectory planning, linear segment with parabolic blend, This paper presents the process of proposing a method of
TIG pipe welding using a configuration of 3 degrees of freedom,
the kinetic problem of setting up a control trajectory is
I. INTRODUCTION researched and developed specifically for the welding
The manufacturing industry uses lots of metal pipes with process for the pipeline. The calculated results are simulated
various diameters and profiles which impose heavy demands and analyzed to evaluate the suitability of the selected
on the skills and concentration of the welder [1]. Due to the configuration for welding the profiles of round pipe joints,
numerous significant requirements to the humans involved ensuring the welding technical requirements.
and therefore the heavy demands on welding flexibility and II. DESIGN AND CALCULATION
welding productivity, welding automatic offers an excellent
potential for several industries to enhance welding rates and A. Blocks used
specifically improve welding quality [2]. Depending on the
type of round stainless steel used, we can change the
thickness of the welds to meet the requirements of the bearing
capacity of their welds but still have high aesthetics.
However, the characteristics of the product depend a lot on
the quality and thickness of the material [3 - 7].
The pipe welding machine is capable of welding round
pipes with a diameter ranging from 50mm to 200mm, the
joints can be changed. For example, we can weld a 200mm
diameter pipe with a 100mm diameter pipe, as long as the
joints are snug. After adjusting the welding parameters, as
well as changing the contact direction of the TIG welding
Figure 2. Schematic of pipe welding machine
machine head, welding can be carried out, with the thickness
and allowed speed. Fig.1 shows the design mechanism.

Fig. 2 illustrated the structure of pipe welding machine, Table 1. Input parameters
including: Specification Value
• Power supply: This block supplies single-phase 220V
X, Y-axis travel 500x400 mm
current for PLC to control the servo motor and for the
TIG welding machine. Maximum running speed of X-axis v1 = 9 m/min
without machining
• PLC: is responsible for controlling the engine and
mechanical structures by pre-programmed programs, Maximum running speed of X-axis when
machining
v1 = 6 m/min
serving to the automatic welding process [8].
Maximum operating acceleration of the a = 0.5g = 0.5m
• Driver: is responsible for receiving the control command system
signal (pulse) from the PLC and transmitting the command
Speed motor NMax = 2000 rpm
to the servo motor and controlling the servo motor to
operate according to the command, and at the same time
Repeating accuracy 0.1
receiving continuous feedback on position and speed servo
motor current from an encoder.
Coefficient of friction  = 0.01
• Servo motor: receive the control command from the
driver and start running by following the flow of programs Operation time (4-6 years) Lt = 17520h
[9].
• Gearbox: provide torque multiplication, speed reduction, B. Kinematic
and inertia matching for motor-driven systems, also The coordinate system on the welding machine in Fig. 3
supplying high precision and stiffness for servo engine. is designed with the table of kinematic parameters of the
machine [1],[11] as shown in Table 2:
• Chuck: the mechanics that use to fix the pipe in particular
contour that is suitable for the welding process. Table 2. D-H table
• TIG welding machine: operated by receiving
220Vcurrent from the power supply, transmitting the Ɵ αi Ai (m) Di (m)
current through the welding wire to arc out from the 0 0 0 0 0
welding nozzle.
• Step motor: the engine that receives the control signal 1 vs 0 ɵ 0 - dx01, dy01, dz01
from the driver to move X-axis which guaranteed the
2 vs 0 0 0 dy02 + u1 dx02, dy02, dz02
distance between the welding nozzle and exhaust pipe.
• Rotary handle: a mechanic that allows individuals to 3 vs 2 0 0 dz23+u2 dx23, dy23, dz23
move or control Y-axis appropriately with the weld
4 vs 3 0 0 0 dx34, dy34, dz34
position by themselves.
With: dx01=0.24, dy01 = 0.144, dz01=0.113, dx02=0,
• Welding nozzle: located exactly the distance that we set
dy02=0.05, dz02=0.046, dx23=0.24, dy23=0.138, dz23=0,
up before. Moreover, it is connected to the welding
u1=0.4, u2=0.34
machine and starts arcing after receiving the electronic
signal from the welding control table. Welding head coordinate matrix relative to the
workpiece:
Operation description:
Pipe welding machine works by 3 main blocks:
electronic, mechanical and control signal. First, the welding
trajectories are designed and converted into joint trajectories
Join
for X, Y and chuck respectively. These trajectories are
2
transferred to the main controller which is converted into
control signals for the drivers of the motors. The motors are Joint 1
controlled in position and speed through the drivers and
integrated with the reducer, which ensures that the torque and
speed parameters of the motor according to different
operating conditions have been pre-calculated. The motors Chuck
drive the X and Y axis and are operated synchronously with
the chuck motor to control the welding nozzle to follow the
given weld trajectory precisely [10]. TIG welding machine is
also controlled synchronously with the welding trajectory.
All configurations and specifications of the pipe welding
machine are as shown in Table 1.
Figure 3. Manipulator and the Coordinate System
386
1 0 0 dx01
0 cos  − sin  dy01
0
A1 =   (1)
0 sin  cos  dz01
 
0 0 0 1 
1 0 0 dx 02  1 0 0 dx 23  1 0 0 dx34 
0 1 0 dx 02 + u1  0 1 0 dy 23  0 1 0 dy34 
0
A 2 A3 A 4 = 
2 3
. 
0 0 1 dz02  0 0 1 dz 23 + u 2  0 0 1 dz34  Figure 4. T-pipe Figure 5. Y-pipe
    
0 0 0 1  0 0 0 1  0 0 0 1 
1 0 0 dx 02 + dx 23 + dx34 
0 + Construct the orbital equation:
1 0 dy02 + dy23 + dy34 + u1 
= (2)
0 0 1 dz02 + dz 23 + dz34 + u 2  Using marking method, the orbital shape of the
  connection of two pipes on the plane can be deduced.
0 0 0 1 
Position of the tip relative to the workpiece: When the coupler works with the reciprocating movement
of the welding head, it will produce a profile similar to that
x0  x3  0  of the joint on the plane.
y  y  0 
 0 0  3 0  
  = DH 4   = DH 4  
z
 0 z
 3 0 

 1  
 
1  
1 

sin()(dy12 + u 2 − dy01 − dy 23 ) + cos()(− dx 01+ dx 23 ) 
 sin()(dx − dx ) + cos()(dy + dy − dy − u ) 
 23 01 01 23 12 2 
=  (3)
 dz 01 − dy12 + u 1 

 1 

Velocity matrix between welding head and workpiece: α

h
H
 f1 f1   x1 x1 
 q ... ...
q n   q1 q n 
f    
1
J(q) = =  ... ... ...  =  ... ... ...  (4) Figure 6. Flat pattern T-pipe
q    
 f m ...
f m   x m
...
x m 
 q1 q n   q1 q n  From Fig. 6 the equation of the orbit can be deduced:
  (t −1).15 D 
-sin() -cos() 0 X
y (t ) = H + h (t ) = H + R − R. cos( (t )) = H + R 1− cos 
 cos() -sin()  (6)
 0 Y
   R 
J=  (5)
 0 0 0 0

 0 0 0 0
 When α rotates from 0o to 180o then distance of
intersection point is changed. That is h(t).
R
X = cos()(dy12 + u 2 − dy01 − dy23 ) − sin()(− dx01+ dx 23 )
Y = cos()(dx 23 − dx01) − sin()(dy01 + dy23 − dy12 − u 2 ) y
III. SIMULATION
Although there are various types of pipe connection, a T-
pipe in Fig. 4 and a Y-pipe in Fig. 5 connector is the most 45
basic one. Therefore, constructing the orbital equation of T- α h

pipe and Y-pipe, the general mathematical expression of all Figure 7. Flat pattern Y-pipe
types of the intersection curve can be obtained [1].
Apply Pythago: ℎ = 𝑅 − 𝑅. cos(𝛼)
Using Linear Segment with Parabolic Blend to calculate
the position of the welding nozzle [12],[13]. Link from From the Fig. 7 the equation of the orbit can be deduced:
Solidwork to Matlab Simulink and apply results from Linear ( R − cos( )  R) + ( R − cos( )  R)  cos(450 ) + h
y (t ) = (7)
Segment with Parabolic Blend to model. cos(450 )
387
Figure 9. Diagram of trajectory according to LSPB
Figure 8. Linear Segment with Parabolic Blend (LSPB)
Trajectory planning using parabolic blend is very

extremely in industries [14 - 16]. As shown in Fig. 8, this
trajectory is split into three parts, constant acceleration or
ramp velocity, zero acceleration and constant deceleration.
Instead of using multiple points, just special ones are used.
Then the trajectory passes almost through the intermediate
point q2, ... qN-1 like Fig. 8, then the trajectory is the line Figure 10. Diagram of velocity and acceleration of T-pipe
segment between the two points, and in the vicinity of the according to LSPD
intermediate point can replace it with a parabola.
Linear segment:
• Trajectory:
qk +1 − qk  dtk   dtk 
qk ( t ) =  t − tk −  + qk  t k + 
tk +1 − tk  2   2 
(8)
 dt dt 
,  tk + k  tk  tk +1 − k +1 
 2 2 
Figure 11. Diagram of velocity and acceleration of Y-pipe according
• Velocity: to LSPD
qk +1 − qk  d dt 
qk = ,  tk + tk  tk  tk +1 − k +1  (9) The corresponding trajectory response for T-pipe and Y-
tk +1 − tk  2 2  pipe according to Linear Segment with Parabolic Blend is
shown in Fig. 9, this trajectory shows that the values satisfy
Parabol segment: the equation (8), (10).
• Trajectory: Although not quite the same trajectory of Flat pattern, but
2 it can accept. Because speed and acceleration must still
qk +1 − qk   dt    dt  
qk +1 ( t ) =
follow trajectory, so LSPB is the best choice. Fig. 10 and Fig.
 t −  tk +1 − k +1  + qk t −  tk +1 − k +1   + qk 11 are also simulating velocity and acceleration, LSPB
2dtk +1   2    2 
generates a trapezoidal velocity response to protect the
 dt dt 
,  tk +1 − k +1  t  tk +1 + k +1  (10) actuator and can provide zero velocity at the endpoint.
 2 2 
• Velocity:
qk +1 − qk   dtk +1    dtk +1 dtk +1 
qk +1 ( t ) = t −  tk +1 − 2   + qk ,  tk +1 − 2  t  tk +1 + 2 
dtk +1     
(11)
qk + 2 − qk +1
qk +1 =
tk + 2 − tk +1
Figure 12. Simplified model
388
control trajectories, velocity and acceleration based on Flat

pattern and Linear Segment with Parabolic Blend were given.
REFERENCES
[1] Yao, T., Gai, Y., Liu, H.: Development of a robot system for
pipe welding. International Conference on Measuring
Technology and Mechatronics Automation 1 (2010)
[2] Lauridsen, J., Kirkegaard: Model based control of a one degree
of freedom workpiece manipulator for welding of nozzles.
Mathematics and Computers in Simulation 41, 407–417 (1996)
[3] Yayla, P.A.Ş .A., Kaluc, E., Ural, K.: Effects of welding
processes on the mechanical properties of HY 80 steel
weldments. Materials & design 28, 1898–1906 (2007)
[4] Jovičić, R.: Effects of welding technology on the occurrence
Figure 13. Model diagram in Matlab of fracture in welded joints. Procedia Structural Integrity 13,
1682–1688 (2018)
[5] Dadfar, M.: Effect of TIG welding on corrosion behavior of
316L stainless steel. Materials Letters 61, 2343–2346 (2007)
[6] Li, H.: The effect of TIG welding techniques on microstructure,
properties and porosity of the welded joint of 2219 aluminum
alloy. Journal of Alloys and Compounds 727, 531–539 (2017)
[7] Lothongkum, G., Viyanit, E., Bhandhubanyong, P.: Study on
the effects of pulsed TIG welding parameters on delta-ferrite
content, shape factor and bead quality in orbital welding of
AISI 316L stainless steel plate. Journal of Materials Processing
Technology 110, 233–238 (2001)
[8] Lei, S., Yong: Servo-control system design of automatic
production line based on PLC and HMI. Applied Mechanics
and Materials 457 (2014)
[9] Chen, H.: A robust visual servo control system for narrow seam
double head welding robot. The international journal of
advanced manufacturing technology 71, 1849–1860 (2014)
[10] Chou, J.J., Yang, D.C.H.: (1991)
[11] Barakat, A.N., Khaled, A., Gouda, K.A., Bozed: Kinematics
analysis and simulation of a robotic arm using MATLAB. 4th
International Conference on Control Engineering & Infor-
mation Technology (CEIT) (2016)
Figure 14. Correlation between desired and actual [12] Al-Khayyt, Saad, Z., Saeed: Creating Through Points in Linear
Function with Parabolic Blends Path by Optimization Method.
Al-Khwarizmi Engineering Journal 14, 77–89 (2018)
Building a geometric model in SolidWorks (Fig. 12) then [13] Abdul-Lateef, W., Essmat, A.F., Huayier, N.H., Farhood:
implementing the model in Matlab/Simulink environment Design of planning Trajectory for the planar Robot Manipulator
[17]. A block scheme is obtained immediately after the using Linear Segments method with Parabolic Blends (LSPB).
implementation. After that, PID controller is used to control Journal of Mechanical Engineering Research and
model with trajectory calculated (Fig. 13) [18],[19]. Developments 44, 159–171 (2021)
[14] Parikh, P.A., Keyur, D., Joshi, R., Trivedi: Vision-Based
From the result of the simulation of trajectory, velocity
Trajectory Planning for a Five Degree of Freedom Assistive
and acceleration in Fig. 14, the actual trajectory is following
Feeding Robotic Arm Using Linear Segments with Parabolic
the desired trajectory. The simulating results based on the Blend and Cycloid Functions. Springer, Cham
PID control, the time of response to velocity is quick and [15] Macfarlane, S., Croft, E.A.: Jerk-bounded manipulator
follows the desired speed. Maximum of velocity is trajectory planning: design for real-time applications. IEEE
approximately 8 rad/s and 17 rad/s, maximum of acceleration Transactions on robotics and automation 19, 42–52 (2003)
is approximately 78 rad/s2 and 100 rad/s2, it is lower than the [16] Gasparetto, A., Zanotto, V.: A new method for smooth
allowable limit value. trajectory planning of robot manipulators. Mechanism and
machine theory 42, 455–471 (2007)
IV. CONCLUSION [17] Cekus, D., Posiadała, B., Warys, P.: Integration of modeling in
In this paper, a welding machine with 3 D.O.F is SolidWorks and Matlab/Simulink environments. Archive of
designed, calculated and simulated in Matlab for the T-pipe Mechanical Engineering 61, 1–1 (2014)
and Y-pipe. The article has designed and calculated a welding [18] Sartika, E., Merry, T.R., Sarjono, D.D., Saputra: Prediction of
robot with a diameter of workpiece from 50mm to 100mm. PID control model on PLC. Telkomnika 17, 529–536 (2019)
[19] Paulusová, J., Körösi, L.: (2009)
Kinematics and trajectory design has been configuration
for multiple trajectories of welding, optimal algorithms to
389
Investigation of the Mechanical Properties of

Lightweight Geopolymer Concrete Using
Keramzite as Coarse Aggregate
Kiet Tran Tuan Duc Nguyen Phan Tuan Le Anh
Faculty of Civil Engineering General II Import & Export Joint Faculty of Civil Engineering
Ho Chi Minh City University of Stock Company (GENERALIMEX Ho Chi Minh City University of
Technology and Education JSC) Technology
Ho Chi Minh City, Vietnam Ho Chi Minh City, Vietnam Vietnam National University Ho Chi
ttkiet@hcmute.edu.vn namnguyen.co@gmail.com Minh City
latuan@hcmut.edu.vn
Khoa Nguyen Tan An Huynh Thao

Institute of Research and Development School of Built Environment, Engineering And Computing
Faculty of Civil Engineering Leeds Beckett University
Duy Tan University Leeds, UK
Da Nang City, Vietnam a.huynh@leedsbeckett.ac.uk
nguyentankhoa@duytan.edu.vn
Abstract: In this study, the mechanical properties of process of geopolymer concrete is different from conventional
lightweight geopolymer concrete using keramzite as coarse concrete which needs water to occur. Thus, there is generally
aggregate were determined. The purpose of using keramzite no portlandite or C-S-H formation at the end of
as coarse aggregate is to reduce the self-weight of geopolymerization process [6].
geopolymer concrete. Two types of keramzite such coating Lightweight concrete made with foaming agent or
and un-coating with geopolymer paste layer were used with lightweight aggregate has been researched and applied in the
fly ash, ground granulated blast furnace slag, alkaline liquids range of both structural and non-structural applications [7, 8].
and fine aggregate to fabricate lightweight geopolymer There are several research which relate to lightweight
concrete. The compressive strength, splitting tensile strength geopolymer concrete (LGC) in the world. For example, LGC
and flexural strength with various alkaline liquid to is manufactured by mixing geopolymer binder with recycled
geopolymer solid ratios were tested. The experimental results lightweight block [9], lightweight aggregate vermiculite [10],
demonstrate that the maximum compressive strength of 38.80 perlite or oil palm shell [11]. These studies indicated that the
MPa, splitting tensile strength of 1.69 – 3.71 MPa and density of LGC is about 1000-1500 kg/m3, compressive
flexural strength of 2.48 – 4.25 MPa. This kind of lightweight strength at 28-day of 2-13 MPa. On the other hand, LGC is
geopolymer concrete can be used as brick for wall and also produced by using foaming agents such as hydrogen
partition or structural members. peroxide, metallic aluminum powder combined with
geopolymer binder. The experimental results showed that the
Keywords: lightweight geopolymer concrete, slag, fly ash, density of this kind of LGC is in a range of 900-1500 kg/m3,
keramzite, mechanical properties while compressive strength at 28-day is of 3-21 MPa [12].
Although there are some studies about lightweight
I. INTRODUCTION geopolymer concrete, the understanding of mechanical
In 1979, Davidovits [1] firstly introduced geopolymer as properties of LGC using light weight aggregate is limited. The
new green material. This material is paid significant attention main purpose of this study is investigating the mechanical
by scientists from all around the world. Geopolymer is a low- properties of lightweight fly ash/GGBS based geopolymer
cost and enviro-friendly material which has great potential to concrete using keramzite as coarse aggregate. The mechanical
replace the conventional materials such as Portland cement properties such compressive strength, indirect tensile, and
[2]. Currently, geopolymer materials, especially geopolymer flexural strength are determined with various ratios of alkali
concrete is researched and applied popularly in construction liquid to geopolymer solids.
building materials [3–5]. Geopolymer material is a type of
inorganic polymers which is synthesized from source II. MATERIALS AND TEST METHOD
materials and alkaline solutions. Geopolymer concrete is
A. Materials
usually made by mixing binders (such: fly ash, ground
granulated blast slag (GGBS), metakaolin,…), aggregate, In this study, the experimental work is performed by
alkaline solutions, and superplasticizer. The hardening using fly ash and GGBS as binder, alkali solution, fine

aggregate and keramzite as coarse aggregate. The specific Fly ash and GGBS were quantified as mix proportion in
gravity and fineness of GGBS is 2.55 g/cm3 and 3600 cm2/g, Table 2. Then, these two materials were mixed for about five
respectively. Class F fly ash [13] with specific gravity of 2.5 minutes, for Na2SiO3 solution was poured into the solids and
g/cm3 is used. The details of chemical compositions of fly ash kept mixing in five minutes. Finally, keramzite and fine
and GGBS are shown in Table 1. aggregate were added to the slurry and continued to mix about
The alkali solution usually is a combination of sodium three more minutes. The fresh geopolymer concrete was cast
silicate and sodium hydroxide solutions. However, the current and compacted into the molds. After finishing this step, the
study used only sodium silicate solution to mix with solid as specimens were cured in the oven at 80oC for 24 hours. The
follows [14]. The modulus ratio (Ms) of the Na2SiO3 solution specimens are tested at 7 days after curing. Compression test,
was 3.3 (where Ms= SiO2/Na2O, Na2O= 8.37%, indirect tensile test, flexural strength test.
SiO2=27.63%). Aggregates, including keramzite as coarse B. Testing method
aggregates (CA) and fine aggregates (FA) were used. They
were mixed with the ratio of 0.45:1 by mass. The specific The testing program used for investigating the mechanical
gravity of keramzite is 600 kg/m3 and 2650 kg/m3 for fine properties of LGC is illustrated in Fig.2, comprising
aggregates. Two types of kermazite such coating and un- compression test, and indirect tensile test. Compressive
coating with geopolymer paste, as given in Fig. 1, were used strength test followed the ASTM C39[15]. Three 150 x 300
in this study. Un-coating keramzite aggregate was immersed mm concrete samples from each group were employed to
in water for 24 hours before mixing. The mix proportions of determine the compressive strength. The indirect tensile test,
LGC using keramzite as coarse aggregate are shown in Table according to ASTM C496 [16], was performed on three 150 x
2. In the mix proportion, the ratio of Na2SiO3 solution (AL) to 300 mm cylindrical specimens for each mixture. Based on
geopolymer solid (GS) ranged from 0.7, 0.8, and 0.9 by mass. ASTM C78 [17], the flexural performance of LGC was
The geopolymer solid is a combination of fly ash and GGBS. evaluated by using specimens 100 x 100 x 400 mm with a span
The ratio between fly ash and GGBS is 2:8, by mass. Noted length of 300 mm.
that, the specific gravity of specimens varies from 1800 – 2000
kg/m3.
Table 1. Chemical compositions of fly ash and cement
Oxide (%) Fly ash GGBS

SiO2 51.7 35.9
Al2O3 31.9 13
a) b) c)
CaO 1.21 38.13
Figure 2. Testing program of this study: a) compression test, b)
MgO 0.81 7.5 indirect tensile strength test, c) flexural strength test
Fe2O3 3.48 -
SO3 0.25 -
K2O + In this research, the mechanical properties include
1.02 1.01 compressive strength, indirect tensile strength and flexural
Na2O
LOI 9.63 1.15 strength were determined. Abbreviations and Acronyms
*LOI: Loss of Ignition C. Compressive strength
The compressive strength of LGC using coating and un-
coating keramzite is illustrated in Fig. 3. The results are shown
that the compressive strength of LGC using keramzite varied
from 10.92 MPa to 38.80 MPa. It can be seen that the
compressive strength of specimens which used coating
keramzite is higher than that of un-coating keramzite
compressive strength about 7-38%. Normally, the
compressive strength of geopolymer concrete decreases when
the AL/GS increases from 0.7 to 0.9. This result is consistent
with the previous research [14] It can be explained that the
a) b) shape of kezamzite is rounded, so when the alkali liquid
Figure 1. Two types of keramzite: a) Coating, b) Uncoating increases, the compressive strength of both groups of samples
decreases. Besides, LGC with geopolymer paste coating
outside has more strength than the un-coating keramzite. The
Table 2. Mix proportions of LGC by ratio geopolymer paste covered the coarse aggregate. This layer
plays two important roles which are preventing the water
Mixture Fly Ash GGBS Na2SiO3 Keramzite FA inside alkali liquid go through keramzites structure and
GC1 0.2 0.8 0.7 0.45 1 making the kezamzites structure stronger.
GC2 0.2 0.8 0.8 0.45 1 D. Indirect tensile strength
GC3 0.2 0.8 0.9 0.45 1 According to Fig. 4, tensile strength of LGC is from 1.69
MPa to 3.71 MPa. And the specimens using coating keramzite
are higher than uncoating keramzite. This trend is followed by
391
the compressive strength. However, the development of easier. Thereby, the structure of LGC using uncoating
tensile strength for two cases is a little different especially for keramzite is strongly built and shows better flexural behavior.
un-coating keramzite specimens. In terms of LGC using On the other hand, the surface of coating keramzite is
uncoating keramzite, the tensile strength decreases 1.79 times smoother due to covered geopolymer paste layer. When the
with the increase of ratio AL/GS from 0.7 to 0.9 as the solution increases, the aggregate could not link together. As a
compressive strength trend. While the tensile strength of result, the structure of LGC using coating cannot resist the
coating specimens increases when AL/GS changes from 0.7 high flexural loading.
to 0.8 and reduces beyond 0.8. Noted that at the ratio AL/GS
of 0.7 and 0.9, the indirect tensile strength of LGC using
coating keramzite is almost same to each other (2.83 MPa
compared with 2.80 MPa). Noted that the indirect tensile
strength of LGC using coating keramzite reaches the highest
value at the ratio AL/GS of 0.8. This is a different trend from
the previous study.
Figure 5. Flexural strength of LGC with various AL/GS
IV. CONCLUSION
This study evaluated the mechanical properties of LGC
using coating and un-coating keramzite as coarse aggregate.
The results and investigations can be summarized as follows:
- Firstly, the compressive strength of LGC varied from
Figure 3. Compressive strength of LGC with various AL/GS 10.92 MPa to 38.80 MPa. The coating keramzite shows higher
performance than un-coating keramzite.
- The tensile strength of LGC follows the trends of
compressive strength. The speci-mens using coating
keramzite has higher strength than the one using un-coating.
- The flexural strength of LGC is in the range of 2.48 MPa
to 4.25 MPa. The flexural strength of LGC using coating
keramzite decreases with the increase of ratio AL/GS from 0.7
to 0.9. While the flexural strength of uncoating specimens
increases when AL/GS changes from 0.7 to 0.8 and reduces
when the ratio AL/GS keeps increasing to 0.9.
REFERENCES
[1] Davidovits J (1979) Synthesis of new high temperature

geopolymers for reinforced plastics-composites. 151–154
Figure 4. Tensile strength of LGC with various AL/GS [2] Turner LK, Collins FG (2013) Carbon dioxide equivalent
(CO2-e) emissions: A comparison between geopolymer and
OPC cement concrete. Constr Build Mater 43:125–130.
E. Flexural strength https://doi.org/10.1016/j.conbuildmat.2013.01.023
Based on Fig. 5, there are two tendencies of flexural [3] Nguyen KT, Ahn N, Le TA, Lee K (2016) Theoretical and
strength of LGC using keramzite as aggregate. Firstly, in the experimental study on mechanical properties and flexural
strength of fly ash-geopolymer concrete. Constr Build Mater
case of LPC using coating keramzite, the flexural strength 106:65–77. https://doi.org/10.1016/j.conbuildmat.2015.12.033
decreases 1.71 times with the increase of ratio AL/GS from [4] Nguyen KT, Le TA, Lee J, et al. (2017) Investigation on
0.7 to 0.9 as the compressive strength tendency. Secondly, the properties of geopolymer mortar using preheated materials and
flexural strength of uncoating specimens increases 8% when thermogenetic admixtures. Constr Build Mater 130.
AL/GS changes from 0.7 to 0.8 and reduces by 29.2% when https://doi.org/10.1016/j.conbuildmat.2016.10.110
the ratio AL/GS keeps increasing to 0.9. For experiment work, [5] Le AT, Nguyen KT, Nguyen TN, et al. (2020) Investigation of
when the gap is lower than 10%, it can be seen as a trivial mechanical properties of fly ash/rfcc based geopolymer
concrete
difference. Generally, the development of flexural strength of
LGC is same as the case of compressive strength. It can be [6] Davidovits J (2011) Geopolymer chemistry and applications,
3rd ed. Geopolymer Institute
explained that the surface of uncoating keramzite is rough.
[7] Yu QL, Spiesz P, Brouwers HJH (2013) Development of
Thus, when the alkali solution increases, the aggregate inside cement-based lightweight composites - Part 1: Mix design
the mixture could transport, arrange, and connect together
392
methodology and hardened properties. Cem Concr Compos weight geopolymers. Ceram Int 40:13891–13902.
44:17–29. https://doi.org/10.1016/j.cemconcomp.2013.03.030 https://doi.org/10.1016/j.ceramint.2014.05.108
[8] Yu QL, Spiesz P, Brouwers HJH (2015) Ultra-lightweight [13] ASTM C618-17a (2017) Standard specification for coal fly ash
concrete: Conceptual design and performance evaluation. Cem and raw or calcined natural pozzolan for use in concrete. Annu
Concr Compos 61:18–28. B ASTM Stand
https://doi.org/10.1016/j.cemconcomp.2015.04.012 [14] Pham KVA, Nguyen KT, Le TA, Lee K (2022) Investigation
[9] Posi P, Teerachanwit C, Tanutong C, et al. (2013) Lightweight of impact behavior of innovative non-curing steel fiber
geopolymer concrete containing aggregate from recycle geopolymer composites. Case Stud Constr Mater 16:e01011.
lightweight block. Mater Des 52:580–586. https://doi.org/10.1016/j.cscm.2022.e01011
https://doi.org/10.1016/j.matdes.2013.06.001 [15] ASTM C39 (2018) Standard test method for compressive
[10] Zuda L, Drchalová J, Rovnaník P, et al. (2010) Alkali-activated strength of cylindrical concrete specimens
aluminosilicate composite with heat-resistant lightweight [16] ASTM C496/C496M - 11 (2011) Standard Test Method for
aggregates exposed to high temperatures: Mechanical and Splitting Tensile Strength of Cylindrical Concrete Specimens.
water transport properties. Cem Concr Compos 32:157–163. In: ASTM International
https://doi.org/10.1016/j.cemconcomp.2009.11.009
[17] ASTM C78 / C78M-18 (2018) Standard Test Method for
[11] Kupaei RH, Alengaram UJ, Jumaat MZ Bin, Nikraz H (2013) Flexural Strength of Concrete (Using Simple Beam with Third-
Mix design for fly ash based oil palm shell geopolymer Point Loading)
lightweight concrete. Constr Build Mater 43:490–496.
https://doi.org/10.1016/j.conbuildmat.2013.02.071
[12] Masi G, Rickard WDA, Vickers L, et al. (2014) A comparison
between different foaming methods for the synthesis of light
393
Researching and Improving the Registration and

Treatment Process in Health Care Facility with
Lean Principles and Ergonomics Standards
Minh-Tai Le
Department of Industrial Systems Engineering
Faculty of Mechanical Engineering
tailm@hcmute.edu.vn
Abstract: The medical field in Vietnam today has been Some researches [1-2] presented an overview of the Lean
faced with many problems. The question here is how to continuous improvement framework's use in the healthcare
develop a model of registration and treatment system to setting. As a result, Lean initiatives in healthcare enhanced
reduce pressure on doctors, decrease patient waiting time but patient outcomes, as well as patient happiness, while
still bring efficiency. Direct observations and patient surveys simultaneously lowering operating costs and boosting
were initially made at Ai Nghia medical clinic - a small one financial performance. For example, two years after
located in Vietnam - to understand the processes and identify implementing lean (in 2002), Virginia Mason Medical
the initial causes of dissatisfaction. A value stream map Center saw a reduction in expenditures (53% less inventory),
(VSM) was then used to further break down the patient lead time (65 percent less), people and product distance (44%
flow’s complexity. A Fishbone diagram was completed to and 72% less, respectively), and setup time (82% less). In
find the root cause leading to the patient's dissatisfaction. The 2018, Alkinaidri et al. [3] claimed that the readiness to act by
information obtained from the Lean tools showed that the hospitalists that use the new principles of Lean within
registration and treatment process generates the duplication hospitals can deliver high-quality and efficient care to
of steps that cause unwanted delays. This study addressed patients. The authors also underlined the cultural and
difficulties regarding patient satisfaction on the registration practical barriers to overcome to spread the use of Lean
and treatment processes including reducing the unwanted techniques.
step, redesigning layout to mitigate these delays and
According to some researchers [4-6], patient waiting time
improving the process. This model improved on the basis of
and duration of stay are two common patient flow issues in
applying Lean Principles and Ergonomics standards. The
health care. Coffin [7] defined that Lean Health Care helps
procedure was performed on an outpatient clinic by modeling
to minimize ambulatory care waiting times and lengths of
and simulating the registration and treatment process using
stay, owing to its focus on identifying and reducing non-
the ARENA modeling. This project illustrated not only the
value-added activities. Despite this, the hospital's operating
application of Lean principles to resolve issues regarding a
procedures are researched and continuously improved as a
central registration and treatment process but also the
result of using Lean Principles. Staff’s advice for decreasing
importance in applications of Ergonomics in medical
the number of paperwork and unnecessary procedures in the
working space in Health Care Centre in general and Ai Nghia
medical process. It's tough to say whether Lean Principles are
Medical clinic in particular.
totally applicable to the medical field. Richard J.
Keywords: lean principles, ergonomics, health care Schonberger's article [8] discussed some of the issues with
facility, registration and treatment process. Lean's poor fit in the medical field. Firstly, customers are
closely involved in the healthcare sector. On the other hand,
I. INTRODUCTION most manufacturing workers have limited, if any, interaction
with real customers. Secondly, the healthcare sector can deal
It is clear that the global environment has created a highly
with a large number of patients at the same time.
competitive atmosphere, which leads all companies to
answer the question of how to achieve rapid development in A paper about the application of Ergonomics in
all areas. Nowadays, organizations are struggling as if to healthcare was published by Coffin and colleagues. This
expand their abilities and to maintain a sustainable overall study defined human Ergonomics and how it has become a
performance with a purpose to meet the best expectation of critical issue in Health Care services, as well as effective
all customers. With the hopes to satisfy the customers, methods to reduce the impact of people's factors in the
companies and organizations have developed many methods present. Chen and Colak [9-10] presented evidence that
and processes. Lean theory is one of the theories which have current patterns of HIT using may pose increased risks of
proved its effectiveness in both the manufacturing and work-related musculoskeletal disorders. Results from this
service fields. review can be used to optimize the implementation of future
HIT initiatives in ways that will benefit user performance

2022 6th International Conference on Green Technology and Sustainable Development (GTSD
while minimizing their injury risks. The ergonomic change applied to the patient to determine what elements became a
needs to involve both sports methods and working big deal from the patient's point of view. The survey was
environments. According to studies on office ergonomics conducted on 50 people, with 3 departments having the most
therapies, measures to reduce upper extremity visitors including: General Pediatric (30%), Diabetes
musculoskeletal complaints are likely to be useful for visual Mellitus (22%), Obstetrics and Gynecology (14%). The
problems as well. number of negative comments surveyed has been divided
into 4 main categories: waiting, too many steps, distance and
Other parts of this study are arranged as follows: Section
other (Table 1).
two gives a brief vision of methods and materials to solve the
issue. Section three is results and discussion, which not only However, this study only analyzed based on two criteria.
details the comparison of the KPIs in applying Lean Firstly, patients went to one of the three departments
principles but also illustrates the observation of how crucial mentioned above and did a complete basic procedure. It is
Ergonomics is in our daily life. The final section will noticeable that three issues of dissatisfaction in the given
summarize all the findings and the benefit of applying these table are interrelated. For the problem of too many steps,
essential principles. repeating the steps made the patient feel confused when they
had to go back and forth many times. The second thing was
II. METHODS waiting. Due to the area limitation of the clinic, the number
Data in this research was obtained from an actual survey of patients coming in is too much leading to congestion.
conducted at Ai Nghia. The location was observed in the Therefore, it could be summed up that reducing unnecessary
overall, as well as the medical assessment and treatment steps would decrease time in procedure.
process. The most complete examination process was
surveyed in 3 departments: General Pediatric, Diabetes Table 1. Unsatisfied categories
Mellitus, Obstetrics and Gynecology. After that, data was Unsatisfied categories
Depart- Total
filtered and analyzed appropriately to serve the research. The -ment
Waiting Many
Distance Others
Total
patient
%
time steps
study used 5 principles of Lean: Define, Value Stream General
3 4 0 5 12 15 80
Mapping, Create Flow, Establish Pull and Improvement Pediatric
Recommendations. Obstetrics and
2 2 2 2 8 11 73
Gynecology
Others 1 3 0 2 6 7 86
The first step was Define. Accurately defining value from Total 29 50 58
the patient’s point of view became an essential beginning and
by finding 3 common dissatisfaction problems. What needs The next thing to be done after determining the value was
to be done after determining the value was to determine the to determine the improvement method. First, the
improvement method for each problem. rearrangement of the layout of the departments reduced the
distance problem, optimizing the time to move between the
The second step was Value Stream Mapping (VSM). The patient's departments. Secondly, reducing unnecessary
goal of this step was to determine which activities create processes was the right direction, eliminating the steps at the
value and which activities are wasteful and should be main counter and carrying out procedures right at each
removed at each step of the value chain. With the help of the department. Last but not least, the distance between
ARENA software, the improved model’s usefulness would departments was not logical. At the Diabetes Mellitus
be further demonstrated. department, there were 2/8 people surveyed not satisfied with
Next step was Create Flow. Following the removal of the current clinic layout. They had to move a lot because of
wastes from VSM, the next step was to ensure that the the distance between the clinic and the laboratory. After
remaining activities in the value chain flow without analyzing all departments, the direction of improvement was
interruption, delay, or congestion, ensuring that the service given to improve the waste in moving and find out the way
reached patients' reach. Fishbone Diagram was built for to shorten the time.
analyzing root unsatisfied problems and making suggestions B. Value Stream Mapping
to solve those problems.
Table 2. Process activity mapping
The fourth step was creating a Pull system. The main
purpose of using Pull System was to limit the number of
queues and some options for creating Pull system were
offered as section III.D.
And the final step was Improvement Recommendations.
5S and Visual Management, Kanban and Ergonomic were all
applied with the purpose to fulfill the conception of Lean’s
method.
A. Define
The research applied 5 Lean’s Principles to clarify and
solve problems. The first stage was Define. A survey was
395
The goal of VSM was to determine which activities As VSM, the non-adding value was 3.5 times as long as
created value and which activities were wasteful and should the adding value time. The hospital's value to the patient was
be removed at each step of the value chain. This study paper low in comparison to the amount of time the patient spent
applied VSM to establish a foundation for better solutions. here.
Process Activity Mapping (Table 2) was built before VSM to
identify phases, intervals, and define actions.
Figure 1. Current VSM
beneficial if other issues such as personnel, management

capacity, and financial security are not considered. With the
help of the ARENA software, the improved model was
further demonstrated (Fig. 2).
C. Create Flow
Following the removal of wastes from Value Stream
Mapping, the next step was to ensure that the remaining
activities in the value chain flow without interruption, delay,
or congestion, ensuring that the service reaches patients.
Lean concepts and eight forms of waste reduced the impact
of these three sorts of problems, which has been discussed in
the first phase. In the case of other problems, many
implementation-based approaches have been proposed.
Patients had difficulty understanding the medical
examination and treatment process and spent a significant
amount of time asking questions about that. To avoid this,
the clinic should provide a quick and easy-to-understand
process guidance for the patient right at the front entrance.
They could also post information on the order of steps on the
clinic's website, or other social media platforms for patients
Figure 2. Layout improvement to refer. The clinic map should also be designed and
displayed in the Registration area so that the patient can
A long process model with many little steps could be easily locate the needed location. The signal problem was
observed. Although it made it easier for the owner to keep further illustrated in the Fishbone chart shown before. The
track of business finances, it inadvertently confused patient, in particular, has difficulty identifying the room label
customers and wasted time. Shortening the procedure by and it should be placed perpendicular to the wall, allowing
allowing fees to be paid at the lab and pharmacy is highly the patient to locate the room from afar.
396
Figure 3. Fishbone diagram
397
D. Establish Pull System personnel. Its benefits in Health Care central include:
The expressions "Push" and "Pull" System have their improves people's comfort, reduces the chance of medical
origins in logistics and supply chain management, but they staff being injured, reduces the chance of error. The first
are now widely used in the marketing and hotel distribution thing to mention is the design of the workspace with
industries as well. However, there are now a few medical maximum natural light. The doctor's working space should
facilities in Vietnam that use the Push-Pull System approach be prioritized where there are many windows to enhance
to improve the medical examination and treatment process alertness, concentration and work efficiency.
while also increasing patient satisfaction. Only “Pull” is The second thing that needs to be discussed is the neutral
impossible in health care. Because of natural variations in posture. Compared with conventional chairs, ergonomic
care procedures, each person's physical condition is unique; chairs will be more comfortable. The seat cushion slides in
no one is the same. Both Pull and Push System must work and out to help the spine stay close to the back of the chair,
together. Pull System, on the other hand, must be preferred adjustable seat height to prevent leg numbness, 3D/4D
over Push System. When a patient comes in and completes armrest suitable for all shoulder sizes and the last thing to
the Check-in procedure, it can predict which departments the mention is the high-quality, convenient, lightweight, airy,
patient will visit, allowing staff to provide the necessary durable material, but still ensures stability for users when
equipment. When the patient arrives in those departments, sitting and moving.
the procedure is performed immediately instead of waiting
for the staff to prepare as usual. Hospital beds are also One of the key principles of Ergonomics is the principle
available in the recovery room for patients who have recently of keeping everything within reach. Frequently used items
completed minor surgery. should be within arm's reach. In some cases, reaching too far
will cause some unnecessary injury: the arm hurts from
In industrialized countries, online medical examination trying to grab something far away from the sitting position
and treatment have become a widespread way of medical but does not want to stand up.
assessment. However, it is not particularly well-known or
One of the factors that lead to reduced human
widely used in Vietnam. The doctor is available 24/7 on the
performance is working in an uncomfortable space. Working
medical facility's website, and patients are able to arrange an
area needs to be set up carefully so that there would be
appointment whenever they need it. In circumstances when
enough space not only to create the feeling of being trapped
samples must be taken for testing, hospital staff is sent to the
in a box but also more likely to create injuries to the body. It
patient's home. Then, using internet platforms, examination
is not possible to design cabinets to place data directly above
and re-examination take place. The patient's prescriptions are
the head of the occupant. This will cause some injuries and
delivered to their home. This service reduces patient
bumps when standing up to move.
movement, waiting time, and contact with others while also
allowing patients to request an examination whenever they The last thing is to reduce the pressure point. One is a
need it. padded mouse pad. A specialized ergonomic mouse pads
usually have padding that protrudes slightly at the bottom
E. Improvement Recommendations
edge to specifically support the wrist. A desk with the right
5S technologies and Visual Management were applied to height will bring the worker a lot of advantages in work, as
give institutions with solutions to increase service quality, well as prevent diseases related to the spine. A desk that is
reduce treatment times, and reduce patient’s dissatisfactions. too low will make you always work with a stooped posture,
Factors were analyzed to remove any unneeded items such and if it is too high, it will be very tiring for your hands, back
as expired paperwork, damaged tools and equipment, useless and shoulders when sitting for a long time, which affects
advertising posters, and so on. To put everything in the work performance.
proper position, in the appropriate place, such as the waiting
area, corridor, doctor's working area, hospital bed area,
medical equipment area, and so on must be clearly
designated. Besides that, arranging organization is extremely
important in sectors like Medical Records and Pharmacy.
Documents in the Medical Record Room must be sorted in
groupings or alphabetical order to ensure accuracy,
rememberability, visibility, and retrievability. Maintain a
clean workplace at all times, devoid of trash and dirt on the
floor, equipment, and especially the patient restroom. There
have been several complaints concerning the lack of light,
odor, ruined facilities, dirty floors, and other issues in the
toilet area. To sustain and expand people's interest in 5S,
senior management should implement a program to
encourage the doctors and staff by organizing 5S emulation Figure 4. Doctor’s working table
and frequently evaluating performance across departments.
Kanban principles is a Lean management method that is
In addition to 5S and Visual Management, Ergonomics intuitive, easy to implement, but ensures the requirements of
was also applied in this study. In a medical setting, Lean including: the right type, the right quantity, the right
Ergonomics is crucial for both patients and medical
398
time. Last but not least, it can reduce the risk of running out Mellitus had a reduction of 56.293 % in Transfer Time.
of stock or having excess inventory causing waste. In this According to the previous data, the Department of Diabetes
study, we will apply the Kanban principle in managing goods Mellitus was strongly affected by Transfer Time. The clinic
in the pharmacy. was relocated to the ground floor, replacing the Medical
Record room, in order to achieve this improvement.
The arrangement of goods on the shelves will follow the
principle of two Kanban bins (two-bin Kanban System). Overall, because the method eliminates the returned steps
There are two identical bins, both bins are filled with goods to the original Registration counter, the Total Time can be
of the same type and the same quantity. Bin 1 stacked in front greatly decreased. The moving distance is cut in half, and the
and will be used first; Bin 2 stacked in the back, it will be wait time is greatly reduced. This aligns with Lean's goals of
used later. When the last goods in box 1 are taken out for use, increased labor productivity and the elimination of
bin 1 is also removed from its position and placed on the top occupations that do not bring value to customers.
shelf of the goods shelf in the “empty bins” position, and at Furthermore, the new process is based on Lean's eight waste
the same time pulls up the second bin replacing the position principles, such as reducing patient movement and waiting
of the empty one. The time when Bin 1 is placed in the time.
"empty bins" position, it will be a visual signal to help
Table 4. KPI (Continue)
pharmacists know that it is about time to add more items with
the right type and the right quantity listed on the Kanban Average WIP Average Usage of Clerks
card. While waiting for the items to be added to bin 1, the Current Improving % Current Improving %
pharmacist will use the ones in bin 2. When the goods have
been taken out of the box but for some reason it is not for the General
4.057 3.286 19 0.315 0.03 90
Pediatric
usage, they need to return the goods to the shelf.
Diabetes
3.535 3.23 9 0.223 0.04 82
Mellitus
Obstetrics
and 2.379 2.056 14 0.156 0.03 80
Gynecology
Reducing the number of turns back to the Registration

counter to complete procedures allows more time for clerks
to work on other assignments. It's important to note that these
statistics only include the use of clerks in the three
departments, as well as the patients who completed the
procedure. According to the survey report, up to 6 patients
complained about concerns with clerks at the Registration
Figure 5. Two – bin Kanban System counter. The time spent per patient increases as a result of the
Table 3. KPI reduced process connected with the Registration counter,
lowering customer complaints. However, the decrease in
Average Total Time Average Wait Time Average Transfer clerks usage has an impact on pharmacists and staff at
(min) (min) Time (min)
Pharmacy and Laboratory. They will be responsible for the
Impro- Impro- Impro-
Current
ving
% Current
ving
% Current
ving
% clerks' jobs, but it will be small because the duty is distributed
between both the Pharmacy and the Laboratory, efficiently
General
162.27 131.45 19 124.5 104.022 16 1.067 0.767 28 rather than having the Registration counter do everything.
Pediatric
About the WIP, after improving the model, General Pediatric
Diabetes dramatically improved the WIP value. The rationale for this
159.08 145.34 7 120.822 116.84 3 2.304 1.007 56
Mellitus is that the number of patients will be spread out, evenly
Obstetrics divided, and not overcrowded at any given time, particularly
and 142.71 123.38 14 105.919 96.306 9 1.286 0.769 40 in the Registration area and in front of each department.
Gynecology
IV. CONCLUSION
ARENA model simulations are done automatically and
result indications are generated to provide a good overview The research used five Lean principles and tools, along
of how the medical examination process is changing. The with Ergonomic principles, to improve the medical
patient's total time in the operation dropped considerably examination and treatment process, as well as the working
after improving the procedure as well as the distance moving, environment in medical institutions. The study has been
especially in the General Pediatric department. In terms of modified and developed to be more relevant to the
Waiting Time, the new method had a substantial influence on Vietnamese context, referencing many materials from many
the same department. The question is why, when compared sources. All of the research's recommendations for
to the other departments, the General Pediatric Department improvement was aimed toward the patient's benefit, with a
has the overall decrease in Total Time. The answer is obvious special focus on lowering patient dissatisfaction, particularly
from the start, General Pediatrics has enormous health needs. in focuses on the three difficulties described in the first
As a result, doing many steps will cause patients to wait principle. Since then, the patient's duration in the procedure
longer than in departments with fewer patients. Diabetes has been greatly reduced, and the doctors' and staff's
399
efficiency has risen as well. However, other [3] A. Alkinaidri, H. Alsulami, “Improving Healthcare Referral
recommendations, such as the online medical examination System Using Lean Six Sigma”, American Journal of Industrial
model, would be prohibitively expensive to implement in the and Business Management, vol. 8, no. 2, pp. 193-206, 2018.
existing model. The cost of developing and maintaining an [4] M. Mannon, “Lean Healthcare and Quality Management: The
online medical examination platform, as well as paying an Experience of ThedaCare”, The Quality Management Journal,
vol. 21, pp. 7-10, 2014.
on-call doctor, is unaffordable for small businesses.
[5] L. Cavalheiro, R. Caserta Eid, C. Talerman, C. Prado, F. Gobbi,
Besides, through all the above analysis, it can be seen that and P. Andreoli, “Design of an Instrument to Measure the
Ergonomics plays an essential role in our lives. By applying Quality of Care in Physical Therapy”, Health Economics and
Ergonomic principles, the workspace will be significantly Management, vol. 13, pp. 260-268, 2014.
improved and the risk will be minimized. Furniture designed [6] C. Nicolay, S. Purkayastha, A. Greenhalgh, J. Benn, S.
according to ergonomic principles will cost more than other Chaturvedi, N. Phillips and A. Darzi, “Systematic Review of the
items. That is why it becomes a dilemma for many places. Application of Quality Improvement Methodologies from the
However, if Ergonomics can be applied to hospital, medical Manufacturing Industry to Surgical Healthcare”, British
Journal of Surgery, vol. 99, pp. 324-335, 2012.
clinic interior designs, it will bring high efficiency to people's
[7] C. Coffin, “The Continuous Improvement Process and
working productivity. And it is obvious that comfort and
Ergonomics in Ultrasound Department”, Radiology
efficiency are always the first priority, no one wants after Management, vol. 35, pp. 22-25, 2013.
spending such a long time for several years devoted to their
[8] R. J. Schonberger, “Reconstituting Lean in Healthcare: From
career to have unwanted diseases like: insomnia, backache, waste elimination toward ‘queue-les’ patient-focused care”,
etc. Business Horizons, Elsevier, vol. 61, no.1, pp. 13-22, 2018.
REFERENCES [9] X. Chen, X. Wang, C.Q. Ai, “Application of Ergonomics in
Health Care”, 2020 International Conference on Advances in
[1] Alkhamis, “A. Health Care System in Saudi Arabia: An Biological Science and Technology, IOP Conference Series:
Overview”, Eastern Mediterranean Health Journal, vol. 18, pp. Earth and Environmental Science.
1078-1080, 2012. [10] S. Colak, A. Tekın Orha, M.D. Yener, T. Colak, B. Bamac, E.
[2] A. Karim and K.A. Arif-Uz-Zaman, “Methodology for Colak, “Musculoskeletal system-related complaint: Is there any
Effective Implementation of Lean Strategies and Its effect of sports ergonomics and lack of core stabilization
Performance Evaluation in Manufacturing Organizations”, exercises?”, Science & Sports, vol. 36, no. 6, pp. 481.e1-481.e7,
Business Process Management Journal, vol. 19, pp. 169-196, 2021.
2013.
400
Application of Lean and Six Sigma Tools to

Improve Productivity and Product Quality at Dien
Quang Company, Vietnam
Minh Tai Le Hoang Khang Lu Nhat Huy Do Huynh
Department of Industrial Systems Department of Industrial Systems Department of Industrial Systems
tailm@hcmute.edu.vn 18104020@student.hcmute.edu.vn 18104016@student.hcmute.edu.vn
Abstract: At present, the market demand is becoming such as: reducing operating costs, reducing waste, reducing
more and more strict according to the trend of global scrap and loss, controlling production warehouses with
socialization. This requires manufacturers to have options in flexible work stations. The results show that 3 factors are
production plans as well as strategies to go beyond important for Kanban implementation and lean
competitors to meet customer needs. Therefore, the efficient manufacturing: responsibilities of the Chief Manager,
use of production resources, reduction of excess wastes, supplier involvement, inventory management, and quality
reduction of production costs but do not want to restructure improvement. Anil S. Badiger et al. [2] identified options for
production too much is a problem for manufacturing improving device performance by implementing Kaizen and
poka-yoke. The study intends to improve overall
companies now. Along with the trend of integrating into the
performance to enhance productivity. The results showed
global economy, improving competitiveness is inevitable,
that OEE increased from 49.9% to 74.68%. Improved OEE
especially for Dien Quang, a leading enterprise in the field leads to better use of resources, increased availability,
of lighting and electrical equipment. In order to improve increased high-quality products, and improved employee
competitiveness, it is necessary to make the transition to a morale and confidence. Kumar et al. [3] applied lean to an
professional production environment, implement automobile company with the aim of reducing the cycle time
improvements to improve productivity and product quality. of truck body assembly. Research is mainly focused on the
Through the above issues, this paper applies Lean and Six assembly line. Cycle time reduction is done in two stages.
Sigma production methods to improve production capacity, First is line balancing, then is applying lean principles like
optimize production resources and meet the fierce demands 5S, VSM, etc. This research helps to identify wasteful
of the market. it doesn't affect the existing production operations and eliminate them step by step, therefore
process much as well as it reduces costs when changing the reducing cycle times by proper line balancing. Tyagi et al. [4]
production structure. implemented a reduction in the lead time of a product using
VSM for lean implementation. They found that VSM is the
Keywords: lean, six sigma, improve, cycle time, takt most important and influential tool for realizing the benefits
time. of lean philosophy in an organization. A gas turbine product
was selected for study and it was run on the workshop floor
I. INTRODUCTION to plot the status quo. Through the analysis of different
Nowadays, the application of tools to improve wastes and their root causes were identified. After
productivity and product quality is not too strange for eliminating waste, a final state map is drawn. The results
companies not only in the manufacturing sector but also in showed that the production time of the product was reduced
the commercial and service sectors. Regardless of the by 50%. Gracanih et al. [5] have shown that reducing lead
industry, these tools share the same basic principle: eliminate times is not enough, but reducing costs is a must for an
all non-value-added and wasteful activities from the business, organization to excel. They combined VSM as the most
while increasing productivity and product quality. Lean is a effective tool for identifying non-value-adding activities and
tool in the lean manufacturing process that eliminates as eliminating them with value stream costs and cost time
much waste as can be eliminated, creates more value but profiles. Value stream costing attempts to eliminate
requires fewer resources. These are 6 typical benefits of Lean: unnecessary costs associated with mapped processes in value
less process waste, reduced production time, less rework, stream mapping. Through those studies, it can be seen that
financial savings, reduced inventory. value stream mapping (VSM) is an essential tool to
implement Lean. By using VSM, it is possible to track the
Rahman et al. [1] showed that, implementing lean product from start to finish. All processes involved in
production with minimum inventory and reducing costs by product manufacturing can be observed and then optimized
Kanban system. The author has conducted the research by by eliminating waste through other lean tools like Kaizen,
implementing the Kanban system taking to many benefits poka-yoke, etc. Just in time (JIT) or Kanban systems also

lead to huge benefits including improved productivity, structured Six Sigma processes into organizations still has
reduced costs, and inventory levels. Kaizen and poka-yoke room for improvement. Those efforts all contribute to
implementations lead to increased overall equipment improving the efficiency and maximizing the value of the
performance (OEE), better resource utilization, and greater Six Sigma method. Effective Six Sigma principles and
safety for workers. Line balancing with other lean tools has implementations are more likely to succeed by continually
been shown to be beneficial in increasing line efficiency. improving organizational culture.
Six Sigma is a tool or a project-oriented management Following the success of projects applying Lean and Six
approach to improve an organization's products, services, Sigma in many different fields. In this article, we apply Lean
and processes by continuously reducing organizational and Six Sigma to improve the productivity and product
defects, improving productivity, product quality, and quality of a company in Vietnam. By evaluating and
financial performance. Process improvement leads to surveying in the field, it is possible to apply Lean and Six
improved customer satisfaction, increased market share, Sigma to improve the company's productivity in the most
profits, etc [6]. Adan Valles et al [7] conducted a Six Sigma effective, optimal, and best quality way. The implementation
project at a semiconductor company that manufactures ink of Lean and Six Sigma also helps to increase the satisfaction
cartridges for inkjet printers. Six Sigma is tested in the final of employees and customers, create a positive working
stage of measuring electrical characteristics to accept or environment, love work.
reject them. Electrical faults account for about 50% of all
failures. The results show that with the correct application of II. LITERATURE REVIEW
this approach and support for the organization's team and
A. Lean Manufacturing
employees, a positive impact on quality (approximately 50%
reduction in electrical failures) and it can be achieved other The term Lean Manufacturing first appeared in the 1990
important features for customer satisfaction. N.Abbes et al. book "The Machine that Changed the World" by authors
[8] developed a new model by integrating DMAIC and Daniel Jones, James Womack, and Daniel Roos on the shift
PDCA approaches and Six Sigma methodology to find the from mass production to lean manufacturing. It is used as
source of the measurement defect cutting unit and improve it. the name for the lean manufacturing method system, which
Its index capabilities at a business specializing in clothing.
The results show that top management involvement and continuously improves business processes. For Womack and
responsibility in performance measurement and Jones, Lean is defined as a five-step process: defining
communication are the most important success factors for the customer value, defining value stream, making it flow,
successful implementation of Six Sigma in SMEs. A study establishing pull, and striving for excellence. The different
was carried out in the garment industry by Abdur Rahman et levels include Lean Manufacturing, Lean Enterprise, and
al [9]. Implemented by applying DMAIC methodology to Lean Thinking. After all, the central goal of Lean is to
reduce product defects in a garment manufacturing produce finished products at the rate required by the
organization in Bangladesh and to investigate defects, root
customer and with minimal or no waste. Lean is defined as a
causes and provide solutions to eliminate these defects.
Analysis from the use of Six Sigma and DMAIC shows that systematic model that includes principles and tools for
broken stitching and open seams affect the number of improvement, from focusing on analyzing customer
defective products. As a result, about 35% of garment defects perspectives to creating value and thereby eliminating
were reduced, which helped the research organization to waste, any rationality is not necessary for an organization's
reduce defects and thus improve its Sigma level from 1.7 to production or service provision. Lean creates a huge benefit
3.4. Ansari et al [10] conducted a study on the application of that is optimally using resources, minimizing the time of a
Six Sigma methodology in the finance department. production cycle. Through the thorough implementation of
Specifically applied in a project to enhance Continuing
Account Reconciliation carried out by the finance continuous improvements to provide services, organizations
department of a major U.S. defense contractor. The goal of can create services and products that meet customer
the project is to streamline and standardize the setting and requirements while minimizing the waste that may occur.
maintaining costs and planning of all business activities The focus is on eliminating “anything other than the
within the current financial management process. The minimum amount of equipment, materials, parts, space, and
implementation of Six Sigma resulted in a significant time that are absolutely necessary to add value to the
reduction in the average cycle time and cost, per unit of product” [11].
activity, required to produce the required financial statements.
Six Sigma focuses not only on the process or product defects B. Six Sigma
but also on improving overall management performance.
Researchers have tried to integrate Six Sigma with other Six Sigma describes a business improvement approach
existing innovation management approaches to make Six that seeks to eliminate defects and errors in manufacturing
Sigma even more appealing to organizations that may not and service processes by focusing on outputs that are
have started yet. start or fully implement the Six Sigma important to customers and clear financial profit for the
methodology. The successful implementation and growing organization [12]. Six Sigma has been proven to be an
interest of organizations in the Six Sigma methodology have effective methodology to achieve significant savings by
exploded over the past few years. It is quickly becoming a
applying the DMAIC (Define, Measure, Analyze, Improve
key driver for many technology-driven, project-driven
organizations. Various approaches to Six Sigma have been and Control) cycle to an existing system. DMAIC is often
adopted to increase the overall performance of different used in Six Sixma to solve problems faced by businesses.
business areas. However, the integration of data-driven This method divides the improvement process into several
402
stages in favor of identification errors and their causes in company encountered many inadequacies leading to low
order to eliminate them. productivity in the production process such as unreasonable
arrangement of workers, and a large inventory of WIP in
C. Takt time many areas,...
Takt time, also known as production pace: is the time it
takes to create a product to satisfy customer requirements.
Takt time is a term used to allocate interrelated production
activities and is also a time criterion used to evaluate the
working capacity of enterprises through controlling time at
stages to be able to create continuous production.
D. Lead time Figure 1. LEDBU10 50W
Lead time, also known as production time: is the time The assembly process of LEDBU10 50W bulbs includes
elapsed at the beginning of a processing process from 13 stages as shown in the diagram below:
placing an order until the customer receives the product.
E. Cycle time
Cycle time is the time interval between two
consecutively created products. In fact, cycle time is the
completion time at each stage measured in many different
cases depending on the property of the production stage.
F. Bottleneck
Bottlenecks occur when workloads arrive too quickly Figure 2. Assembly process of LED bulb BU10-50W
and exceed production capacity. Inefficiencies caused by The required production quantity of the company is 5000
bottlenecks often create delays and higher production costs. bulbs/day and the working time of workers in a day is 8
hours and a 30-minute lunch break.
G. LOB
𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑡𝑖𝑚𝑒 1 𝑑𝑎𝑦 7.5 ∗ 60 ∗ 60
Line balancing is one of the typical Lean tools whose 𝑇𝑎𝑘𝑡 𝑡𝑖𝑚𝑒 = =
𝑅𝑒𝑞𝑢𝑒𝑠𝑡 𝑎 𝑑𝑎𝑦 𝑜𝑟𝑑𝑒𝑟 5000
main effect is to balance the volume performed at stages
= 5.4 (𝑠𝑒𝑐𝑜𝑛𝑑)
according to the flow of materials. The main purpose of
LOB is to balance the cycle time at all stages so that it is
less than or equal to the allowed takt time to avoid the
following phenomena: congestion, instantaneous machine
stop, wip,…
Formula for calculating line efficiency:

∑𝑛𝑖=1 𝑡𝑖
𝐸=
𝑁. 𝑅𝑡𝑡
Inside:
∑ 𝑡𝑖 is the total time of the work steps Figure 3. Line balancing performance diagram before improvement
N is the number of workplaces
𝑅𝑡𝑡 is the actual Takt time Line balancing performance:
With: ∑13
𝑖=0 𝑡𝑖 67.8
𝑅𝑡𝑡 = 𝑚𝑎𝑥{𝑅𝑖 } 𝐿𝑂𝐵 = = = 35.4%
𝑛 ∗ 𝑇𝑚𝑎𝑥 13 ∗ 14.7
𝑅𝑖 is the cycle time in the balanced workgroups. Large line unbalance: 64.6%
III. IMPLEMENTATION AND RESULT After analyzing and evaluating the current line balancing
performance, the authors found the problem point that needs
H. Productivity Improvement to be improved as follows:
The authors selected the LEDBU10 50W lamp (Fig 1),
• Improving the Bottleneck point at the E27 lamp
which is a typical product line of Dien Quang Lamp Joint
head mounting station with cycle time = 14.7
Stock Company and is produced in the highest volume of the
seconds.
year with about 150000 products/year. With such high sales,
this is a remarkable product and if it can be improved better, • Improving working operation at 2 stages: installing
it will bring significant profits to the company. Although the PCB and soldering because cycle time is larger
being a leading company in the production of light bulbs, the than takt time.
403
• Rearranging materials at 2 stages: installing driver Table 2. Current status of bottleneck stage after improvemnt
and aluminum heat sink because the unreasonable Result Image
layout affects the performance of the operation.
• Reducing the backlog of WIP at the label printing 5S is done better, WIP are no longer
stage because of narrowing the working area. stacked.
• Balancing performance improvements to enhance
productivity. Workers do not need to use too much
force to twist the lamp head, work
1) Improved the Bottleneck point productivity is high, errors at the later
stage are improved.
Cycle time at high E27 lamp head station with
Tmax = 14.7s.
The operation is performed
professionally, redundant operations
Table 1. Current status of bottleneck stage before improvemnt are limited, the material flow is not
Issue Image stagnant.
There is a large backlog of WIP at The distance between workers is

the stage of installing E27 lamp reasonably arranged, limiting
heads. collisions with each other when
working, easy to perform operations.
At the end of working hours, the 2) Standardizing working operations at 2 stages:

workers' hands hurt. Installing the PCB and soldering
The cycle time of the 2 stages of installing the PCB and
soldering is 8.1s and 5.9s respectively, which is 5.4s larger
The redundant operation, the time than the standard takt time. The cause is that workers
to complete this stage is quite high.
perform redundant operations and repeat the operation many
times for a product. After observing and finding out the
The number of workers at the stage cause, the authors proposed solutions: re-instructing
consists of 6 people, sitting close to working operations, visualizing standard operations at work,
each other, creating a small organizing a skills contest,... The results show that the cycle
working area that makes it difficult
to manipulate when working.
time after improvement in the 2 stages of installing the PCB
and soldering is 4.4s and 5.3s, and there is no longer a
From the problems described (Table 1), the authors have backlog of WIP (Fig. 4).
found the cause and proposed the following solutions:
• The WIP are backlogged because the production
process is not strict, 5S has not been implemented.
To improve this problem, the authors have used the
5S tool to rearrange the inventory of WIP in rows,
creating a partition between the layers to avoid
collisions between the WIP.
• The worker’s hands being hurt is due to the fact that
during the working process, the operation is mainly Figure 4. Before and after pictures of improvements in PCB
manual, there is no machine to support. Therefore, installation and soldering
the authors have proposed using a tool to twist the 3) Rearranging materials at 2 stages: installing driver
lamp head instead of manual operation. and aluminum heat sink
• Workers have to perform many operations, which For the driver installation stage: materials are stacked on
are difficult to identify and standardize. Therefore,
top of each other, difficult to perform operations (Fig. 5a);
the authors have standardized operations and divided
work according to one-piece flow. materials placed away from the worksite (Fig. 5b). The
cause is the improper manipulation of workers in the
• There are no regulations and standards on working
distances, places to place materials. The authors have implementation process; due to unreasonable arrangement
proposed to regulate the sitting distance between 2 of the working area.
people next to each other is at least 1 arm, and
materials will be placed in the middle position.
After applying the recommendations given in the
improvement process, the following results are obtained
(Table 2). Cycle time at bottleneck was reduced from 14.7s
to 5.7s.
Figure 5. The current status before the improvement of the driver
installation stage
404
For the installation of the aluminum heatsink: The than the standard takt time, and the line balancing efficiency
aluminum heatsinks are scattered on the conveyor, causing is increased by 40%.
the material flow to be blocked (Fig. 6). The source of this
problem is that the stage of supplying raw materials is not
reasonable.
Figure 6. The current status before the improvement of the

heatsink aluminum installation
Figure 9. Current status of the label printing stage after
The solution for these two stages is: arranging the sitting improvement
position and material storage area reasonably, standardizing
the working operation, and supplying the right materials I. Quality Improvement
with the current quantity. 1) Define
In the process of productivity improvement research, this
study found problems affecting the quality of BU 10-50W
LED products.
2) Measure
The paper identified 13 main defects affecting the
Figure 7. The current status after the improvement of the driver
installation stage (a) and the heatsink aluminum installation (b)
quality of the BU 10-50W LED as follows: the dented lamp
head, the installed wrong lamp head, the inactive lamp,
After implementing improvement measures, the results exploding lights during testing, the aluminum heatsink is
show that the materials are arranged neatly, no longer vibrated, the driver fell from the lamp body, short circuit,
overflowing out of the conveyor belt, the workers are easier the light is dimly lit, blinking when testing light, the screw
to manipulate to take materials (Fig. 7). is not in the hole, touching the board, solder PCB lacks lead,
2 contacts touch each other, heat sink glue smeared outside
4) Rearranging materials at 2 stages: installing driver
the lamp housing, the DC wire is not outside the aluminum
and aluminum heat sink
heatsink, DC and PCB wire are not tightly sticky, blurry
At this point, the authors have performed 5 times label, printed in the wrong position.
random sampling of WIP, 10 minutes for each time. The
results show that the total number of WIP in stock in a day The authors have taken 10 samples in 9 different times
is 315 products. This narrows the working area of the corresponding to the established time points within 1 week.
workers (Fig 8). The cause is overproduction, not capturing The results show that the total number of defects is 1218
the quantity done at a later stage. The solution is to produce and the average production quantity is 4392 products.
just enough according to order requirements. Figure 10 shows the defects rate of BU 10-50W LED.
Number of detected defects ∗ 106

𝐷𝑃𝑀𝑂 =
Number of chances of a defect
6
1218 ∗ 10
= ≈ 21332
4392 ∗ 13
From that result, the current sigma level of the product
is:  = 3.5
Figure 8. Current status of the label printing stage before
improvement
5) Balancing of Line
After improving the stages with a cycle time greater than
the standard Takt time, the improvement reduced the
amount of WIP on the stage, some manipulation and re-
arrangement problems at work positions. Re-measuring the
average cycle time and redraw the line balancing
performance diagram after improvement (Fig. 9). Then,
recalculate the line balancing efficiency and get LOB = Figure 10. The Pareto chart shows the order of materiality of
75%. In general, most of the cycle time at the stages is less defects
405
3) Analyze The lamps that don’t work (Fig. 15) accounted for about
16.26% of the total defects in the stages and occurred in the
The lamp head is dented after performing a single-pins
light test stage.
operation (Fig. 11).
Figure 11. Dented lamp head Figure 15. The lamp does not work
Figure 12 indicates Fishbone diagram that is used to Figure 16 indicates Fishbone diagram that is used to
determine the root cause of the dented lamp head: determine the root cause of the inactive lamp
Figure 12. Fishbone diagram analyzes the defect of the dented lamp Figure 16. Fishbone diagram analyzes the defect of the lamp head
head installed wrong
The wrong installed lamp head (Fig. 13) accounted for 4) Measure
16.58% of the total defects of the stages and occurred at the
Through the root causes of product defects, it was found
stage of installing the E27 lamp head.
that improper working practices of workers and substandard
materials are the two main causes affecting the quality of the
BU 10-50W LED. Therefore, the authors made some
suggestions for improvement as follows: upgrading the
inspection procedure of the QC department, visualizing
operations.
Figure 13. The lamp head installed wrong

Figure 14 indicates Fishbone diagram that is used to
determine the root cause of the wrong installed lamp head
Figure 17. The graph shows the number of defects that appear
before and after the improvement
After implementing improvement solutions, the results
showed that the number of product defects decreased to 925
(76% reduction). This shows that the proposed solutions
have brought significant efficiency (Fig. 17).
Number of detected defects ∗ 106

𝐷𝑃𝑀𝑂 =
Figure 14. Fishbone diagram analyzes the defect of the installed
Number of chances of a defect
6
wrong lamp head 925 ∗ 10
= ≈ 16200
4392 ∗ 13
Sigma level after improvement:  = 3.63
406
5) Control K. Quality improvement

After making improvements to reduce the defect rate, the • The defects rate was reduced in 3 categories: the
authors continue to control, track the defects occurring dented lamp head decreased by 26%, the wrong
within 2 weeks and obtain the control chart (Fig. 18) as installed lamp head decreased by 17%, the inactive
shown below. lamp decreased by 12%.
• The inspection capacity of the QC department is
improved, and the data is objectively aggregated,
with the high reliability.
After implementing the improvement solutions, the
results show that there is a positive change, the company's
productivity is improved. This gives a high competitive
advantage over competitors in the same industry.
REFERENCES
[1] Rahman, N. A. A., Sharif, S. M., Esa, M. M. (2013). Lean
Manufacturing Case Study with Kanban System
Implementation, International Conference on Economics and
Business Research, vol. 7, 2013, 174-180.
[2] Badiger, A. S., Gandhinathan, R., Gaitonde V. N. and Jangaler,
R. S. (2010). Implementation of Kaizen and Poka-yoke to
Enhance Overall Equipment Performance - A Case Study,
Manufacturing and Industrial Engineering, vol. 1, Issue 4, 24-
29.
[3] Kumar, S. S., Kumar, M. P. (2014). Cycle Time Reduction of a
Truck Body Assembly in an Automobile Industry by Lean
Principles, International Conference on Advances in
Manufacturing and Materials Engineering, vol. 5, 2014, 1853-
1862.
[4] Tyagi, S., Choudhary, A., Cai, X., Yang, K. (2015). Value
stream mapping to reduce the lead-time of a product
development process, International Journal of Production
Economics, vol. 160, 202-212.
[5] Gracanih, D., Buchmeister, B., Lalic, B (2014). Using Cost
Time Profile for Value Stream Optimizaton, 24th DAAAM
International Symposium on Intelligent Manufacturing and
Automation, vol. 69, 2014, 1225-1231.
[6] Antony, J. and Banuelas, R. (2002). Key ingredients for the
effective implementation of Six Sigma program, Measuring
Business Excellence, 6(4), 20–27.
Figure 18. The control charts of defects
[7] Adan Valles (2009) Implementation of Six Sigma in a
IV. CONCLUSION Manufacturing Process: A Case Study, International Journal of
Industrial Engineering, 16(3), 171-181.
J. Productivity Improvement [8] N.Abbes, N.Sejri, Y.Chaabouni, and M.Cheikhrouhou (2018).
• Improved the Bottleneck point at the E27 lamp head Application of Six Sigma in Clothing SMEs: A case study,
mounting station reduced from 14,7s to 5,7s. Materials Science and Engineering, vol. 460, 20-22.
[9] Abdur Rahman, Salaha Uddin Chowdhury Shaju, Sharan
• The operation at the two stages of installing the Kumar Sarkar, Mohammad Zahed Hashem, S. M. Kamrul
PCB and soldering is standardized, so the cycle Hasan, Ranzit Mandal, Umainul Islam (2017). A Case Study of
time at these two stages is reduced and is smaller
Six Sigma Define-Measure-Analyze-Improve-Control
than the standard takt time.
(DMAIC) Methodology in Garment Sector, Independent
• Materials at two stages of installing the driver and Journal of Management & Production, vol. 8, No. 4, 1309-
installing heatsink aluminum are arranged 1323.
reasonably, easy to operate, and timely feeding time. [10] A. Ansari, Diane Lockwood, Emil Thies, Batoul Modarress,
Jessie Nino. Application of Six-Sigma in finance: a case study,
• Semi-finished products remaining on the label
Journal of Case Research in Business and Economics.
printing and other stages are optimally reduced.
[11] Dumitrescu & Marilena Dumitrache (2015). The Impact of
• The line balancing efficiency reached 75%, an Lean Six Sigma on the Overall Results of Companies,
increase of 35% compared to before the Economia. Seria Management, vol. 14(2), 535-544.
improvement. [12] Evans, J. R. (2016). Quality and Performance Excellence,
Quality Management Journal, vol. 22(4), 6-17.
407
Hydrogen Plasma Annealed Gallium and Aluminum

co-doped Zinc Oxide Films Applied
in Low-emissivity Glass
Shang-Chou Chang1,2,3 Yun-Che Tsai Huang-Tian Chan
1
Department of Electrical Department of Electrical Green Energy Technology Research
Engineering Engineering Center
2
Green Energy Technology Research Kun Shan University Kun Shan University
Center Tainan City, Taiwan Tainan City, Taiwan
3
University Social Responsibility gb122082099@yahoo.com.tw a106000057@g.ksu.edu.tw
(USR) team
Kun Shan University
Tainan City, Taiwan
jchang@mail.ksu.edu.tw
Jian-Liang Lai1,2 Jian-Syun Wu1,2 Wei-Min Wang1,2
1 1
1
Department of Electrical Department of Electrical Department of Electrical
2 2
2
University Social Responsibility University Social Responsibility University Social Responsibility
(USR) team (USR) team (USR) team
Kun Shan University Kun Shan University Kun Shan University
Tainan City, Taiwan Tainan City, Taiwan Tainan City, Taiwan
abc86159263@gmail.com kevin88774917@gmail.com w5210@mail.ksu.edu.tw
Abstract: Low emissivity (Low-e) glass has high visible light application fields is a way to save energy [1]. Many countries
transmittance, high infrared reflectance and provides energy provide product standards of windows (such as ISO 9050, EN
saving. This study deposited gallium and aluminum co-doped 1096-3:2012 et al.) used for building applications [2]. Low-e
zinc oxide (GAZO) films on glass by in-line sputtering as low-e glass often used in modern buildings, is high thermal
glass. This paper study of the effect of substrate temperature and resistance and visible light transmission. Low-e glass save
post-annealed on the structure, optical properties, electrical energy and may also provide enough ambient light and
properties, and emissivity of GAZO films. The GAZO films comfort [1, 3].
were deposited on glass substrates. The substrate temperature
varied from room temperature to 200 degrees Celsius. After that, At the moment, transparent conducting oxide materials
GAZO films were post-annealed in hydrogen plasma. such as aluminum-doped zinc oxide (AZO) are applied to
Experiment results indicate hydrogen plasma annealed method Low-e material, Low-e glass, etc [4, 5]. Emissivity is an
of GAZO films indeed influences the structure, optical, and important index for low-e glass. High infrared reflectivity
electrical properties. The optimum processing condition is corresponds to low emissivity. According to the Hagen-
hydrogen plasma treatments for the GAZO films deposited at Rubens relationship [6], a correlation between emissivity and
200 degrees. The electric resistivity and average optical electrical resistivity, the lower the electrical resistivity, the
transmittance in the visible range are 13.4 ×10−4 Ωcm and 90%, lower the emissivity. Recent studies have shown that
respectively, for the GAZO films deposited at 200 degrees. After hydrogen plasma treatment and annealing treatment could
plasma treatment, the electric resistivity and average optical reduce the electrical resistivity of AZO films [7-9].
transmittances are 7.8 ×10−4 Ωcm and 96%, respectively. The Improvement in electrical properties could be ascribed to the
decrease of electric resistivity for post-annealed possible causes desorption of oxygen and/or the replacement of Zn ions with
includes the substitution of Al and Ga ions at Zn ions or Al ions [7, 8].
desorption of oxygen species in GAZO films. The emissivity of
GAZO films is accompanied by a decrease in electrical On the other hand, Gallium and Aluminum co-doped Zinc
resistivity. At all substrate temperatures, the emissivity in post- Oxide (GAZO) can have the advantages of both AZO and
annealed GAZO films is under 0.18 which decreases at least Gallium doped Zinc Oxide (GZO). GAZO had good electrical
37% compared with that of the as-deposited ones. The hydrogen properties compared to AZO and that was the lowest
plasma treatments GAZO films can be applied to Low-E glass. resistivity of GAZO [10-12]. S.-C. Chang et al. reported that
This paper contributes to the industry of energy-saving the structural, electrical and optical properties of the GAZO
buildings. films prepared at various substrate temperatures were
Keywords: gallium and aluminum co-doped zinc oxide, investigated [12]. In addition, hydrogen plasma annealing and
hydrogen plasma, low-emissivity, sputter deposition, energy- vacuum annealing of GAZO could improve the electrical
saving buildings. properties of GAZO [13].
In this paper, GAZO films were prepared on glass
I. INTRODUCTION substrates via in-line sputtering technique. The substrate
Low emissivity (low-e) technology having wide temperature varied from room temperature to 200 degrees
Celsius. After that, GAZO films were post-annealed in

hydrogen plasma. It causes good electrical properties and

optical properties for GAZO films. The structure, electrical
properties and optical properties of GAZO films were
measured. This study surveyed post-annealed GAZO films for
low-e glass.
II. EXPERIMENTS
The GAZO films were deposited on glass substrates by
employing the in-line sputtering system. The prepared GAZO
films have a thickness is 500 nm. The GAZO films were
sputtered using a GAZO target (ZnO:Ga2O3:Al2O3 = 97:2:1
wt.%) at DC power of 7 kW, with Ar flow of 500 sccm and
process pressure of 3 mTorr. The substrate temperature is
between room temperature (RT) and 200°C. After deposition,
all produced GAZO films were post-annealed in hydrogen
plasma. The hydrogen plasma treatment was operated with a
processing power of 600 W at a pressure of 25 Torr for 10
min. The GAZO films properties are measured before and
after hydrogen plasma treatment. The crystal structure was
checked by an X-ray diffraction meter (XRD, MO3XHF22,
MAC-Science). Surface morphologies were observed with the Figure 1. XRD patterns of the as-deposited and hydrogen plasma
scanning electron microscopes (SEM, JEOL 6700). Electrical annealed GAZO films: deposited at RT before and after annealed
resistivity, hall mobility, and carrier density were obtained (top view); (b) deposited at 200°C before and after annealed
using a Hall measurement system (Ecopia HMS-3000). The (bottom view)
sheet resistance of GAZO films was obtained using a four-
point probe. The emissivity values were calculated using the
following Equation (1) [14].
ε=0.0129×R-6.7×10-5×R2 ()
where R is sheet resistance of GAZO films.
The XRD spectra of as-deposited and hydrogen plasma
treatment GAZO films prepared at different substrate
temperatures were shown in Fig. 1. All the GAZO films
(a)
sample were oriented preferentially along (002) of zinc oxide
direction from the XRD patterns. For before hydrogen plasma
treatment samples, the (002) peak shifts toward a higher
diffraction angle with enhancing deposition temperature
(substrate temperature). The peak with respect to (002)
increases when after different deposition temperatures, which
may imply that the distance between crystal plane reduces of
GAZO films has changed. It could be related to the
phenomenon that residual stress of the GAZO films or Al
atoms replaced Zn in the ZnO crystal lattice [12].
The SEM micrographs of the as-deposited and the
hydrogen plasma treatment GAZO films were shown in Fig. (b)
2. The SEM micrographs of as-deposited GAZO films were
similar for substrate temperatures of RT and 200°C. After
hydrogen plasma treatment, there were visible changes in the
morphology. Fig. 2 (b) and (c) displayed that the surface
grains of the hydrogen plasma treatment films were apparently
etched.
Optical transmittance spectra of GAZO films before and
after hydrogen plasma treatment were shown in Fig. 3. The
average transmittance in the visible region (400~800 nm)
were shown in Table 1. For before hydrogen plasma treatment
samples, an increase in the transmission is observed with an
increase in substrate temperature. The average transmittance (c)
for the films deposited at substrate temperatures RT and Figure 2. Scanning electron micrograph of the as deposited and
200°C were 76 and 90%, respectively. When after annealing hydrogen plasma annealed GAZO films: (a) as-deposited at RT;
in hydrogen plasma the average transmittances were 97% and (b) deposited at RT and annealed at hydrogen plasma; (c)
96%, respectively. deposited at 200°C and annealed at hydrogen plasma
409
plasma treatment, the electric resistivity and average sheet

resistance are 7.8 ×10−4 Ω-cm and 15.6Ω/cm2, respectively.
The decrease of electric resistivity for post-annealed (plasma
treatment) possible causes includes the desorption of oxygen
species in GAZO films. These results are similar to those of
S.-C. Chang et al. [13].
As Table 1 shown, the resistivity decreases and the
emissivity decreases for plasma annealed GAZO films. The
emissivity for the films deposited at substrate temperatures RT
and 200°C were 0.31 and 0.30, respectively. When after
annealing in hydrogen plasma the average transmittances
were 0.17 and 0.18, respectively. At all substrate temperatures,
the emissivity in post-annealed (plasma treatments) GAZO
films is under 0.18 which decreases at least 37% compared
with that of the as-deposited ones. It could be observed that
the plasma treatment reduces the electrical resistivity and
lowers the emissivity of GAZO films. The phenomenon
Figure 3. Optical transmission spectra of the as deposited and follows the Hagen-Rubens relationship [6]. The emissivity of
annealed GAZO films: deposited at RT and annealed at hydrogen GAZO films reduces when after hydrogen plasma annealing,
plasma; deposited at 200°C and annealed at hydrogen plasma which may imply that the GAZO films can be applied in low-
e glass.
Table 1. Structure, optical properties, electrical properties and
emissivity of the before and after hydrogen plasma treatment IV. CONCLUSION
processed of GAZO films
In this paper, in the case of hydrogen plasma annealed
Substrate temperature samples, the electrical and optical properties improved in
comparison with as-deposited. After plasma annealing, the
Room 200°C
temperature
emissivity decreases with decreased electrical resistivity of
GAZO films. After annealing, transmittance of all films
boosted in the visible range, and film prepared at 200°C
Hydrogen Plasma substrate temperature showed the highest value at 96%. The
Before After Before After
Treatment minimum emissivity can be less than 0.18. The produced films
will be a good performance of low emissivity and high
Average transmittance. This work contributes to the energy saving
transmittance in buildings.
visible region 76 97 90 96
(400~800 nm) ACKNOWLEDGMENT
(%)
This work was supported in part by the Ministry of Science
Resistivity and Technology of Taiwan, ROC, under MOST 110-2637-E-
13.9 7.2 13.4 7.8 168-004, and in part by the Green Energy Technology
(10-4Ω-cm)
Research Center, Department of Electrical Engineering, Kun
Carrier Density Shan University, Tainan, Taiwan, through The Featured Areas
8.3 9.8 8.5 7.3
(1020cm-3) Research Center Program within the framework of the Higher
Education Sprout Project by the Ministry of Education (MOE)
Hall Mobility
5.6 9.2 5.7 11 in Taiwan. We also thank University Social Responsibility
(cm2/Vs)
(USR) team for their support to this study.
sheet resistance
27.8 14.4 26.8 15.6 REFERENCES
(Ω/cm2)
[1] G. Ding and C. Clavero, “Silver-Based Low-Emissivity
Coating Technology for Energy- Saving Window
Emissivity 0.31 0.17 0.30 0.18 Applications”, in Modern Technologies for Creating the Thin-
Film Systems and Coatings, N. N. Nikitenkov, Eds.
The electrical properties and emissivity of films deposited IntechOpen: London, UK, 2017, Chapter 20, pp. 409-431.
at different temperatures and post-annealed were shown in [2] M. Kamalisarvestani, R. Saidur, S. Mekhilef, F. S. Javadi,
Table 1. For before hydrogen plasma treatment samples, an “Performance, materials and coating technologies of
increase in carrier density is observed with an increase in thermochromic thin films on smart windows”, Renew. Sustain.
Energy Rev., vol. 26, pp. 353-364, October 2013.
substrate temperature. The plasma treatment significantly
[3] S. Somasundaram, A. Chong, Z. Wei, “Energy saving potential
decreases electrical resistivity and lowers the sheet resistance of low-e coating based retrofit double glazing for tropical
of GAZO films. A decrease in electric resistivity for plasma climate”, Energy Build., vol. 206, pp. 109570, January 2020.
annealing could be ascribed to some Zn atoms being [4] T. Horiuchi, and K. Yoshimura, “Solar heat gain coefficient
substituted by Al atoms in GAZO films and/or desorption of and heat transmission coefficient of Al-doped ZnO thin-film
oxygen species in GAZO films. coated low-emissivity glass”, J. Ceram. Soc. Japan., vol. 128,
pp. 220-223, April 2020.
The optimum processing condition is hydrogen plasma [5] E. Ando, and M. Miyazaki, “Moisture resistance of the low-
treatments for the GAZO films deposited at 200 °C. After emissivity coatings with a layer structure of Al-doped
410
ZnO/Ag/Al-doped ZnO”, Thin Solid Films, vol. 392, pp. 289- [10] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, X. Ma,
293, July 2001. H. Song, “Comparative study of the sintering process and thin
[6] E. Hagen, H. Rubens, “Über Beziehungen des Reflexions- und film sputtering of AZO, GZO and AGZO ceramics targets”,
Emissionsvermögens der Metalle zu ihrem elektrischen Ceram. Int., vol. 40, pp. 12905-12915, September 2014.
Leitvermögen”, Ann. Phys., vol. 11, pp. 873-901, April 1903. [11] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, L.
[7] H. Tong, Z. Deng, Z. Liu, C. Huang, J. Huang, H. Lan, C. Zhang, R. Wu, “Investigation of aluminum–gallium Co-doped
Wang, Y. Cao, “Effects of post-annealing on structural, optical zinc oxide targets for sputtering thin film and photovoltaic
and electrical properties of Al-doped ZnO thin films”, Appl. application”, J. Alloys Compd., vol. 575, pp. 174-182, October
Surf. Sci., vol. 257, pp. 4906-4911, March 2011. 2013.
[8] W. Yang, Z. Wu, Z. Liu, A. Pang, Y. L. Tu, Z. C. Feng, “Room [12] S.-C. Chang, “In-line sputtered gallium and aluminum codoped
temperature deposition of Al-doped ZnO films on quartz zinc oxide films for organic solar cells”, Int. J. Photoenergy,
substrates by radio-frequency magnetron sputtering and effects vol. 2014, pp. 1-6, August 2014.
of thermal annealing”, Thin Solid Films, vol. 519, pp. 31-36, [13] S.-C. Chang, “Post-annealed gallium and aluminum co-doped
October 2010. zinc oxide films applied in organic photovoltaic devices”,
[9] F. H. Wang, H. P. Chang, C. C. Tseng, C. C. Huang, “Effects Nanoscale Res. Lett., vol. 9, pp. 562, October 2014.
of H2 plasma treatment on properties of ZnO: Al thin films [14] J. Szczyrbowski, A. Dietrich, K. Hartig, “Bendable silver-
prepared by RF magnetron sputtering”, Surf. Coat. Technol., based low emissivity coating on glass”, Sol. Energy Mater.,
vol. 205, pp. 5269-5277, September 2011. vol. 19, pp. 43-53 September 1989.
411
Study on Pectinase Production by Bacillus subtilis

in Molasses and Its Application for Coffee
Fermentation
Ngan K. Le Duy Q. Nguyen Nhi Y. Dinh
Department of Food Technology, School Faculty of Chemical and Food Department of Food Technology,
of Biotechnology Technology School of Biotechnology
International University _ Vietnam Ho Chi Minh City University of International University _ Vietnam
National University Technology and Education National University
lkngan9698@gmail.com duynq@hcmute.edu.vn dynhi1417@gmail.com
Phu H. Le*
Department of Food Technology, School of Biotechnology
International University _ Vietnam National University
lhphu@hcmiu.edu.vn
*
Corresponding author
Abstract: Pectinase is also a well-known term for range of bacteria, yeast, and fungi [5]. Submerged
commercial enzyme preparation that has a wide application; fermentation is one of the two primary types of fermentation
however, the high cost of industrial extracellular enzymes has depending on the kind of substrate utilized during
led to the high cost of finished industrial products. This fermentation [6].
research aimed to study the suitable conditions to produce
pectinase by using Bacillus subtilis fermentation by using Bacilli are particularly important in the fermentation
sector because they produce a wide range of valuable
molasses medium for the purpose of reducing pectinase
production expense. The enzyme activity was determined by enzymes and antibiotics [6]. This study surveyed Bacillus
measuring the absorbance at λ = 540 nm with 3, 5-DNS subtilis for pectinase production in submerged fermentation.
reagent. As the result, pectinase was achieved from molasses Furthermore, investigation of its effectiveness in the viscous
with the supplement of 1% of (NH4)2SO4 and 3% of pectin layer removal in coffee beans. Besides that, this study used
substrate, initial pH 5 and incubated at 37oC and for 5 days cheap raw materials such as molasses (sugar by-product) to
had the highest pectinase activity at 76.91±0.15 (U/ml). Then, reduce the cost of production.
the crude pectinase was applied to the de-mucilage of coffee II. MATERIALS AND METHODS
layers; investigated conditions of coffee fermentation for the
highest soluble solid content (TSC) extraction. For coffee A. Materials
treatment, the highest amount of removed mucilage was Bacillus subtilis strain was supplied by International
52.53%, with 3% crude enzymes (v/w) incubated at 35oC for University – Vietnam National University, HCMC. The
2 days (48 hours) and was 2.4 times higher than these amount isolated strain was inoculated into a sterilized molasses
of the control. The highest soluble solid content was medium (37oC) to reproduce the pectinase. Molasses and
4.01±0.08 (oBrix) with 50% moisture content of coffee beans, barley malt were bought from the supermarket with high
3% enzyme (v/w), and incubated at 35oC for 36 hours. This quality. Robusta coffee cherries were bought and transported
result can be used as a reference for application on pilot-scale from Kontum City, Gia Lai Province, Vietnam. Qualities of
or further study. the beans were uniform in size and shape with no
contamination of insects and broken beans.
Keywords: bacillus subtilis, coffee beans fermentation,
molasses, pectinase, total soluble solids content B. Methods
1) Preparation of malt extract agar (MEA) and culturing
I. INTRODUCTION B. subtilis
Coffee is one of the most important and widely used
commercial products in the world. Hence, coffee should be Prepared MEA with 200 g of barley malt, 1 L of distilled
investigated to improve the quality and variety of its products water, and ground well by blender. Filtered through filter
for both health and economic reasons. In green coffee, pectin cloth three times. Then put the solution in the water bath at
covers about 33% of beans [1] in the mucilage layer and 52oC (20 min), 63oC (30 min), 73oC (30 min), and 100oC (15
causes problems with its elimination. Enzymatic treatment by min). Adjusted Brix degree to 10° by D-glucose [7], pH 5.5.
using pectinase is an effective way to reduce pectin because After that, added 1% of (NH4)2SO4, 20 g of agar powder/ 1 L.
pectinase has the ability to hydrolyze pectin and cause The MEA solution was then poured 10 ml to test tubes and
pectin–protein complexes to flocculate [2, 3, 4]. Pectinase sterilized at 121oC for 20 min. All the test tubes were placed
synthesis via fermentation has been reported from a wide

in slant until they were well solidified [8]. B. subtilis was

slant cultured on MEA and incubated at 37oC for 7 days [9]. 60
2) Pretreatment of molasses 47.67d
Pectinase activity (U/ mL)

50
Crude molasses was diluted by distilled water until it
40
reached 11oBrix [10], then added 3.5 ml of concentrated
30.72c
sulfuric acid per 1 L of crude molasses before shaking (125 30.62c
30
rpm, room temperature, 24 hours). After that, it was filtered 23.71b
by vacuum pump with filter paper two times and adjusted to 20 16.02a
pH 5.5 by NaOH 6 M. Then, it was sterilized at 121oC for 20
minutes before use [11]. 7.72e
10
3) Production of pectinase by B. subtilis
0
a) Effect of substrate concentration 0 1 2 3 4 5
Substrate concentration (%)
In the 250 ml Erlenmeyer flask containing 75 ml of
treated molasses, 1% (NH4)2SO4 for nitrogen source, the
initial pH was 5, 75 ml of pectin substrate solution from 1 to Figure 1. Effects of substrate concentration to pectinase
5% of pectin. Pectin substrate was the solution of citrus pectin
(w/v) mixed in acetate buffer 0.1 M, pH=4. The control was activity
the sample without adding pectin. After sterilizing at 121 oC (Note: Data with the same letter were not significantly different
for 20 min, added 5 ml of a starter culture of B. subtilis (p>0.05))
(around 107 cells/ml). Finally, they were incubated at 37oC catalyzes µmol of galacturonic acid per minute (µmol. min−1)
for 2 days (48 hours) [10]. under the assay conditions.
b) Effect of duration of incubation U
The cultured medium was prepared by using optimal Pectinase activity ( )
ml
concentration pectin in molasses obtained in the previous 1000 1
result. The experiments were conducted at different time = mg of galacturonic acid x x xD
212.2 5
periods, ranging from 24 to 144 hours [12]. The other
D: Dilution factor.
conditions were kept unchanged. All media were sterilized at
121oC for 20 minutes before inoculating 5 ml of a starter 212.12: Molecular weight of galacturonic acid (mg/mol);
culture of B. subtilis with an initial cell concentration of
around 107 cells/ml. Then, they were incubated at 37oC. 5: Reaction time (minutes);
4) Crude enzyme extract 1000: Converting factor from mmol to µmol
After finishing the incubation, the solution was filtered by
cloth and transferred the liquid into 50 ml falcon tubes. Then 6) Application of enzyme on coffee
it was centrifuged at 7,000 rpm and at room temperature for After experiment II.B.5 was done, pectinase with the
30 min. The crude enzyme was harvested by filtration with highest activity was chosen to treat coffee beans for
the sterilized membrane filter having a pore size of 0.2 µm to increasing the extraction of soluble solid content.
eliminate all microorganisms that remained in the medium.
The supernatant was collected as a source of the crude a) Effects of enzyme concentration
enzyme [13]. 50 grams of coffee beans were soaked in water for 2 hours
to get the 50% of moisture content [15]. The beans were
5) Determination of pectinase activity
treated with crude pectinase at 5 different treatments: 1, 2, 3,
Pectinase activity was determined based on the 4, and 5% (v/w), then added 100 ml of sterilized distilled
determination of reducing sugars produced as a result of water. The control was a sample without enzyme treatment.
enzymatic hydrolysis of pectin by using dinitrosalicylic acid All controls and samples were incubated at 35oC for 32 hours
reagent (DNS) method [14]. Prepared acetate buffer (0.1 M, [9].
pH 4.4). For enzyme assay, 1 ml of the sample included
collected crude enzyme and commercial enzyme b) Effects of fermentation duration
(Viscozyme®L) were diluted 10 times by acetate buffer. In 50 grams of soaked coffee beans were treated by pectinase
addition, the substrate was prepared by mixing 1% citrus with the optimal ratio (result from previous experiments).
pectin in the prepared buffer (w/v). There were three clean Then the samples were incubated at 35oC with 5 different
labeled test tubes; one for the crude enzyme, one for the durations: 12, 24, 36, 48, and 60 hours, and the control was
commercial enzyme, and one for blank. All test tubes at 0 hour.
contained 1 ml prepared substrate. Then, 0.5 ml of diluted
sample was added to each labeled tube and 0.5 ml of distilled 7) Measurement of the total soluble solid content
water to the blank. Then, the test tubes were incubated at 40oC After treatment, fermented coffee beans were washed
for 5 minutes in the water bath. After incubation, 3 ml of 3,5- three times to remove all remaining silver skin. Then they
DNS was added to all test tubes, to stop the reaction. Then, were dried in an oven at 60oC for 6 hours until reaching the
they were placed in a boiling water bath (90∘C) for 10 min. initial moisture content of 7.75%. Then dried coffee beans
Finally, the tubes were cooled, and optical density (OD) was were roasted at 240oC by roaster for 15 minutes and ground
measured using spectrophotometer (UV/Vis) at 540 nm [8]. into powder. A sample of 10 grams of coffee powder was put
The enzyme unit was defined as the amount of enzyme that
413
90
76.91e 4,5 4,09c
80
68.14d 4 3,65b 3,79b
70
Pectinase activity (U/mL)
62.82c
58.88f 3,5 3,14a
60 52.49b 2,98a
3 2,72a
TSC ( OBrix)
50
2,5
40
30.20a 2
30
1,5
20
1
10 0,5
0 0
24 48 72 96 120 144 0 1 2 3 4 5
Incubation time (hours) Concentration of crude enzyme (%)
Figure 2. Effects of incubate duration on pectinase activity Figure 3. Effect of enzyme concentration on TSC
(Note: Data with the same letter were not significantly different (Note: Data with the same letter were not significantly different
(p>0.05)) (p>0.05))
in a beaker and 70 ml of boiled distilled water was added. The III. RESULTS AND DISCUSSION
mixture was shaken at 175 rpm for 10 min. After cooling
C. Effects of substrate concentration on pectinase activity
down to the room temperature, they were filtered by filter
paper to collect the supernatant. Finally, the soluble solid Molasses was utilized as a culture medium for
content of the extracted solution was measured by a microorganism growth, and pectin was used as a substrate for
refractometer (°Brix). the pectin biosynthesis activator. Initially, the B. subtilis
strain used sugar, a readily digestible component in the
8) Application of produced enzymes into de-mucilage culture medium. The microbe then released enzymes to
layers of coffee breakdown other substances into sugar in order to survive
when sugar was used up. However, increasing substrate
There were 3 treatments in this experiment: crude enzyme
concentrations did not mean the synthesis of pectin was
produced by B. subtilis, Vicozyme®L solution (commercial
increased. To be more precise, if the substrate concentration
enzyme), and the blank (no enzyme). The ripened fresh
increased over a point of the limited range, enzyme activity
coffee had to be cleaned and removed pulp (wet parchment).
would decrease because of two reasons. Firstly, the substrate
100 g of pulped coffee was soaked in 100 ml water for 2
gave the pressure to reduce the pectinase synthesis process.
hours. Then, each sample was washed 3 times. After that, the
Secondly, the substrate increases while the ability of the
coffee bean was put in flasks with 100 ml of distilled water
microorganism in enzyme secretion is constant.
and 3% of pectinase (v/w). Did them in triplicate. All samples
were incubated at 35oC for 24 hours. The treatment was From Fig. 1, the highest enzyme activity was 47.67±0.13
stopped by washing with water. The beans were dried in an (U/ml) and observed at 3% pectin. Compared to [17],
oven at 60oC for 6 hours. Finally, the dried beans were cooled pectinase activity reached 63.83 U/ml at 2% of pectin.
in desiccators. However, due to the mixture of both microorganism strains
(Candida boidinii and Bacillus subtilis), it can take up to 3%
9) Nutritional values of enzyme-treated coffee (w/v) to produce the best performance. When pectin was used
a) Carbohydrate and reducing sugar as the substrate, the pectinase activity was at its peak [18]. As
Carbohydrate determination was carried out based on [16] a result, pectinase has a high affinity for pectin when
whilst reducing sugar content was determined following [8] compared to the other enzymes used in this analysis.
with the measurement of absorbance at the wavelength 490 D. Effect of duration of incubation on pectinase activity
and 540 nm, respectively, by UV–Vis spectrophotometer.
Pectinase synthesis by B. subtilis was found to be
b) Caffeine and protein content in coffee extracellular according to Kelly and Fogarty's criteria, as this
Coffee powder of two samples (no enzyme treatment, enzyme was present in the culture supernatant and no
fermentation by the crude pectinase) in duplicated was sent pectinase activity was detected in the cell lysate or cell
to CASE center (No 2, Nguyen Van Thu Street, Da Kao washings of intact cells [19]. Data showed that pectinase
Ward, District 1, HCMC) to check for caffeine and protein production increased gradually during the fermentation
content by the TCVN 9723: 2013 (Vietnamese standard). period and reached its maximum value (76.91±0.15 U/ml)
after the fifth day (120 hours) of initial incubation (Fig. 2). It
10) Statistical analysis was found that the best production of pectinase enzyme by B.
subtilis in this study with 3% of pectin as substrate was 120
All measurements were carried out in triplicate. ANOVA
hours. However, after 48 and 72 hours of incubation, the
test for data analysis, means, and standard deviations were
pectinase activity was 52.5±0.07 (U/ml) and 62.83±1.6
run with Turkey’s SPSS software.
(U/ml). According to the paper [20], after 18 hours of
incubation, the enzyme production began, peaked at 24 hours
(110 U/ml), and then began to decline. Incubation time is
414
4,5 4.01c 70 62.97c

3.85d
% Removal mucilage of coffee

4 3.54b
60 52.53b
3,5 3.42b
3.00a 50
3
TSC ( oBrix)
2,5 40
2
30
1,5 21.57a
1 20
0,5 10
0
12 24 36 48 60 0
Control Crude enzyme Viscozyme L
Fermentation duration (hours)
Enzyme Treatment
Figure 4. Effects of fermentation duration on TSC

Figure 5. Effects of treatment on de-mucilage layers of coffee
(Note: Data with the same letter were not significantly different (p>0.05))
(Note: Data with the same letter does not differ significantly (p>0.05))
determined by the microorganism's growth rate and pattern duration, the amount of TSC could be released and dissolved
of enzyme development. Different incubation times are in medium. Therefore, the suitable time that brings the
attributable to different strains of Indian Bacillus from this highest TSC during coffee fermentation was 36 hours. In
strain, all of which were grown in submerged fermentation. other research, the optimal fermentation duration is 20 hours
The amount of enzyme production was quite high in a much for the highest extracted soluble solid content [13]. The
shorter time than 120 hours. Thus, the time of incubation reason for that result is the crude enzyme from this research
should be kept at 48 hours when apply to a real situation. The was incorporated cellulase and pectinase. Murthy and Naidu
time had been shortened by 2.5 times, but the amount of (2011) also stated that Robusta coffee was completely
enzyme pectinase obtained was still relatively high. demobilized after 36 hours of incubation [23]. The enzyme
E. Effects of enzyme concentration on the extraction of total treatment in this study speeds up fermentation and avoids
soluble solid content coffee quality degradation caused by conventional coffee
processing.
During coffee fermentation, pectinase plays a crucial
function in degrading pectin structure to increase soluble G. Application of produced enzymes into de-mucilage
solid content (TSC). It would be insufficient to hydrolyze layers of coffee
pectin levels in coffee beans if pectinase concentrations After experiment II.B.5 was done. Each 100 g prepared
were low. The enzyme could break down coffee beans and sample of coffee beans was treated with pectinase enzymes
dissolve soluble solid content into the water phase if the (v/w): crude enzymes (pectinase activity 76.91 ± 0.15 (U/ml)
pectinase ratio was high. The highest concentration of TSC and commercial enzyme (pectinase activity 99.63 ± 1.73
was 4.09oBrix at 3% of the crude enzyme (Fig. 3), U/ml). Fig. 5 illustrates the mucilage layer removal after the
increasing 0.44oBrix (10.75%) compared to the sample fermentation process. With adding a crude enzyme, the
treated by 2% of crude enzymes. To explain this result, less removal of pulp is 52.53% of weight, 62.97% of weight for
than 2% of crude enzymes might not have enough quantity Viscozyme®L whereas for the control sample, it was only
for the extraction of TSC while more than 2% of enzymes 21.57% weight of beans were removed. The highest pectinase
would dissolve TSC into the medium. In the other study, the activity was Viscozyme®L treatment. While the treatment
highest TSC was also extracted in the sample at 4 % of crude variations gave a significant effect (p<0.05). In the control
enzymes [21]. It increased by 33.49% compared to the sample, the pulp of beans was slowly removed. When using
control. The TSC increased as the enzyme concentration the commercial enzyme to ferment, the pulp of beans was
removed rapidly compared to crude enzymes produced by B.
increased [22]. According to the experiments, the TSC was
subtilis treatment. However, the effectiveness of the crude
found to be highest at 3% of enzyme concentration. The
enzyme was acceptable when the percentage of mucilage
explanation for these, more enzymes may be needed during removal was more than one and a half of weight. Although
fermentation, depending on the viscous layer and the nature the efficiency of crude enzymes was not equal to that of
of the original coffee bean from different regions. commercial enzymes, this study also opened the way for
F. Effects of fermentation duration on the extraction of further development of low-cost pectinase production.
total soluble solid content H. Nutrient content of coffee after fermentation
The fermentation duration is one of the most important After fermentation, the caffeine content was decreased,
factors that affect the quality of the coffee bean while it is other nutrients remain unchanged in coffee powder. The
fermented. If the fermentation time was not long enough, the fermentation treatment in this article has been relatively
enzyme did not work at its limit with compatible substrates. successful. The nutritional quality of coffee has improved
As the result, the TSC of treated samples was 4.09 ± gradually not only in the process but also in farming
0.08oBrix corresponding to 36 hours (Fig. 4). If the duration techniques.
was too short, the enzyme would not have enough time to
degrade pectin molecules leading a lower concentration of
extracted TSC. By contrast, with too long a fermentation
415
Table 1. Nutrient content in fermented coffee fermentation”. World Journal of microbiology and
Biotechnology, vol. 18, no. 9, pp. 835-839, 2002.
[6] David Dubnau, 1982. Molecular Biology of the Bacilli: Bacillus
subtilis. Published by Academic Press Inc, ISBN 10: 0122227018
[7] Dao, T.T.Q., and Phu, L.H. Enzyme and Food fermentation –
Laboratory manual, Food Technology department, School of
Biotechnology, International University – Vietnam National
University – Ho Chi Minh, 2014.
[8] Phu H. L., “Production of microbial enzymes pectinase, cellulase
and their application on production of green coffee by
fermentation”, PhD Thesis, University of Science _ Vietnam
National University, Ho Chi Minh City 2012.
[9] Quynh X. T. Nguyen. “Pectinase Production by Bacillus Subtilis
on Different Fermentation Modes and Its Application on Coffee
Treatment”. BSc Thesis, International University, 2017.
IV. CONCLUSION [10] Nga N.T.Q. “Production of pectinase and cellulase Aspergillus and
Bacillus subtilis cultured using molasses”, BSc Thesis, Food
The study of producing pectinase by Bacillus subtilis was Technology department, School of Biotechnology, International
successfully investigated. The highest pectinase activity was University – Vietnam National University-Ho Chi Minh, 2014.
76.91±0.15 (U/ml) achieved with 3% of pectin substrate, 1% [11] Yekta, G., Ksungur, A.U, “Production of Pullulan form beet
ammonium sulfate in molasses medium, pH=5, for 5 days of molasses and synthetic medium by Aureobasidium pullulans”.
incubation. As a result, the cost of production of microbial Turkish Journal of Biology, vol. 28, no. 1, pp. 23 – 30, 2004.
[12] Irshad, M., Anwar, Z., Mahmood, Z., Aqil, T., Mehmmod, S., &
enzymes can be reduced compared with commercial Nawaz, H. “Bio-processing of agro-industrial waste orange peel
products. It was observed that the pectinase from the Bacillus for induced production of pectinase by Trichoderma viridi; its
subtilis strain has huge promising potential in the removal of purification and characterization”. Turkish Journal of
viscous from coffee beans. The mucilage layer had been Biochemistry/Turk Biyokimya Dergisi, vol. 39, no. 1, 2014.
removed 52.53% after fermentation with the crude enzyme. [13] Duy N.Q., Huyen D. N. H., Phu H. L. “Study on the production of
Cellulase and Pectinase by using Aspergillus Oryzae in molasses
The highest TSC was 4.09±0.08 oBrix (respectively, p<0.05) and their application for the extraction of soluble solid content
from fresh coffee at 50% moisture content, and 3% (v/w) from coffee”. International Journal of Modern Engineering
concentration of enzyme at 35oC for 36 hours. In comparison Research, vol. 6, no. 11, pp. 41-45, 2016.
with natural fermentation, the enzyme treatment in this study [14] Miller, G. L. “Modified DNS method for reducing sugars”. Anal
sped up fermentation and avoided coffee quality degradation Chem, vol. 31, no. 3, pp. 426-428, 1959.
caused by conventional coffee processing. The product had [15] Chien N. V. “Improvement of soluble solids extraction of Coffee
robusta using combinated Bio-products from aspergillus niger,
standard caffeine (2.48%); some nutrients have been Bacillus subtilis HNL01 and issatchenkia orientalis”. BSc Thesis,
identified to help better understand the current coffee quality. Food Technology department, School of Biotechnology,
The green coffee beans were treated by enzyme extracted International University – Vietnam National University – Ho Chi
from Bacillus subtilis so there was no concern about Minh City, 2010.
aflatoxin. Therefore, crude pectinase can be applied to [16] Nielsen S. S. Protein nitrogen determination, Ch.5.In: Nielsen SS,
Food analysis laboratory manual, 2ndedn. Springer, New York,
produce coffee products with the improvement of quality, 2010.
safety, and cost reduction. [17] Phuong N T.T. “Study on the microbial symbiosis of Candida
Boidinii and Bacillus subtilis for improvement of soluble solid
ACKNOWLEDGMENT content extraction of Robusta coffee by fermentation”. BSc Thesis,
We would like to express our faithful thanks to the Food Food technology department, School of Biotechnology,
International University Vietnam National University – Ho Chi
Technology Department of International University which Minh City, 2016.
provided valuable support in both facilities and spiritual [18] Oumer, O. J., & Abate, D. “Characterization of pectinase from
perspectives during the accomplishment of this project. We Bacillus subtilis strain Btk 27 and its potential application in
also appreciate everyone who is involved directly and removal of mucilage from coffee beans”. Enzyme research, 2017.
indirectly in this study. [19] Kelly, C. T., & Fogarty, W. M. “Production and properties of
polygalacturonate lyase by an alkalophilic microorganism Bacillus
REFERENCES sp. RK9”. Canadian Journal of Microbiology, vol. 24, no. 10, pp.
1164-1172, 1978.
[1] Clarke R. J., Coffee, volume 2: Technology, vol. 2, 1987, pp. 64. [20] Ahlawat, S., Dhiman, S. S., Battan, B., Mandhan, R. P., & Sharma,
[2] Chye, M. L., Sin, S. F., Xu, Z. F., & Yeung, E. C. “Serine J. “Pectinase production by Bacillus subtilis and its potential
proteinase inhibitor proteins: exogenous and endogenous application in biopreparation of cotton and micropoly fabric”.
functions”. In Vitro Cellular & Developmental Biology-Plant, vol. Process Biochemistry, vol. 44, no. 5, pp. 521- 526, 2009.
42, no. 2, pp. 100-108, 2006. [21] Quynh N.N. “Isolation and identification yeasts and bacteria that
[3] Liew Abdullah A. G, Sulaiman N. M, Aroua M. K, Mohd Noor M. produce cellulose and pectinase activities from coffee plant”. BSc
J. “Response surface optimization of conditions for clarification of Thesis, International University, 2014.
carambola fruit juice using a commercial enzyme”. J Food Eng, [22] Joshi V.K., Parmar M., Rana N.S. “Pectin esterase production from
vol. 81, pp. 65–71, 2006. apple pomace in solid-state and submerged fermentations”. Food
[4] Sun, C., Liu, R., Liang, B., Wu, T., Sui, W., & Zhang, M. Technol Biotechnol, vol. 44, pp. 253- 256, 2006.
“Microparticulated whey protein-pectin complex: A texture- [23] Murthy, P. S., & Naidu, M. M. “Improvement of Robusta coffee
controllable gel for low-fat mayonnaise. Food Research fermentation with microbial enzymes”. European Journal of
International”, vol. 108, pp. 151-160, 2018. Applied Sciences, vol. 3, no. 4, pp. 130-139, 2011.
[5] Sunnotel, O., & Nigam, P. “Pectinolytic activity of bacteria
isolated from soil and two fungal strains during submerged
416
Environmental Sustainability: Exploring Managers’

Attitudes and Behaviours at High-End
Accommodation Businesses in Vietnam
Thanh-Nhan Nguyen Quang-Tri Tran Tho Alang
Fashion and Tourism Faculty Fashion and Tourism Faculty School of Business
Ho Chi Minh City University of Ho Chi Minh City University of Ho Chi Minh City International
Technology and Education Technology and Education University
nhanntt@hcmute.edu.vn tritq@hcmute.edu.vn alangtho@hcmiu.edu.vn
Abstract: Sustainability in general and environmental strengthens peaceful relations all over the world [5]. But the
sustainability, in particular, have been studied and put into rapid growth of the tourism industry in general and hotels,
practice in developed countries for decades. However, resorts in particular, contribute significantly to environmental
developing countries are still struggling with many obstacles, degradation worldwide, as hotels consume a large amount of
from academic research to practical implications. In the energy and water for heating, cooling and lighting [6]]7]. So
ongoing consequences of climate change and the greenhouse given that businesses consider sustainability and green
effect, the government of Vietnam has committed to zero practices in their operations [8][9][10]. One of the three
waste by 2050 at the 26th United Nations Climate Change dimensions of sustainability that have been paid attention to
by the public is environmental sustainability with the meaning
Conference (COP26). This determination from the
of using the natural resources effectively in input (renewable
government requires the hospitality industry to urgently
or nonrenewable resources) and output rules (reduction of
sustainable actions. This study aims to explore the attitudes pollution and climate change) [11][12]. With its vital role as
and behaviours of departmental managers about their the measure of assessing sustainable policy's success [13].
environmental sustainability practices at high-end Very few countries have achieved environmental
accommodation establishments in Vietnam. Primary data are sustainability because the quality of the environment has
obtained from semi-structured interviews with 22 declined [14] and the problems are most relevant to
respondents from the 4-5 stars hotels and resorts in HCMC, environmentally irresponsible human attitudes and behaviours
Nhatrang, and Hanoi. These data are analyzed with thematic [4][15][16].
approaches. Findings show that implementing environmental
sustainability has positive attitudes through raising awareness The growing demand from customers for eco-friendly
hotels has encouraged hospitality sectors worldwide to change
of pro-environmental after the Covid-19 pandemic and
their attitudes and behaviours about redesigning their structure
COP26 summit; and the prevalence of the 17 sustainable and organizational culture toward an increasingly
development goals and Triple bottom line (TBL) principles environmental sustainability system [3][17][18]. In previous
in sustainable reports. Additionally, behaviours on studies of environmental sustainability in the hospitality
environmental sustainability are also expressed through industry, researchers agree that environmental issues can be
design and construction; green operational plan; and the managed and resolved by correcting human attitudes and
assessment of energy, water and waste metrics. behaviours in an environmentally sustainable way [17].
Okumus [19] proved that hotel employees' attitude to
Keywords: sustainability, environmental sustainability, environmental practice includes three elements as
Vietnam, hospitality, attitude, behaviour environmental knowledge, environmental awareness, and
environmental concern. These items affected ecological
I. INTRODUCTION behaviours on practising green and Shin [20] stated that hotel
The world population has increased to approximately 7.8 customers' attitudes focused also on environmental concern,
billion currently with a growth rate of about 14% [1]. In environmental knowledge. International hotel employee
replying to human progress and evolution, industrial actions attitudes toward environmental management were explored
have affected negatively the environment leading to social under vital items on the vigorous support of top management;
damage such as problems of poverty and inequity, climate active hotel's green committee; environmental training
change [2] greenhouse effect, air/water/soil pollution, programs; supervise and incentive systems, and leadership
extinction/loss of species, and exhaustion of natural resources style [21,22]. Attitudes on environmental protection navigate
[3][4]. The tourism sector is not an exception, it has been the behaviour of implementing environmental sustainability
considered to be one of the main sectors in developed and positively or negatively.
developing countries with an important role in creating job
opportunities, making a small gap in poverty, improving Particularly, the change in individuals’ consumption
gender equality and protecting natural resources. It behaviours (approaching, buying, and consuming the products

in a pro-environmental manner) are considered to be critical practices in the end-high accommodation sector in three
solutions for environmental sustainability [3]. Articles in the regions Ho Chi Minh, Ha Noi and Nha Trang in Vietnam. As
world showed the environmental sustainability behaviours. summarized by the authors, Most of the previous studies on
Jones [23] proved that environmental sustainability practice attitudes and behaviours used the quantitative method
has solved problems of constructing green buildings, [22][33], the use of the qualitative method is still limited in the
management of energy, water, and waste, purchasing research on environmental sustainability, where some
procurement in the hospitality industry. They have applied it perspectives on attitudes and behaviours toward
in MICE through specific indicators. With energy environmental sustainability is a new phenomenon. Therefore,
management and conservation, the hospitality industry and the qualitative method is required to analyze attitudes and
MICE venues have practised solutions of energy-saving as behaviours toward environmental sustainability in end-high
installing energy-efficient appliances and equipment (LED accommodation facilities to achieve objectives in the new
lamps, heat pump systems, motion sensors, timers to reduce context.
light-ing), energy star-qualified products in offices, public
areas, and guest rooms, kitchen, and building [5][23][24][25]. Samples were selected based on snowball sampling
because it is difficult to approach and invite them to join the
This research aims to enrich the extant hospitality interviews in Vietnam. Authors contacted managers via social
sustainability literature on managers’ attitudes and behaviours media and networks to find the interview consensus.
of environmental sustainability practice by filling the gaps Therefore, selecting and interviewing some interviewees with
described subsequently. While scholars have found ecological initial convenience conditions, then asking them to provide
behaviour of guests of hotels [26][27][28]; relationships additional respondents who have managerial positions. The
between the environmental attitudes and their ecological respondents in this study are managers from 4,5-star hotels
behaviours at the employee level in Hotels [19], others looked and resorts in Ho Chi Minh, Ha Noi and Nha Trang in
to investigate the travellers’ attitude-behaviour gap in eco- Vietnam, who have working experience in environmental
hotel [29][30] and a few papers researched managers' attitudes sustainability practices. Twenty-two managers from the end-
and behaviours on environmental sustainability. Previous high accommodation sector in Vietnam were selected as the
papers have applied the quantitative method in the research to interviewees for this study. Collecting and reading the
understand sustainable attitudes and behaviours. [9],[16],[21], sustainable reports of organizations that participants have
[25],[26],[28],[36], [41],[43] but few ones used the qualitative been working on was to validate data analysis.
method for that. Ali [31] showed that the Covid-19 pandemic
has a significantly positive impact on environmental A semi-structured interview questionnaire with open-
awareness, sustainable consumption and social responsibility ended questions used in this study contained three sections.
The first section featured questions related to attitudes toward
and it moderated the relationship between Covid-19 and its
impact on sustainable behaviours. So call for investigating the environmental sustainability, the second, included behaviours
reasons for changes in managers’ attitudes and behaviours on toward environmental sustainability, and the final section
environmental sustainability in Vietnam in the Covid-19, contained questions about the demographic profiles of the
where green hotels seem to be a new concept and the green interviewees. All questions in the interview were designed
transition in the lodging sector in Vietnam is being taken on based on the existing literature [19][ 33][35], and the
slowly [32]. interview questions were reviewed by one researcher and one
practicer. These main research questions included:(a) What
This study aims to explore managers' attitudes on pro- are the attitudes of managers of high-end accommodation
environmental and defines the environmental sustainability businesses towards environmental sustainability after the
practice in high-end accommodation businesses through two Covid-19 pandemic? (b) What solutions (behaviours) have
research questions: What is the perception of managers of high-end accommodation businesses achieved in
high-end accommodation businesses towards environmental environmental sustainability practice? In addition, some
sustainability after the Covid-19 pandemic? What solutions follow-up questions were asked in the interview
have high-end accommodation businesses achieved in
environmental sustainability practice? The study results can Twenty-two interviews were carried out from November
help managers understand the proper meaning of 2021 to February 2022, and the duration of each interview
environmental sustainability, also develop and implement ranged from 30 to 50 min. One of the authors conducted and
plans to apply environmental sustainability practices more recorded the interviews via face-to-face communication and
effectively. To make data for the policymaker to be more internet communication in Vietnam. The primary language
active in running campaigns to promote environmental used in the interviews was English, although both Vietnamese
sustainability practices in high-end accommodation and English were used depending on the preference of the
businesses. interviewees. The interview stopped at the 20th person as
achieving the saturation point, no new information was
The first section provides a critical review of attitudes and discovered. However, the authors still took more information
behaviour toward environmental sustainability in the about the last two persons from these two interviews to make
hospitality industry. The second states the method employed the confidential data. The process was recorded to collect the
in this study. In the third section, the findings are presented interview results and the interviews were transcribed
and discussed. Finally, the conclusion illustrates the immediately to ensure the accuracy of the content. The
impressive problems and offers suggestions for future profiles of interviewees were illustrated in Table 1.
research.
After the interviews were recorded and transcribed, the
II. METHODOLOGY corresponding information was inputted into NVivo 12.0 for
further analysis. Sustainability reports were adopted along
The objective of this study is to explore managers' with NVivo 12.0 to analyze the qualitative data. Two authors
attitudes and behaviours toward environmental sustainability coded and categorized the information from interviews during
418
the beginning stages of collecting data. Other checked data to Themes Subthemes
The exemplar of coding
triangulate the investigators’ framework. Specifically, the categories and Selected
relevant quotations
authors discussed results until they reached mutual
agreements on the emerging themes as in Table 2. Green operational Energy management: LED
plan lights, Dimmer / Motor PWM
Table 1. Background information of the key interviewees Speed Controller, heat pump
system, turn off unused
Interviewees Gender Ages Busines Ownership Position
equipment, set the room
temperature.
(Inter) s type
Water conservation: reuse their
towels, automatic sensor faucet
Inter 1 Male 55 5-SHa Pb- FIc Md. Ee system, use and reuse water.
Waste management: replacing
Inter 2 Male 34 5-SH P-FI M. FOf
the plastic product, purchasing
Inter 3 Male 40 5-SH P-FI M.E non-biodegradable products,
waste separation
Inter 4 Male 45 5-SH JS M.FBg The assessment of Evaluation, monitoring, metrics
h
energy, water and of energy, water and waste
Inter 5 Female 37 4-SR Private - JS M. SM waste metrics reduction
i
Inter 6 Male 58 5-SR P M. SM
III. RESULTS
Inter 7 Male 53 5-SH P- FI M. E
Two main themes and five sub-themes were covered and
Inter 8 Male 36 5-SH P-FI M. FO identified on attitudes and behaviours toward environmental
Inter 9 Male 43 4-SH Private M. FB sustainability in this article. The themes include (1)
Awareness of pro-environmental after the Covid-19
Inter 10 Male 60 5-SH State M. FB pandemic, (2) The prevalence of the 17 sustainable
Inter 11 Female 32 5-SR P M. FO development goals and Triple Bottom Line (TBL) in the
environmental sustainability reports, (3) Design and
Inter 12 Male 34 5-SH P M. SM
construction, (4) Green operation plan, (5) The assessment of
Inter 13 Male 30 4-SH P-FI Sk. FO energy, water and waste metrics.
Inter 14 Female 58 5-SH State S. Hj A. Attitudes toward environmental sustainability
Inter 15 Male 30 4-SH Private M. FO After Covid-19, most participants were aware of the
emergency on pro-environment and understood clearly the
Inter 16 Female 61 5-SH State M. FO
negative effects on environmental destruction of people and
Inter 17 Male 30 5-SR P M.FB society. The subthemes under this theme were raising
Inter 18 Male 41 5-SH P-FI M. SM
awareness of pro-environment after the Covid-19 pandemic
and COP 26 conference and The prevalence of the 17
Inter 19 Female 32 4-SH Private M. FB sustainable development goals and Triple Bottom Line (TBL)
Inter 20 Male 55 5-SH P-FI M. E in the environmental sustainability reports.
Inter 21 Male 37 5-SR P M. SM 1) Raising awareness of pro-environmental after the
Covid-19 pandemic and COP 26 submit
Inter 22 Male 59 5-SH Private M. H
a
During the time of the Covid-19 pandemic, the managers
star hotel,bPartnership,cForeign investment,dManager,
e
Engineering, fFront office, gFood and Beverage, hsales and marketing, istar
of the high-end accommodation facilities saw positive sights
resort, jHousekeeping,kSupervisor. in improving environmental quality when the activity of the
whole society slowed down, which shows that Day-to-day
Table 2. Themes and coding categories environmental management activities at the facility are urgent
The exemplar of coding and this awareness is even stronger with the government
Themes Subthemes
categories and Selected commitment at the COP26 conference. In the current study,
relevant quotations most of the managers recognized the importance of protecting
Awareness of pro- Environmental quality. green the environment and environmental sustainability is one of the
environmental after certificates. Environmental vital aspects to support long-term sustainable profit for
the Covid-19 indicators, innovation;
businesses.
pandemic and
COP26 summit Interviewee 1: Actually, solutions to environmental
Attitudes
The prevalence of The guideline, 17 sustainable protection have been practised in our business operations
the 17 sustainable developments goals, TBL(triple
development goals bottom line), sustainable report for many years, we got some green certificates such as ISO
and Triple Bottom 14001…. However, during Vietnam's lockdown due to the
Line (TBL) in the Covid-19 pandemic, I recognize clearly the impact of
sustainability reports economic and social activities on the environment through
Design and Green building, cost-saving, positive indicators in the news when socio-economic
construction installing frequency converters,
Behaviours
water pumping systems, solar
activities slow, also feel myself (the same comment with
energy, green technology and interviewees 4,12)
equipment
419
Interviewee 2: I think 5-star hotels have paid more attention of the green facilities at the beginning period or continuous
to environmental protection solutions. Especially, we are a improvement on the old base. This solution will solve
luxury resort, …. we strive for our customers. When I read extremely cost saving on energy management, water
information about the commitment of the Vietnamese conservation and waste management.
government at the COP26 conference on Zero waste to
Interviewee 1: I don't deny all solutions for environmental
2050, we recognized the advantages of our business and
protection are necessary, but building green hotels or
having more motivation to innovate in a pro-environment.
innovating builds to be green is the most important. we
(Interviewees 16, 18 made the same comment).
innovated our facility by installing frequency converters
2) The prevalence of the 17 sustainable development (VSD) for the water pumping system based on the
goals and Triple Bottom Line (TBL) in the sustainability calculation of the minimum water flow of 28 litres / s and
reports the maximum of 40 litres /s. The efficiency is very clear,
In the world, many organizations have granted green hotel saving 55-59% of energy, the total amount of electricity
certifications also green hotel guidelines (LEED, Green saved is about 361,720KW / year (the same comment with
Globe, TripAdvisor Green Leaders, Green Key Global, interviewee 5,6).
Audubon Green Lodging Program...). Luxury hotels and
resorts in Vietnam have achieved green certifications from Interviewee 6: Nowadays, solar energy and modern pieces
Green lotus, ISO 14001, ASEAN green hotels, EarthCheck of equipment contribute largely to building and improving
etc. A finding is interestingly found in the data analysis hotels to be green. We installed solar energy and equipped
procedure, the framework of the sustainable report or the green technology equipment, tools and machines with
environmental sustainability is written based on 17 sustainable the ENERGY STAR® label in our design (the same
developments goals [45] and TBL (triple bottom line). comment with interviewees 2,4).
Interviewee 14: As you know, we are business managers, Interviewee 20: Our energy-saving goal has been interested
it is clearly difficult to understand the definition of in the hotel since its design and construction began. Take
sustainability or environmental sustainability. Luckily, we advantage of natural conditions, directions, colours,
based on the standards of green hotel guidelines to employ materials, glass doors, using LED lights, using Ozone
indicators for environmental protection and now we report technology in the washing process, taking advantage of the
our sustainability results based on three dimensions excess heat of boiler exhaust to make hot water, optimizing
(economy, environment and social) and 17 sustainable the combustion process of the boiler (the same comment
development goals in tourism like hospitality groups in the with interviewees 16,18).
world (the same comment with interviewees 6,17).
2) Green operational plan
Interviewee 6: I think we should publish a sustainable Besides new green building design and innovation on old
report presented to three dimensions (economy, hotels, behaviours toward environmental sustainability are
environment and social) and 17 sustainable development included in the green operational plan as energy saving, waste
goals because UNWTO [45] makes this standard which management, and water conservation.
aimed toward unity format and nowadays many large a) Energy management.
hospitality businesses reported that form (the same When asked about solutions to implement environmental
comment with interviewees 1,18,10). protection policies in this study, most participants consultant
B. Behaviours on environmental sustainability practice solutions to save energy effectively for reducing costs and C02
emission
In the previous studies, the adoption of environmental
sustainability practises is not only moral but also beneficial in Interviewee 17: We do energy-saving practice at all, even
saving costs, improving image, creating market the smallest tasks, replace and install all LED lights, using
differentiation, and increasing CSR [35]. To aim the Dimmer / Motor PWM Speed Controller in all meeting
environmental sustainability, hotels or resorts managers have rooms with adjusting proper light, replace the closed
applied many solutions such as purchasing environmentally cooling tower system to open circulation system, with low
friendly products, the use of environmentally friendly cost. We innovate continuously by installing a heat pump
cleaning products, energy and water-saving practices, and the system, taking advantage of the high temperature of the
selective collection of solid waste [35][36]. Results of this machines and laundry areas to make hot water and provide
study, there are three themes emerging in behaviours of cold air to cool the machines and laundry area (the same
environmentally sustainable practices as design and comment with interviewee 21, 11).
construction; green operational plan; the assessment of
energy, water and waste metrics Interviewee 11: In our hotel, we have the policy to apply
all staff turn off computers, monitors, printers, copiers, and
1) Design and construction
lights in the office when not in use, for example, a
The green building design before doing operation or computer is set to “idle” when it is not used for 20 minutes
innovation of traditional hotels creates the advantage or less; turn off lights when not needed; set electrical
competitive and helps to increase customers' attraction and equipment into eco mode if possible, set the computer
satisfaction [37]; it is impressive to performance of saving as screen to "sleep" mode or turn off the screen when leaving
30 to 50 per cent in energy use, 35 per cent in carbon the desk for more than 10 minutes, for example going to
emissions, 40 per cent in water emissions, and 70 per cent in lunch, meeting; use the stairs instead of the elevator when
solid waste expenses [38]. In this article, most of the possible, it is good for health and also saves energy; set the
managers presented their opinion on the design and building
420
room temperature between 24°C and 26°C. With an Interviewee 2: As you know, the environmental
optimum temperature of 25°C (the same comment with sustainability report showed the innovation by a specific
interviewees 7,12). number, for instance, with a construction density of about
20%, the percentage of trees is estimated at 80%, so the
b) Water conservation
ambient air temperature of the hotel is reduced by 1-2oC
In the operational plan of hotels and resorts, water thanks to the system of lakes and green parks.
reservation (reuse and water-saving) recommended by
managers depends on the characteristics and positions of Interviewee 4: In the past 7 years, the electricity
facilities. Participants present their effort to reduce water cost/revenue ratio has decreased from 8.1% in 2014 to
consumption. 5.0% in 2020. In 2020, due to the impact of the Covid-19
pan-demic, our hotels only use nearly 11.2 million kWh,
Interviewee 2: we ask our customers to reuse their towels;
the rate of energy-saving reaches 10%.
apply to circulate irrigation technology, an intelligent
automatic sensor faucet system to avoid unnecessary water IV. DISCUSSION
waste; use and reuse water as much as possible before
Attitudes toward environmental sustainability affect
disposing of it; check, repair, and maintain the water
sustainable practice behaviours in the business operation of
system regularly to ensure no leaks (the same comment
the hospitality industry under pressure of green customers, the
with interviewee 17,21). requirement of making a competitive advantage by marketing
Interviewee 22: Besides concentrating on energy saving, also coping with climate change, CO2 emission, disasters and
the hotel has also built water-saving options, making the diseases. End-high accommodation facilities in developing
most of wastewater for toilet flushing and cooling tower countries have strived to change attitudes and behaviours
cooling, saving significant wastewater treatment costs. The toward environmental sustainability after Covid-19 pandemic
hotel has also in-stalled sanitary equipment with water- and governmental commitment to Zero waste at Cop 26
saving features (the same comment with interviewees conference in 2021. The findings indicate that the attitudes of
departmental managers of 5-star hotels and resorts are
10,20).
positively aware of the value and importance of pro-
Interviewee 3: Our resort Install a smart automatic sensor environment after Covid-19 and governmental commitment
faucet system to avoid unnecessary water waste, on Zero waste at Cop 26 conference also attention to
Wastewater after treatment is reused to water plants and presenting sustainability report which follows 17 sustainable
lawns throughout the resort (the same comment with development goals and TBL. Under positive attitudes, it
interviewees 11, 17). explored through many environmental sustainability practice
behaviours such as the design and construction of hotel and
c) Waste management resort buildings, green operational plan (energy management,
Solid waste and wastewater are sources that damage water conservation and waste management), and evaluating
community health and the environment. Although waste and monitoring effectively environmental sustainability in
separation (sorting) of people is practised extremely slow, metrics of energy, water and waste reduction.
high-end accommodation facilities have sorted inside and Kollmuss and Agyeman [39] defined environmental
most of them focused on some solutions of replacing plastic awareness as “Knowing of the impact of human behaviour on
products and non-biodegradable products the environment”. Pro-environmental practice is usually
Interviewee 3: In recent years, our resort has strived to impacted by three elements as environmental knowledge,
eliminate plastic waste as Inventory of all plastic items in environmental awareness and environmental concern [19].
operations; eliminating plastic water bottles, straws and Contrary to Thao [26 ]’s and Nhan [40]’s findings stated that
single-use plastic (the same comment with interviewee 5). Vietnam’s hospitality industry has been changing slowly and
managers who are being charged with MICE venues was lack
Interviewee 17: In the past many years, solid waste of perspective on sustainable business in the hospitality
reduction is an emergency issue. We eliminated plastic industry, this study is similar to findings of [41] and [42]
items such as plastic cups from guest rooms, plastic demonstrated the need for environmental protection, that
packaging around toothbrushes, and cotton buds from their managers' attitude presented, also managers in chain-affiliated
bathrooms; eliminating plastic items such as straws, hotels were generally more likely to pay attention to
stirrers, cutlery, plates, and cups from their food and environmental issues and more innovation. In fact, 5-star
beverage services hotels and resorts have practised environmental sustainability
for many years under customer pressure and social, they
Interviewee 11: Besides the practice of sorting solid waste, employed most solutions for cost-saving, customer
our office has replaced paper documents with files and satisfaction, promotion of brand name [21].
emails for communication, printing two sides of the paper,
and reusing stationary. This motivation showed that they were not aware of
sufficient meaning, importance, and benefits of environmental
3) The assessment of energy, water and waste metric sustainability because their staff did not effectively practice
Performance assessment is essential to evaluate and pro-environmental measures with the perspective of saving
monitor effectively environmental sustainability in metrics of cost, profit [43]. It is not persuaded by them in practice. After
energy, water and waste reduction. In this study, many the Covid-19 pandemic, both managers and staff have clearly
participants showed the attitudes of high consensus on recognized the harmful effects of human and social activities
presenting data in 17 sustainable development goals and on the environment, and under the pressure of the
principles of TBL through actions of measuring index. government's commitment to zero emissions by 2050,
421
attitudes toward environmental awareness enhanced contribute to zero waste governmental commitment until
environmental protection, as a basis for businesses to create 2050, the Vietnam tourism industry needs to put effort into
many improved behaviours in the environmental protection applying green technology; green procurement and recycling;
performance of the hospitality industry. motivating startups on green solutions, green transportation.
Especially, regulation on sustainable practice from the
Leading hospitality facilities reported on the initiatives of government is necessary.
environmental protection policies through their websites to
show their commitment to environmental protection and In actions of assessing the performance of energy, water
practice. Hsieh [44] illustrated that about 46 per cent of the and waste metrics, this finding showed years' comparison by
hotels used web pages to post information related to specific numbers that businesses achieve. This result is like
environmental issues on their public websites. It is interesting Jorge’s [50] statement on evaluating and monitoring the
to bring the hospitality industry into this study, attitudes of activity of water, waste and energy management by metrics
managers on report form on websites based on the principles that support decision making, and encourage continuous
of 17 sustainable development goals [45] and three aspects of improvement. Although most of the managers who participate
the economic, social and environment. Actually, hotels and in this study have been working for hotels and resorts under
resorts have applied many guidelines of green certificates
the management of multinational hospitality groups showed
organizations and environmental sustainability such as ISO
that all metrics presented in leading hospitality group reports
14001; EarthCheck; the European Union’s EMAS; Reports
from the Coalition for Environmentally Responsible but few hotels or resorts publish these sustainable
Economics (CERES); the Global Reporting Initiative (GRI); performance metrics on their own websites. It is extremely
and the ICC Business Charter for Sustainable Development necessary to accommodate facilities’ practice of
(ICC) [44] and they presented the vital results under major environmental sustainability by metrics and public on the
environmental focus areas (energy management, website in order to attract the customers' loyalty and to make
environmental education for customers and employees, water the motivation to continuous improvement
conservation, waste management (recycling and reuse), and
green supply and purchase) [21][26]. These findings show that V. CONCLUSION
most reports presenting data follow 17 sustainable This article was mainly examined to explore the attitudes
development goals and the TBL principle. Seen that the and behaviours of environmental sustainability practice in
tourism and hospitality industry has found common principles end-high accommodation facilities in Vietnam from the
to achieve sustainable development and direct all tourism over departmental managers' perspective using the qualitative
the world with a high consensus on pro-environmental method. These findings of the study provided theoretical and
practice. empirical implications for implementing environmental
sustainability. From the theoretical implication, the previous
In the developing country context, manager attitudes on studies showed attitudes and behaviours toward
environmental awareness are positive after Covid-19 and zero environmental sustainability from tourists and employees, but
waste governmental commitment at the Cop16 conference. little research has investigated managers' attitudes and
This positive awareness motivated and transferred into behaviours in developing countries after the Covid-19
solutions to environmental sustainability practices. These pandemic. Attitudes and behaviours studies on hospitality
findings on behaviours or actions of environmental sustainable practices from managers focused on main areas,
sustainability have been practised in its contextual limitation. including organizational, operational and strategic [51]. This
Most managers recognized that building green facilities or research on attitudes and behaviours of environmental
innovating the green building on an old base is a priority sustainability implementation in Vietnam from the managers’
measure aimed at energy saving. Green buildings have perspectives has been overlooked.
brought in both energy conservation and environmental
benefits: capacity of an annual energy saving of 1.49 million Thus, this study can enrich the theoretical framework
TCE plus capacity of an annual carbon dioxide reduction of dealing with attitudes and behaviours on environmental
1.45 million ton-c [46], they make a Competitive Advantage sustainability practices in the hospitality sector. The results
for Hotels in the Tourism Industry [37]. The previous study revealed that attitudes toward raising environmental
showed that green building is one of the environmental awareness, applying 17 sustainable development goals and
sustainability practices [47][48] but few researchers analysed TBL principles in sustainable reports and behaviours to
the priority in sustainable practice for hotel green building. practising environmental sustainability toward design and
This result is the same as Chan [47] and Alexand [48] on the construction of the green building, green operational plan and
benefits of green building. the Vietnamese government issued performance assessment of energy, water, waste metrics in
the policy to support companies to install renewable energy Vietnam context after Covid-19 and COP 26 conference.
resources in production and business to make advantages for After analyzing the findings, we found the design and
hospitality go forward environmental sustainability. construction dimension and performance assessment was the
most important theme that could support the improvement of
In the operational plan for environmental sustainability implementing environmental sustainability practices in this
practices, interviewees introduced many solutions for sector, followed by the awareness, green operational plan and
reducing energy, water conservation and waste reduction. sustainable report following 17 sustainable development
Measures mainly eliminate plastic products, install and use goals. In addition, from the practical aspect, the identified
energy-saving equipment, and reuse wastewater and water- attitudes and behaviours on environmental sustainability can
saving tools. However, solutions to environmental practices in assist policymakers in making guidelines, incentives, planning
developed countries [5][50] are much more than in developing for raising environmental awareness in managers, students,
countries in hospitality operations because the developing employees in the hospitality industry and practitioners who
countries lack the completion of a sustainable strategy. To
422
understand insight on measures of environmental [7] E. Dimara, E. Manganari and D Skuras, “Don’t change my
sustainability practice. towels please: Factors influencing participation in towel reuse
programs”, Tour. Manag. 2017, 59, 425–437, 2017
VI. RECOMMENDATIONS AND LIMITATIONS [8] Y. J. Jang, T. Zheng, and R .Bosselman, “Top managers’
environmental values, leadership, and stakeholder engagement
This study also proposed several corresponding in promoting environmental sustainability in the restaurant
suggestions based on attitudes and behaviours of industry”, International Journal of Hospitality Management,
environmental sustainability practice, which can provide 63(), 101–111, 2017, doi:10.1016/j.ijhm.2017.03.005
practical insights for hospitality policymakers and [9] H. Han, & H.Yoon, “Hotel customers’ environmentally
responsible behavioral intention: Impact of key constructs on
practitioners. decision in green consumerism”, International Journal of
First, The Vietnam tourism government should make Hospitality Management, 45, 22–33, 2015,
https://doi.org/10.1016/j.ijhm.2014.11.004
policies, provide more incentives, improve monitoring to
[10] H.-Y. Park, D.-K. Kim, & B.S Kim, “A study on guideline
address problems in environmental sustainability practices, development for green mice facilities focusing on location and
also promote communication to raise environmental building management of exhibition & convention centers”,
awareness among stakeholders in the hospitality industry. Korean Journal of Hospitality Administration, Vol. 24 No. 1,
Second, ensuring consistency of green quality of service pp. 23-37, 2019
provided to the customers, establishing a formal [11] A.Niñerola,M.V.Sánchez-Rebull,A.B.Hernández-Lara,
environmental management system by metrics with “Tourism Research on Sustainability: A Bibliometric
Analysis”, Sustainability, 11(5), 1377, 2019,
sustainable report publication, and providing more effective doi:10.3390/su11051377
green training opportunities for employees to raise their [12] R. Goodland, “The Concept of Environmental Sustainability.
knowledge and awareness on pro_environment. Third, Annual Review of Ecology and Systematics”, 26(1), 1–24,
seeking more collaboration with green suppliers, green 1995, doi:10.1146/annurev.es.26.110195.000245
technology, green transportation and installing the recycling [13] Howes, M, “Politics and the Environment: Risk and the Role
station. of government and Industry”, Allen & Unwin: Sydney,
Australia, 2005.
This study is a preliminary attempt to identify attitudes [14] M.Howes, L.Wortley, R. Potts, A. Dedekorkut-Howes, S.
toward behaviours on environmental sustainability. However, Serrao-Neumann, J. Davidson, T.Smith and P.Nunn,
it still has several limitations that require further “Environmental Sustainability: A Case of Policy
investigations. First, the sample size of this study is too small. Implementation Failure?”, Sustainability, 9(2), 165, 2017,
doi:10.3390/su9020165
Many managers of luxury hospitality are difficult to approach
for interviews. Future studies can incorporate data from other [15] D.Hopkins, “Sustainable mobility at the interface of transport
and tourism”, Journal of Sustainable Tourism, 28(2), 129–143,
sizes of hotels and resorts, including 1- 4 star hotels to increase 2020, https://doi.org/10.1080/09669582.2019.1691800
the sample size. Second, this study only focused on the [16] J.Wu, X.Font, & J.Liu, “Tourists’ pro-environmental
managers’ perspectives. The perspectives of customers, behaviors: Moral obligation or disengagement?”, Journal of
frontline employees, and the government were not presented Travel Research, 2020,
in this research. Hence, these stakeholders can be investigated https://doi.org/10.1177/0047287520910787
in future studies. Third, this study aimed to explore the high- [17] H. Han, “Consume behaviouror and environmental
end accommodation sector in Vietnam; hence, the results sustainability in tourism and hospitality: a review of theories,
concepts, and latest research”, Journal of Sustainable Tourism,
might not represent the state of environmental sustainability 29(7), 1021–1042, 2021,
practices in the hospitality industry in other regions. Future doi:10.1080/09669582.2021.1903019
studies could thus examine the hospitality sectors in other [18] O. Fatoki, “Hotel employees’ pro-environmental behaviour:
regions of the world. Effect of leadership behaviour, institutional support and
workplace spirituality”, Sustainability, 11(15), 4135, 2019,
REFERENCES https://doi.org/10.3390/su11154135
[1] 2021 World Population Data Sheet. Population Reference [19] F. Okumus, M.A. Köseoglu, E. Chan, A.Hon, U.Avci, “How
Bureau 2021. [Online]. Available: do hotel employees' environmental attitudes and intentions to
https://interactives.prb.org/2021-wpds. implement green practices relate to their ecological behavior?”,
Journal of Hospitality and Tourism Management, 39(), 193–
[2] R. Dhingra, R. Kress and R. Upreti, “Does lean mean green?”, 200, 2019, doi:10.1016/j.jhtm.2019.04.008.
Journal of Cleaner Production, 85(), 1–7, 2014,
doi:10.1016/j.jclepro.2014.10.032 [20] Y. H. Shin, J. Im, S. E. Jung, & K. Severt, “Consumers'
willingness to patronize locally sourced restaurants: The
[3] S.Wang, J.Wang, J.Li, & F.Yang, (2019) “Do motivations impact of environmental concern, environmental knowledge,
contribute to local residents’ engagement in pro-environmental and ecological behavior”, Journal of Hospitality Marketing &
behaviours? Resident-destination relationship and pro- Management, 26(6), 644–658, 2017
environmental climate perspective”, Journal of Sustainable
Tourism, 28(6), 834–852, 2019, [21] E. S. Chan, & R. Hawkins, “Attitude towards EMSs in an
https://doi.org/10.1080/09669582.2019.1707215 international hotel: An exploratory case study”, International
Journal of Hospitality Management, 29(4), 641–651, 2010
[4] F.Xu, L.Huang and L.Whitmarsh, “Home and away: Cross-
contextual consistency in tourists’ pro-environmental [22] E. S. W. Chan, A. H. Y. Hon, W.Chan, & F. Okumus, “What
behavior”, Journal of Sustainable Tourism, 28(10), 1443–1459, drives employees' intentions to implement green practices in
2020, https://doi.org/10.1080/09669582.2020.1741596 hotels? The role of knowledge, awareness, concern and
ecological behaviour,” International Journal of Hospitality
[5] A. Abdou, H. Hassan, H. Thowayeb, H. El Dief and Management, 40,20–28, 2004
M.Mohammed, “A Description of Green Hotel Practices and
Their Role in Achieving Sustainable Development”, [23] P.Jones, D.Hillier, & D.Comfort”, Sustainability in the
Sustainability, 12(22), 9624–. doi:10.3390/su12229624 hospitality industry: some personal reflections on corporate
challenges and research agendas,” International Journal of
[6] A.Abdel-Maksoud, H.Kamel, S.Elbanna, “Investigating Contemporary Hospitality Management, 28(1), 2016
relationships between stakeholders’ pressure, eco-control
systems and hotel performance”, International Journal of [24] K.T.Huang, J.C.Wang, Y.C Wang, “Analysis and
Hospitality Management, 59(), 95–104, 2020, benchmarking of greenhouse gas emissions of luxury hotels”,
doi:10.1016/j.ijhm.2016.09.006 Int. J. Hosp.Manag. 2015, 51, 56–66.
423
[25] V.Pereira, G.M.Silva, Á. Dias, “Sustainability Practices in Ekonomska Istraživanja, 28(1), 1–30,
Hospitality: Case Study of a Luxury Hotel in Arrábida Natural 2015, doi:10.1080/1331677x.2014.995895
Park”, Sustainability, 13, 3164, 2021, [38] J.Butler, “The Compelling "Hard Case" for "Green" Hotel
https://doi.org/10.3390/su13063164 Development”, Cornell Hospitality Quarterly, 49(3), 234–244,
[26] Thao, N.T.T. “The Relationship between Eco-friendly 2018, doi:10.1177/1938965508322174
Practices and Attitudes toward Green Hotels for Domestic [39] A. Kollmuss, J Agyeman, “Mind the Gap: Why do people act
Tourists”, VNU Journal of Science: Economics and Business, environmentally and what are the barriers to pro-environmental
Vol. 33, No. 2, 101-111, 2017 behavior?,” Environmental Education Research, 8(3), 239–
[27] J. A. Campos-Soria, A. García-Pozo, A. J. Marchante-Mera, 260, 2002, doi:10.1080/13504620220145401
“Explaining tourists´ attitudes to environmental support: a [40] N.T.T. Nhan, T.Q. Tri, A. Thoi, “Toward Sustainable Business
multilevel approach”, Journal of Sustainable Tourism, (), 1– Practice of MICE Host Organizations: Exploring the Role of
20,2018, doi:10.1080/09669582.2018.1435667 Managers”, The 4th Conference on Economics, Business, and
[28] J. López-Bonilla, M. Reyes-Rodríguez, L. López-Bonilla, “The Tourism (4th CEBT-2021) in Ho Chi Minh City, Vietnam in
Environmental Attitudes and Behaviours of European Golf July 17, 2021. ISBN: 978-604-73-8449-5
Tourists”, Sustainability, 10(7), 2214, 2018, [41] M .Wei, “An Analysis on Environmental Awareness and
doi:10.3390/su10072214 Behavior in Chinese Hospitality Industry––A Case of Xiamen
[29] G. Agag, & N. Colmekcioglu, “Understanding guests’ City”, 5(none), 1126–1137, 2011
behaviour to visit green hotels: The role of ethical ideology and doi:10.1016/j.egypro.2011.03.198
religiosity”, International Journal of Hospitality Management, [42] P. Bohdanowicz,” Environmental awareness and initiatives in
91, 102679, 2020, doi:10.1016/j.ijhm.2020.102679 the Swedish and Polish hotel industries—survey results”,
[30] R.Eid, G.Agag, & Y. M. Shehawy, “Understanding Guests’ 25(4), 0–682, 2006, doi:10.1016/j.ijhm.2005.06.006
Intention to Visit Green Hotels”, Journal of Hospitality & [43] A.Kasim,“Managerial attitudes towards environmental
Tourism Research, 109634802094780, management among small and medium hotels in Kuala
2020, doi:10.1177/1096348020947800 Lumpur,” Journal of Sustainable Tourism, 17:6, 709-725,
[31] Q. Ali, S. Parveen, H.Yaacob, Z. Zaini, & N. A. Sarbini, 2009, DOI: 10.1080/09669580902928468
“COVID-19 and dynamics of environmental awareness, [44] Y.C, Hsieh, “Hotel companies' environmental policies and
sustainable consumption and social responsibility in practices: a content analysis of their web pages”, International
Malaysia”, Environmental Science and Pollution Research, Journal of Contemporary Hospitality Management, Vol. 24 Iss
2021 doi:10.1007/s11356-021-14612-z 1 pp. 97 121, 2012
[32] H.L.T.Trang, J-S. Lee, H. Han,” How do green attributes elicit [45] UNWTO, Indicators of sustainable development for tourism
pro-environmental behaviours in guests? The case of green destinations: A guidebook. Madrid: World Tourism
hotels in Vietnam”, Journal of Travel & Tourism Marketing, (), Organization, 2004
1–15, 2018, doi:10.1080/10548408.2018.1486782
[46] X.Pengpeng, H.W. C. Edwin, K. Q Queena, “Success factors
[33] E. Myung, “Environmental knowledge, attitudes, and of energy performance contracting (EPC) for sustainable
willingness to pay for environmentally friendly meetings – An building energy efficiency retrofit (BEER) of hotel buildings in
exploratory study”, Journal of Hospitality and Tourism China”, 39(11), 7389–7398, 2011,
Management, (), S1447677016302285, 2017, doi:10.1016/j.enpol.2011.09.001
doi:10.1016/j.jhtm.2017.03.004
[47] W.W. S. Chan, E.S.W. Yeung, Chan and D. Li, “Hotel heat
[34] S. Mohd, A.Mohd, P .Justin 2022. “Eco-friendly hotel stay and pumps hot water system”, nt. J. Contemp. Hospital. Manage.
environmental attitude: A value-attitude-behaviour 25(3): 428-446, 2013
perspective”, International Journal of Hospitality Management
100, 103094, 2022 https://doi.org/10.1016/j.ijhm.2021.103094 [48] S. Alexander, and C. Kennedy, “Green hotels: Opportunities
and resources for success Zero Waste”, Alliance, pp: 2-9, 2002
[35] T. Kularatne, C.Wilson, J. Månsson, V. Hoang & B. Lee, “Do
environmentally sustainable practices make hotels more [49] H. Han, & S. Hyun, “What influences water conservation and
efficient? A study of major hotels in Sri Lanka”, Tourism towel reuse practices of hotel guests?”, Tourism Management,
Management, Volume 71, Pages 213-225, 2019, 64, 87–97, 2018
doi:10.1016/j.tourman.2018.09.009 [50] C.Jorge, M. Almeida, C. Do & D. Covas, “Performance
[36] P.D.Modica, L.Altinay, A.Farmaki, D.Gursoy, M.Zenga, Assessment System for Energy Efficiency in Wastewater
“Consumer perceptions towards sustainable supply chain Systems”, Water, 13(13), 1807, 2021, doi:10.3390/w13131807
practices in the hospitality industry”, Curr. Issues Tour. 2020, [51] S. H. Kim, K.Lee, & A.Fairhurst, “ The review of “green”
23, 358–375. research in hospitality, 2000-2014: Current trends and future
[37] S. Hashemkhani Zolfani, M. Sedaghat, R. Maknoon, E. K, research directions”, International Journal of Contemporary
Zavadskas, “Sustainable tourism: a comprehensive literature Hospitality Management, 29(1), 226–247, 2017
review on frameworks and applications”, Economic Research-
424
Research on Optimal Algorithms Using

Experimental Planning to Improve Shoe Production
Line Productivity
Minh Tai Le Thi Cam Duyen Doan Huynh Thao Vy Nguyen
Department of Industrial Systems Department of Industrial Systems Department of Industrial Systems
University of Technology and University of Technology and University of Technology and
tailm@hcmute.edu.vn 18104004@student.hcmute.edu.vn 18104061@student.hcmute.edu.vn
Abstract: In production operation, line balancing is a top Therefore, in this study, the problem of balancing the chain
priority that is promoted by manufacturing plants to aim to to increase the productivity of the line is analyzed and
improve productivity. The optimal algorithms are used in solved. The first step to optimization is to screen out the
experimental planning to improve the productivity of shoe factors that affect the production line the most. There has
production lines. The Taguchi and RSM methods and the been a lot of research done to optimize each factor while
factors directly affecting the line balancing performance others have remained the same and that is the simplest way
were analyzed in this study. From experimental results, the to do it.
optimal condition of Line of Balancing (LOB) was Line Balancing means the leveling or assigning working
calculated by the surface response method (RSM) and the operations to workstations and workload across all
central composite design (CCD). The results showed that operations along the production line in such a way that the
the optimization process achieved an optimal efficiency of assignment is optimal and to remove bottenecks. [5]
94.9% corresponding to the following conditions: Lead time
In this study, the Taguchi method was used with the aim
and number of stages were 300s and 38 stages, respectively, of optimizing the parameters affecting the line balancing
and CT max was 8.32s. In this work, the RSM method is performance in the complete shoe assembly line [6]. Design
proposed to increase the productivity of shoe production of Experiment (DOE) is a tool to help investigate and
lines, and the performance shows that RSM is 1.81% better analyze simultaneously the impact of input variables on
than other methods. output variables. Currently, DOE is the main tool used in the
improvement step of the 6 Sigma project. DOE is also very
Keywords: line balancing, Taguchi method, RSM, shoes commonly used in product design or in setting operational
production line, CCD, BBD parameters of a manufacturing process [7].
I. INTRODUCTION Design experiments were performed by selecting the
Line balancing is the term used when everything is most suitable orthogonal array, determining the appropriate
working together in a balanced manner. Everyone is doing number of columns, and describing the combinations of
the same amount of work required by the client. The single experiments. These experiments consist of a series of
transformation was smooth, no one was overwhelmed, and runs or tests. The data set will be collected at each run.
no one had to wait. Line balancing helps companies achieve After the initial screening step, optimization by response
all of these through quantitative methods [1]. surface method (RSM) with the structural alternatives CCD
In Vietnam, Binh Tien Consumer Goods Production Co., (Central Composite Design) and BBD (Box-Behnken
Ltd specializes in manufacturing shoes with diverse and Design) is the second method used. used to optimize the
prestigious designs for users. Quality and aesthetics are the value of the factors that directly affect the LOB coefficient of
top criteria to attract customers, and the price will also the line [8]. This is a multifactor experimental planning
determine the revenue and profit of the company. method based on a mathematical model. The basic principle
Bottlenecks are one of the most controversial concerns in of RSM is to complete the experiment at some point of the
industrial processes. Reducing congestion and waiting time factors to describe the result by a regression equation in the
are two important factors [2-3]. Optimizing production lines form of a quadratic polynomial.
solves some production or output defects. Line balancing to This optimization method has been implemented in
increase productivity is one of the effective methods to Vietnam as well as in the world. A case study was carried out
optimize production lines that have not reached the highest in the assembly line of a paint manufacturer in Iran [9], they
productivity. Thus, improving the production line to optimize proposed RSM and DOE to improve productivity in the
production is an urgent need to have enough capacity to meet manufacturing industry with the optimum yield achieved at
orders. Besides, some studies show that improving line 93.5%.
balance can reduce 30% production time, equipment uptime,
40% inventory cost, 17% workers, 80% idle time idle, ... [4].

In this study, Taguchi and RSM methods are proposed to No A B C LOB (%) S/N (%)
improve shoe production line productivity. The testing 4 290 37 10 78.38 37.88
5 290 38 11 69.38 36.82
process was carried out at a company in the South of 6 290 39 9 82.62 38.34
Vietnam, with the optimum yield achieved at 93.09% for the 7 290 37 11 71.25 37.06
Taguchi method and 94.9% for the RSM method. 8 290 38 9 84.80 38.57
9 290 39 10 74.36 37.43
II. METHODS 10 300 38 9 87.72 38.86
11 300 39 10 76.92 37.72
A. Taguchi Method 12 300 37 11 73.71 37.35
Table 1. Input factors and their extent 13 300 38 10 78.95 37.95
14 300 39 11 69.93 36.89
Factor input Ref. code Level 1 Level 2 Level 3 15 300 37 9 90.09 39.09
Lead time A (s) 290 300 310 16 300 38 11 71.77 37.12
17 300 39 9 85.47 38.64
Number of stages B (stages) 37 38 39 18 300 37 10 81.08 38.18
19 310 39 9 88.32 38.92
CT max C (s) 9 10 11 20 310 37 10 83.78 38.46
21 310 38 11 74.16 37.40
The levels above are the values of the factors that affect 22 310 39 10 79.49 38.01
23 310 37 11 76.17 37.64
the performance of the line balancer. Table 1 has shown the 24 310 38 9 90.64 39.15
factors affecting the LOB rate. Firstly, Lead time is the total 25 310 39 11 72.26 37.18
time of the stages in the production line which is 290 seconds 26 310 37 9 93.09 39.38
(level 1), 300 seconds (level 2), 310 seconds (level 3), 27 310 38 10 81.58 38.23
respectively. The second influencing factor is that the
number of stages must be minimized excessively or Analysis of Variance (ANOVA):
combined together to reduce the transition time from one
stage to another, the corresponding number of stages is The ANOVA analysis method was used to determine the
specified as 37 stages ( level 1), 38 stages (level 2), 39 stages significance of each parameter in the designed experimental
(level 3) in the experiment. The third influencing factor, CT study.
max acting as the bottleneck, is the cycle time with the We rely on the p and F values in the ANOVA table to
highest value. The corresponding CT max value is given as 9 determine the importance of each factor:
s (level 1), 10 s (level 2) and 11 s (level 3). Maximum
criterion of quality characteristics was the aim of this study. • If p value is less than 0.05 with 95% confidence,
So the following objective function is used: this factor has statistical significance to LOB
performance.
1 n 1 
S / N = −10log   2  (1) • If F value is less than [F] then this factor is not a
 n i =1 yi  strong influence on LOB performance.
where: B. RSM – CCD Method
The S / N ratio (Signal-to-Noise ratio) is a parameter that Number of experiments is calculated with Nu = 20 with t
evaluates the degree of influence of internal and external =3
factors on the output. It can be understood as a quantity to
measure the influence of factors. Nu = 2t + 2t + 6 (4)
• n : number of experiments.
where, t is the number of independent variables (number
• i : number of test runs. of factors) and 2t is the number of additional experiments at
the star point. Distance from center to star point α = 2t /4 ( α =
• yi : value of observation data. 1.68 with t = 3). All studies were performed at five levels (-
Setting production line efficiency (LOB) as the α, -1, 0, +1, +α). Table 3 pointed out that 3 factors that have
calculation objective for the objective function, the the greatest influence on survey results were selected to
equilibrium coefficient of line 3 is calculated according to conduct experiments according to RSM - CCD with 5 levels
the formula (2) & (3): and 20 experiments. In this study, the survey domain was
selected as follows: Lead time (283.18 - 316.82), number of
(T1 + T2 + T3 + ...Tn ) stages (36 - 39), CT max (8.32 - 11.68).
T= (2)
n
Table 3. Lead Time, Number of Stages and Ct Max Used in RSM
T – CCD Method
H= 100% (3)
Tmax Code Levels
Factor
Table 2 presented Signal-to-Noise parameters for 27 − -1 0 +1 +
experiments (L27). Lead time A 283.18 290 300 310 316.82
Table 2. Value of LOB, S/N Number of stages B 36.32 37 38 39 39.69

CT max C 8.32 9 10 11 11.68
No A B C LOB (%) S/N (%)
1 290 37 9 87.09 38.80
2 290 38 10 76.32 37.65
3 290 39 11 67.60 36.60
426
C. RSM – BBD Method and the calculation results of these two methods showed no
In this study, experimental design was collected and significant difference.
analyzed using Minitab and Design Expert software. Where
Lead time has a range of 290 - 310 seconds, stage number is Table 5. Extent and Influence of Inputs
37 - 39 and CT max is 9 - 11 seconds shown in Table 4. Lead time Number of CT max (C)
below. (A) stages (B)
Level 1 37.68 38.20 38.87
Table 4. Lead Time Value, Number of Stages and Ct Max Used in Level 2 37.99 37.97 37.95
Surface Response Method (RSM) – Box Behnken (BBD) Option Level 3 38.18 37.62 37.12
Factor Code Test value Affect 0.50 0.59 1.75
Lower (-1) Center (0) Higher (1)
Rating 3 2 1
Lead time A 290 300 310
Number of stages B 37 38 39
CT max C 9 10 11 Table 6 confirmed that the number of stages and CT max
In the experimental design, Nu is randomized and BBD is were the most significant factors because the p-values were
optimized based on the cubic design as presented in equation 0.034 and 0.012, respectively. Lead time is less significant
(5). because the p-value is 0.064 which is greater than 0.05.
N u = 2t (t − 1) + C p (5) Table 6. Extent and Influence of Inputs

Source DF Adj SS Adj MS F-value p-value
where t is the number of levels and Cp is the number of A 2 1.452 0.726 233.850 0.064
repetitions of the center points. The control experiment of B 2 0.813 0.407 130.930 0.034
this study was performed randomly and independent of the C 2 13.978 6.989 2251.120 0.012
pre-conditions of the prediction runs. An important Error 20 0.062 0.003 - -
assumption is that all operating variables are continuous, Total 19 0.023 0.001 0.030 1.000
measured and controlled by experiments with negligible *Significant at reliability = 0.05
errors in order to maximize the three-factor line-equilibrium
performance at three levels, t = 3 and Cp = 3 were used. This Table 7. Model Summary
produces 15 experiments, five of which are central points
S R-sq (%) R-sq(adj) (%) R-sq(pred)(%)
(0,0,0).
0.0557192 99.62 99.50 99.35
III. RESULTS AND DISCUSSIONS R2 values above 50% are statistically significant.
D. Screening of factors that most impact LOB The results of the correlation coefficient R2 are shown in
Table 7. The high correlation coefficient R2 (99.62%) proves
From Figure 1, the optimal levels for LOB are at the that the model can be used to predict experimentally and vice
third level of Lead time, the second level of the number of versa. This shows that the regression equation accurately
stages and the first level of CT max at 310 s, 37 stages and describes the experimental data. On the basis of experimental
9s, respectively. results obtained from RSM methods - Box Benken and
Central Composite Design, Design Expert software was used
to analyze and propose solutions to achieve the highest LOB
based on three survey factors. The results are shown in Table
8 & Table 9.
Table 8. Analysis of Variance (CCD)

Source Sum of df Mean F-value p-value
Squares Square
Model 1048.850 9 116.5400 15732.8300 0.0001
A-A 95.7800 1 95.7800 12930.0400 0.0001
B-B 59.7900 1 59.7900 8071.9100 0.0001
C-C 882.7700 1 882.7700 1.192E+05 0.0001
AB 0.0392 1 0.0392 5.2900 0.0442
AC 0.5660 1 0.5660 76.4200 0.0001
Figure 1. Sensitivity analysis of LOB using L27 orthogonal array
BC 0.3528 1 0.3528 47.6300 0.0001
with larger the better option
A² 0.0001 1 0.0001 0.0182 0.8953
B² 0.0385 1 0.0385 5.2000 0.0457
Table 5 has shown the results after performing the C² 9.4400 1 9.4400 1274.5100 0.0001
experiments to evaluate the performance of the line balance Residual 0.0741 10 0.0074
for the optimization process. The remaining values are Lack of 0.0741 5 0.0148
calculated based on the experimental results and predicted Fit
responses. For the Taguchi method, this section includes Pure Error 0.0000 5 0.0000
standard analyzes such as the main effect plot of the signal- Cor Total 1048.920 19
to-noise ratio (S/N ratio) and the settings to minimize the *Significant at reliability =0.05
variation of the response function. The data processing was
performed and calculated on Excel and Minitab software,
427
Table 9. Analysis of Variance (BBD) influence of independent variables such as Lead Time,
Source Sum of df Mean F-value p-value number of stages and CT max, only 0.05% of variation is
Squares Square due to unknown factors (random error). In addition, the low
Model 602.8600 9 66.9800 1.544E+05 0.0001 coefficient of variation demonstrates that the experiments
A-A 56.0000 1 56.0000 1.291E+05 0.0001 were performed accurately and have high repeatability
B-B 34.9300 1 34.9300 80517.04 0.0001
(0.1082%).
C-C 509.0900 1 509.0900 1.174E+06 0.0001
AB 0.0192 1 0.0192 44.28 0.0012 In the survey area, the regression equation shows that CT
AC 0.2826 1 0.2826 651.53 0.0001 max has the greatest negative effect on both first- and
BC 0.1764 1 0.1764 406.64 0.0001
A² 2.819E- 1 2.819E- 0.0006 0.9806
second-order line balance performance. This can be
07 07 explained by reducing the maximum production time of the
B² 0.0112 1 0.0112 25.74 0.0039 process (CT max) resulting in a rapid increase in efficiency.
C² 2.35 1 2.3500 5416.46 0.0001 Lead time had the lowest impact of all three factors including
Residual 0.0022 5 0.0004 first and second order. Figure 2 shows the mutual influence
Lack of 0.0022 3 0.0007 between the factors of the CCD method as well as the BBD
Fit
Pure 0.0000 2 0.0000
method. So, the optimal conditions for performing LOB are
Error as follows: Lead time: 300 (s), number of stages: 38 (stages),
Cor Total 602.87 14 CT max: 8.32 (s).
*Significant at reliability =0.05
Equation of RSM Box Behnken experiment:
E. Optimization of Elements LOB = 78.95 + 2.65*A - 2.09*B -
Since all the factors used have a significant effect on the 2.98*C - 0.0693*AB – 0.2658*AC +
S/N ratio, instead of being 37.97, φpredicted= 37.97 + (38.18- (7)
0.2100*BC – 0.0003*A*A + 0.0550*B*B +
37.97) + (38.20 - 37.97) + (38.87 – 37.97) = 39.31dB; where 0.7977*C*C
ξ is the average of the sum of the S/N ratios, and A3, B1, C1
are the S/N ratios over a significant level. According to the equation of RSM-CCD, the experiment
The predicted value φpredicted is 39.31 and the optimal value in and the graph in Figure 3 show the pair of Lead Time -
the Taguchi experimental table is 39.38 (the largest S/N number of stages (A*B), Lead time - CT max (A*C), number
value). of stages - CT max (B*C) all negatively affect LOB where
Lead Time - CT max(A*C) pair affects most negatively in
39.38 − 39.31 the same way as BBD method.
=  100% = 0.178%  10%
39.31 The optimal conditions for LOB performance are as
Table of values [F] [11] shows that [F]=3.52, = 0.05 follows: Lead time: 310(s), number of stages: 38 (stages),
(95% confidence level), t1 = 2 (Degree of freedom of each max CT: 9(s).
element) and t2 = 19 (Sum of degrees of freedom). The high correlation coefficient R2 (99.99%) proves that
Table 10 proved that the model is suitable to predict the the model can be used to predict experimentally and vice
value of the optimal set of parameters (A3, B1, C1). versa. This shows that the regression equation accurately
describes the experimental data.
Table 10. Model Summary
Due to the similarity between the results of the two
Variable Level LOB S/N Ratio
experimental methods, the comparison table of ANOVA
In the A1 B1 C1 87.09% 38.80 values is presented in Table 11.
beginning
Predictive A3 B1 C1 93.09% 39.38 Table 11. Comparing Anova Values for Quadratic Equation
optimization Between CCD and BBD Method
At the same time, in all three factors A, B and C, the p - value R2 (%) CV (%)
CCD 0.0001 99.95 0.1082
value of F is greater than the value of [F] = 3.52, so these
BBD 0.0001 99.99 0.0262
parameters are effective parameters of the line-balancing
performance. To optimize the line balance coefficient, in this study,
The influence of independent factors can be explained two methods are proposed, Taguchi and RSM. The
based on the regression equation obtained from Design experimental results show that the line balance performance
Expert software; the equation of the RSM - CCD of the two methods is not significantly different. This is
experiment: shown in Table 12 with LOB performance from Taguchi
reaching 93.09%, while RSM reaching 94.9%. Table 12
LOB = +78.95 + 2.65*A - 2.09*B - shows LOB of the different approaches, it shows the
8.04*C - 0.0700*AB - 0.2660*AC + effectiveness of the RSM method in a line balancing,
(6)
0.2100BC - 0.0031*A*A + 0.0517*B*B + namely application in a shoe production line.
0.8094*C*C
Table 12. Comparison LOB Value with Different Approaches
The p-value is related to the model's F-test (F-test or
Fisher's test) in Table 11 with a value less than 0.05 (p = Approaches LOB(%)
Zahraee et al.[8] 93.5
0.0001) which shows the compatibility of the regression X et al.[10] 77.30
equation with the experiment, thereby showing the statistical Our_Taguchi 93.09
reliability. The correlation coefficient R2 indicates that Our_RSM 94.90
99.95% of variation in LOB performance is due to the
428
a b
d
c
e f
Figure 2. The interaction between 3 elements Lead time, number of stages and CT max of CCD method
IV. CONCLUSION research should be carried out looking at the factors that
directly affect line performance. The main objective of this
In industrial problems, waiting times and bottlenecks in paper is to identify a suitable method, namely the response
manufacturing companies have continuous and complex surface method (RSM) combined with the Central
production processes with many workers, negatively Composite Design (CCD) to improve productivity and
impacting productivity as well as increasing production performance in the shoe industry in a cost-effective manner.
costs. In order to achieve a high level of profitability,
429
This study compared the fit and optimal model between [5] Mozdgir, A., Mahdavi, I., Badeleh, I. S., & Solimanpur, M.,
Taguchi and RSM methods. After conducting the “Using the Taguchi method to optimize the differential
evolution algorithm parameters for minimizing the workload
experiments, three factors Lead time, number of stages smoothness index in simple assembly line balancing”,
(stages), CT max(s) were determined as essential factors that Mathematical and Computer Modelling, vol. 57, no. 1-2, pp.
directly affect the line balancing coefficient. 137-151, 2013.
[6] Antony, J., Coleman, S., Montgomery, D. C., Anderson, M.
The input parameters set with different levels are J., & Silvestrini, R. T., “Design of experiments for non-
screened and the most optimal set of parameters for the line manufacturing processes: benefits, challenges and some
balancing performance for the RSM method are defined as examples”, Proceedings of the Institution of Mechanical
Lead time (300 seconds), number of stages (38) and CT max Engineers, Part B: Journal of Engineering Manufacture,
vol. 225, no. 11, pp. 2078-2087, 2011.
(8.32 seconds) with the most optimal efficiency of 94.9%.
[7] Pereira, L. M. S., Milan, T. M., & Tapia-Blácido, D. R.,
REFERENCES “Using response surface methodology (RSM) to optimize 2G
bioethanol production: a review”, Biomass and Bioenergy,
[1] Germanes, J. S., Puga, M. F., Sabio, R. B., Sanchez, E. M., & vol. 151, pp. 1-14, 2021.
Hugo, J. C., “Improving Efficiency of Shoe Manufacturer [8] Zahraee, S. M., Rohani, J. M., & Wong, K. Y., “Application
through the Use of Time and Motion Study and Line of computer simulation experiment and response surface
Balancing”, Journal of Industrial and Intelligent Information, methodology for productivity improvement in a continuous
vol. 5, no. 1, pp. 16-22, 2017. production line: Case study”, Journal of King Saud
[2] Chen, C. C., & Chuang, M. C., “Integrating the Kano model University-Engineering Sciences, vol. 30, no. 3, pp. 207-217,
into a robust design approach to enhance customer 2018.
satisfaction with product design”, International journal of [9] Betterton, C. E., & Silver, S. J., “Detecting bottlenecks in
production economics, vol. 114, no. 2, pp. 667-681, 2008. serial production lines–a focus on interdeparture time
[3] Zahraee, S. M., Golroudbary, S. R., Hashemi, A., Afshar, J., variance”, International Journal of Production Research,
& Haghighi, M., “Simulation of manufacturing production vol. 50, no. 15, pp. 4158-4174, 2012
line based on Arena”, Advanced Materials Research-Trans [10] Germanes, J. S., Puga, M. F., Sabio, R. B., Sanchez, E. M., &
Tech Publications, vol. 933, pp. 744-748, 2014. Hugo, J. C. (2017). Improving Efficiency of Shoe
[4] Budiman, I., Sembiring, A. C., Tampubolon, J., Wahyuni, Manufacturer through the Use of Time and Motion Study and
D., & Dharmala, A., “Improving effectiveness and efficiency Line Balancing, Journal of Industrial and Intelligent
of assembly line with a stopwatch time study and balancing Information, vol. 5, no. 1.
activity elements, Journal of Physics: Conference Series,
vol. 1230, no. 1, pp. 1-5, 2019.
430
Application of Multivariable Linear Regression

Algorithm to Support Inventory Management
Trung Tin Ngo Minh Tai Le
Department of Industrial Systems Engineering Department of Industrial Systems Engineering
19104052@student.hcmute.edu.vn tailm@hcmute.edu.vn
Nguyen Kim Thoa Vo Thanh Nam Luong

Department of Industrial System Engineering Department of Industrial System Engineering
19104044@student.hcmute.edu.vn 19104027@student.hcmute.edu.vn
Abstract: Talking about how to bring multivariable changing data, a system that can automatically update and
linear regression algorithm in machine learning to build analyze to support decision making with the highest
inventory prediction models for businesses and putting accuracy needs to be implemented.
artificial intelligence into inventory management systems The application of machine learning to the warehouse
become a branch of the intelligent inventory management management system will help businesses proactively
system. Compared with conventional management, the approach the industrialization process faster, modernize
introduction of machine learning will reduce the time to and improve professional qualifications, promote
sort as well as calculate the order quantity for the sustainable development, business stability. The change
warehouse, optimize the storage space of the warehouse. and application of technology in the 4.0 era will be the
This paper proposes to use the prediction algorithm from leading factor to help businesses compete with rivals in the
the multivariable linear regression equation to calculate the market.
number of goods to be stocked based on the input criteria:
In addition to Multiple Linear Regression, the
the minimum number of warehouses, the maximum Regression model also has Simple linear regression:
number of warehouses, current row count, safe threshold
level. The used platform is Google Colab as well as default
language used here is Python. The results obtained are quite
good with an almost correct prediction rate when applying
the classical regression model. Along with that is the
selection of suitable data files for input standards, the
parameters must be logical and fit the model to make the
best prediction. The prediction results bring back positive
effects based on the parameters of the data file are shown Figure 1. Regression models
clearly and accurately. From there, it can be developed and
used for other applications. Simple linear regression is the simplest form of linear
regression. It is used when user want to predict one variable
Keywords: Google colab, multiple linear regression, (dependent variable/result variable) according to the value
demand forecasting, inventory, artificial intelligent of the other variable (independent variable/explanatory
variable). However, multivariable linear regression is seen
I. INTRODUCTION as another form of the regression equation. The variable we
want to predict is called the dependent variable (or
Currently, inventory management is interested by most
sometimes, the outcome, objective, or criterion variable).
businesses and considers it a top priority because it has a
Considering the purpose of the research topic, the design
great influence on the overall performance of the business.
and development of a warehouse management system need
If the inventory management process is effective, it will
more than an ordinary input parameter. Therefore, this
bring many benefits to businesses such as: response the
model applies a multivariable linear regression algorithm to
maximum demands for customer satisfactions, help
predict the amount of inventory that needs to be reset for
businesses easily balance and keep high inventory levels
businesses focusing on the warehouse sector.
thus avoiding expired goods and products, eliminating
inventory wastes, and minimizing unnecessary arising in Nicholas Ampazis et al [1] used Artificial Neural
logistic and supply chain management. However, to Networks (ANNs) trained with an effective second order
monitor and handle large quantities of goods and constantly algorithm, and Support Vector Machines (SVMs) for

regression. With the aim of highlighting the effective predicts the outcome variable without explicitly
application of machine learning approached as an effective programming to perform specific tasks.
predictive method for predicting customer needs at the first
organizational level of the supply chain, where products The difficulties encountered by hard-coded data-driven
were displayed and sold to customers. Lolli Francesco et al systems suggest that AI systems need to be able to acquire
[2] searched for the reordering policies of a sample item by their own knowledge, by extracting patterns from sample
the lowest cost classification of in-sample items achieved. data. This competence is known as machine learning.
The experimental analysis of two data sets showed Machine learning starts with data - text, photos, such as
excellent accuracy, which showed that machine learning pictures of car objects, traffic lights, pictures of people or
can be effectively deployed in advanced inventory even bakery items, repair records, time-series data from
classification systems. Praveen K B et al [3] presented the sensors, or sales reports. The data is collected and used for
challenges to build an inventory management system and the purpose of training the model to learn and improve,
discussed the design decisions. With forecasting thereby giving the best results.
techniques, excess and inventory of items would be
reduced and minimized inventory costs. Samiul Islam et al The function of machine learning can be described as
[4] used two machine learning techniques, the Distributed the system using data to explain what has happened or will
Random Forest (DRF) and Gradient Boosting Machine be predicted, in other words the system will use the data to
(GBM) to predict the back orders of products. The forecast and suggest actions to take.
performance of the machine learning model was improved
by 20% using this range approach when the dataset was
highly biased with random error. Overall, the prediction
accuracy improved significantly, but they were still
struggling with highly skewed real data and many factors
that were not considered.
Some other studies on the application of artificial neural
networks such as Tereza Šustrová et al. [5] studying the
most suitable artificial neural network methods and their
applications in supply chain management. The study
successfully addressed the issues of inventory management,
batch sizing, and shipping costs. Galina Merkuryeva et al.
[6] also created a case study of the supply of
pharmaceutical products from a retailer to a distributor in
an emerging market. They compared and used methods
such as Simple Moving Average, Multiple Linear
Regression and Symbolic Regression with genetic Figure 2. Subfield of artificial intelligence
programming to find a method to predict demand with high
accuracy. Hironobu KAWAMURA et al. [7] studied the 2) Definition Multiple Linear Regression (MLR)
effectiveness of multiple regression analysis and ANN in Regression model is used to describe the relationship
determining appropriate safety stock levels and inventory between variables by showing it as a graph with a
costs using demand forecasting based on beer business data regression line drawn based on prediction results or
of a Japanese hotel. The study provided empirical analysis calculated parameters demonstrating that this model scale
of inventory management and showed that establishing a is significant. Regression allows to estimate how a
secure warehouse seems suitable for combining demand dependent variable changes when the independent
forecasting with customer information and management variable(s).
experience.
Linear regression is typically a function that allows an
This study presents the effectiveness of the Multiple analyst or statistician to make a prediction about a variable
Linear Regression model application applied on Google based on known data about one or more other variables.
Collab in the warehouse management system and predicts Linear regression can only be used when a problem has two
quite accurately about the order demand based on the input continuous variables - an independent variable and a
data: Minimum holding capacity of depot, Balance point dependent variable. The independent variable is the
capacity of depot, Maximum holding capacity of depot, parameter used to calculate the dependent variable or the
Closing Stock. outcome.
II. METHODOLOGY The formula for a multivariable linear regression is:
A. Multiple Linear Regression in Machine Learning y = β0 + β1X1 + … + βnXn
1) Definition Machine Learning y: The predicted value of the dependent variable
Machine learning (ML) is a branch of artificial
intelligence (AI), it is a field of study that allows computers β0: The intercept
having the ability to improve themselves based on training β1X1: The regression coefficient (β1) of the first
data or training data based on experience (what has been independent variable (X1) (the effect of increasing the value
learned). It builds the mathematical model over the sample of the independent variable on the predicted y value)
data and tries to find the trend of the sample data. It
βnXn: The coefficient of the last independent input variable.
432
B. Data processing and Model

1) Data processing
In the data processing step, the bases and input
parameters from the data file will be shown and screened.
When looking for data, quality and quantity are what we
aim for. Initially the file has a total of 10 columns with 10
different variables. This selection is based on our original
criteria: “Predicting inventory orders based on space and
warehouse conditions”. When inspecting the warehouse
with its storage capacity, it's really necessary to consider
Figure 3. Operation chart of Multiple Linear Regression in the maximum and minimum capacity that the warehouse
Machine Learning can hold products, the reordering index is also a point to
note when below the safety threshold. The output variables
3) The application of Multiple Linear Regression of the model and their meanings are described in the Table
1 below.
(MLR)
In business, this dependent variable can also be called a
Table 1. Explain variables
predictor or factor of interest, e.g., product sales, price, risk,
weather, stock market volatility, etc. The independent Column Name Meaning
variables are also known as explanatory variables because MinDOC (Hl) Minimum holding capacity of depot for products
it explains the causes impacting the dependent variable as Reorder Point
Balance point capacity of depot for products
(Hl)
well as the degree of influence that can be calculated using MaxDOC (Hl) Maximum holding capacity of depot for products
“parameter estimates” or “coefficients”. These coefficients Closing Stock Current stock at the depot for products
are tested for statistical significance by building confidence Demand Quantity of product that needs to be distributor
intervals around them so that the model that is built is
statistically robust and based on objective data. The Table 2 is an example of how to test the data of the
elasticity based on the coefficient can tell the extent to input variables, with the parameters getting from the
which a certain factor explains the dependent. Furthermore, reality, the model will apply the multivariable regression
a negative coefficient can be interpreted to have an inverse algorithm and give predictions based on the correlation of
relation with the dependent variable and a positive input variables and output variables from the collected data.
coefficient can be said to have a positive influence. The key This gives a more objective view of the applicability of the
element in any statistical model is a proper understanding model, relevance and its practicality.
of the domain and its application in practice.
This topic is not able to cover all its applications in a Table 2. Example of standardized input-output dataset
few lines. But we can consider a few typical and familiar MinDOC Reorder MaxDOC Closing Demand
examples, such as often hear TV forecasts about inflation, (Hl) Point (Hl) Stock
the country's GDP growth rate or reports on the market (Hl)
demand of an enterprise to prepare a business production 19218.6 70565.6 157453.7 49875 102745
3228.3 40347.9 54904.3 6618 9053
plan. In finance we can forecast stock prices and financial 139.5 532.3 1231.8 1065 1208
indices based on linear regression. Even in biology, linear 57.7 182.3 408.9 181 322
regression algorithms have been used such as finding the 223.7 969.4 2430.6 2314 2351
relationship between blood cholesterol (x) and carotid
intima (NTM) thickness (y). However, there are often many 2) Model
different factors that lead to a phenomenon. For example, In this model, the inputs are MinDOC (Hl), Reorder
in biomedicine, carotid artery thickening is not only caused Point (Hl), MaxDOC (Hl), Closing Stock and the output is
by blood cholesterol but also by many other factors such as Demand. The input parameters are filtered and selected
genetics, race, cardiovascular disease, age, gender, BMI, from the data file with additional criteria for the output
hypertension, diabetes... Therefore, a multivariable linear parameters, MinDOC(HI) and MaxDOC(HI) tell the
regression model is required to solve the inventory ordering minimum and maximum quantity of stock, respectively,
problem. From the above parameters, the authors have may contain. In warehouse management, knowing the stock
somewhat obtained the basic points to give a logical level is quite important for businesses when the demand for
prediction result that is the necessary level of demand. goods fluctuates from day to day, the control of spatial
indexes is absolutely necessary to order and manufacture.
The Reorder Point (HI) will let know when the item needs
to be re-ordered and will alert if it's under-stocked. And an
important and obvious parameter is the current Closing
Stock that will help optimize the storage capacity and solve
the inventory ordering problem. So, from the above
parameters, there are some basic points to make a logical
prediction with a necessary level of demand.
Figure 4. The applications of Multiple Linear Regression
433
Figure 5. Diagram of the inputs and outputs of constructed MLR
Based on the inputs and outputs in Figure 5, the demand

regression equation looks like this:
Demand = β0 + β1* MinDOC (Hl) + β2* Reorder Point (Hl)
+ β3*MaxDOC (Hl) + β4*Closing Stock
The correlation is a capacity of linear combination
between two variables. The values of the correlation
coefficient are always between -1 and +1. A correlation
coefficient of +1 illustrates that two variables are perfectly
relevanted in a positive linear sense, a correlation
coefficient of -1 illustrates that two variables are perfectly
relevanted in a negative linear sense, and a correlation Figure 7. Chart from seaborn
coefficient of 0 illustrates that there is no linear relationship
between two variables. III. RESULTS AND DISCUSSION
It is only possible to tell how or to what extent variables A. Results
are linked. The correlation coefficient in the regression
After adjusting the input and output variables based on
equation only measures the degree of linear association
Machine Learning, intercept values and coefficients have
between two variables. In Figure 6, the correlation metrics
been found in the Table 3. The probability values (p-value)
show that the input and output variables are highly
are all very low (mostly 0.0), which indicates that the
correlated, with the largest correlation of 0.99 and the
results of this model are statistically significant. The p-
smallest correlation of 0.67.
value of β2 is 0.052 but it's still within acceptable range
(Table 4), and the standard error (std err) also shows the
values quite well.
Table 3. Results of model

Value p-value std err
β0 -624.732727518287 0.002 310.934
β1 -1.18801497 0.0 0.098
β2 0.00723547 0.052 0.016
β3 0.81304756 0.0 0.016
β4 0.18758236 0.0 0.010
Table 4. Meaning of p-value

p - value Interpretation
0.001
0.01
Highly Significant
0.02
Figure 6. The correlation of variables 0.03
0.04
Figure 7 shows the seaborn diagram indicating the high Significant
0.049
correlation close to the regression line between the input 0.050 Redo calculations
variables, as well as between the output variable and the 0.051
On the edge of significant
input variable together and demonstrating continuity and 0.06
0.07
compatibility with the model. It can be seen that the 0.08
correlation between ReorderPoint and MaxDoc with the Highly suggestive, significant at the p < 0.10 level
0.09
output variable Demand has a certain convergence when 0.099
the demand for products increases, leading to the priority of >= 0.1 Meaningless
storage space and the reset point increasing accordingly.
From this basis, the model will follow along with other As a result (Table 3), with the values of β0, β1, β2, β3,
input parameters to make predictions. and β4, the formula of the demand regression looks as
follows:
434
Demand -624.732727518287 - 1.18801497 * MinDOC Table 6 shows some examples of comparisons between
(Hl) + 0.00723547*Reorder Point (Hl) + original demand and predicted demand. It shows that the
0.81304756*MaxDOC (Hl) + 0.18758236*Closing Stock accuracy is quite high, although there are cases where the
prediction is slightly wrong. In the 3rd row, when the
The above formula shows the model can predict original demand value is 2041, the model only predicts
demand from input variables such as MinDOC (Hl), demand at approximately 1418. This is easy to understand
Reorder Point (Hl), MaxDOC (Hl) and Closing Stock. when applying a classical linear regression model to
The model results also show very a high accuracy as predict, the results from the model will not completely
well as the mean absolute error has a good value (Table 5). match the parameters in the data file. However, most of the
The accuracy of Training set is 97%, while the accuracy of demand prediction results show high accuracy when
Testing set and Predict vs Original demand are 98% compared to the original demand. For example, in the 4th
showing that the model is very accurate, but not absolute. row, when the value of original demand was 8592, the
With high accuracy, the proposed model demonstrates that model predicted demand as ~8343. Therefore, linear
the data file has logic and uniformity between input and regression is difficult to predict with 100% accuracy, in
output, and the Figure 6 has shown this. Before the input some cases there will be a difference between the
and output variables are fit together, the data has gone prediction and the original.
through a cleaning process, removing the extraneous data.
The mean absolute error of the model is approximately Table 6. Compare results
3114, which is not too large. Original Demand Predict Demand
8490 7595.921050
The resulting indicators have shown that multiple linear 2589 2604.004210
regression in Machine Learning may be the most 2041 1418.395567
appropriate method for predicting data with very high 8592 8343.801069
accuracy. 53552 46284.413407
Table 5. Model accuracy As described in Section A, part 3.1, this model allows
The mean absolute
import data, then the model will process input and export
The accuracy output (Table 7). Applying the multivariable linear
error
Training Testing Predict vs Demand
3114.1950158229974
regression algorithm and running the β variables from the
97% 98% 98% data file entered into the algorithm and output the
B. Discussion prediction results clearly. The model also presents
flexibility in forecasting and convenience for managers and
The results of this model are really amazing with a high companies by importing input values, the model will
prediction accuracy rate. Figure 8 has shown that the ratio automatically export the results with high accuracy.
between Original demand and demand Prediction demand
does not differ much (with Original being red, and Table 7. Process input and output of the model
Prediction being blue). The graph shows the coverage of
blue on red, except in some places (but it's range very Input
MinDOC (Hl) 1167.3
small, not a concern). This proves that the model is capable Reorder Point (Hl) 3102.9
of making appropriate predictions for the business based on MaxDOC (Hl) 11036.8
the given input parameters. Previous tests with other Closing Stock 7247
models such as OLS, neural networks gave different results Output
but were not very positive when the original prediction Demand 8343.801010991714
index was quite different from the model's prediction. The
introduction of different methods will help optimize the In this model, linear regression in machine learning has
purpose of this study, which is to put Machine Learning well solved the problem of demand forecasting from input
into the warehouse management system with the desired variables. In this point, all of the steps are completed, the
results.
model is performed and optimized and it can be concluded
that linear regression in machine learning is probably one
of the best statistical and forecasting methods.
IV. CONCLUSION
Machine Learning techniques are researched and
applied in the warehouse design management system to
optimize the operational indicators of the warehouse
system. This study uses the Multiple Linear Regression
model implemented on Google Collab platform through the
collected data of MinDOC (Hl), Reorder Point (Hl),
MaxDOC (Hl), Closing Stock variables as input, from the
simulated statistics. From there, predicting the quantity of
demand to order is expressed through the function Demand.
Figure 8. Original demand and demand Prediction chart The goal of the article is to maximize customer satisfaction,
reduce inventory and inventory waste, so that inventory
435
management can be improved as a part of supply chain [3] Praveen K B, Prateek J, Pradyumna Kumar, Pragathi G,
management. Prof. Madhuri J., “Inventory Management using Machine
Learning”, International Journal of Engineering Research
The model was built and trained to give results with a & Technology (IJERT), vol. 9, pp. 866- 869, 2020.
forecasting accuracy of 98%. These results show that the [4] Samiul Islam, Saman Hassanzadeh Amin, “Prediction of
probable backorders scenarios in the supply chain using
basic model has met the goal of predicting customers’ Distributed Random Forest and Gradient Boosting Machine
demand through the initial variables quite accurately. The Learning techniques”, Journal of Big Data, vol. 7(1), pp. 1-
research is just completed in its early stages, still needs 22, 2020.
further development. For future research, the model will be [5] Tereza Šustrová, “A Suitable Artificial Intelligence Model
for Inventory Level Optimization”, Trends Economic and
deployed and developed with a larger source of multiple Management, vol. 25(1), pp. 48-55, 2016.
input variables. The prediction speed of the model will be [6] Galina Merkuryeva, Aija Valberga, Alexander Smirnov,
improved, and the accuracy reaches the maximum level. “Demand forecasting in pharmaceutical supply chains: A
The model can be combined with many tools and software case study”, ICTE in Transportation and Logistics 2018
to bring wider, closer, and more convenient application. (ICTE2018), vol. 149, pp. 3-10, 2019.
[7] Hironobu KAWAMURA, Keisuke NOMOTO, Enchih
KUO, “Inventory Management Based on Demand
REFERENCES Forecasting Using Ryokan’s Beer Sales Data”, Innovation
[1] Nicholas Ampazis, “Forecasting Demand in Supply Chain and Supply Chain Management, vol. 9(4), pp. 127-135,
Using Machine Learning Algorithms”, International 2015.
Journal of Artificial Life Research, vol 5, pp. 56-57, 2015.
[2] Lolli Francesco, Balugani Elia, Ishizaka Alessio, Gamberini
Rita, Rimini Bianca, Regattieri Alberto, “Machine Learning
for multi-criterio inventory classification applied to
intermittent demand”, Production Planning and Control,
vol. 30(1), pp. 1-14, 2019.
436
Ethanol in Gasoline Fuel Blends: Experimental

Investigation Effects on Exhaust Emission of the
Homogeneous Charge Compression Ignition Engine
Minh Xuan Le Thanh Tuan Nguyen*
Automotive Engineering Dept. Automotive Engineering Dept.
Dong A University Nha Trang University
Da Nang, Vietnam Khanh Hoa, Vietnam
xuanlm@donga.edu.vn nguyenthanhtuan@ntu.edu.vn
Abstract: The study was carried out on the HCCI engine ethanol mixing with gasoline fuel can improve the octane
converted from a primitive diesel engine. The fuel used for number. A 3% reduction in fuel consumption and a 10%
this study is a mixture of gasoline and alcohol fuel with reduction in carbon dioxide emissions are achieved in the case
different blends of alcohol such as E0, E10, E20, and E30. of the combined use of both fuels [4–6].
The value of the con-centration of toxic substances obtained The HCCI engine (Homogeneous Charge Compression
from the experiment can confirm that, with the higher ethanol Ignition - HCCI) is a type of internal combustion engine, with
fuel composition in the mixture, the concentration of harmful the fuel and air mixture being mixed first, then compressed to
substances released into the environment decreases. The HC, the point of self-ignition. The type of combustion model with
CO, NOx ratio decreased the most when the engine used E30 HCCI has many advantages in terms of high thermal
gasoline compared with E0 was 53%, 59%, and 50%, efficiency and very small NOx and PM emissions, ensuring
respectively. Also, in the study, the value of the air residue strict requirements on emission and pressure reduction for
coefficient was changed by values of 1 and 0.7. The rich fuel petroleum-based fuels because of the type of engine. This
mixture increases the concentration of HC and CO. In engine is suitable for using alternative fuels of biological
contrast, the concentration of NOx decreases sharply. This is origin [7]. So HCCI is one of the research directions to
due to the lack of oxygen in the fuel, incomplete combustion, develop engines soon. HCCI is a combination of advantages
and reduced combustion temperature. of both diesel engines (thermal efficiency) and gasoline
engine (emissions). The engine operates with a light mixture.
Keywords: HCCI engine, emission concentration, ethanol These factors help to improve thermal efficiency [6]. There
are no local areas with high fuel density when creating a
I. INTRODUCTION homogenous mixture. The intense combustion diffusion
The demand for fossil fuels is increasing day by day due process does not take place, reducing PM emissions. In
to the fantastic development of both society and industry. The addition, combustion takes place entirely, and in the entire
issue of ensuring energy security has always been focused on cylinder, so the combustion temperature is reduced, NOx
by countries, with gasoline prices being too high, energy emissions are reduced. However, this type of engine still has
shortages due to depletion of reserves, and political conflicts many problems that need to be solved, such as the
in the Gulf region, especially in the Gulf of Tonkin. An concentrated working area at small loads and the difficulty of
escalation of tension and war between Russia and Ukraine has directly controlling the combustion process. The working area
just occurred. Therefore, alternative fuel sources are needed to of the HCCI engine is limited by two factors: non-combustion
reduce dependence on fossil fuels. and detonation [8]. The mixture is more challenging to self-
ignite at high speeds due to insufficient time to react. The
Ethanol is considered one of the essential alternative fuels combustion process of HCCI engine is very complicated and
with the potential to replace fossil fuels because of its ease of is determined by the reaction dynamics, so it is necessary to
production and the availability of agricultural products. ensure the properties of the mixture so that the combustion
Ethanol is liquid and has chemical properties similar to time starts near the top dead center [9]. The problem of
conventional fuels [1, 2]. The octane number in ethanol is controlling the self-ignition moment on HCCI engines is not
higher than in gasoline, which is why it is used in engines to as simple as on gasoline and diesel engines [10–12], with the
improve detonation [5]. Many researchers have published the ignition time being calculated from the time the spark plugs
study of ethanol fuel blending. Yusuf and Inambao did ignite or the time when diesel fuel is injected into the
experimental research on a single-cylinder, four-stroke, SI combustion chamber. In addition, starting is also one of the
engine equipped with an electronic fuel injection system. The main problems with this type of engine.
fuel is pure gasoline (E0) mixed with bio-alcohol at the rate of
5%, 10%, and 15% ethanol. The results show that emissions Published studies have shown that combustion control in
are lower when the engine runs on E10 and E15, and NOx HCCI engines effectively combines changes in compression
formation rates increase higher with E5 and E10. CO2 and CO ratio and intake air temperature [5, 6, 9]. Combustion on HCCI
emissions for E15 decreased and overgrew with E5 and E10, engines has more suitable characteristics when using bio-fuels.
while HC emissions decreased at all injections for E5, E10, In this study, an experiment was conducted to determine the
and E15 [3]. Celik and Agarwal have shown that increased influence of the fuel mixture composition on the fuel

consumption rate and the concentration of toxic substances The schematic diagram of the test system is shown in Fig.
discharged into the environment by the engine. 1. The engine is connected to the power test bench, this
device can measure up to 500 HP. Use class K thermistors to
II. THE APPARATUS AND PROCEDURES OF measure intake air, fuel, and coolant temperatures,
EXPERIMENT thermocouple temperature range from 0 to 600°C. The
A. Fuel Test cooling water temperature is controlled within the limit of
Experimental engines were conducted using five types of 75oC. The amount of fuel consumed is measured by a flow
fuel. The first is gasoline A95 (E0), other fuel blends meter. The injection time is controlled by an electronic
containing 10%, 20%, and 30% ethanol by volume (E10, E20, controller designed and connected to the engine speed sensor.
and E30, respectively). The main fuel characteristics of all the Exhaust gas measurement equipment is used EKOS MOD-
test blends and pure ethanol are shown in Table 3. It clearly 9000. Analyzers measured emissions: non-dispersive
explains that, with the addition of ethanol to the fuel, the infrared (NDIR) analyzed CO emissions, a flame ionization
oxygen content and density of the ethanol/gasoline fuel, the detector (FID) analyzed HC emissions, and a
studied octane number increased, while the carbon content, chemiluminescence detector (CLD) measured NOx
calorific value, and isostatic air/fuel ratio decreased emissions.
simultaneously. Fuel properties are shown in Table 1.
Table 1. Fuel Properties of Gasoline, Ethanol and Ethanol in

Gasoline Fuel Blends [13]
Properties E0 E10 E20 E30 Ethanol
LHV (MJ/kg) 43.3 42.3 40.5 38.5 27
Stoichiometric A/F 14.6 14.1 12.6 11.7 9

(kg/kg)
Density (kg/m3) 745 750 755 760 792
Octane number 95 96.7 98.8 100.8 115

(RON)
Carbon content 86.6 81.2 75.5 72 52.2

(%wt)
Oxygen content 0 5.4 10.2 14.4 34.7

(%wt)
Figure 1. The homogeneous charge compression ignition (HCCI)
engine test bench.
B. Engine Test and Experimental Apparatus
When the engine starts working, to start the HCCI engine
D28 diesel engine converted to the HCCI engine in this at low speed and warm-up, the engine initially uses spark plug
study. The fuel supply system has been reworked, with fuel ignition mounted on the top of the engine cover. After running
injectors in the intake manifold. The intake manifold installed when stabilizing at low speed, the engine will cut off the
a blower to push air into the combustion chamber, similar to ignition source to run on the HCCI principle.
the turbocharger installed on the engine. The rotary blower is
regulated to maintain the intake air pressure at 3.5 bar. After The experiment was conducted when in two modes of rich
conversion, the engine image is mounted on an iron frame for and standard fuel mixture, respectively, with air residue
adjustment and operation (Figure 1) before being completed coefficient (λ) of 0.7 and 1. The experimental fuels used were
and put on a test platform. The engine parameters before and E0, E10, E20, E30, respectively. The engine rotation speed
after conversion are shown in Table 2. and compression ratio were kept the same in the test modes
with values of 1300 rpm and 14, respectively. This is the most
Table 2. parameters of the Aspirated engine and the engine relevant value for this HCCI engine in the previous study. The
converted into HCCI atmospheric temperature during the experiment was 30oC.
The steady-state tests were repeated at least twice to ensure
Parameter Aspirated engine HCCI engine that the results were reproducible within the uncertainty of the
test.
Engine CI, Single cylinder HCCI, Single cylinder
Max power 17.2 (kW) - III. RESULT AND DISCUSSION

Fig. 2 shows the experimental results of measuring the
Max engine speed 2200 (rpm) 1600 rpm
concentration of HC at the engine rotation speed of 1300 rpm,
Bore x stroke 125 x 120 (mm) 125 x 120 (mm) when the engine operates in two modes of mixed enrichment
and mixture at equilibrium. The fuels used are E0, E10, E20,
Displacement 1,472 (l) 1,472 (l) E30. It can be seen that the more HC is released into the
environment when the air residue coefficient is small, then the
Compression ratio 17 14
excess fuel component has not entirely burned due to the lack
Specific fuel 257 (g/kW.h) - of oxygen. When increasing the ethanol content in fuel, the
consumption concentration of HC tends to decrease gradually because the
molecular structure of ethanol contains fewer hydrocarbon
438
components than gasoline. The most significant reduction in mixture is rich but reduces the air residue coefficient. It will
HC emissions is 53% when the engine uses E30 fuel (40 ppm) affect the concentration of HC and CO emissions, and the
compared to pure gasoline E0 fuel (85 ppm). combustion pressure will decrease and reduce engine power.
Figure 2. Effect of ethanol composition mixed with gasoline on

HC emission concentration Figure 4. Effect of ethanol composition mixed with gasoline on
NOx emission concentration
Like CO concentration, with an air residue coefficient of
0.7, CO concentrations from the engine are higher when IV. CONCLUSION
operating with an air residue coefficient of 1 (Fig. 3). When Due to rising fuel prices, increasing energy demand, and
the mixture is rich, due to lack of oxygen, non-combustible stricter regulations on air pollution in the road transport sector,
areas will appear, leading to a lower combustion temperature researchers worldwide are working to develop new fuels.
and reduced burning rate. This is the cause of incomplete Among the available alternative fuels, ethanol in gasoline fuel
combustion, low combustion efficiency, and high CO mixtures is one of the most widely used renewable fuels for
concentration. Meanwhile, CO concentration also decreased internal combustion engines. The present study
gradually when increasing ethanol content in gasoline fuel. experimentally investigated the effect of the ethanol/gasoline
This reduction may be attributed to the increase in oxygen mixture on emissions of the HCCI engine. The results
content in the ethanol/gasoline blended fuels, which improves obtained in this study show that all emissions are reduced
the oxidation of CO in the fuel-rich regions of the combustion when ethanol is more present in the fuel mixture. When using
chamber, ensuring complete combustion during the engine E30 and E0, the concentration of toxic substances HC, CO,
running. The most significant reduction in CO emissions is 59% and NOx decreased by 53%, 59%, and 50%, respectively.
when the engine uses E30 fuel (0.9% vol) compared to pure
gasoline E0 fuel (2.2% vol). The reduction in CO and HC emission factors is related to
the high oxygen content of the ethanol molecule (which
improves the oxidation of incomplete combustion products in
the fuel-rich regions of the combustion chamber).
Furthermore, the addition of ethanol in the gasoline fuel
mixture produces a lower flame temperature resulting in a
lower exhaust gas temperature, making lower NOx emissions.
Although the NOx concentration is greatly reduced when
operating with a fuel-rich mixture, further studies are needed
to evaluate the affected engine performance parameters. This
study is the basis for other studies to perfect the operation of
the HCCI engine using gasoline mixed with ethanol.
REFERENCES
Figure 3. Effect of ethanol composition mixed with gasoline on
[1] N. T. Tuan, “Applicability and development LPG vehicles in
CO emission concentration Vietnam. In: XLI. international scientific conference of Czech
and Slovak university departments and institutions dealing with
For NOx emissions (Fig. 4), when the ethanol composition the research of combustion engines”, 2010, pp. 303–315.
is higher, the calorific value of the fuel mixture will decrease. [2] N. T. Tuan, “Exhaust emission of motorcycles engines running
At the same time, increase the latent heat of vaporization of on ethanol blended gasoline”, xlii. International scientific
the mix. At the end of the intake stroke of the engine cycle conference of Czech and Slovak university departments and
then, the temperature of the air/fuel mixture decreases. In this institutions dealing with the research of combustion engines,
2011.
condition, the combustion temperature will also drop
[3] P. Iodice, G. Langella, A. Amoresano, “Ethanol in gasoline fuel
gradually, which will affect the NOx heat generation process. blends: Effect on fuel consumption and engine out emissions
This is what causes NOx emissions to decrease as the ethanol of SI engines in cold operating conditions”, Appl. Therm. Eng,
content of the blended fuel increases. The most significant 2018, 130, 1081–1089.
reduction in HC emissions is 50% when the engine uses E30 [4] C. Cinar, A. Uyumaz, H. Solmaz, F. Sahin, S. Polat, E. Yilmaz,
fuel (100ppm) compared to pure gasoline E0 fuel (200ppm). “Effects of intake air temperature on combustion, performance
NOx concentration decreases sharply when the combustion and emission characteristics of a HCCI engine fueled with the
439
blends of 20% n-heptane and 80% isooctane fuels”, Fuel [9] T. K. Sharma, G. A. P Rao, K. M. Murthy, “Homogeneous
Process. Technol, 2015, 130, pp. 275–281. Charge Compression Ignition (HCCI) Engines: A Review,”
[5] D. H. Qi and C. F. Lee, “Combustion and emissions behaviour Arch. Comput. Methods Eng, 2015, pp. 623–657.
for ethanol–gasoline-blended fuels in a multipoint electronic [10] N. T. Tuan and N. P. Dong, “Improving performance and
fuel injection engine”, International Journal of Sustainable reducing emissions from a gasoline and liquefied petroleum
Energy, 2014, 35, pp. 323–338. gas bi-fuel system based on a motorcycle engine fuel injection
[6] H. A. K. Shahad, S. K. Wabdan, “Effect of Operating system”, Energy Sustain. Dev, 2022, 67, pp. 93–101.
Conditions on Pollutants Concentration Emitted from a Spark [11] N. T. Tuan, and N. P. Dong, “Theoretical and experimental
Ignition Engine Fueled with Gasoline Bioethanol Blends”, J. study of an injector of LPG liquid phase injection system”,
Renew. Energy. 2015, pp. 1–7. Energy Sustain. Dev, 2021, 63, pp. 103–112.
[7] M. K. Mohammed, H. H. Balla, Z. M. H. Al-Dulaimi, Z. S. [12] T. N. Thanh, and D. N Phu, “A Study of CNG Fuel System
Kareem, M. S. Al-Zuhairy, “Effect of ethanol-gasoline blends Uses Mixer for Engine of the Suzuki Viva Motorcycle. Int. J.
on SI engine performance and emissions”, Case Stud. Therm. Mech. Eng. Robot. Res, 2022, 11, pp. 37–42.
Eng, 2021, 25. [13] M. Amine and Y. Barakat, “Properties of gasoline-ethanol-
[8] J. E. Tibaquirá, J. I. Huertas, S. Ospina, L. F. Quirama, J. E. methanol ternary fuel blend compared with ethanol-gasoline
Niño, “The effect of using Ethanol-Gasoline blends on the and methanol-gasoline fuel blends”, Egypt. J. Pet, 2019, 28, pp.
mechanical”, Energy and Environmental Performance of In- 371–376.
Use Vehicles. Energies, 2018, Vol. 11, pp. 221-232.
440
The Impact of the Industrial Revolution 4.0 on

Logistics Businesses: A Case in Mekong Delta
Ngo Hong Ngoc Huang Ying Yin
Department of Industrial Engineering & Management Department of Industrial Engineering & Management
National Taipei University of Technology National Taipei University of Technology
Taipei, Taiwan Taipei, Taiwan
nhngoc@ctuet.edu.vn yingyinhuang@mail.ntut.edu.tw
Abstract: The 4th Industrial Revolution is the engines were utilized to mechanize production during the first
manufacturing movement that is reshaping the business of industrial revolution, which began in the late eighteenth
building products and strongly impacting related industries. century. The second revolution took place at the end of the
As in the case of most industries, logistics in Vietnam is twentieth century thanks to the application of electricity to
encountering sustained and profound changes, which mass production. Electronics and information technology are
demands a rapid shift in processes to accommodate the used to automate production during the third revolution [1] ,
fundamental of Industry 4.0. This study aims to examine the [2]. Unlike previous revolutions, Industry 4.0 is not associated
impact of the 4th Industry Revolution (Industry 4.0) on the with the birth of a specific technology but rather integrates
multiple industries with an emphasis on application [2] [3] [4].
logistics industry and enterprises in the Mekong delta. The
As a result of this revolution, industries in society must shift
article also gives an overview in terms of activities of
to become "smarter" and more efficient through boosting
logistics service within the Mekong Delta region and suggests automation in connection with the digital system. This is to
possible solutions to improve logistics services in Industry relieve individuals from manual and intellectual labor in order
4.0 era. to fulfill increasing demands in today's society. The speed and
scale of impact of modern technologies have accelerated
Keywords: Industry 4.0, enterprises, Logistics 4.0, socio-economic development in the industrial and service
Mekong delta. sectors [2].
I. INTRODUCTION In general, Industry 4.0 emphasizes the role of digital
production, often known as smart factories [4]. Smart
The 4th industrial revolution (Industry 4.0) is now owning
manufacturing is not about replacing people at work; instead,
every corner of life, strongly affecting every industry, every
it is about preventing errors and speeding up production by
field, every business (enterprise), creating a big change in the
collaborating to share knowledge or data effectively [1] [4].
way production mode, the convergence of applied physics and
All information and data-sharing processes are managed by
digital to form the emergence of the Internet of Things (IoT)
humans [1] [4]. Data is the lifeline created in all
that changes rapidly, the entire width of the value chain from
manufacturing processes, from product development to
R&D to manufacturing, logistics Customer service, reduced
production to after-sales service, and the amount of data
transaction costs, and shipping come magically to production
generated is enormous (Figure 1) [1] [2]. In the
and capacity operations. The competitiveness of enterprises is
implementation of advanced business models, adaptable data
assessed by applying technology in production and
that reflects compatibility between customers and suppliers, if
management.
systematically and intelligently integrated, can provide great
The Mekong Delta is identified as a key economic region efficiency [1]. The emergence of several digital technologies
of Vietnam with the main advantage of agricultural production such as Artificial Intelligence (AI), as well as the integration
that has not fully promoted its role to contribute to the of the Internet of Things (IoT) and Big Data or Cloud
promotion of production and export, especially the operation computing, are indeed implications. The core of digital in the
of the product business, which is the strength of the region. In 4.0 industrial revolution has formed a special production
the face of fierce competition at the school, especially the platform, that is smart manufacturing. The application of
Industrial Revolution 4.0, in order to adapt to the trend of information and communication technologies, as well as
international economic integration, the development of computer science IoT (Internet of Things) and IoS (Internet of
logistics services in the Mekong Delta region will be an Services), would have an impact on every aspect of the
important factor in optimally developing the network of modern production process, boosting the operational
supply and application of goods of this rich land with potential efficiency of the entire supply chain. This helps to transform
for economic development. existing working methods, develop numerous innovative
products with different and outstanding quality, reduce time,
II. THE 4TH INDUSTRIAL REVOLUTION (INDUSTRY and protect the environment. [1] [2] [3] [5]. As a result, supply
4.0) chain management activities, particularly product and
information flow management (logistics), are introduced in
A. Industry 4.0
this article as "Logistics 4.0".
In the history of technological and industrial development,
mankind has witnessed three industrial revolutions. Steam

C. Impact of Industry 4.0 on logistics

Industry 4.0 has been identified as having 4 main impacts
on businesses, including i) customer expectations, ii) product
quality improvement, iii) new author change, and iv) format
organization.
For industry logistics, Industry 4.0 is the core foundation
for future development. Show the development trends of
logistics can be divided into 3 areas: IT and software (IT &
software), robotics and sensor engineering (robot & sensor
technology), and network systems (network). The
combination of three elements forms the core foundation for
the different directions in logistics 4.0 [9].
Industry 4.0 helps logistics businesses improve customer
service efficiency while minimizing order fulfilment costs and
time, opening new facilities and business types in the field.
Figure 1. The combined data and digital technology of the value
this. To achieve that, ask businesses that need to change and
chain form a smart factory reorient their development strategies, specifically:
(Source:https://www.lgcns.com/LGCNS.GHP.Main/Solution/SmartFactory
_En) • Data automation and transparency: New technologies
applied to data collection and analysis (which are a
B. Logistics 4.0 central part of logistics) provide an opportunity for
Logistics is frequently classified based on the type and expert logistics professionals to understand the status,
level of logistics services provided by a logistics firm (1PL, data analysis is a fine way to get the title item. Those
2PL, 3PL, 4PL, 5PL) [6]. However, in this case, the technologies are decimal systems and data
development process in logistics is examined. Logistics 1.0 processing, logistics control, artificial intelligence
refers to the period of transport mechanization that began in (AI), ... [3]
the second half of the nineteenth century when ships and • Utilizing innovative modes of transportation,
steam engines replaced people and animals as the primary including self-driving cars and drones, to assemble
modes of transport [3]. Logistics 2.0 (from the early twentieth and transport goods by automating devices or
century to 1960) was an automation era in logistics that vehicles. Warehouse handling machines, delivery
included freight and warehouse mechanization, automated drones, automatic delivery trucks, smart forklifts, and
handling systems, mechanical loading and unloading systems so on... [3]
(electric power invention), and mass manufacturing [3] [7].
Logistics 3.0 (1960 - 2000), a stage of information technology • The use of digital-based devices in the logistics
(IT) application to standardize logistics management. system. Examples, Wi-Fi and Bluetooth-enabled
Warehouse management systems (WMS) and Transportation devices for tracking, positioning, navigation, and
Management Systems (TMS), for example, which automate observation; online barcode scanning apps in
and efficiently handle logistics, inventory, and transportation, warehouse management; cloud-based inventory
have greatly evolved and improved [3], [7]. Logistics 4.0 optimization... [3]
(2000 - present) is the logistics development phase in the • Latest production methods in the future in terms of 3D
Industry 4.0 era, mainly focused on the development of the printing's influence: can change traditional logistics
Internet of Things (IoT) and Big Data. The primary purpose services, enhance product output to the market [6]
of logistics 4.0 is to reduce labor costs and standardize the [10] [6, 10].
workforce in supply chain management [8]. Warehouse robots
and autonomous vehicles are attempting to replace procedures Furthermore, Industry 4.0 enables businesses to start up
that do not require human input or decision-making. The aim with less initial investment and earn higher profits in a shorter
is to obtain the ideal balance of automation and mechanization time. Firms can effectively access each business unit with the
[3] [7]. help of IoT, monitoring and evaluating their performance in
real-time. Hyperlinking, on the other hand, raises
cybersecurity concerns.
III. THE CURRENT STATE OF LOGISTICS IN THE
MEKONG DELTA
D. Opportunities
• Government’s support policies
Joining international trade agreements has created chances
to increase the number of commodities imported and exported
by the whole country, and by the Mekong Delta in particular.
With the benefit of diversified development in agriculture,
fisheries, and tourism, the government has adopted policies to
Figure 2. The history of Logistics 4.0's development
(Source: [4]) invest in the country's transport infrastructure connecting with
other regions.
442
The Prime Minister's approving the long-term strategy for logistics services as well as the correlating resource
the development of the logistics compartment in the entire requirements, to ensure meeting future development trends.
country to 2020, with orientation to 2030, would determine
sub-regional economic development. The Mekong Delta's
central economy will have one Grade II logistics center with a
minimum scale of 30 hectares by 2020 and more than 70
hectares by 2030. The scope of strategy mainly includes the 9
provinces and cities in the Mekong Delta, connecting with
inland ports, river ports (Can Tho, My Thoi), airports, railway
stations, bus stations, industrial parks, and border gates (in
Kien Giang and An Giang provinces) [11].
Moreover, according to Official Dispatch No.
7709/VPCP-CN, dated September 15, 2020, on the request
"There’s a must-have separated cargo airline (Cargo Airlines)
serving Vietnamese agricultural products, for separate routes,"
[12] several VLA enterprises have aided capital to establish
Asian Cargo Gateway Joint Stock Company (ACG), with the
first flight on March 6, 2021, providing fixed weekly air Figure 3. Transport model – Forecasting the increase of inland
freight services from Ho Chi Minh City to several large Asian waterway transport 2030 - 2040
markets, including – Jakarta, Bangkok, and Incheon. As the (Source: [13])
largest granary of rice and seafood in the country, it is
regarded not only as a significant opportunity for agriculture • Market potentials
in the Mekong Delta but also as a big motive to impulse the Currently, the country has almost 300,000 logistics
development of the region's logistics industry. businesses, with a total workforce of approximately 1.5
• Strategic position million workers. Meanwhile, the number of logistics firms
operating in the Mekong Delta region is modest, with no
The Mekong Delta offers key locations that connect significant firms directly participating in this sector.
transportation and have enough room to establish a very Meanwhile, the demand for logistics services has constantly
convenient logistics center. Furthermore, because of the expanded, both in quantity and quality, particularly since
geographical features of the Mekong Delta, which is in an area Vietnam's recent membership in FTAs, EVFTAs, and
with international maritime and aviation traffic routes worldwide economic activities.
connecting South Asia and Southeast Asia, as well as
Australia and other Pacific Ocean archipelagos. To According to the Annual Economic Report of the Mekong
agriculture, the system not only provides water for production Delta in 2020, the Mekong Delta currently contributes about
activities but also has enormous implications for product 40% of agricultural production value, over 54% of rice
transportation. In comparison to roadways, inland waterways production, which is 90% of those is exported, 65% of aquatic
provide more efficient delivery with huge volumes of products, and 70% of fruit production of the country [14]. The
agricultural products. region's potential for agricultural and aquatic product export
indicates that there is an urgent need for developing the
In general, the Mekong Delta's waterway and land logistics service industry to serve the entire region's
transportation systems are dense and evenly developed. The agricultural and aquatic products, with major logistics services
Mekong Delta's port system spreads across the Hau and Tien such as transportation and warehousing. goods, preserving
rivers. The rivers and canals system are 28,000 km long, with goods, and value-added services. Services such as cold
23,000 km capable of transportation, accounting for 70% of storage, irradiation, and autoclaving for fruit products, for
the country's river length. There are two vital waterways from example, to ensure the quality of exported goods as well as
Ho Chi Minh City to Kien Giang and Ca Mau, as well as the merchandise distributed for domestic consumption.
Quan Chanh Bo canal, which allows large vessels to enter and
exit the Hau River. There are five road corridor routes that Furthermore, the Mekong Delta's economy is thought to
connect the Mekong Delta to the Southeast and the rest of the be growing steadily, and the business environment is
country. Furthermore, the Mekong Delta's airport system has improving to attract investors. As a result, this is a very
been upgraded and developed so that it can connect to promising investment market.
international flights. • Opportunities provided by Industry 4.0
With the advantages of inland waterway transport and the Industry 4.0 enables firms, particularly logistics
development trend of the Mekong Delta's inland waterway enterprises in the Mekong Delta, to expand their markets,
transport infrastructure system, an increase in inland which is the driving force of corporate development. To reach
waterway transport is expected in 2030-2040, with an average Industry 4.0 standards, businesses must invest in machinery,
growth rate of 5%. Report on the Feasibility Study Activity technology, and automation in production and management
Development Project Southern Regional Logistics and operations such as applying digital technology and integrating
Waterway Corridor of the Ministry of Transport, funded by smart technologies to optimize process and mode of operation.
the World Bank (WB). The implementation survey
demonstrates the transformation from road transportation to Automated technology also helps enterprises in combating
inland waterway transport in the Mekong Delta regions, fragmentation, instability, inconsistent product quality, and
consequently increasing the need for inland waterway noncompliance with quality standards such as Global GAP.
Following that, businesses in this region would confidently
443
sign contracts to supply items to big clients with guaranteed Delta is currently hampered by issues such as a lack of
quality. investment budget, high transportation costs for enterprises
due to a lack of key logistics centers and other satellite center
Vietnam is one of the five nations in the world with the systems, a lack of empty container yards, a warehouse system
fastest growth in over 30 years. Industry 4.0, like earlier in ports, and constrained road and port traffic systems, among
industrial revolutions, promises to deliver significant benefits others. The mentioned challenges have blocked the logistics
and prospects to the logistics industry, leading to shorter system, resulting in supply chain congestion.
delivery times, reduced shipping expenses, and lower
communication costs, then optimizing business costs. At the Moreover, a report in 2022 shows that, even though
same time, it will help the logistics system and supply chain approximately 123 factories are producing and processing the
of companies be more transparent. main agricultural and aquatic products (rice, fruit, aquatic
products), the Mekong Delta accounts for only 30% of the
E. Difficulties cold storage in the Southern region, primarily concentrated in
a) Hard logistics infrastructure Long An, Can Tho, and Hau Giang [16].
• The Mekong Delta's logistics infrastructure remains In the context of the Covid-19 pandemic, the Ministry of
deficient. Industry and Trade has issued recommendations for logistics
companies, particularly those involved in cold storage, to
According to the report for the period 2014-2018, the
support agribusinesses by prioritizing the preservation of
Mekong Delta region's export growth rate is 10.57 percent,
agricultural products that are struggling to export to China and
with exports increasing by an average of 9.43 percent per year
lowering storage and handling costs.
and imports increasing by 13.1 percent per year. The region's
average annual export volume of products is around 17-18 This fact demonstrates the critical need to develop cold
million tons. However, up to 70% of cargo volume must be storage infrastructure in the Mekong Delta region in order to
transported by road to Ho Chi Minh City and Cai Mep ports. improve the region's ability to respond to any arising incidents
This increases shipping costs by 10% to 40% per cargo. and ensure that they can store and preserve agricultural and
aquatic products after harvest while reducing spoilage rates
The Mekong Delta is home to 12 seaports, 35 wharves,
and be proactive in production, export, and distribution for
and 4.9 kilometres of wharves. These seaports function as
domestic consumption.
satellite docks for the ports of Ho Chi Minh City and Cai Mep
- Thi Vai, as well as for short-distance inter-regional b) Soft logistics infrastructure
transportation. Because of the dense waterway traffic, it is
considered to have the advantage of waterways for • Enterprises in the industry lack interaction with
transporting goods. However, at present, inland waterway farmers and industry linkages
transport in the Mekong Delta still has to compete with road It is estimated that more than 90% of farmers consider
transportation due to limited infrastructure, which varies traders as their sole source of markets, as well as their
depending on the water and the static height of the bridge. As exclusive buyer of agricultural products. Therefore, traders act
a result, the average transit time for the routes Can Tho - Cat as sellers and manipulate prices while businesses are unaware
Lai and Can Tho - Cai Mep Thi Vai is 18 hours and 36 hours, of the supply situation. As a result, businesses lack sufficient
respectively, while the average road traffic on the same route information to make accurate forecasts and supply planning
is 5 hours and 8 hours. The freight of refrigerated containers decisions. If the output enterprise in a supply chain does not
by inland water vessel is significantly higher than that of land master the plan, the efficiency of the entire supply chain
transport (70 percent higher per 40 feet container, and about 9 suffers. The Mekong Delta is the country's largest rice and
percent higher per 20 feet container) [15]. The above issues aquatic manufacturer. According to a survey conducted by the
are huge barriers to promoting transport development of Vietnam Logistics Research and Development Institute
inland waterways that connect the Mekong Delta with the (2019) among logistics enterprises and shippers about
main ports of the southern region. logistics challenges serving agricultural and aquatic products.
This situation is due to the fact that the Mekong Delta still Some surveyed groups said they handled logistics themselves,
lacks seaports and deep-water ports capable of serving the which indicates the significance of outsourcing costs,
growing number of modes of transportation. A major logistics apprehension about information leakage, and a lack of trust in
center and satellite center systems have yet to be built; there logistics companies. The reason is that the region's logistics
are no container yards or cold storage systems for agricultural enterprises continue to operate independently and without the
and aquatic products at ports or airports, which are primarily necessary connections. While the current logistics trend is
manufacturer's depots, ... collaboration and outsourcing, each company should focus on
its strengths and consider outsourcing services that are not
Furthermore, the transport infrastructure between strengths.
localities is underdeveloped and asynchronous, particularly in
rural areas, making it difficult for heavy-loaded vehicles to • The application of information technology (IT) in
enter and pick up goods. enterprises is still limited
• Lack of key logistics center Most websites of logistics companies simply introduce
their services, lacking the utilities that customers require such
Located in a geographically advantageous place with a as track and trace tools, shipping schedules, e-booking,
lengthy and high-quality waterway network capable of tracking of documents, particularly visibility (the ability to see
exploiting and conveying two artery waterways for ships from and control orders). The application of modern IT in
Ho Chi Minh City to many Mekong Delta provinces. businesses, such as warehouse management software (WMS),
Nevertheless, the growth of logistics systems in the Mekong radio identification technology (RFID), or logistics cloud, is
444
also quite limited. The reason, in this case, is a lot of logistics In the Mekong Delta, there are only 6 universities that offer
enterprises in the Mekong Delta primarily provide basic advanced education in Logistics and Supply Chain. This
logistics services. amount is insufficient in comparison to societal needs.
Furthermore, most logistics enterprises in the Mekong IV. CONCLUSION
Delta have difficulty accessing specialized management
software such as warehouse management software (WMS) Industry 4.0 brings significant benefits as well as
and transportation system management software (TMS), challenges to logistics service providers in Vietnam in general,
because they frequently have to buy those products from and in the Mekong Delta in particular. With an abundant
foreign software developers. The highlight barriers are the source of raw materials and a large output, the Mekong Delta
cost of buying from overseas, the highly problematic is the country's largest production center for food, aquatic
installation and commissioning process, and the connection products, and tropical fruit trees. As a result, Mekong Delta's
standards are not synchronized internally and with customers. logistics development is vital to the socio-economic
Therefore, logistics businesses frequently prefer to use development of the Mekong Delta, as well as the economic
domestic transport management systems, vehicle tracking, development of the Mekong Delta and the overall
and traditional forwarding service management tools development of Vietnam's logistics industry. Thus, forming
developed by domestic suppliers. However, this rate is less the logistics industry strategy in an organized manner
than 10% of all businesses. A majority of businesses remain following established requirements of the era is not only the
dependent on Microsoft Excel software for management responsibility of enterprises but also demands significant
because of no adequate system software for small and support from the Government and authorities. Innovating
medium-sized businesses. business processes to assess the benefits of Industry 4.0, as
well as planning a cohesive infrastructure network in
• A lack of highly qualified logistics specialists management are the fundamental solutions to develop
potential business industries such as logistics in The Mekong
Delta in the Industry 4.0 era.
REFERENCES
[1] C. J. Bartodziej, "The concept industry 4.0", in The concept

industry 4.0, Springer, 2017, pp. 27-50.
[2] T. M. Tâm, "Gaining knowledge about the Fourth Industrial
Revolution", Journal of Van Lang University, vol. 6, pp. 34-
40, 2017.
[3] D.T.Phuong, "Logistics in Vietnam in industrial revolution
4.0: Opportunities and challenges", in Conference of
Transportation Science and Technology, Hanoi, 2018.
[4] L. Barreto, A. Amaral, T. Pereira, "Industry 4.0 implication in
logistics: An Overview", Procedia Manufacturing, vol. 13,
pp. 1245-1252, 2017.
[5] H. Kagermann, W. Wahlster, and J. Helbig,
Figure 4. Skilled logistics human resources availability
"Recommendations for implementing the strategic initiative
in Vietnam – 2020 Industrie 4.0: Final report of the Industrie 4.0 working group",
(Source: [17])
Forschungsunion, Berlin, Germany, 2013.
This is not only the circumstance in the Mekong Delta [6] D. Horenberg, Applications within Logistics 4.0: A research
region but also in the country generally. According to statistics conducted on the visions of 3PL service providers, University
from the Vietnam Logistics Business Association (VLA), of Twente, 2017.
Vietnam's logistics industry has lacked approximately [7] L. D. Galindo, The Challenges of logistics 4.0 for Supply
300,000 high-quality workers with professional qualifications chain management and Information Technology, NTNU,
and skills between 2021 to 2025. According to a VLA survey, 2016.
the annual growth rate of human resources in logistics
[8] K. Wang, "Logistics 4.0 Solution - New Challenges and
enterprises in Vietnam is estimated to be around 7.5 percent Opportunities", in the 6th International Workshop of
in the upcoming years. According to Figure 4, approximately Advanced Manufacturing and Automation (IWAMA),
60-80 percent of enterprises interviewed reported an Manchester, England, 2016.
inadequate overall average of skilled workforce availability in
[9] A. S. Kuhlmann and M. Klumpp, "Digitalization of logistics
Vietnamese enterprises, ranging from direct workers to
processes and human perspective", in Proceeding of the
administrators. This situation demonstrates that businesses Hamburg International Conference of Logistics (HICL),
and colleges have not made a strong bond and effective link. Hamburg, Germany, 2017.
According to a report from the Vietnam Ministry of [10] M. Attaran, "Additive manufacturing: the most promising
Industry and Trade [18], there are currently about 49 technology to alter the supply chain and logistics", Journal of
universities, 56 colleges, and 11 intermediate schools, which Service Science and Management, vol. 10, pp. 189-205, 2017.
provide specialized logistics training, and each year provide [11] T. P. Minister, Decision 1012/QD-TTg: On approving the
approximately 10,560 employees with over 300,000 master plan on development of logistics center system in the
recruitment needs. If higher education graduates are counted whole country to 2020, with orientation to 2030, Hanoi,
in, the whole country would only meet 3% of labor demand. Vietnam, 2019.
445
[12] T. P. Minister, "Official Dispatch no. 7709/VPCP-CN", Products and Aquatic Products of the Mekong Delta, Can
Government Office, 15 September 2020. [Online]. Available: Tho, Vietnam, 2019.
https://thuvienphapluat.vn/cong-van/Thuong-mai/Cong-van- [16] L. Vu, "Dien Dan Doanh Nghiep", 26 05 2022. [Online].
7709-VPCP-CN-2020-Hang-nong-san-Viet-Nam-kho-vuon- Available: https://diendandoanhnghiep.vn/logistics-cho-
ra-the-gioi-vi-chi-phi-logistics-452998.aspx. [Accessed 28 03 nong-san-dbscl-can-co-che-dot-pha-thu-hut-dau-tu-phat-
2022]. trien-logistics-223807.html. [Accessed 25 06 2022].
[13] Ministry of Transport, World Bank, "The Project of [17] VIRAC, "The current status of human resources in Logistics
Feasibility Study for Corridor Development to Southern industry: Lack of high quality manpower, human capital on
Region Waterways and Logistics", Vietnam, 2017. course to decline post-Covid-19", VIRAC, 2020.
[14] N. P. Lam, V. T. T. Anh, et al., "The Annual Economic Report [18] Ministry of Industry and Trade, "Vietnam Logistics Report
of the Mekong Delta in 2020", VCCI & FSPPM, Can Tho, 2021 - Developing Logistics Human Resource", Industry and
Vietnam, 2021. Trade Publishing, Ha Noi, 2021.
[15] H. T. T. Hoa, B. T. B. Lien, T. Q. Dao, T. D. Khiem, "Enhance
the interconnection of logistics services infrastructure for
agricultural and fishery products from the Mekong Delta -
From mindsets to actions", The Conference of Increasing
Connection for the Value Chain Improvement in Agricultural
446
Enhancement in Dielectric Constant of Poly Vinyl

Alcohol by Loading of Strontium Titanate for
Supercapacitor Applications
Anju Yadav Dinesh Kumar Yadav Khushbu Meena
Department of Physics Department of Physics Department of Physics
University of Rajasthan University of Rajasthan University of Rajasthan
Jaipur, India Jaipur, India Jaipur, India
Kiran Devat Narendra Jakhar Rajesh Sahu

Department of Physics Department of Physics Department of Physics,
University of Rajasthan University of Rajasthan School of Basic Sciences
Jaipur, India Jaipur, India Manipal University Jaipur
Jaipur, India
S.K. Jain Balram Tripathi

Department of Physics, School of Basic Sciences Department of Physics
Manipal University Jaipur S S Jain Subodh PG (Auto.) College
Jaipur, India Jaipur, India
sushilkumar.jain@jaipur.manipal.edu balramtripathi1181@gmail.com
Abstract: The polymer nanocomposites have extensive Ue = ½ ϵo ϵr E2bd. Where ϵ is dielectric permittivity and Ebd
applications in electrical and electronic industry owing to is breakdown strength of material. So in view of increasing
their amazing dielectric properties. In the present work, demand of energy storage, materials with high dielectric
strontium titanate (ST) dispersed polyvinyl alcohol (PVA) constant along with high breakdown strength are desired to
nanocomposites (100 μm) have been synthesized using store more energy. The traditional ceramic materials such as
solution cast method. The concentration of filler ST varied BaTiO3 (1700), SiO2 (3.9), Al2O3 (9), SrTiO3 (300),
from 1 to 5 mwt% to improve and optimize the dielectric PbZrTiO3 (1250) etc have been widely used as dielectric
properties of PVA for energy storage applications. X-ray materials owing to their high dielectric constant [5].
diffraction (XRD), Fourier transform infrared (FT-IR) Although dielectric permittivity of ceramics is high but the
spectroscopy, Scanning electron microscopy (SEM) and ceramic capacitors have low power density due to less
Impedance dielectric spectroscopy were employed to study breakdown field strength of ceramics. Furthermore, it is
the structural, bonding, surface morphology and dielectric difficult to manufacture the very thin film of ceramic
constant of all synthesized materials. SEM images show that materials. Whereas, polymer materials have high dielectric
ST/PVA composites make a porous structure. The dielectric strength, flexibility, low cost and easily processed into thin
films with large area, but their low dielectric constants result
constant of composite ST5%PVA95% has been found to be
into low energy density. In order to upgrade dielectric
increased 2.6 times to that of pristine PVA at 1 kHz
properties of polymers, the polymer nanocomposites are
frequency. Such ST/PVA composites having high dielectric developed [6-8]. In this work, ceramic ST is selected as
constant with porous structure are suitable for the electrode filler due to its amazing dielectric properties. It exhibits
material of supercapacitors to store more energy. large values of dielectric constant which is desired for high
energy density applications. Polymer PVA is used as matrix
Keywords: dilectric constant, polyvinyl alcohol, due to its flexibility, low cost, non-toxic nature, high
Polymers. dielectric strength and excellent film-forming capability [9-
10]. Polymer nanocomposite films of ST/PVA are
I. INTRODUCTION
synthesized to improve and optimize the dielectric
Electrical energy storage system plays an important role properties of PVA for energy storage applications. The
in hybrid electric vehicles, standby power systems, cell concentration of filler ST was kept very low (1 to 5 mwt%).
phones and pulse power applications in form of batteries, Although a high concentration of ceramic desired to
fuel cells and supercapacitors [1-2]. Supercapacitors store increase the dielectric constant of polymer composite, but
energy in the form of charge on the electrode surface, since flexibility and processibility decreases and the dielectric
in the supercapacitors the charging – discharging process losses increases with high concentration of ceramic [11-12].
happens only on the electrode surface instead of bulk
material as in batteries, so they possess comparatively high
power density due to release of energy at high rates.
Volumetric energy density stored by a capacitor is [3-4] –

II. EXPERIMENTAL CHARACTERIZATION corresponding to crystal plane (101), reflecting its semi
crystalline nature [13-14]. The hydrogen bond present
A. Materials between PVA chains is responsible for its semi crystalline
PVA powder (99% pure) with density 1.19 g/cm3 was structure [15-16]. The peak becomes more broadened with
bought from Akshar Chem, India. ST powder with more the concentration of filler pointing to the presence of
than 99.5% purity purchased from Sigma-Aldrich, China. amorphous phase of PVA [17]. The diffraction peaks
The solvent Dimethyl formamide with more than 99% appeared in the XRD pattern of ST are indexed to the
purity was acquired from Shri Maruti Chem Enterprises perovskite type cubic structure, in agreement with JCPDS
Private Limited, Mumbai of brand Emplura, Germany. Card no.…... The most intense peak appears at 2Ɵ = 32.52
corresponding to the plane (110). No any impurity peak
B. Synthesis of Samples
appears in the XRD patterns of ST and ST/PVA composites
Sample films of pristine PVA and ST/PVA composites which reveal presence of single phase of samples.
were synthesized via solution casting method. Polymer Moreover, the intensity of diffraction peaks in the
polyvinyl alcohol was dissolved in the solvent dimethyl composites ST/PVA increases with the concentration of ST
formamide using magnetic stirier at 420 rpm. At the same reflecting well dispersion of ST particles in matrix PVA.
time filler ST in DMF was sonicated at 200 V. Both the The lattice parameters calculated are a = b = c = 0.389 nm
solutions were added and further sonicated to disperse the which are as per cubic structure of strontium titanate [18].
filler ST in matrix PVA homogeneously. This solution was Average crystallite size of ST in composites ST/PVA was
poured into a balanced pert dish and dried at room found to be 41.6 nm.
temperature to make films with uniform thickness.
E. FT-IR spectroscopy
C. Characterization
To determine crystallite size and crystal structure, X-ray 0.999
diffraction patterns of pristine PVA and ST/PVA 0.962
0.925
composites were recorded using Cu Kα radiation of Bruker 0.888

1.026 1000 2000 3000
ST5%PVA95%
4000
D8 ADVANCE X-ray diffractometer. Scans were taken in 0.972
the range 2Ɵ = 15o to 70o with step resolution 0.02o. To 0.918
0.864 ST4%PVA96%
evaluate the crystallite size origin software and Scherer’s 1000 2000 3000 4000
Transmittance (%)
0.968
equation were used. To identify chemical bond present in 0.880
0.792
the composites FT-IR spectroscopy was carried out using 0.704

1.012 1000 2000 3000
ST3%PVA97%
4000
Alpha Bruker FT-IR Spectrometer (ATR-ECO mode) in 0.968
wave number range 4000 – 400 cm-1. Scanning electron

0.924
0.880 ST2%PVA98%
microscopy was employed to study the surface morphology 0.999

1000 2000 3000 4000
of the polymer nano composites using scanning electron 0.972
0.945
microscope JSM-7610F Plus, JEOL Ltd Tokyo, Japan at 5.0 0.918

1000 C-H wagging 2000
ST1%PVA99%
3000 O-H stretching 4000

1.000
kV. Dielectric measurements were taken by impedance 0.975
analyzer WAYNE KERR Electronics (65ooB series), 0.950
0.925 C-H rocking mode

C-H bending
C=O stretching
C-C stretching vibration

C-O stretching vibration
C-H stretching
PVA
London, UK in the frequency range 1 kHz to 1MHz. 1000 2000 3000 4000
Wave number (cm-1)

D. X-ray diffraction Figure 2. FTIR spectra of pristine PVA and all synthesized
ST/PVA composites
(110)
ST5%PVA95%
(200)
Fig. 2 displays FTIR spectra of pristine PVA and

(111)
(211)
(220)
ST/PVA composites at different concentration of filler ST.

Various absorption peaks appear in the wave number range
(110)
ST1%PVA99%
(4000-400 cm-1) investigated. A wide absorption peak at
(111)
(200)
Intensity (a.u.)
(211)
3274 cm-1 is corresponding to the O-H stretching mode of

(220)
vibrations [19]. Absorption band at 2921cm-1 is related to C-

(110)
ST H stretching vibrations. The peak appeared at 1716 cm-1 is

(200)
assigned to C=O stretching vibration of acetate group [20-

(211)
(111)
(220)
(100)
21]. The peak corresponding to C-H bonding mode

vibration appears at 1424 cm-1 [22]. Absorption peaks at
1373 cm-1,1243 cm-1,1085 cm-1,836 cm-1 are found to be
Pristine PVA consistent with C-H wagging, C-C stretching, C-O
stretching and C-H rocking mode vibrations respectively
[23-24]. Broad band at 511-596 cm-1 is consistent with
15 20 25 30 35 40 45 50 55 60 65 70 stretching vibrations of Ti-O bond [25]. A slight shifting and
Diffraction angle (2Ɵ)
change in intensity of peaks in the composites reflects the
Figure 1. XRD patterns of pristine PVA, ST, ST1%PVA99% and good interaction between matrix PVA and filler ST.
ST5%PVA95% composites
F. Scanning Electron Microscopy
Fig. 1 displays XRD patterns of pristine PVA, ST and Fig. 3 shows SEM micrographs of pristine PVA and
ST1%PVA99% and ST5%PVA95% composites. XRD pattern of ST1%PVA99% and ST5%PVA95% composite films at two
pristine PVA shows a characteristic peak at 2Ɵ = 19.8 different magnifications. The SEM images of pristine PVA
448
[Fig. 3 (a) (b)] shows almost plane and smooth surface [26]. The values of ɛr increasing proportionally with
The white dots in these micrographs are due to the residual concentration of ST reflect well dispersion of filler ST in
solvent. The surface morphology of ST/PVA [Fig. 3 (c) (d) host matrix PVA. Low values of dielectric constant at higher
& (e) (f)] composite films are totally different. Porous frequency are due to dielectric relaxation of ST ceramics
structure is originated by adding filler strontium titanate in [28]. At these frequencies dipoles leg behind the applied
matrix PVA. It was also observed that more porous structure electric field hence reduction in dipolar polarization. In
emerges as the concentration of strontium titanate increases addition to this, due to the long relaxation time space charge
which is suitable to trap charge carriers and to store more polarization doesn’t happen. So both the dipolar and space-
energy density [27]. charge polarization don’t contribute at higher frequencies.
As a result of which the values of dielectric constant falls
down [29].
Table 1. Frequency vs dielectric constant of ST/PVA composites

Dielectric Constant(ɛr)
S. Frequency
No. (Hz) ST1% ST2% ST3% ST4% ST5%
PVA
PVA99% PVA98% PVA97% PVA96% PVA95%
1 1 kHz 6.43 10.89 12.20 13.87 15.49 16.99
2 10 kHz 5.68 9.56 10.36 11.73 13.19 14.87
3 100 kHz 4.91 8.19 8.85 9.97 10.93 12.94
1 MHz
Figure 3. SEM Images of pristine PVA [ (a), (b)], ST1%PVA99% 4 4.09 6.81 7.38 8.28 9.29 11.01
[(c), (d)] and ST5%PVA95%[(e), (f)]
G. Dielectric spectroscopy Enhancement in dielectric constant with ST content

Fig. 4 presents the variation of dielectric constant (ɛr) as could be explained as intrinsic dipole polarization and
a function of frequency for pristine PVA and ST/PVA interfacial polarization at the boundary of ST and PVA. In
composites with different concentration of ST. The the composite materials charge entrapped at the
dielectric response changes with frequency in similar heterojunction boundaries of ST/PVA consequently, large
dipoles were formed. So by addition of filler ST in polymer
manner for all composite materials. The value of ɛr drops
PVA, more and more polarization takes place which in turn
down continuously with increasing the frequency. For
increase the dielectric constant of ST/PVA composites [30-
pristine PVA maximum value of dielectric constant
32]. Hence, the ST/PVA polymer composites give better
observed is 6.43 at 1 kHz frequency which reaches to 4.09
results to that of pristine PVA for dielectric constant.
at 1MHz. For the composites ST 5%PVA95%, ST4%PVA96%,
ST3%PVA97%, ST2%PVA98%, and ST1%PVA99% these ɛr values IV. CONCLUSION
vary from 16.99 to 11.01,15.49 to 9.29, 13.87 to 8.28, 12.20 In summary, composite ST/PVA with different mwt% of
to 7.38, and 10.89 to 6.81 respectively as shown in Table-1. ST were synthesized via solution casting method. X-ray
Thus a remarkable increase in dielectric constant was diffraction and Fourier transform infrared spectra confirm
observed over the entire frequency range investigated. formation of ST/PVA polymer ceramic nanocomposites.
The XRD pattern of ST showed a perovskite type cubic
20 Pristine PVA structure with lattice parameter a = 0.389 nm. Composites
ST1%PVA99%
ST2%PVA98%
make a porous structure which was confirmed by surface
Dielectric Constant
ST3%PVA97% micrographs. The dielectric behavior of pristine PVA and

15 ST4%PVA96%
ST5%PVA95%
ST/PVA composites revealed that it increases with content
ST and decreases with increasing the frequency. Maximum
value of dielectric constant observed for the composition
10 ST5%PVA95% at 1 kHz. It is almost 2.64 times larger than
that of pristine PVA at 1 kHz frequency. The porous
structure and high dielectric constant of ST/PVA composites
make them a suitable material for capacitive energy storage
5
applications.
3 4 5 6 ACKNOWLEDGEMENTS
log f (Hz) The authors are greatly acknowledged to DBT New
Delhi for research grant (BT/HRD/11/23/2019) under DBT
Figure 4. Dielectric constant vs frequency curve of pristine PVA Star Scheme, Central analytical facilities, Manipal
and ST/PVA composites University Jaipur, for providing FT-IR & SEM
characterization facilities.
449
REFERENCES [18] Jarabana, Kanaka & Mishra, Ashutosh & Bisen, Supriya.
Structural and Optical properties of poly-crystalline BaTiO 3
[1] Nalwa, H. Handbook of Low and High Dielectric Constant and SrTiO 3 prepared via solid state route. Journal of Physics:
Materials and Their Applications. Academic Press: London, Conference Series 755 (2016) 012020.
UK, 1999.
[19] M.H. Makled, E. Sheha, T.S. Shanap, M.K. El-Mansy,
[2] Osaka, T., Datta, M. Energy Storage Systems for Electronics; Electrical conduction and dielectric relaxation in p-type
Gordon and Breach. Amsterdam, The Netherlands, 2001. PVA/CuI polymer composite, J. Adv. Res., 4 (2013), pp. 531-
[3] Wenjing Ji, Hua Deng, Cong Guo, Chengxiao Sun, Xuan 538.
Guo, Feng Chen, Qiang Fu. The effect of filler morphology [20] Shujahadeen B. Aziz, M.A. Brza, M.H. Hamsan, M.F.Z.
on the dielectric performance of polyvinylidene fluoride Kadir, S.K. Muzakir, Rebar T. Abdulwahid. Effect of ohmic-
(PVDF) based composites, Composites Part A 118, 2019, pp. drop on electrochemical performance of EDLC fabricated
336–343. from PVA:dextran:NH4I based polymer blend electrolytes.
[4] Dang Z.M., Yuan J.K., Yao S.H., Liao R.J. ChemInform Journal of Materials Research and Technology, Vol. 9, No. 3,
abstract: flexible nanodielectric materials with high 2020, pp. 3734-3745, ISSN 2238-7854.
permittivity for power energy storage. Adv. Mater. 2014, [21] Anju Yadav, Dinesh Kumar Yadav,Jitendra Kumar Mishra,
45(3):6334–65. Rajesh Sahu,S.K. Jain,Shalini Dixit,Garima Agarwal,
[5] Barber, P., Balasubramanian, S., Anguchamy, Y., Gong, S., Narendra Jakhar,Balram Tripathi, Effect of
Wibowo, A., Gao, H., Ploehn, H.J., Zur Loye, H.-C. Polymer BaTiO3 Nanofillers on the Energy Storage Performance of
Composite and Nanocomposite Dielectric Materials for Pulse Polymer Nanocomposites, Macromol. Symp. 2021, 399,
Power Energy Storage. Materials, 2009, 2, pp. 1697-1733. 2100024.
[6] Li, J.Y., Zhang, L., Ducharme, S. Electric energy density of [22] R. Ashiri, Detailed FT-IR spectroscopy characterization and
dielectric nanocomposites. Appl. Phys. Lett., 2007, 90, (13), thermal analysis of synthesis of barium titanate nanoscale
p. 132901. particles through a newly developed process, Vibrational
[7] Tang, H., Lin, Y., Andrews, C., et al. ‘Nanocomposites with Spectroscopy, Vol. 66, 2013, pp. 24-29.
increased energy density through high aspect ratio PZT [23] Tang, C.-M., Tian, Y.-H., & Hsu, S.-H. (2015). Poly(vinyl
nanowires’, Nanotechnology, 2011, 22,(1), p. 015702. alcohol) Nanocomposites Reinforced with Bamboo Charcoal
[8] Yu-Chen Lu, Shuhui Yu, Xierong Zeng, Rong Sun, Ching- Nanoparticles: Mineralization Behavior and Characterization.
Ping Wong, High energy density polymer nanocomposites Materials, 8(8), pp. 4895–4911,
with Y-doped barium strontium titanate nanoparticles as http://dx.doi.org/10.3390/ma8084895.
fillers, IET Nanodielectr., 2018, Vol. 1, No. 4, pp. 137-142. [24] Jipa, I., Stoica, A., Stroescu, M., Dobre, L.-M., Dobre, T.,
[9] Shujahadeen B. Aziz, Rebar T. Abdulwahid, Mariwan A. Jinga, S., & Tardei, C. (2012). Potassium sorbate release from
Rasheed, Omed Gh. Abdullah and Hameed M. Ahmed, poly(vinyl alcohol)-bacterial cellulose films. Chemical
Polymer Blending as a Novel Approach for Tuning the SPR Papers, 66(2), http://dx.doi.org/10.2478/s11696-011-0068-4.
Peaks of Silver Nanoparticles, Polymers 2017, 9, p. 486. [25] Pratap, Arun & Joshi, Nirav & Rakshit, Priya & Grewal, G. &
[10] Srivastava, Nisha & Patodia, Tarun & Yadav, Anju & Jakhar, Shrinet, Vagish. (2013). Dielectric Behavior Of Nano Barium
Narendra & Sharma, K.B. & Tripathi, Balram. 2020. Titanate Filled Polymeric Composites. International Journal
Graphene derivative based high dielectric constant polymer of Modern Physics Conference Series. 22.
composite electrodes for supercapacitors. Materials Today: 10.1142/S2010194513009859
Proceedings. 42. [26] Ma, Pibo, Wang, Xin, Xu, Weilin, Cao, Genyang. (2009).
[11] Yan Wang, Minggang Yao, Rong Ma, Qibin Yuan, Desuo Application of Corona Discharge on Desizing of Polyvinyl
Yang,Bin Cui, Chunrui Ma, Ming Liu, Dengwei Hu, Design Alcohol on Cotton Fabrics. Journal of Applied Polymer
strategy of barium titanate/polyvinylidene fluoride-based Science - J Appl Polym Sci. 114, pp. 2887-2892.
nanocomposite films for high energy storage, J. Mater. Chem. [27] T. Badapanda, V. Senthil, S. Anwar, L.S. Cavalcante, N.C.
A, 2020, 8, pp. 884-917. Batista, E. Longo. Structural and dielectric properties of
[12] Rajesh Sahu, Tarun Patodia, Dinesh Yadav, S.K. Jain, Balram polyvinyl alcohol/barium zirconium titanate polymer–ceramic
Tripathi, Visible-light induced photo catalytic response of composite. Current Applied Physics, Vol. 13, No. 7, 2013, pp.
MWCNTs-CdS composites via efficient interfacial charge 1490-1495.
transfer, Materials Letters: X, Vol. 13, 2022, 100116, [28] Dinesh Kumar Yadav, Anju Yadav, Khushbu Meena, Kiran
[13] Aziz, S.B., Abdullah, O.G., Hussein, S.A., Ahmed, H.M. Devat, Jitendra K. Mishra, Rajesh Sahu, Sushil K. Jain,
Effect of PVA Blending on Structural and Ion Transport Ambesh Dixit, Nisha Srivastava, Tarun Patodia, Narendra
Properties of CS:AgNt-Based Polymer Electrolyte Jakhar, and Balram Tripathi. Study of CNT Intercalated
Membrane. Polymers 2017, 9, p. 622. Bi2O3/PVDF Composite for Super Capacitors Applications.
Macromol. Symp. 2021, 399, 2100022.
[14] Aziz, S.B. Modifying poly(vinyl alcohol) (PVA) from
insulator to small-bandgap polymer: A novel approach for [29] J. E. Garcia, R. Pérez and A. Albareda. High Electric Field
organic solar cells and optoelectronic devices. J. Electron. Measurement of Dielectric Constant and Losses of
Mater. 2016, 45, pp. 736–745. Ferroelectric Ceramics. Journal of Physics D: Applied
Physics, Vol. 34, No. 22, 2001, pp. 3279-3284.
[15] M. Hema, S. Selvasekerapandian, G. Hirankumar, A.
Sakunthala, D. Arunkumar, H. Nithya. Structural and thermal [30] Vo, Hung T., and Frank G. Shi. Towards model-based
studies of PVA:NH4I. Journal of Physics and Chemistry of engineering of optoelectronic packaging materials: dielectric
Solids, Vol. 70, No. 7, 2009, pp. 1098-1103, ISSN 0022- constant modeling. Microelectronics Journal 33.5-6 (2002),
3697. pp. 409-415.
[16] Nangia, Rakhi, Shukla, Neeraj, Sharma, Ambika. Frequency [31] G.C Psarras, E Manolakaki, G.M Tsangaris, Electrical
and temperature-dependent impedance spectroscopy of relaxations in polymeric particulate composites of epoxy resin
PVA/PEG polymer blend film, High Performance Polymers, and metal particles, Composites Part A: Applied Science and
1–9, 2019. Manufacturing, Vol. 33, No. 3, 2002, pp. 375-384.
[17] Aziz, S.B., Kadir, M.F.Z., Hamsan, M.H. et al. Development [32] Chakraborty, G. & Meikap, A. & Babu, Ramesh & Blau,
of Polymer Blends Based on PVA:POZ with Low Dielectric Werner. (2011). Activation behavior and dielectric relaxation
Constant for Microelectronic Applications. Sci Rep 9, 13163 in polyvinyl alcohol and multiwall carbon nanotube
2019. composite films. Solid State Communications, 151, pp. 754-
758.
450
Free Vibration Analysis of Smart FG Porous Plates

Reinforced by Graphene Platelets Using
Isogeometric Approach
Lieu B. Nguyen Duc-Thien Pham
Faculty of Civil Engineering Faculty of Civil Engineering
Education Education
lieuntb@hcmute.edu.vn thienpd@hcmute.edu.vn
Abstract: In this paper, an effective computational reinforcing even a minute amount of graphene to matrix
formulation based on isogeometric method to analyze free material can significantly improve its properties as thermal
vibration of functionally graded (FG) porous plates properties, high stiffness, high strength but low mass density
reinforced by graphene platelets (GPLs) bonded with and electrical too [10]. King et al. [12] realized that Young’s
piezoelectric layers is investigated for the first time. It is modulus of epoxy reinforced nanocomposites rises
called as SFGP-GPLs for short. The plate like as a sandwich approximately 0.64 GPa by enhancing 6.0 wt% of GPLs as the
structure is created by a core layer, which has many porosities reinforcement in the composite plates. So, in recent years
and is reinforced by GPLs, and two piezoelectric layers materials with many internal empty holes hardened by GPLs
have received much attention from scientists for their
perfectly glued on the upper and lower surfaces of the host
outstanding properties [13-15]. For both GPLs and porosities,
plate. The effective mechanical properties of core layer are
the mechanical properties of the plate are drastically
estimated by the modified Halpin–Tsai micromechanical appreciated but still retain their advantages for lightweight
model. Besides, law of electric potential varied linearly along structures [16]. There are many studies being shown to study
the thickness is considered for the piezoelectric layers. A the impacts of GPLs and porosities on the performances of
generalized higher-order shear deformation theory (GHSDT) structures under different conditions. Using the first-order
combined with isogeometric analysis (IGA) is proposed for shear deformation theory (FSDT) and Chebyshev-Ritz
solving the problems. The influence of parameters including method, the buckling and free vibration responses of
the coefficient and distribution types of porosity, weight functionally graded (FG) porous plates reinforced by
fractions and dispersion patterns of GPLs as well as input graphene platelets, FGP-GPLs, were also investigated by
electrical voltages on natural frequencies also is presented Yang et al. [17]. Additionally, alumina ceramic composites
and discussed through various numerical examples. reinforced with GPLs are investigated by Liu et al. [18]. In
order to study the static, free vibration and buckling of FGP-
Keywords: piezoelectric materials, FG-porous plate, GPLs, Li et al. [19] used IGA with both FSDT and TSDT.
graphene platelet reinforcements, isogeometric analysis, free Free and forced vibrations of functionally graded polymer
vibration composite plates reinforced with graphene nanoplatelets are
also examined by Song et al. [20].
I. INTRODUCTION
In this paper, the functionally graded porous plate
Smart or intelligent structures, especially piezoelectric reinforced by GPLs attached to two piezoelectric layers is
elements, are nowadays extensively applied in a multiplicity considered. Mechanical and electrical properties are radically
of engineering purposes such as vibration control, structural studied. Three dispersion patterns of weight fractions of GPLs
health monitoring, shape control, energy harvesting and are considered through the thickness direction whereas three
microelectro mechanical systems [1]. Although piezoelectric types of porosity distribution in the core layer are assumed.
materials have been introduced for more than a century, they Piezoelectric materials are assumed to be absolutely bonded
also have an importance over the last few decades due to their to the top and the bottom of host plates. To estimate the
high potential in engineering fields. For the plate combined efficient material properties, modified Halpin-Tsai
with piezoelectric material, there are different mathematical micromechanics [21] is used. An isogeometric finite element
methods being presented to clarify their behaviors. formulation based on the generalized higher-order shear
Porous materials have been used in various applications deformation theory (GHSDT) with five variations per one
due to outstanding properties such as lightweight, excellent control point is performed. Some related works are presented
energy absorption and heat resistance [2-5]. However, the in [22-24] for analysing the layered FG-GPLs (no porosities)
presence of inner pores leads to a noteworthy decrease the composite laminated plates with piezoelectric layers.
stiffness of structures [6]. In order to get the better of this However, the topic of free vibration analysis of SFGP-GPLs
weakness, carbon nanotubes (CNTs) [7-9] and graphene is still very limited. Furthermore, the effects of various
platelets (GPLs) [10-11] often are chosen to reinforce for the parameters like distribution of pattern, porosity coefficient,
composite materials. Though, in comparison between carbon different porosity distribution types, as well as various
nanofillers and graphene platelets researchers stated that boundary conditions and external electric voltages are
considered in some numerical examples.

II. MATERIAL PROPERTIES OF A SFGP-GPLS dispersion patterns are represented by 𝑗 = 1,2,3 ,

PLATE respectively.
A three-lamina sandwich plate has length a, width b and
total thickness of ℎ = ℎ𝑐 + 2ℎ𝑝 (Fig. 1). The thicknesses of
core layer and the piezoelectric face layers are ℎ𝑐 and ℎ𝑝 ,
respectively.
There are three types of porosity distribution versus the
thickness direction of the core layer. They are the non‐uniform
symmetric 1,2 (NUPD1, NUPD2) and a uniform porosity
distribution (UPD) shown in Fig. 2. It can be explained that
for UPD 𝐸’ is Young’s modulus, 𝐸1′ and 𝐸2′ represent the
maximum and minimum Young’s moduli of the NUPD
without GPLs, respectively. Moreover, three GPL dispersion
patterns shown in Fig. 2 are presented for each porosity
distribution. In one pattern, the GPL volume fraction VGPL is
seen to vary smoothly along the thickness direction. The
symbol 𝑆ij are the maximum volume fraction, where 𝑖 =
1,2,3 are three porosity distribution types and three Figure 1. Configuration of a PFGP- GPLs plate
z z z z
h ' h h h
E 1
2 2 2 2
x VGPL VGPL VGPL
E2' S11 S12 S13
h h h h
− E1' − − −
2 2 2 2
Pattern A Pattern B Pattern C
(a) NUPD 1
z z z z
h ' h h h
E
1
2 2 2 2
x VGPL VGPL VGPL
S 21 S 22 S 23
h h h h
− E2' − − −
2 2 2 2
(b) NUPD 2
z z z z
h h h h
E'
2 2 2 2
x VGPL VGPL VGPL
S31 S32 S33
h h h h
− − − −
2 2 2 2
(c) UPD
Figure 2. Three types of porosity distributions and three types of GPL distribution
The material properties for different porosity distribution where

types can be described as equation (1), in which 𝐸(𝑧), 𝐺(𝑧)
and 𝜌(𝑧) are Young’s moduli, shear modulus and mass 𝑐𝑜𝑠( 𝜋𝑧/ℎ𝑐 ), (𝑁𝑈𝑃𝐷1)
density, respectively. 𝜆(𝑧) = { 𝑐𝑜𝑠( 𝜋𝑧/2ℎ𝑐 + 𝜋/4), (𝑁𝑈𝑃𝐷2) (2)
𝜆, (𝑈𝑃𝐷)
𝐸(𝑧) = 𝐸1 [1 − 𝑒0 𝜆(𝑧)],
{𝐺(𝑧) = 𝐸(𝑧)/[2(1 + 𝑣(𝑧))], (1)
𝜌(𝑧) = 𝜌1 [1 − 𝑒𝑚 𝜆(𝑧)],
452
The porosity coefficient 𝑒0 can be expressed by 2𝑙𝐺𝑃𝐿 2𝑤𝐺𝑃𝐿

𝜁𝐿 = , 𝜁𝑊 = ,
𝑡𝐺𝑃𝐿 𝑡𝐺𝑃𝐿
𝑒0 = 1 − 𝐸2′ /𝐸1′ (3) 𝜂𝐿 =
(𝐸𝐺𝑃𝐿 /𝐸𝑚 )−1
, 𝜂𝑊 =
(𝐸𝐺𝑃𝐿 /𝐸𝑚 )−1
, (13)
(𝐸𝐺𝑃𝐿 /𝐸𝑚)+𝜁𝐿 (𝐸𝐺𝑃𝐿 /𝐸𝑚 )+𝜁𝑤
where 𝐸1 = 𝐸1′
is used for non‐uniform porosity
distribution and 𝐸1 = 𝐸 ′ for uniform porosity distribution. where 𝑤𝐺𝑃𝐿 , 𝑙𝐺𝑃𝐿 and 𝑡𝐺𝑃𝐿 mean the average width, length
and thickness of GPLs, respectively; 𝐸𝐺𝑃𝐿 and 𝐸𝑚 symbolize
Using Gaussian Random Field (GRF) scheme [25], Young’s moduli of GPLs and metal matrix, respectively.
equation described mechanical characteristic of closed‐ cell
cellular solids is shown as The mass density 𝜌1 and Poison’s ratio 𝜈1 of the GPLs
reinforced for porous metal matrix can be determined
𝐸(𝑧) 𝜌(𝑧)/𝜌1 +0.121 2.3 according to the rule of mixture as
=( ) (4)
𝐸1 1.121
𝜌(𝑧) 𝜌1 = 𝜌𝐺𝑃𝐿 𝑉𝐺𝑃𝐿 + 𝜌𝑚 𝑉𝑚 , (14)

for (0.15 < < 1)
𝜌1 𝜈1 = 𝜈𝐺𝑃𝐿 𝑉𝐺𝑃𝐿 + 𝜈𝑚 𝑉𝑚 (15)
where 𝜌1 is the maximum value of mass density where 𝜌𝐺𝑃𝐿 , 𝜈𝐺𝑃𝐿 and 𝑉𝐺𝑃𝐿 imply the mass density,
Then, the coefficient of mass density 𝑒𝑚 in Eq. (1) is Poisson’s ratio and volume fraction of GPLs, respectively;
possibly defined as while 𝜌𝑚 , 𝜈𝑚 and 𝑉𝑚 = 1 − 𝑉𝐺𝑃𝐿 indicate the mass density,
Poisson’s ratio and volume fraction of metal matrix,
2.3
1.121(1− √1−𝑒0 𝜆(𝑧))
respectively.
𝑒𝑚 = (5)
𝜆(𝑧) III. THEORY AND FORMULATION OF SFGP-GPLS
Also according to the closed‐cell GRF scheme [25], PLATE
Poisson’s ratio 𝜈(𝑧) is derived as
A. The generalized higher-order shear deformation theory
𝑣(𝑧) = 0.221𝑝′ + 𝑣1 (0.342𝑝′2 − 1.21𝑝′ + 1), (6) (GHSDT)
where 𝜈1 is the Poisson’s ratio of the metal matrix without According to the generalized higher order shear
internal pores and 𝑝′ is defined as deformation theory [26], the displacement field of a point in
the plate is formulated as follows:
𝑝′ = 1.121(1 − 2.3√1 − 𝑒0 𝜆(𝑧)) (7)
The mass per unit of surface 𝑀 is given to be equivalent 𝒖̄ (𝒙, 𝒚, 𝒛) = 𝒖𝟏 (𝒙, 𝒚) + 𝒛𝒖𝟐 (𝒙, 𝒚) + 𝒇(𝒛)𝒖𝟑 (𝒙, 𝒚) (16)
and is formulated by
ℎ /2 where
𝑀 = ∫−ℎ𝑐 𝜌(𝑧)𝑑𝑧 (8)
𝑐 /2 𝑤,𝑥
𝑢̄ 𝑢 𝛽𝑥
Then, for uniform porosity distribution 𝜆 in Eq. (1) can be 𝒖̄ = { 𝑣̄ } ; 𝐮1 = { 𝑣 } ; 𝐮2 = − {𝑤,𝑦 } ; 𝐮3 = {𝛽𝑦 } (17)
defined as (𝜆 is the coefficient of porosity) 𝑤̄ 𝑤 0 0
1 1 𝑀/𝜌1 ℎ + 0.121 2.3 where 𝑢, 𝑣, 𝑤, 𝛽𝑥 and 𝛽𝑦 are the in-plane, transverse and
𝜆= − ( ) (9)
𝑒0 𝑒0 0.121 rotation displacements, respectively; 𝑓(𝑧) is a continuous
The volume fraction of GPLs for three dispersion patterns 4𝑧 3
function and 𝑓(𝑧) = 𝑧 − is chosen in this study.
shown in Fig.2 can be assumed as 3ℎ2
𝜋𝑧 Strain components can be achieved as follows:

𝑆𝑖1 [1 − cos ( )] , Pattern A
ℎ𝑐
𝜺 = {𝜀𝑥𝑥 𝜀𝑦𝑦 𝛾𝑥𝑦 }𝑇 = 𝜺𝟏 + 𝑧𝜺𝟐 + 𝑓(𝑧)𝜺𝟑 and
𝑉𝐺𝑃𝐿 = 𝑆 [1 − cos ( 𝜋𝑧 𝜋
+ )] , Pattern B (18)
𝑖2 (10)
2ℎ𝑐 4 𝜸 = {𝛾𝑥𝑧 𝛾𝑦𝑧 }𝑇 = 𝑓 ′ (𝑧)𝜺𝒔
{𝑆𝑖3 , Pattern C
where
The volume fraction 𝑉𝐺𝑃𝐿 and weight fractions 𝛬𝐺𝑃𝐿 is set 𝑢,𝑥 𝑤,𝑥𝑥
by 𝟏 𝑣 𝟐 𝑤
𝜺 ={ ,𝑦 }; 𝜺 = − { ,𝑦𝑦 };
ℎ𝑐
𝛬𝐺𝑃𝐿 𝜌𝑚 𝑢,𝑦 + 𝑣,𝑥 2𝑤,𝑥𝑦
𝛬𝐺𝑃𝐿 𝜌𝑚 +𝜌𝐺𝑃𝐿 −𝛬𝐺𝑃𝐿 𝜌𝐺𝑃𝐿 −
∫ 2ℎ𝑐[1 − 𝑒𝑚 𝜆(𝑧)] 𝑑𝑧 = (19)
ℎ𝑐
2
(11) 𝛽𝑥,𝑥
𝟑
𝛽𝑥
∫ 2
𝑉𝐺𝑃𝐿 [1 − 𝑒𝑚 𝜆(𝑧)] 𝑑𝑧. 𝜺 ={ 𝛽𝑦,𝑦 }; 𝜺𝒔 = { }
ℎ
− 𝑐 𝛽𝑦
2 𝛽𝑥,𝑦 + 𝛽𝑦,𝑥
Using the Halpin‐Tsai model [21], Young’s modulus 𝐸1 is
defined as in which 𝑓 ′ (𝑧) is the derivation of the function 𝑓(𝑧)
′ 𝑑𝑓
(𝑓 (𝑧) = ).
3 1+𝜁𝐿 𝜂𝐿𝑉𝐺𝑃𝐿 5 1+𝜁𝑤 𝜂𝑤 𝑉𝐺𝑃𝐿 𝑑𝑧
𝐸1 = ( ) 𝐸𝑚 + ( ) 𝐸𝑚 , (12)
8 1−𝜂𝐿 𝑉𝐺𝑃𝐿 8 1−𝜂𝑤 𝑉𝐺𝑃𝐿 B. Garlerkin weak forms of SFGP-GPL plates
in which The piezoelectric constitutive equations are linear and can
be stated as follow [27]
453
𝝈 𝒄 −𝒆𝑇 𝜺̄ The electric potential adaptation is supposed to be linear

[ ]=[ ][ ] (20) in each sublayer and is approximated through the thickness of
𝑫 𝒆 𝒈 𝑬
piezoelectric layer, please see [22] for more details.
where 𝜀̄ and 𝝈 demote the strain vector and the stress
vector, respectively; 𝒄 is the elastic constant matrix. D. Governing equations of motion
The elementary governing equation of motion for
𝑨 𝑩 𝑳 𝟎
𝑩 𝑮 𝑭 𝟎 vibration analyzing problem can be performed in the
𝑐=[ ] (21) following form
𝑳 𝑭 𝑯 𝟎
𝟎 𝟎 𝟎 𝑫𝑺 𝐌 𝟎 𝐝̈ 𝐊 𝑢𝑢 𝐊 𝑢𝜙 𝐝 𝟎
where [ 𝒖𝒖 ][ ] + [ ][ ] = [ ] (24)
𝟎 𝟎 ϕ̈ 𝐊 𝜙𝑢 −𝐊 𝜙𝜙 𝛟 𝟎
where
(𝐴𝑖𝑗 , 𝐵𝑖𝑗 , 𝐺𝑖𝑗 , 𝐿𝑖𝑗 , 𝐹𝑖𝑗 , 𝐻𝑖𝑗 ) =
ℎ/2 𝐊 𝑢𝑢 = ∫ 𝐁𝑢𝑇 𝐜𝐁𝒖 dΩ ; 𝐊 𝑢𝜙 = ∫ 𝐁𝒖𝑻 𝐞̃𝑇 𝐁𝜙 dΩ

∫ 2
(1, 𝑧, 𝑧 , 𝑓(𝑧), 𝑧𝑓(𝑧), 𝑓 ̄𝑏
2 (𝑧))𝑄
𝑑𝑧, Ω Ω
(25)
𝑖𝑗
−ℎ/2 𝐊 𝜙𝜙 = ∫ 𝐁𝜙𝑇 𝐠𝐁𝜙 dΩ ; 𝐌𝑢𝑢 ̃ 𝑇 𝐦𝐍
= ∫𝐍 ̃ dΩ ;
(22)
Ω Ω
𝑖, 𝑗 = 1,2,6
with
ℎ/2
𝑖𝑗
𝐷𝑠 = ∫ [(𝑓 ′ (𝑧))2 ]𝑄̄𝑖𝑗 𝑑𝑧,
𝑠
𝑖, 𝑗 = 4,5 𝐁𝑢 = [𝐁 𝟏 𝐁 𝟐 𝐁 𝟑 𝐁 𝒔 ]𝑇 ; 𝐍̄ = [0 0 𝑅𝐴 0 0];
(26)
−ℎ/2 𝐞̃ = [𝐞𝑇𝑚 z𝐞𝑇𝑚 f(z)𝐞𝑇𝑚 f ′ (z)𝐞𝑇𝑠 ]
The electric field vector E, can be described as and
0 0 0 0 𝑒15
𝑬 = −grad𝜙 = −𝛻𝜙 (23)
𝒆𝑚 = [ 0 0 0 ] ; 𝒆𝑠 = [𝑒15 0] (27)
𝑒31 𝑒32 𝑒33 0 0
C. Approximation of mechanical displacement and electric
potential field The global mass matrix 𝑴𝑢𝑢 is written as
1) Mechanical displacement field
𝐍 0 𝑇 𝐈𝟏 𝐈𝟐 𝐈𝟒 𝐍0
With the Non-Uniform Rational Basis functions, NURBS,
the mechanical displacement field of the plate can be written 𝐌𝑢𝑢 = ∫ {{𝐍1 } [𝐈𝟐 𝐈𝟑 𝐈𝟓 ] {𝐍1 }} 𝑑Ω (28)
as follows
Ω 𝐍 2 𝐈𝟒 𝐈𝟓 𝐈𝟔 𝐍2
𝑚×𝑛 where
ℎ (𝜉,
𝒖 𝜂) = ∑ 𝑅𝐴 (𝜉, 𝜂) 𝒅𝑨 (21)
𝐴
𝑅𝐴 0 0 0 0
where m×n is the number of basis functions. 𝐍0 = [ 0 𝑅𝐴 0 0 0] ;
Meanwhile𝑅𝐴 (𝜉, 𝜂) indicates a NURBS basis function and 0 0 𝑅𝐴 0 0
𝒅𝑨 = [𝑢0𝐴 𝑣0𝐴 𝑤𝐴 𝛽𝑥𝐴 𝛽𝑦𝐴 ]𝑇 is the vector of nodal 0 0 𝑅𝐴,𝑥 0 0
degrees of freedom related to control point A. 𝐍1 = − [0 0 𝑅𝐴,𝑦 0 0] ; (29)
The in-plane and shear strains can be rewritten as 0 0 0 0 0
𝑚×𝑛 0 0 0 𝑅𝐴 0
[𝜺 𝜸]𝑇 = ∑ [𝑩𝟏𝑨 𝑩𝟐𝑨 𝑩𝟑𝑨 𝑩𝒔𝑨 ]𝑇 𝒅𝑨 (22) 𝐍2 = [0 0 0 0 𝑅𝐴 ]
𝐴=1 0 0 0 0 0
where
IV. NUMERICAL RESULTS
𝑅𝐴,𝑥 0 0 0 0
E. Convergence and verification studies
𝑩𝟏𝑨 = [ 0 𝑅𝐴,𝑦 0 0 0] ;
𝑅𝐴,𝑦 𝑅𝐴,𝑥 0 0 0 The truthfulness and reliability of the present method are
proved through a numerical example reported by Li et al. [28].
0 0 𝑅𝐴,𝑥𝑥 0 0 A sandwich square FGP-GPLs plate with simply supported
𝑩𝟐𝑨 = − [0 0 𝑅𝐴,𝑥 0 0]; boundary condition (BCs-SSSS) is considered. The sandwich
plate has three layers including two isotropic metal face layers
0 0 2𝑅𝐴,𝑥𝑦 0 0 (23) (Aluminum) and a core layer with porosities. The core layer
0 0 0 𝑅𝐴,𝑥 0 is created by the uniformly distributed porous and is
𝑩𝟑𝑨 = [0 0 0 0 𝑅𝐴,𝑦 ] ; reinforced by uniformly distributed GPLs along the thickness.
The input data of plate is given as: a = b =1 m, h =0.005a,
0 0 0 𝑅𝐴,𝑦 𝑅𝐴,𝑥
ℎ𝑓 = 0.1ℎ, ℎ𝑐 = 0.8ℎ, 𝑒0 = 0.5. Material matrix is copper
0 0 0 𝑅𝐴 0 (Cu) and material parameters are given Table 1. And for the
𝑩𝒔𝑨 = [ ] GPLs 𝑙𝐺𝑃𝐿 = 2.5𝜇𝑚 , 𝑤𝐺𝑃𝐿 = 1.5𝜇𝑚 , 𝑡𝐺𝑃𝐿 = 1.5𝑛𝑚 and
0 0 0 0 𝑅𝐴
𝛬𝐺𝑃𝐿 = 1.0𝑤𝑡. %.
2) Electric potential field
454
We use cubic basis function (p=3) for mesh levels of 7x7, middle area of the plate which has more harm than other areas
11x11, 15x15, 17x17 and 19x19 elements. The frequencies [29]. As observed in Table 2 obtained results agree well with
obtained from the presented method are compared with [29] for all selected modes. Furthermore, at mesh levels of
analytical solutions reported by Li et al. [28] based on CPT. 17x17 and 19x19 elements, there is the same accuracy of
Table 2 shows the natural frequencies of the first four m and n natural frequency and the difference between the two mesh
values with various control mesh. It can be explained why levels is not noteworthy. So, the mesh of 17x17 cubic
choose these modes because of the active vibration in the elements is used for all numerical examples.
Table 1. Material properties (𝜀0 = 8.85𝑥10-12 F/m)
Alumium oxide Al Al2O3 Cu GPL PZT-4
Elastic properties E11 (GPa) 320.24 70 380 130 1010 81.3

E22 (GPa) 320.24 70 380 130 1010 81.3
E33 (GPa) 320.24 70 380 130 1010 64.5
G12 (GPa) - - - - - 30.6
G13 (GPa) - - - - - 25.6
G23 (GPa) - - - - - 25.6
𝜈12 0.26 0.3 0.3 0.34 0.186 0.33
𝜈13 0.26 0.3 0.3 0.34 0.186 0.43
𝜈23 0.26 0.3 0.3 0.34 0.186 0.43
3
Mass density p (kg/m ) 3750 2702 3800 8960 1062.5 7600
Piezoelectric coefficients d31 (m/V) -1.22 x10-10
d32 (m/V) - - - - - -1.22 x10-10
Electric permittivity p11 (F/m) - - - - - 1475𝜀0
p22 (F/m) - - - - - 1475𝜀0
p33 (F/m) - - - - - 1300𝜀0
Table 2. Convergence of the natural frequency (rad/s) of a square F. FGP-GPLs plates

sandwich FGP-GPLs for SSSS BCs with different control meshes
Consider a square FGP-GPLs plate with matrix material
Mesh Mode (m,n) Present Li et al.[28] made of copper for simply support SSSS and fully clamped
(1,1) boundary conditions CCCC. The data of GPLs is similar to
161.2154 160.6964
4.1 sub-section. The fundamental frequency is defined as𝜔∗ =
(1,3) 853.9223 803.4820 2)
7x7 𝜌𝑚 (1−𝜈𝑚
(3,3) 1536.2419 1446.2676 𝜔𝑎√ . Table 3 shows the frequency parameters of a
𝐸𝑚
(3,5) 2886.1548 2731.8389 square FGP plate versus the length-to-thickness ratio (a/h),
(1,1) 160.7702 160.6964 weight fraction 𝛬𝐺𝑃𝐿 = 1 𝑤𝑡. % and porosity coefficient
𝑒0 = 0.5
(1,3) 822.1501 803.4820
11x11
(3,3) 1466.3852 1446.2676 We compare the present results to the obtained results in
[17] studied by Yang et al. which used the Chebyshev-Ritz
(3,5) 2799.8712 2731.8389
method. There is a good similarity between obtained results
(1,1) 160.7068 160.6964 and reference solutions. Clearly, with a value of weight
(1,3) 811.5503 803.4820 fraction the stiffness of the plate decreases outstandingly when
15x15
(3,3) 1455.5842 1446.2676 the length-to-thickness ratio increases. It can be seen that
increasing leads to a lessening of the fundamental frequency
(3,5) 2766.2133 2731.8389
in case of NUPD 1 and UPD for both SSSS and CCCC BCs.
(1,1) 160.7010 160.6964
We continue studying the impact of the coefficient of
(1,3) 810.1398 803.4820
17x17 porosity 𝑒0 , the GPLs dispersion patterns and weight fraction
(3,3) 1450.6178 1446.2676 𝛬𝐺𝑃𝐿 on the natural frequencies of FGP-GPLs plate for three
(3,5) 2745.0773 2731.8389 porosity distribution types as exhibited in the Table 4. It can
(1,1) 160.6970 160.6964 be recognized that these quantities have a significant influence
on the results of the study. Of course, as the internal pore size
(1,3) 810.1320 803.4820
19x19 increases, the plate stiffness decreases and leads to a reducing
(3,3) 1452.6037 1446.2676 in the natural frequencies of FGP-GPLs plate. When
(3,5) 2753.0870 2731.8389 observing more carefully, it can be seen that the obtained
455
results for NUPD 1 slightly decrease. However, for NUPD 2 porosities, GPL A is the best one and pattern GPL B is the
and UPD, they fall more sharply. In addition, when weight worst one among three distinctive GPL distribution patterns.
fraction 𝛬𝐺𝑃𝐿 increases from 0.5 𝑤𝑡. % to 1 𝑤𝑡. % , the As a result, the GPL A produces the best reinforcing
obtained results also increase significantly because the plate performance for the free vibration analysis of FGP-GPLs
becomes harder. plate. Besides, for the NUPD 1 frequency is the biggest means
that the best stiffness is achieved for this case. A great
The careful observation shows that, frequency of case combination between NUPD 1 and GPL A will create the
pattern A supplies the greatest values, next is GPL C and the outstanding behaviors for FG porous structures compared
final is GPL B. It means that for the plates with many with other considered combinations.
Table 3. Dimensionless fundamental natural frequencies of a square FGP-GPLs plate with several the ratios a/h (𝛬𝐺𝑃𝐿 = 1 𝑤𝑡. %, 𝑒0 =
0.5)
NUPD 1 UPD
a/h Methods
GPL A GPL B GPL C GPL A GPL B GPL C
SSSS 20 Yang et al. [17] 0.3958 - 0.3574 0.3627 - 0.3252
Present 0.3960 0.3498 0.3571 0.3625 0.3189 0.3240
30 Yang et al. [17] 0.2657 - 0.2397 0.2433 - 0.2179
Present 0.2660 0.2342 0.2391 0.2432 0.2135 0.2172
40 Yang et al. [17] 0.1997 - 0.1801 0.1828 - 0.1637
Present 0.2002 0.1762 0.1800 0.1828 0.1604 0.1630
50 Yang et al. [17] 0.1600 - 0.1442 0.1464 - 0.1311
Present 0.1604 0.1411 0.1441 0.1464 0.1285 0.1309
CCCC 20 Yang et al. [17] 0.7022 - 0.6366 0.6456 - 0.5814
Present 0.6958 0.6190 0.6322 0.6398 0.5682 0.5770
30 Yang et al. [17] 0.4783 - 0.4324 0.4387 - 0.3938
Present 0.4771 0.4223 0.4311 0.4374 0.3859 0.3915
40 Yang et al. [17] 0.3616 - 0.3265 0.3313 - 0.2971
Present 0.3616 0.3190 0.3259 0.3305 0.2909 0.2959
50 Yang et al. [17] 0.2904 - 0.2620 0.2659 - 0.2383
Present 0.2906 0.2561 0.2615 0.2651 0.2330 0.2365
Table 4. Studying of porosity coefficient, type of porosity distribution, weight fraction 𝛬𝐺𝑃𝐿 and GPLs dispersion patterns on
dimensionless natural frequencies of a square FGP-GPLs plate
𝒆𝟎 NUPD 1 NUPD 2 UPD
GPL A GPL B GPL C GPL A GPL B GPL C GPL A GPL B GPL C
𝛬𝐺𝑃𝐿 = 0.5𝑤𝑡%
0.1 0.6135 0.5774 0.5745 0.6065 0.5663 0.5684 0.5995 0.5587 0.5593
0.3 0.6089 0.5677 0.5699 0.5872 0.5427 0.5475 0.5882 0.5510 0.5512
0.5 0.6041 0.5651 0.5678 0.5558 0.5125 0.5189 0.5717 0.5398 0.5378
𝛬𝐺𝑃𝐿 = 1.0𝑤𝑡%
0.1 0.7059 0.6285 0.6372 0.6985 0.6200 0.6315 0.6975 0.6210 0.6300
0.3 0.6992 0.6233 0.6339 0.6720 0.5925 0.6080 0.6705 0.5960 0.6052
0.5 0.6945 0.6189 0.6321 0.6371 0.5580 0.5784 0.6396 0.5872 0.5770
G. SFGP-GPLs plates dispersion pattern A offers the best reinforced candidate while
Next, an FGP-GPLs plate incorporated with piezoelectric the pattern B is the lowest. Furthermore, the porosity
layers in the top and bottom surface of the plate, SFGP‐GPLs distribution with metal foams distributed symmetry (NUPD 1)
with SSSS BCs is considered. Piezoelectric layer is made from supplies the fundamental natural frequency greater than the
PZT-4 material and material properties are also given in Table other for two cases of electrical boundary.
1. Geometrical dimensions and the dimensions of GPLs of the Next, Table 6 shows the five first natural frequencies of
plate are similar to the above example. The thickness of the the SFGP‐GPLs plate with 𝛬𝐺𝑃𝐿 = 1 .0wt.% and 𝑒0 = 0.3
core and piezoelectric layers is 50 mm and 1 mm, respectively. under different electrical boundaries where the thickness of
Table 5 shows the first natural frequencies of the core layer is 10 mm and the thickness of each piezoelectric
piezoelectric FGP-GPLs for two types of porosity distribution layer PZT-4 is 1 mm. We also present the natural frequency
and electrical boundary, different porosity coefficients, increment between C-C and C-O condition and calculated as
𝜔 −𝜔
dispersion pattern and weight fraction of GPLs. Clearly, as 𝛥 = 𝑜𝑝𝑒𝑛 𝑐𝑙𝑜𝑠𝑒 . The careful examination also shows that
𝜔𝑐𝑙𝑜𝑠𝑒
enhancing a small amount of GPLs into the metal matrix, the the mishmash between NUPD 1 and GPL 𝐴 creates the best
natural frequencies rise due to the stiffness of the plate structural behaviors among all considered combinations.
structure is drastically enriched. It can be seen that the
456
Table 5. The first natural frequencies (Hz) of a square SFGP-GPLs plate with SSSS BCs
Electrical
𝒆𝟎 𝛬𝐺𝑃𝐿 = 0.5𝑤𝑡% 𝛬𝐺𝑃𝐿 = 1.0𝑤𝑡%
condition
GPL A GPL B GPL C GPL A GPL B GPL C
Porosity distribution 1 (NUPD 1)
0 C-C 226.501 211.675 211.548 260.177 231.016 234.183
O-C 227.665 212.939 212.811 261.215 232.226 235.349
0.2 C-C 224.940 210.362 210.348 258.177 229.306 232.774
O-C 226.180 211.711 211.683 259.295 230.614 234.030
0.4 C-C 223.975 209.755 209.899 256.728 228.290 232.202
O-C 225.322 211.219 211.345 257.930 229.703 233.549
0.6 C-C 224.052 210.403 210.804 256.212 228.490 233.096
O-C 225.518 211.997 212.375 257.533 230.048 234.570
Uniformly distribution (UPD)
0.2 C-C 219.381 204.350 205.053 251.831 222.454 226.875
O-C 220.659 205.751 206.429 252.980 223.817 228.164
0.4 C-C 210.551 195.500 197.156 241.342 212.163 213.736
O-C 212.006 197.110 198.715 242.648 213.732 219.455
0.6 C-C 198.540 183.963 186.655 226.840 198.890 206.174
O-C 200.281 185.904 188.503 228.420 200.791 207.906
Table 6. The first five natural frequencies (Hz) of the square SSSS piezoelectric FGP-GPLs with different conditions
Mode Pattern A Pattern B Pattern C

C-C O-C 𝜟(%) C-C O-C 𝜟(%) C-C O-C 𝜟(%)
Porosity distribution 1 (NUPD 1)
1 53.524 54.678 2.173 48.620 49.925 2.684 49.118 50.390 2.578
2 134.010 136.900 2.160 121.810 125.060 2.668 123.055 126.207 2.561
3 134.055 136.950 2.160 121.856 125.107 2.668 123.100 126.253 2.561
4 214.732 219.335 2.143 195.290 200.464 2.649 197.282 202.300 2.544
5 268.692 274.421 2.133 244.440 250.890 2.637 246.928 253.181 2.532
Uniformly distribution (UPD)
1 51.699 52.910 2.339 46.818 48.189 2.928 47.483 48.796 2.765
2 129.470 132.491 2.325 117.315 120.731 2.912 118.986 122.255 2.747
3 129.514 132.525 2.325 117.310 120.725 2.911 119.030 122.300 2.747
4 207.510 212.300 2.307 188.110 193.548 2.893 190.780 195.987 2.729
5 259.680 265.644 2.296 235.478 242.261 2.880 238.820 245.309 2.717
graded materials for biomedical applications”, Materials

V. CONCLUSION Science and Engineering: A, 362 (1-2), 2003, pp. 40–60.
[3] L.P. Lefebvre, J. Banhart, D. C. Dunand, “Porous metals and
A valuable numerical model using IGA combined with the metallic foams: current status and recent developments”,
GHSDT has been suggested for the free vibration responses Advanced Engineering Materials, 10(9), 2008, pp. 775–787.
of SFGP-GPLs plates. By analyzing the vibration, the [4] C. Betts, “Benefits of metal foams and developments in
influence of dissimilar parameters counting as porosity modelling techniques to assess their materials behaviour: a
coefficients, external electric voltages, porosity distribution review”, Materials Science and Technology, 28(2), 2012, pp.
types, dispersion patterns and weight fractions of GPL on the 129–143.
behaviors of SFGP-GPLs plates are carefully studied. [5] B. Smith, S. Szyniszewski, J. Hajjar, B. Schafer, S. Arwade,
Remarkably, the achieved results match well with reference “Steel foam for structures: A review of applications,
manufacturing and material properties”, Journal of
studies in the literature. Additionally, we obtained numerical Constructional Steel Research, 71, 2012, pp. 1–10.
results for SFGP-GPLs, while analytical results for them have [6] X. Xia, X. Chen, Z. Zhang, X. Chen, W. Zhao, B. Liao, B. Hur,
not been achieved yet. Then, they have been used for “Effects of porosity and pore size on the compressive
references for related works. properties of closed-cell mg alloy foam”, Journal of
Magnesium and Alloys, 1(4), 2013, pp. 330–335.
REFERENCES [7] S. Iijima, “Helical microtubules of graphitic carbon,” Nature,
[1] Benasciutti D, Moro L, Zelenika S, et al., “Vibration energy 354 (6348) 56, 1991.
scavenging via piezoelectric bimorphs of optimized shapes”, [8] K. Liew, Z. Lei, L. Zhang, “Mechanical analysis of
Microsystem Technologies, 16(5), 2010, pp. 657–668. functionally graded carbon nanotube reinforced composites: a
[2] W. Pompe, H. Worch, M. Epple, W. Friess, M. Gelinsky, P. review,” Composite Structures, 120, 2015, pp. 90–97.
Greil, U. Hempel, D. Scharnweber, K. Schulte, “Functionally [9] Nguyen Van Thanh, Nguyen Dinh Khoa, Ngo Duc Tuan,
Phuong Tran, Nguyen Dinh Duc, “Nonlinear dynamic response
457
and vibration of functionally graded carbon nanotube- [21] Halpin, J. C., and Kardos, J. L., “The halpin-tsai equations: A
reinforced composite (FG-CNTRC) shear deformable plates review”, Polymer Engineering and Science, 16, 1976, pp. 344–
with temperature-dependent material properties and 352.
surrounded on elastic foundations,” Journal of Thermal [22] Lieu B. Nguyen, Nam V. Nguyen, Chien H. Thai, A.M. J.
Stresses, 40, 2017, pp. 1254-1274. Ferreira, H. Nguyen-Xuan, “An isogeometric Bézier finite
[10] G. Mittal, V. Dhand, K. Y. Rhee, S.-J. Park, W. R. Lee, “A element analysis for piezoelectric FG porous plates reinforced
review on carbon nanotubes and graphene as fillers in by graphene platelets”, Composite Structure, 214, 2019, pp.
reinforced polymer nanocomposites”, Journal of Industrial and 227-245.
Engineering Chemistry, 21, 2015, pp. 11–25. [23] Nam V. Nguyen, Lieu B. Nguyen, H. Nguyen-Xuan, Jaehong
[11] D. G. Papageorgiou, I. A. Kinloch, R. J. Young, “Mechanical Lee, “Analysis and active control of geometrically nonlinear
properties of graphene and graphene-based nanocomposites”, responses of smart FG porous plates with graphene
Progress in Materials Science, 90, 2017, pp. 75–127. nanoplatelets reinforcement based on Bézier extraction of
[12] King JA, Klimek DR, Miskioglu I, Odegard GM, “Mechanical NURBS”, International Journal of Mechanical Sciences, 180,
properties of graphene nanoplatelet/epoxy composites”, J Appl 2020, 105692.
Polym Sci 128(6), 2013, pp. 4217–4223. [24] Quang Huan Nguyen, Lieu B. Nguyen, Hoang B. Nguyen, H.
[13] C. Betts, “Benefits of metal foams and developments in Nguyen-Xuan, “A three-variable high order shear deformation
modelling techniques to assess their materials behaviour: a theory for isogeometric free vibration, buckling and instability
review”, Materials Science and Technology, 28(2), 2012, pp. analysis of FG porous plates reinforced by graphene platelets”,
129–143. Composite Structures, 245, 2020, 112321.
[14] L.P. Lefebvre, J. Banhart, and D. C. Dunand, “Porous metals [25] A. P. Roberts, E. J. Garboczi, “Elastic moduli of model random
and metallic foams: current status and recent developments”, three-dimensional closed-cell cellular solids”, Acta Materialia,
Advanced Engineering Materials, 10(9), 2008, pp. 775–787. 49(2), 2001, pp. 189–197.
[15] H. N. Wadley, N. A. Fleck, A. G. Evans, “Fabrication and [26] Thai CH, Ferreira AJM, Bordas SPA, Rabczuk T, Nguyen-
structural performance of periodic cellular metal sandwich Xuan H, “Isogeometric analysis of laminated composite and
structures”, Composites Science and Technology, 63(16), sandwich plates using a new inverse trigonometric shear
2003, pp. 2331–2343. deformation theory,” European Journal of Mechanics -
A/Solids, 43, 2001, pp. 89-108.
[16] I. Duarte, E. Ventura, S. Olhero, J. M. Ferreira, “An effective
approach to reinforced closed-cell al-alloy foams with [27] P. Phung-Van, T. Nguyen-Thoi, T. Bui-Xuan, Q. Lieu-Xuan,
multiwalled carbon nanotubes”, Carbon, 95, 2015, pp. 589– “A cell-based smoothed three-node Mindlin plate element (CS-
600. FEM-MIN3) based on the C0-type higher-order shear
deformation for geometrically nonlinear analysis of
[17] J. Yang, D. Chen, S. Kitipornchai, “Buckling and free vibration laminated”, Computational Materials Science, 96, 2015, pp.
analyses of functionally graded graphene reinforced porous 549-558.
nanocomposite plates based on chebyshev-ritz method”,
Composite Structures, 193, 2018, pp. 281–294. [28] Q. Li, D. Wu, X. Chen, L. Liu, Y. Yu, W. Gao, “Nonlinear
vibration and dynamic buckling analyses of sandwich
[18] Liu J, Yan H, Jiang K, “Mechanical properties of graphene functionally graded porous plate with graphene platelet
platelet-reinforced alumina ceramic composites”, Ceram Int reinforcement resting on winkler-pasternak elastic
39(6), 2018, pp. 6215–6221. foundation”, International Journal of Mechanical Sciences,
[19] K. Li, D. Wu, X. Chen, J. Cheng, Z. Liu, W. Gao, M. Liu, 148, 2018, pp. 596-610.
“Isogeometric analysis of functionally graded porous plates [29] S. Kudus, Y Suzuki, M. Matsumura, K. Sugiura, “Vibration -
reinforced by graphene platelets”, Composite Structures, 204, response due to thickness loss on steel plate excited by
2018, pp. 114–130. resonance frequency”, In IOP Conerence Series: Earth and
[20] Song, M., Kitipornchai, S., and Yang, J., “Free and Forced Environmental Science, IOP Publishing, 140(1), 2018, pp. 012-
vibrations of functionally graded polymer composite plates 123.
reinforced with graphene nanoplatelets”, Composite
Structures, 159, 2017, pp. 579–588.
458
Free Vibration of MSGT Porous Metal Foam

Microplates Using a Moving Kriging Meshfree
Approach
P. T. Hung
hungpht@hcmute.edu.vn
Abstract: Based on a Moving Kriging (MK) meshfree viscoelastic foundation. On the other hand, the polygonal
method, free vibration of the porous metal foam microplates finite element formulation has been utilized by Nguyen et al.
was investigated in this article. The pores are distributed [12] to study the active vibration control of FG porous metal
through plate thickness in three types: uniform, symmetric foam plates reinforced with graphene platelets (GPL). The
and asymmetric. A combination of the modified strain nonlinear buckling and post-buckling of metal foam plates
gradient theory (MSGT) and two variable refined plate theory based on the CPT were presented by Tu et al. [13]. Based on
is used for the investigation of the size effect of the the analytical method, Wang et al. [14] investigated the
microplates. According to Hamilton’s principle, the vibration of porous metal circular cylindrical shells
reinforced by GPL according to Donnell’s nonlinear shell
governing equations of the microplates are derived. The
theory. Also, Pham et al. [15] investigated mechanical
natural frequencies of the porous metal foam plates are found
buckling of porous metal cylindrical shells subjected to axial
by solving the explicit governing equation based on the MK compression based on the analytical method.
meshfree method. The influence of the length scale
parameter, porosity distribution, porosity coefficient and The general higher-order stress theory was firstly
geometry on the frequencies of the microplates are studied. proposed by Midlin [16] with five length-scale parameters
(LSPs). After that, Lam et al. [17] developed this theory
Keywords: porous metal foam plates, Moving Kriging namely modified strain gradient theory with only three LSPs.
meshfree method, porosity distributions, two variable refined Many researchers used MSGT to investigate the behaviors of
plate theory the microstructures. Based on MSGT, Sahmani et al. [18]
presented the nonlinear vibration of FG microbeams
I. INTRODUCTION according to the analytical method. According to the
Porous structures have been intensively investigated in differential quadrature method, Singh et al. [19] used the
recent years, because of their outstanding mechanical MSGT to study the free vibration of the microbeams.
properties. Based on the Timoshenko beam theory, Chen et Moreover, Zanoosi [20] calculated the analytical natural
al. [1] analyzed the nonlinear vibration of the sandwich beams frequencies of the microbeams under thermal effects based on
with functionally graded (FG) porous metal foam core. the MSGT. Via an MSGT, Zhang et al. [21] carried out the
Furthermore, Magnucka [2] studied the buckling analysis of analytical bending, free vibration and buckling of the FG
the sandwich plates with metal foam core based on an microplates base on the RPT. Using the MSGT combined
analytical method. The classical plate theory (CPT) and the with an analytical method, Ansari et al. [22] studied the
analytical method are used by Jabbari et al. [3, 4] to bending, vibration and buckling of the FG microplates.
investigate the mechanical and thermal buckling of the porous Besides, Jamalpoor et al. [23] investigated analytical free
metal foam plates. Based on the first-order shear deformation vibration of orthotropic multi-visco-elastic microplates via an
plate theory (FSDT), Rezaei et al. [5] found the analytical MSGT and Kirchhoff plate theory. Using the same method,
natural frequencies of the FG plate made of porous materials. these authors [24] found the natural frequencies and the
Barati et al. [6] utilized the analytical free vibration and buckling load of the electro-magneto-elastic microplates
buckling analyses of functionally graded piezoelectric porous resting on the Visco-Pasternark subtract.
(FGPP) plates. Also, the nonlinear forced vibration of FGPP As we see in the above literature reviews, there are many
plates was studied by Wang and Zu [7]. Keddouri [8] used the investigations of MSGT microplates using the analytical
refined plate theory (RPT) to investigate the effect of the method. However, these methods are only suitable for simple
porosity distribution and porosity coefficient on the analytical problems with simple boundaries. For real structure in
deflection and the stresses of the FG sandwich plates with practice, numerical methods such as FEM, isogeometric
porosities. According to the quasi-3D shear deformation analysis and meshfree method are the best choices. The MK
theory, Zenkour et al. [9, 10] solved the bending of FG and meshfree method was firstly proposed by Gu [25] for solving
FG sandwich porous plates based on the analytical method. the weak form of the two-dimensional boundary value
In addition, Ebrahimi et al. [11] studied the analytical free problems. After that, based on the MK meshfree approach, Bui
vibration of the porous metal foam plates resting on a [26] carried out the free vibration of the Kirchhoff plates. The

RPT and MK meshfree method were utilized by Nguyen et al. The Poisson’s ratio  is assumed to be a constant across
[27] for the calculation of the nonlinear deflection and the plate thickness [30].
frequencies of the FGM plates. Recently, according to the MK
meshfree approach, Thai et al. [28, 29] researched the size- The definition of e0 and em is described as follows:
dependent analysis of the microplates. As we see, from the
above-mentioned literature, there is no investigation of the
 E2 G2
vibration of the porous metal foam microplates using the MK e0 = 1 − E = 1 − G , 0  e0  1
meshfree method. Therefore, in this article, via an MK  1 1
 (5)
e = 1 −  2 , 0  e  1
meshfree approach, we investigate the size-dependent free
vibration of the metal foam MSGT microplates based on two 
variables refined plate theory. 
m
1 m
II. THE BASIC EQUATIONS The variation of Young’s modulus of the uniform,
symmetric and asymmetric porous metal foam microplates
A. Material Properties of The Metal Foam Microplates with e0 = 0.2 were presented in Figure 1.
Let us consider the porous metal foam microplates with
the pores are distributed through the plate thickness in three
forms: uniform, symmetric and asymmetric. The material
properties of the microplates are presented as follows [30]:
➢ Uniform distribution:
 E ( z ) = E1 (1 − e0 ) ;

G ( z ) = G1 (1 − e0 ) ;

(
  ( z ) = 1 1 − 1 − e0

) (1)
 1 12 2 
2
 = −  − − +
 
1 e 1
e0 e0  
0

➢ Symmetric distribution:
    z  Figure 1. Young’s modulus of the porous metal foam plate (e0 =
 E ( z ) = E1 1 − e0 cos    ; 0.2)
   h 
    z  (2)
G ( z ) = G1 1 − e0 cos    ; B. A Two Variables Refined Plates Theory
   h  The displacement fields at any point of the microplates
  z  according to the two variables refined plate theory [31] can
  ( z ) = 1 1 − em cos    be presented as follows:
   h 
u = − zwb , x + f ( z ) ws , x
➢ Asymmetric 
v = − zwb , y + f ( z ) ws , y
distribution: (
6)
    2 z  
 E ( z ) = E1 1 − e0 cos  + 1 ;  w = wb + ws
  4  2h  
 (3) where wb and ws are bending and shear deflection through the
    2 z 
G ( z ) = G1 1 − e0 cos  + 1  ; z-axis, respectively; the symbols ‘,x’ and ‘,y’ are the derivative
  4 h  of any function with x and y directions; f ( z ) = −4 z 3 / 3h2 is
   2z 
  ( z ) = 1 1 − em cos  + 1  distributed function.

  4  h  C. The Modified Strain Gradient Theory
The weak form for the microplate is presented as follows:
where
U −  K − V = 0 (7)
em = 1 − 1 − e0 (4) where U, K and V are strain energy, kinetic energy and work
done by external loads, respectively.
in which E1 , G1 , 1 are the maximum values of Young’s
Based on the MSGT [17], the strain energy for the linear
modulus, shear modulus, and mass density of the plate, elastic solid is described by
respectively, while E2 , G2 , 2 are the minimum values of
Young’s modulus, shear modulus and mass density, U =  ( σε + mχ + pζ + τη )dV (8)
V
respectively; e0 and em are the porous coefficient and
porosity density coefficient, respectively. where ε is the strain tensor, σ is Cauchy stress tensors χ is the
tensor of symmetric rotation gradient, ζ is the tensor of
460
dilatation gradient, η is the tensor of deviatoric stretch  −2ws , xy 

gradient, m, p and τ are the high-order stresses tensors. 1  1 − ws , y 
 =  2ws , xy  ; 1s = 
b
2 
The strain tensor is defined as 4  4  ws , x 
 ws , xx − ws , yy 
1
ε = u + ( u ) 
T
(9) In addition, the components of the dilatation gradient
2  
tensor take the forms
  
with  =   , and  x   xx , x +  yy , x +  zz , x 
 x y z     
ζ =  y  =  xx , y +  yy , y +  zz , y  =
    +  +   (16)
u   z   xx , z yy , z zz , z 
 
u =  v  = u 0 + zu1 + f ( z ) u 2 ; = ζ1 + zζ 2 + f ( z ) ζ3 + f  ( z ) ζ 4
 w
  where
(10)
 0   wb , x   ws , x 
       0 
u 0 =  0  ; u1 = −  wb , y  ; u 2 =  ws , y 
 
w + w   0   0  ζ1 =  0 ;
 b s    
− w − w 
Inserting Eq. (10) to Eq. (9), the strain tensor reforming  b, xx b , yy 
as follows:  wb , xxx + wb , xyy   ws , xxx + ws , xyy 

   
ζ 2 = −  wb , yyy + wb , xxy  ; ζ 3 =  ws , yyy + ws , xxy  ; (17)
 xx     
  b 
b
  0   0 
ε =  s  ;  =  yy  = z1b + f ( z ) 22 ;
     0 
2 xy  (11) 
ζ4 =  0


2 xz 
  w + w 
εs =   = (1 + f ( z ) ) γ  s , xx s , yy 
s
 
 yz 
2  Furthermore, the deviatoric stretch gradient tensor is
where expressed as follows:
 w,bxx 
 b  b  s 
 w,sxx 
 w,sx  ijk =
1
( ij ,k +  jk ,i +  ik , j ) −
 = −  w, yy  ;  2 =  w, yy  ; γ s =  s 
b
1 (12) 3
 w, y 
2wb 
 , xy 
 2 ws 
 , xy  −
1
15
( )
 ij ( mm, k + 2 mk , m ) +  ik ( mm, j + 2 mj ,m ) − (18)
Besides, the tensor of symmetric rotation gradient χ is

defined as follows:
−
1
15
(
 jk ( mm,i + 2 mi , m ) )
1 in which, δ denotes the Kronecker’s delta.
χ = θ + ( θ )  ;
T
2  Inserting Eq. (6) to Eq. (18), the tensor of the deviatoric

 w, y − v, z  (13) sketch gradient is reformed as follows:
1 
θ =   u = u, z − v, x   ηb 
2  η =  s;
v, x − u, y   η 
Substituting Eq. (6) to Eq. (13), tensors of bending and  xxx 
shear rotation are presented as follows:  
 yyy 
  xx   
 yyx 
 χ b  b   ηb =   = zη1 + f ( z ) η2 + f  ( z ) η3 ;
b b b
χ =  s  ;  =   yy  = 1b + f  ( z )  b2 and 
 xxy 
 χ    (19)
 
  xy  (14)
 zzx 
  xz 
   zzy 
 
 s =   = f  ( z ) 1s
  yz 
   zzz 
 
 xxz 
where η =  = η1 + f  ( z ) η2
s s s

 yyz 
 4wb, xy + 2ws , xy   xyz 
1    
 = 
b
1 −4wb, xy − 2ws , xy ; (15)
4  where
2 ( wb, yy − wb, xx ) + ( wx , yy − ws , xx ) 
461
 −2 wb , xxx + 3wb , xyy   2 ws , xxx − 3ws , xyy   xxx  1 0 0 0 0 0   xxx 

−2 w   2w    0  
 b , yyy + 3 w b , xxy   s , yyy − 3ws , xxy   yyy   1 0 0 0 0   yyy 
1  wb , xxx − 4 wb , xyy  b 1   − ws , xxx + 4 ws , xyy 
  yyx   0   yyx 
η1b =   ; η2 =  ; 2 0 0 1 0 0
5  wb , yyy − 4 wb , xxy  5 − ws , yyy + 4 ws , xxy    = 2Gl3   ;

 xxy  0 0 0 1 0 0   xxy 
 wb , xxx + wb , xyy   − ws , xxx − ws , xyy 
     zzx  0 0 0 0 1 0   zzx 
 b , yyy
w + w b , xxy   − ws , yyy − ws , xxy 
       (25)
 zzy  0 0 0 0 0 1   zzy 
 −3ws , x 
−3w   zzz  1 0 0 0   zzz 
    
 0   xxz 
s, y 
 xxz  2 0 1 0

1  s , x 
− w   = 2Gl3   
ηb3 = 
15  − ws , y 
  yyz  0 0 1 0   yyz 
 xyz   
(20)   0 0 0 1   xyz 

 4 ws , x 
 
 4 ws , y  where l1 , l2 , l3 are three LSPs.
3wb , xx + 3wb , yy − 3ws , xx − 3ws , yy 
 −4w + w + 4w − w  Substituting Eqs. (11), (14), (16) and (19) into Eq. (8), the
1  s , yy  virtual strain energy can be rewritten as follows:
η1s = 
b , xx b , yy s , xx
;
15  −4wb , yy + wb , xx + 4ws , yy − ws , xx 
 U =   ( ε b ) Db ε b d +   ( γ s ) Ds γ s d +
T T
 −5wb , xy + 5ws , xy 
   
−6 ws , xx − 6 ws , yy 
+   ( χ b ) Dbr Γbr χ b d +   ( χ s ) Drs Γ rs χ s d +
T T
 8w − 2 w   
1  s , xx s , yy  (26)
η2s =  
15  8ws , yy − 2 ws , xx  +   ζ D ζd +   ( η
T dil b T
) Ddeb Γ deb ηb d +
 
 10 ws , xy 
 
+  ( η )
s T
D Γ ηs d
des des

Based on the Hook’s law, the classical stress tensor is
presented as follows: where
0  x  ε b = 1b b2  ; χ b = 1b b2  ;

T T
 Q11 Q12 0 0
Q  
 21 Q22 0 0 0    y 
  ζ = ζ1 ζ 4  ; ηb = η1b  η3b 
T
σ= 0 0   xy  = Cε
T
0 Q66 0 (21) ζ2 ζ3 ηb2
  
0   xz
 0 0 0 Q55
 η2s  ;
T
  ; ηs = η1s
 0 0 0 0 Q44   yz 
 Ab Bb 
in which Db =  b ;
B Db 
E (z)  E ( z)
Q11 = Q22 = ; Q12 = Q21 = ;  A di B di Cdi E di 
1 − 2
1 − 2  di 
(22) Ddi F di Ldi 
E (z) E (z) B
Ddi =  di ;
Q66 = Q55 = Q44 = ; =G = C F di H di O di 
2(1 +  ) 2(1 +  )  di  (27)
 E Ldi O di P di 
According to the MSGT [17], the higher-order stress
components are defined as follows:  A deb B deb Cdeb Edeb 
 deb 
B Ddeb F deb Ldeb 
 mxx  1 0    xx  Ddeb =  deb ;
0 0 0 C F deb H deb O deb 
m  0  
 yy   1 0 0 0    yy   deb 
     E Ldeb O deb P deb 
 mxy  = 2Gl1 0
2
0 1 0 0    xy  (23)
m   
0 0 0 1 0    xz   Ab
Dbr =  br
Bbr 
=
 A rs B rs 
 xz   
s
 ; D  s ;
0 1    yz 
r
 m yz  0 0 0  Br Dbr   Br Drs 
A des
B des
 px  1 0 0   x  Ddes =  des ;
  2    B Ddes 
 p y  = 2Gl2  0 1 0   y  (24)
p   
 z  0 0 1   z 
462
 Γb 0  Γrs 0  u 0  I 0 0 0
Γbr =  r  Γ s
=  s  
u = u1  ; I m =  0 I 0 0  ;
; r ;
 0 Γr   0 Γr 
b
u   0 0 I 0 
 Γ deb 0 0 0   2
 
(31)
0 Γ deb
0 0   I1 I 2 I 4 
Γ deb = 
I 0 =  I 2 I 3 I 5 
;
 0 0 Γ deb 0 
   I 4 I 5 I 6 
 0 0 0 Γ deb 
 Γdes 0  where
Γdes =  ;
Γdes 
 0 ( I1 , I 2 , I3 , I 4 , I 5 , I 6 ) =
Γ = diag (1,1, 2) ; Γ = diag ( 2, 2 ) ;
b s h
  ( z ) (1, z, z )
r r 2 (32)
Γdeb = diag (1,1,3,3,3,3) ; Γdes = diag (1,3,3,6 ) , f ( z ) , zf ( z ) , f ( z ) dz
2
= 2
−h
in which 2
(A ,B ,D ) =
b b b
D. A Moving Kriging Meshfree Method
 Q11 Q12 0  We now consider the 2D domain Ω with n nodes,
boundary Γ, and support domain of point x, Ωx ∈ Ω, as shown
= ( z , z f ( z), f ( z ) ) Q21 Q22 0  dz
h/ 2
2 2
−h/ 2 in Figure 2.
 0 0 Q66 
Γ
2 Q 0  Support domain
D =  (1 + f ( z ) )  44
h/2
s
Q55 
dz ;
−h/ 2
 0
(A ,B ,D ) =  2Gl12 (1, f ( z ), f 2 ( z ) ) I3x3dz
h/ 2
b b b
r r r −h/ 2 Ωx x
(A ,B , D ) =  (1, f ( z), f  ( z) ) I
h/ 2
s s s 2 2
r r r 2Gl 1 2x2 dz
−h/ 2
(A di
, B di , Ddi , Cdi , E di , F di ) = Ω
(28)
= 2Gl2 2 (1, z , z 2 , f ( z ), f ( z ), zf ( z ) ) I 3x3dz
h/2
−h/ 2
(L , H
di di
, Odi , P di ) = Figure 2. Support domain of MK meshfree method
2Gl2 2 ( z f ( z ), f 2 ( z ), f ( z ) f ( z ), f 2 ( z ) ) I3x3dz
h/2
 The MK interpolation of the approximation vector
u ( x ) ,  x  x is formulated as
−h/ 2
h
(A deb
, B deb , Ddeb , Cdeb , E deb , F deb ) = n
u h ( x ) = pT ( x ) A + rT ( x ) B  u =  Ni ( x ) ui (33)
= 2Gl32 (1, z , z 2 , f ( z ), f ( z ), zf ( z ) ) I 6x 6 dz
h/2
i =1
−h/ 2
with Ni(x) is the Moving Kriging shape function. The vectors
(L deb
, Hdeb , Odeb , P deb ) =
p ( x ) , r ( x ) , matrices A, B and shape function Ni(x) are
2Gl32 ( z f ( z ), f 2 ( z ), f ( z ) f ( z ), f 2 ( z ) ) I 6x6dz
h/ 2
−h/ 2
determined by
(A , Bdes , Ddes ) =  2Gl32 (1, f ( z), f 2 ( z) ) I 4x4dz

h/ 2
p ( x ) =  p1 ( x ) p2 ( x )  pm ( x )  ;
des T
−h / 2
r ( x ) =  R ( x1 , x ) R ( x 2 , x )  R ( x n , x )  ;
T
in which I2x2 , I3x3 , I4x4 , I6x6 are the identity matrices of size
of 2×2, 3×3, 4×4 and 6×6, respectively.
( )
−1 (34)
A = PT R −1P PT R −1 ; B = R −1 ( I − PA ) ;
The virtual kinetic energy  K is presented by
m n
 K =   u T  ( z ) u dV Ni ( x ) =  p j ( x ) Aji +  rk ( x ) Bki .
(29) j =1 k =1
V
where a dotted variable indicates its time derivative where I is a n×n unit matrix and
Integrating Eq. (29) by part with respect to time, the

virtual kinetic energy is rewritten as
 K =   uT I m ud (30)

in which
463
 p1 ( x1 ) pm ( x1 )   4 Ni , xy 2 Ni , xy 
   
P ( x) =  ;
1
Bbri1 =  −4 Ni , xy −2 Ni , xy  ;
 p1 ( xn ) pm ( xn ) 4 
 (35)  2 ( Ni , yy − Ni , xx ) Ni , yy − Ni , xx  (42)
 R ( x1 , x1 ) R ( x1 , x n ) 
  0 −2 Ni , xy 
R ( x) =  1  1 0 − Ni , y 
 Bbri2 = 0 2 Ni , xy  ; B ris =  
 R ( x n , x1 ) R ( x n , x n ) 4 4 0 Ni , x 
 0 Ni , xx − Ni , yy 
 
In this study, p(x) and R(x) are chosen as
Similarly, substituting Eq. (38) into Eq. (16), the
components of the dilatation tensor are rewritten as follows
p ( x ) = 1 x
T
y x2 xy y 2  ;
ζ = ζ1 ζ4  =
T
  xi − x j 
2
(36) ζ2 ζ3
− 
( )
  (43)
=  B1dili 4i  qi =  Bi qi
a0 n n
R xi , x j = e   T
B dil
2i B3dili B dil dil
i =1 i =1
where γ is the correction factor chosen by 1 as observed in the
where
previous studies [32, 33], and a0 is the maximum distance
between a pair of nodes in the support domain.  0 0
 
In this study, the radius of the support domain dm is B = dil
1i 0 0 ;
defined as follows.  − Ni , xx − Ni , yy 0
 
dm =  dave (37)  − N i , xxx − N i , xyy 0 
 0  ;
in which dave is an average distance between nodes and α is a 2 i =  − N i , yyy − N i , xxy
B dil
scale parameter. The scale parameter can be chosen  0 0 
differently between 2.0 and 4.0 [34], dependent on different (44)
problem classes. In this article, the scale parameter is taken 0 N i , xxx + N i , xyy 
B3i = 0 N i , yyy + N i , xxy  ;
by 2.8. dil
The displacement fields based on the MK basic functions 0 

0
are presented as follows
0 0 
uh ( x ) =   i
( ) 0  wbi  = n N x d
n N x
 
   ( ) i B 4i = 0
dil
(38) 0 
i =1 
 0 Ni ( x )   wsi  I =1 i 0 Ni , xx + Ni , yy 
 
Inserting Eq. (38) into Eq. (11), the components of
The deviatoric gradient components after substituting Eq.
bending and shear strain are reformed as
(38) into Eq. (19), can be obtained by
n
  B d ηb = 1b b3  =
n n
ε =  B1bi di =  Bib di ; γ s =
T s T
Bb2i i i (39) b2
i =1 i =1 i=1 ;
=  B1deb i  qi =  Bi qi
n T n
in which i B deb
2i B3deb deb
i =1 i =1 (45)
 Ni , xx 0 0 Ni , xx 
ηs = η1s η2s  =  B1des 2i  qi =  Bi qi
T n T n
  b   B des des
B = −  Ni , yy
b
1i 0  ; B 2i = 0 Ni , yy  ; i
i =1 i =1
 2 Ni , xy 0  0 2 Ni , xy 
   (40) where
 0 N 
Bis = 
i, x
0 N 
 i ,y 
 −2 N i , xxx + 3Ni , xyy 0
Substituting Eq. (38) into Eq. (14), the rotation gradient  −2 N 0 
 i , yyy + 3 N i , xxy
components are rewritten by
1  N i , xxx − 4 N i , xyy 0 (46)
B1deb =  ;
5  N i , yyy − 4 Ni , xxy
χ =   = B 
n n i
0
b b
 b T b1
B b2 T
qi =  B qi ; b
 N
1 2 ri ri ri
i , xxx + N i , xyy 0
i =1 i =1
(41)  
n  N i , yyy + Ni , xxy 0
12 =  B ris qi
i =1
where
464
0 2 N i , xxx − 3N i , xyy  III. NUMERICAL RESULTS

0 2 N  Let us consider the porous metal foam rectangular
 i , yyy − 3 N i , xxy 
microplates with the length a, width b, thickness h. For a
1 0 − N i , xxx + 4 Ni , xyy  numerical example, we assume that the three LSPs are the
B deb =  ;
5 0 − N i , yyy + 4 Ni , xxy  same ( l1 = l2 = l3 = l ) and the boundary conditions (BCs)
2i
0 − N i , xxx − Ni , xyy  are used as follows

 
0 − N i , yyy − Ni , xxy  ➢ Fully simply supported (SSSS):
0 −3Ni , x  ( wb , ws ) x = 0, a ; y = 0,b
= 0;
0 −3N 
 i, y  ➢ Simply supported at x =0, a; y=0 and free at y = b
1 0 − Ni , x  (SSSF): ( wb , ws ) = 0;
B3deb =  ; x = 0, a; y = 0
15 0 − Ni , y 
i
➢ Simply supported at x =0, a and free at y = 0, b
0 4 Ni , x  (SFSF): ( wb , ws )
  = 0.
x = 0, a
0 4 Ni , y 
Firstly, the free vibration of the FG microplates is
 3N i , xx + 3N i , yy −3Ni , xx − 3Ni , yy  presented to verify the accuracy and accordant of the present
 4 Ni , xx − Ni , yy  method. The material properties is taken [21], [35]:
1 −4 N i , xx + N i , yy
B1des =  ; Em = 70 GPa , vm =0.3 , m = 2702 kg/m3 and
15  −4 N i , yy + Ni , xx 4 Ni , yy − Ni , xx 
i
  Ec = 380 GPa , vc =0.3 , c = 3800 kg/m3 . Table 1 presents

 −5 N i , xy 5 N i , xy 
the dimensionless first natural frequency  =  h c / Ec of
0 −6 N i , xx − 6 N i , yy 
  the SSSS FG microplate with various power index and scale-
1 0 8 N i , xx − 2 Ni , yy 
B des =  to-thickness ratios (l/h). As we see in Table 1, the
15 0 8 N i , yy − 2 Ni , xx 
2i
dimensionless natural frequency match very well with those
  of references in [21] and [35]. The comparison results in
0 10 N i , xy  Table 1 show that the proposed method is accurate and
Besides, the components of displacement can be rewritten accordant.
as follows
Table 1. The first dimensionless natural frequency of the Ssss Fg
u = u1 u3  = square microplate with various power index and scale-to-thickness
T
u2
ratios
n n (47)
=  N1i N 3i  q i =  N i q i
T
N 2i n
i =1 i =1
a/h l/h Reference
0 1 2 5
where Present 0.2113 0.1634 0.1476 0.1367
0 Ref. [21] 0.2113 0.1631 0.1472 0.1378
0 0  − Ni , x 0 0 N i , x  Ref. [35] 0.2113 0.1631 0.1472 0.1358
N1i =  0 
0  ; N 2i = − Ni , y 
0 ; N3i = 0 Ni , y  (48) Present 0.2271 0.1775 0.1606 0.1478
0.1 Ref. [21] 0.2273 0.1774 0.1603 0.1470
 Ni Ni   0 0 0 0 
Ref. [35] 0.2264 0.1766 0.1596 0.1463
Present 0.2687 0.2140 0.1941 0.1767
Substituting Eqs. (39), (41), (43), (45) and (47) into the 5
0.2 Ref. [21] 0.2698 0.2144 0.1944 0.1764
Eq. (26), the weak forms for the free vibration and buckling
Ref. [35] 0.2666 0.2118 0.1920 0.1740
analyses of the metal foam microplate respectively reformed Present 0.4632 0.3785 0.3462 0.3106
as follows 0.5 Ref. [21] 0.4688 0.3823 0.3490 0.3117
( K −  M) d = 0
2
(49)
Ref. [35]
Present
0.4566
0.8480
0.3729
0.6963
0.3402
0.6383
0.3029
0.5702
1 Ref. [21] 0.7011 0.5832 0.5270 0.4553
where K and M are the global stiffness matrix and mass Ref. [35] 0.8344 0.6867 0.6266 0.5540
matrix respectively, expressed as follows Present 0.0577 0.0443 0.0402 0.0379
0 Ref. [21] 0.0577 0.0442 0.0401 0.0377
( ) D B d +  ( B ) D B d +
T T
K =  Bb b b s s s
Ref. [35] 0.0577 0.0442 0.0401 0.0377
 
Present 0.0619 0.0480 0.0436 0.0407
+  ( B ) D Γ B d +  ( B ) D Γ B d +
T T
b b b b s s s s
 r r r r  r r r r 0.1 Ref. [21] 0.0619 0.0480 0.0435 0.0405
Ref. [35] 0.0617 0.0478 0.0434 0.0403
+  ( B ) D B d +  ( B ) D Γ B d +
T T
dil dil dil deb deb deb deb
  (50) Present 0.0729 0.0578 0.0525 0.0481
0.2 Ref. [21] 0.0730 0.0578 0.0524 0.0478
+  ( B ) D Γ B d
T 10
des des des des

Ref. [35] 0.0725 0.0573 0.0520 0.0474
Present 0.1254 0.1026 0.0933 0.0829
M =  NT I m Nd ; d = dei t 0.5 Ref. [21] 0.1258 0.1028 0.0933 0.0825
 Ref. [35] 0.1240 0.1013 0.0918 0.0810
Present 0.2297 0.1900 0.1728 0.1520
where ω, d denote natural frequencies and mode shapes, 1 Ref. [21] 0.2309 0.1907 0.1731 0.1514
respectively. Ref. [35] 0.2268 0.1873 0.1698 0.1482
465
Table 2. The five first dimensionless natural frequencies Ω of the Table 4. The first dimensionless frequenciy Ω of the square
square metal foam microplates ( e0 = 0.1 ; a/h = 10) symmetric porous microplates with various length-to-thickness
l/h ratio and scale-to-thickness ratios ( e0 = 0.2 )
Type BCs
0 0.1 0.2 0.5 1
l/h
0.5733 0.6131 0.7192 1.2256 2.2385 BCs a/h
1.3670 1.4631 1.7178 2.9233 5.3272 0 0.2 0.4 0.6 0.8 1
10 0.5796 0.7205 1.0332 1.4073 1.8051 2.2139
SSSS 1.3680 1.4655 1.7241 2.9512 5.3983
SSSS 20 0.1491 0.1849 0.2646 0.3602 0.4618 0.5663
2.0986 2.2505 2.6516 4.5421 8.3016 30 0.0667 0.0826 0.1182 0.1608 0.2062 0.2528
2.5558 2.7416 3.2277 5.5020 10.0175 10 0.3457 0.4322 0.6226 0.8495 1.0904 1.3378
0.3415 0.3661 0.4313 0.7400 1.3545 SSSF 20 0.0879 0.1098 0.1580 0.2156 0.2768 0.3396
0.7932 0.8502 1.0015 1.7194 3.1494 30 0.0392 0.0489 0.0704 0.0961 0.1234 0.1514
Uniform SSSF 1.1527 1.2350 1.4523 2.4751 4.5104 10 0.2866 0.3574 0.5136 0.6999 0.8978 1.1012
1.6093 1.7261 2.0347 3.4875 6.3747 SFSF 20 0.0727 0.0906 0.1302 0.1774 0.2277 0.2793
1.6857 1.8079 2.1320 3.6652 6.7157 30 0.0324 0.0404 0.0580 0.0791 0.1015 0.1245
0.2831 0.3032 0.3567 0.6100 1.1151
0.4656 0.5004 0.5923 1.0249 1.8823 Table 5. The first dimensionless frequenciy Ω of the square
SFSF 1.0353 1.1108 1.3113 2.2601 4.1465
asymmetric porous microplates with various length-to-thickness
1.0942 1.1720 1.3773 2.3442 4.2689 ratio and scale-to-thickness ratios ( e0 = 0.2 )
1.2925 1.3873 1.6370 2.8074 5.1299
l/h
0.5811 0.6205 0.7257 1.2310 2.2453 BCs a/h
1.3833 1.4786 1.7322 2.9395 5.3548 0 0.2 0.4 0.6 0.8 1
SSSS 10 0.5671 0.7100 1.0245 1.3989 1.7962 2.2040
1.3843 1.4811 1.7388 2.9673 5.4238
SSSS 20 0.1457 0.1821 0.2625 0.3585 0.4604 0.5650
2.1208 2.2720 2.6728 4.5709 8.3548 30 0.0651 0.0813 0.1173 0.1601 0.2056 0.2523
2.5809 2.7661 3.2520 5.5392 10.0925 10 0.3379 0.4258 0.6177 0.8451 1.0860 1.3332
0.3463 0.3707 0.4353 0.7429 1.3577 SSSF 20 0.0858 0.1081 0.1569 0.2147 0.2760 0.3389
0.8037 0.8601 1.0104 1.7270 3.1592 30 0.0383 0.0482 0.0699 0.0957 0.1231 0.1511
Symmetric SSSF 1.1670 1.2485 1.4645 2.4880 4.5318 10 0.2801 0.3521 0.5095 0.6963 0.8943 1.0975
1.6279 1.7439 2.0514 3.5071 6.4087 SFSF 20 0.0710 0.0892 0.1292 0.1767 0.2270 0.2787
1.7053 1.8268 2.1497 3.6849 6.7469 30 0.0316 0.0397 0.0576 0.0787 0.1012 0.1242
0.2871 0.3070 0.3600 0.6124 1.1176
0.4721 0.5065 0.5977 1.0289 1.8867
SFSF 1.0488 1.1236 1.3227 2.2700 4.1590
1.1079 1.1849 1.3890 2.3563 4.2888
1.3082 1.4023 1.6506 2.8221 5.1545
0.5750 0.6148 0.7207 1.2270 2.2405
1.3707 1.4667 1.7212 2.9274 5.3342
SSSS 1.3717 1.4691 1.7275 2.9553 5.4049
2.1038 2.2555 2.6567 4.5491 8.3147
2.5617 2.7474 3.2335 5.5110 10.0355
0.3426 0.3671 0.4323 0.7408 1.3555
0.7956 0.8524 1.0035 1.7214 3.1523
Asymmetric SSSF 1.1559 1.2381 1.4551 2.4784 4.5159
1.6136 1.7302 2.0386 3.4924 6.3833
1.6901 1.8123 2.1361 3.6701 6.7239
0.2840 0.3041 0.3575 0.6107 1.1159 Figure 3. The first dimensionless natural frequency of the SSSS
0.4671 0.5018 0.5936 1.0260 1.8837 rectangular uniform porous microplate with different width-to-
SFSF 1.0383 1.1137 1.3140 2.2627 4.1501 length ratios and porosity coefficient (l/h = 0.1, a/h =10)
1.0973 1.1749 1.3800 2.3473 4.2741
1.2961 1.3908 1.6402 2.8112 5.1363
Table 3. The first dimensionless frequenciy Ω of the square

uniform porous microplates with various length-to-thickness ratio
and scale-to-thickness ratios ( e0 = 0.2 ).
l/h
BCs a/h
0 0.2 0.4 0.6 0.8 1
10 0.5630 0.7062 1.0207 1.3946 1.7912 2.1982
SSSS 20 0.1445 0.1811 0.2616 0.3575 0.4593 0.5638
30 0.0646 0.0809 0.1168 0.1597 0.2052 0.2518
10 0.3354 0.4236 0.6155 0.8427 1.0833 1.3301
SSSF 20 0.0852 0.1075 0.1563 0.2142 0.2754 0.3382
30 0.0380 0.0479 0.0697 0.0955 0.1228 0.1508
10 0.2780 0.3503 0.5078 0.6944 0.8921 1.0950
SFSF 20 0.0704 0.0887 0.1287 0.1762 0.2265 0.2781 Figure 4. The first dimensionless natural frequency of the SSSF rectangular
uniform porous microplate with different width-to-length ratios and
30 0.0314 0.0395 0.0574 0.0785 0.1010 0.1240
porosity coefficient (l/h = 0.1, a/h =10)
466
ACKNOWLEDGMENTS
This work belongs to the project grant No: T2022-152 funded
by Ho chi Minh city university of Technology and Education,
Vietnam.
REFERENCES
[1] D. Chen, S. Kitipornchai, and J. Yang, "Nonlinear free
vibration of shear deformable sandwich beam with a
functionally graded porous core", Thin-Walled Structures,
vol. 107, pp. 39-48, 2016.
[2] E. Magnucka-Blandzi, "Mathematical modelling of a
rectangular sandwich plate with a metal foam core", Journal
of Theoretical and Applied Mechanics, vol. 49, pp. 439-455,
2011.
[3] M. Jabbari, M. Hashemitaheri, A. Mojahedin, and M. Eslami,
"Thermal buckling analysis of functionally graded thin
Figure 5. The first dimensionless natural frequency of the SFSF circular plate made of saturated porous materials", Journal of
rectangular uniform porous microplate with different width-to- Thermal Stresses, vol. 37, pp. 202-220, 2014.
length ratios and porosity coefficient (l/h = 0.1, a/h =10 [4] M. Jabbari, A. Mojahedin, A. Khorshidvand, and M. Eslami,
"Buckling analysis of a functionally graded thin circular plate
Next, some novel results for free vibration of porous metal made of saturated porous materials", Journal of Engineering
foam microplates with materials properties E1 = 200GPa, Mechanics, vol. 140, pp. 287-295, 2014.
1 = 0.33, 1 = 7850kg/m3 are investigated. The influence [5] A. Rezaei, A. Saidi, M. Abrishamdari, and M. P. Mohammadi,
"Natural frequencies of functionally graded plates with
of the scale to thickness ratios and porosity distribution on the porosities via a simple four variable plate theory: an analytical
five first dimensionless natural frequencies approach," Thin-Walled Structures, vol. 120, pp. 366-377,
 = 10 h 1 E1 of the SSSS square porous plates with e0 2017.
= 0.1 and a/h =10 are tabulated in Table 2. It is can be seen [6] M. R. Barati, H. Shahverdi, and A. M. Zenkour, "Electro-
mechanical vibration of smart piezoelectric FG plates with
that from Table 2 the natural frequencies of porous metal
porosities according to a refined four-variable theory",
foam plates increase with an increase of the scale-to-thickness Mechanics of Advanced Materials and Structures, vol. 24, pp.
ratios. Also, the natural frequencies of the symmetric porous 987-998, 2017.
plate are the biggest, while the frequencies of the uniform [7] Y. Q. Wang and J. W. Zu, "Porosity-dependent nonlinear
porous plate are the smallest. Figure 4- Figure 5 show the forced vibration analysis of functionally graded piezoelectric
effect of the porosity coefficient and the width-to-length smart material plates", Smart Materials and Structures, vol.
ratios (b/a) on the first natural frequencies of uniform porous 26, p. 105014, 2017/09/05 2017.
rectangular microplates with different BCs. The length-to- [8] A. Keddouri, L. Hadji, and A. Tounsi, "Static analysis of
thickness ratio and scale-to-thickness ratio are taken by 10 functionally graded sandwich plates with porosities",
and 0.1, respectively. As we see in Figure 4 - Figure 5, the Advances in materials Research, vol. 8, pp. 155-177, 2019.
natural frequency of the porous metal foam microplate [9] A. M. Zenkour, "A quasi-3D refined theory for functionally
increases with an increase in the porosity coefficient and the graded single-layered and sandwich plates with porosities",
length-to-width ratios. Finally, the effect of length-to- Composite Structures, vol. 201, pp. 38-48, 2018.
thickness ratios of the square porous metal foam plates with [10] A. Zenkour, "Quasi-3D refined theory for functionally graded
different scale-to-thickness ratios is shown in Table 3-Table 5. porous plates: displacements and stresses", Physical
A rise in the length-to-thickness ratios makes a decrease in Mesomechanics, vol. 23, pp. 39-53, 2020.
the natural frequency of the porous microplate. [11] F. Ebrahimi, A. Dabbagh, and M. Taheri, "Vibration analysis
of porous metal foam plates rested on viscoelastic substrate",
IV. CONCLUSION Engineering with Computers, pp. 1-13, 2020.
In this article, based on a Moving Kriging meshfree [12] N. Nguyen, J. Lee, and H. Nguyen-Xuan, "Active vibration
method and RPT with two variables were applied to study the control of GPLs-reinforced FG metal foam plates with
free vibration of the porous metal foam microplates with piezoelectric sensor and actuator layers", Composites Part B-
uniform, symmetric and asymmetric porosity distributions. engineering, vol. 172, pp. 769-784, 2019.
The explicit governing equations of the microplates are [13] T. M. Tu, L. K. Hoa, D. X. Hung, and L. T. Hai, "Nonlinear
derived according to the MSGT and Hamilton’s principle. buckling and post-buckling analysis of imperfect porous
The size effect and the influence of the pores, width-to-length plates under mechanical loads", Journal of Sandwich
ratio, length-to-thickness ratio and the boundary conditions Structures & Materials, vol. 22, pp. 1910-1930, 2020.
on the natural frequency of the metal foam microplates have [14] Y. Q. Wang, C. Ye, and J. W. Zu, "Nonlinear vibration of
been presented and discussed. The results of the present study metal foam cylindrical shells reinforced with graphene
show that the length-to-thickness ratio makes a rise in the platelets", Aerospace Science and Technology, vol. 85, pp.
natural frequencies of the porous metal foam plates. Also, the 359-370, 2019.
plate’s stiffness increase with an increase in the porosity [15] P. Toan Thang, T. Nguyen-Thoi, and J. Lee, "Mechanical
coefficient. Besides, numerical results in this paper can be stability of metal foam cylindrical shells with various porosity
considered as benchmark solutions for free vibration analyses of the distributions", Mechanics of Advanced Materials and
porous metal foam microplates. Structures, vol. 27, pp. 295-303, 2020.
467
[16] R. D. Mindlin, "Microstructure in linear elasticity", Columbia [25] L. Gu, "Moving kriging interpolation and element ‐ free
Univ New York Dept of Civil Engineering and Engineering Galerkin method", International Journal for Numerical
Mechanics1963. Methods in Engineering, vol. 56, pp. 1-11, 2003.
[17] D. C. Lam, F. Yang, A. Chong, J. Wang, and P. Tong, [26] T. Q. Bui and M. N. Nguyen, "A moving Kriging
"Experiments and theory in strain gradient elasticity", Journal interpolation-based meshfree method for free vibration
of the Mechanics and Physics of Solids, vol. 51, pp. 1477- analysis of Kirchhoff plates", Computers & Structures, vol.
1508, 2003. 89, pp. 380-394, 2011.
[18] S. Sahmani, M. Bahrami, and R. Ansari, "Nonlinear free [27] T. N. Nguyen, C. H. Thai, H. Nguyen-Xuan, and J. Lee,
vibration analysis of functionally graded third-order shear "Geometrically nonlinear analysis of functionally graded
deformable microbeams based on the modified strain gradient material plates using an improved moving Kriging meshfree
elasticity theory", Composite Structures, vol. 110, pp. 219- method based on a refined plate theory", Composite
230, 2014. Structures, vol. 193, pp. 268-280, 2018.
[19] S. S. Singh, D. K. Nair, A. Rajagopal, P. Pal, and A. K. [28] C. H. Thai, A. Ferreira, J. Lee, and H. Nguyen-Xuan, "An
Pandey, "Dynamic analysis of microbeams based on modified efficient size-dependent computational approach for
strain gradient theory using differential quadrature method", functionally graded isotropic and sandwich microplates based
European Journal of Computational Mechanics, vol. 27, pp. on modified couple stress theory and moving Kriging-based
187-203, 2018. meshfree method", International Journal of Mechanical
[20] A. A. P. Zanoosi, "Size-dependent thermo-mechanical free Sciences, vol. 142, pp. 322-338, 2018.
vibration analysis of functionally graded porous microbeams [29] C. H. Thai, T. Tran, and P. Phung-Van, "A size-dependent
based on modified strain gradient theory", Journal of the moving Kriging meshfree model for deformation and free
Brazilian Society of Mechanical Sciences and Engineering, vibration analysis of functionally graded carbon nanotube-
vol. 42, pp. 1-18, 2020. reinforced composite nanoplates", Engineering Analysis with
[21] B. Zhang, Y. He, D. Liu, L. Shen, and J. Lei, "An efficient Boundary Elements, vol. 115, pp. 52-63, 2020.
size-dependent plate theory for bending, buckling and free [30] M. R. Barati, "Nonlocal-strain gradient forced vibration
vibration analyses of functionally graded microplates resting analysis of metal foam nanoplates with uniform and graded
on elastic foundation", Applied Mathematical Modelling, vol. porosities", Advances in nano research, vol. 5, p. 393, 2017.
39, pp. 3814-3845, 2015. [31] R. P. Shimpi, "Refined plate theory and its variants," AIAA
[22] R. Ansari, R. Gholami, M. F. Shojaei, V. Mohammadi, and S. Journal, vol. 40, pp. 137-146, 2002.
Sahmani, "Bending, buckling and free vibration analysis of [32] C. H. Thai, V. N. Do, and H. Nguyen-Xuan, "An improved
size-dependent functionally graded circular/annular Moving Kriging-based meshfree method for static, dynamic
microplates based on the modified strain gradient elasticity and buckling analyses of functionally graded isotropic and
theory", European Journal of Mechanics-A/Solids, vol. 49, sandwich plates", Engineering Analysis with Boundary
pp. 251-267, 2015. Elements, vol. 64, pp. 122-136, 2016.
[23] A. Jamalpoor, M. Bahreman, and M. Hosseini, "Free [33] C. H. Thai, A. Ferreira, and H. Nguyen-Xuan, "Naturally
transverse vibration analysis of orthotropic multi-viscoelastic stabilized nodal integration meshfree formulations for
microplate system embedded in visco-Pasternak medium via analysis of laminated composite and sandwich plates",
modified strain gradient theory", Journal of Sandwich Composite Structures, vol. 178, pp. 260-276, 2017.
Structures & Materials, vol. 21, pp. 175-210, 2019.
[34] T. Belytschko, Y. Y. Lu, and L. Gu, "Element‐free Galerkin
[24] A. Jamalpoor, A. Ahmadi-Savadkoohi, and S. Hosseini-
methods", International Journal for Numerical Methods in
Hashemi, "Free vibration and biaxial buckling analysis of
Engineering, vol. 37, pp. 229-256, 1994.
magneto-electro-elastic microplate resting on visco-Pasternak
substrate via modified strain gradient theory", Smart [35] S. Thai, H.-T. Thai, T. P. Vo, and V. I. Patel, "Size-dependant
Materials and Structures, vol. 25, p. 105035, 2016. behaviour of functionally graded microplates based on the
modified strain gradient elasticity theory and isogeometric
analysis", Computers & Structures, vol. 190, pp. 219-241,
2017.
468
Impacts of Adverse Weather on Mode Choice

Behavior: A Case Study in Hanoi City, Vietnam
Binh Nguyen Mai Thi Anh Hong Nguyen
Project Implementation Department No.1 Graduate School of Advanced Science and Engineering
Metropolitan Railway Management Board Hiroshima University
Hanoi, Vietnam Higashihiroshima, Japan
nguyenmb.mrb@gmail.com hongnguyen@hiroshima-u.ac.jp
Akimasa Fujiwara Canh Do

Graduate School of Advanced Science and Engineering Graduate School of Advanced Science and Engineering
Hiroshima University Hiroshima University
Higashihiroshima, Japan Higashihiroshima, Japan
afujiw@hiroshima-u.ac.jp canhdo@hiroshima-u.ac.jp
Abstract: Despite the increasingly adverse weather Hanoi suffered a severe heatwave for one week (from May
affecting travel behavior due to climate change, few studies 31st to June 6th) with temperatures up to 42.5°C, which is the
have explored the impacts of adverse weather on mode choice record temperature in history. In addition, the frequency of
behavior in the context of a motorbike-dependent city in a heavy rains and floods in Hanoi has been increasing in recent
developing country in Asia. Here, we investigate factors years. As a result, it can be seen quite clearly that two weather
influencing mode choice behavior with a focus on adverse factors (i.e., high temperatures and extreme rainfall) strongly
weather in Hanoi City (Vietnam). Using data collected from influence people's lives in their daily activities and travel
386 respondents living in Hanoi City in 2018, a mode choice mode choice decisions. This would make them have to change
their travel habits to adapt, such as gradually shifting from
model was developed based on a multinomial logit model to
using motorcycles to using private cars for daily commuting.
understand the impact of adverse weather on the mode choice
In fact, the use of private cars, taxis, and buses increased on
decision of commuters in Hanoi City. Our findings showed rainy days or on the days predicted to be very hot were
that people tend to use cars and urban railways instead of recorded on many major roads in Hanoi.
motorbikes under adverse weather conditions of extreme
temperature and heavy rain. Impacts of adverse weather on commuters vary across
travel modes. Motorbike users are the most sensitive and
Keywords: adverse weather, mode choice behavior, vulnerable to adverse weather among the travel modes.
Hanoi City. Notably, under extreme weather conditions, motorbike drivers
usually face eyes vision, dazzling, heat shock, and difficulties
I. INTRODUCTION in driving, leading to adverse effects on their health and traffic
Like other cities in developing countries in Asia, private safety. Although the bad weather seems less influential on car
vehicles are the main travel modes of residents living in Hanoi drivers than motorbike users, car users often have to suffer
city. In general, the main means of transport within Hanoi city from traffic jams due to the increase in the number of cars on
are motorbikes, followed by cars, buses, and taxis. The roads. For public transport users, despite public bus users still
number of cars has witnessed the most notable change in the being influenced by traffic congestion due to the extreme
past five years since many citizens have purchased cars for weather, the weather seems not to affect those using urban
their daily commuting, and they are considered as a symbol of railways or metro or bus rapid transit with exclusive lanes.
high status in Vietnamese society. The increased number of Almost all studies examining travel mode choice behavior
cars is the leading cause of traffic congestion because roads have focused on conventional factors such as travel time,
and infrastructure in the older parts of Hanoi were not travel cost, frequency, and psychological aspects. Also, global
designed to accommodate them. For public transport, in climate change is still happening complicatedly, affecting
addition to public buses, there is a metro line under many regions and countries in terms of economic
construction and a light rail transit operating in 2021 in Hanoi. development and the daily lives of people. As a result, there
On July 4th, 2017, the Hanoi government voted to ban have been increasing studies focusing on the effects of
motorbikes entirely by 2030 to mitigate traffic congestion, weather on people's travel behavior in recent years.
reduce pollution, and encourage the expansion and use of Specifically, these studies have shown that there are certain
public transport. effects of adverse weather conditions (i.e., extremely high or
Along with cities like Shanghai, Hong Kong, Calcutta, low temperature, heavy rain, snow, and so forth) on the travel
Mumbai, Dhaka, and Jakarta, Hanoi City is also among the behavior of people, especially mode choice behavior.
top most threatened cities in Asia by climate change. In early However, such studies are getting less attention in Vietnam.
June 2017, with being affected by El Niño around the world, In fact, the discovery and measurement of the factors related
to weather conditions affecting the mode choice behavior of

Hanoi people are important and necessary tasks. Once these are demographic and socioeconomic characteristics, vehicle
factors are pointed out, they might be critical factors for ownership characteristics, trip characteristics, environmental
developing and attracting Hanoi residents to use public awareness, attitude and individual vehicle restriction policy
transport modes. [10].
This study aims to explore potential factors that might To sum up, this study aims to examine the impact of
affect people living in Hanoi in choosing their travel modes adverse weather on travel mode choice for commuting trips.
for commuting purposes. Using data collected from 386 To our knowledge, this is the first study exploring weather
respondents living in Hanoi (Vietnam) in 2018, a mode choice impact on mode choice behavior of commuters in Hanoi City,
model is developed based on a multinomial logit model with a typical big city in a developing country suffering from
a focus on the weather factors to understand the impact of climate change in recent years.
adverse weather on the mode choice decision of commuters in
Hanoi City. III. METHODOLOGY
II. LITERATURE REVIEW A. Survey
Some studies examine the impact of adverse weather on The paper-based survey with 300 respondents and the
mode choice behavior. In the study by Khattak and De Palma web-based survey with 86 respondents were conducted from
[1] about the impact of adverse weather conditions on the September 27th to October 26th in 2018 in Hanoi City. The
propensity to change travel decisions, the authors showed that targeted respondents were those over 16 years old and living
in unfavorable weather conditions, the number of people using in Hanoi. The questionnaire was designed into two main parts:
public transport might increase due to the shift from other Revealed Preferences (RP) and Stated Preferences (SP) parts.
types of modes such as cars, bicycles or walking. According The RP part refers to actual and current activities and
to Singhal, Kamga and Yazici [2], weather conditions are information that survey participants take in daily life. This part
treated as exogenous factors that indirectly affect the demand consisted of three sections aimed to collect (1) individual
for transit. Weather is likely to influence commuters' decisions characteristics (i.e., gender, age, occupation, education,
in vehicle selection, route selection, and travel time. monthly income, having a driving license, vehicle ownership,
Specifically, the weather can increase the access time and the traffic conditions), (2) current mode choice behavior (i.e., trip
transfer time between vehicles and could cause schedule length, travel cost, travel time, adverse weather events
changes. In addition, Liu, Susilo and Karlström [3] mentioned influencing their travel decisions such as temperature above
that there is evidence that the weather has a significant 35°C or below 10°C, trip-related information sources such as
influence on travel demand and traffic flow and shapes an weather and traffic), and (3) intention to use urban railway
individual's travel pattern. However, most previous studies services in the future with a focus on weather conditions.
only focused on weather effects on traffic flow and traffic
safety. This leads to a lack of understanding of the impact of
weather on an individual's travel behavior, an important
element in designing proper transportation policies to reduce
the negative effects of adverse weather conditions on transport
infrastructure and make predictions in travel demand more
accurately.
Lots of weather conditions in daily life have been
considered by researchers in previous studies. Kim,
Mahmassani and Dong [4] found that rain significantly
changes car trip schedules, while Winters, Friesen, Koehoorn
and Teschke [5] found that wind significantly influences
cycling. In addition, Stover and McCormack [6] stated that
high winds, cold temperatures, and rain have negative effects
on bus ridership in Washington. Also, Kalkstein, Kuby,
Gerrity and Clancy [7] showed the impact of air mass on rail
transit in three US cities that a number of rail ridership on dry Figure 1. An example of traffic scenarios in the SP part
and comfortable days is significantly higher than on cold and
humid days. According to Sabir, Koetse and Rietveld [8], The SP part refers to hypothetical situations with some
people in the Netherlands switched from cycling to private pre-established conditions. It is commonly used to research
cars and public transport on shallow temperatures and rainy new products or services that have not appeared on the market
days. They prefer walking and cycling when the temperature yet. The SP part in this study was built to focus on adverse
increases. weather conditions considered in commuters' travel mode
choices. The respondents were randomly classified into two
There have been a few studies related to mode choice groups in the SP part. The 189 respondents belonged to the
behavior with a case study in Vietnam. According to Le [9], group commuting under normal weather, and 197 respondents
the five factors that influence the intention of the people in Da were in the group commuting under unfavorable weather
Nang City to use the rapid bus system are awareness of conditions. Each respondent was asked to answer four
usefulness, awareness of the ease of use, subjective norm, scenarios designed with origins at homes and destinations at
habits of using private vehicles, and quality of service. Also, workplaces. More specifically, both origin and destination
in the analysis of mode choice behavior of urban citizens in were assumed to be within 500m from the nearest urban
Can Tho City and Ho Chi Minh City, the mode choice railway station or a ten-minute walk from home/workplace to
behavior of an individual can be explained by five factors that
470
stations (see Fig. 1). Based on attributes of travel mode that their mode choice behavior was always affected by
choices (i.e., travel cost, travel time, frequency, station weather conditions.
facilities, and weather conditions, including temperature, rain,
and road disruption), the respondents made their decisions on The weather conditions affecting the respondents' mode
the most preferred means of transportation among the three choice decisions are shown in Fig. 4. Generally speaking,
types of pre-designated travel modes (i.e., car, motorbike, and most respondents could change their mode choices under the
urban railway) in each scenario. extreme weather conditions of temperature and/or heavy rain.
Specifically, approximately 6% (22 people) said that their
B. Data transportation mode choice was only affected by outdoor
The total respondent was 386, of whom 61.66% were temperature. In contrast, around one-fourth (99 people)
male, 1.29% were under 18 years old, 68.65% were from 19 answered that their mode choices were only affected during
to 30 years old, and 27.72% were in their 30s and 40s. For heavy rainy days. In addition, around 60% (230 people) could
driving licenses, around 90% had a driving license. For the change their mode choices under the extreme weather
occupation, approximately 67% were office staff and conditions of temperature and heavy rain.
university students who had relatively stable travel patterns in 230
250
terms of time and route, and means of transport that they used
for commuting. For the individual monthly income, 28.42% 200
were low income (i.e., below 3 million VND), 51.31% were 150 99
middle income (i.e., from 3 to 6 million VND), which is
100
reasonable because around two-thirds of the respondents were 34
students and office workers. In terms of the number of private 50 22
1
vehicles per household, on average, each household owned 0
two or three motorbikes.
324
350
300
250
200
150
100 19 8 10 14
50 1 5 1 4
0
Figure 4. Weather conditions affecting mode choice behavior
Each respondent answered the four hypothetical scenarios

related to travel mode choices with attributes focusing on
adverse weather conditions in the SP part as mentioned above.
Figure 2. Modal share Based on the SP data, the modal share shows that the highest
mode choice belonged to urban railway (54.53%), followed
Fig. 2 shows the modal share of respondents in this study. by motorbikes (32.32%) and cars (13.15%).
In general, approximately 85% of respondents used
motorbikes as the main travel mode for commuting. The C. Modeling Mode Choice
percentage of the respondents who used public buses was less In order to examine the impact of adverse weather on
than 4%—noted that public bus was the only type of public travel mode choice under the control of other influential
transport at the survey time in Hanoi City. factors, a mode choice model is developed based on a
multinomial logit model by using the SP data with a focus on
200 181 commuting trips. The dependent variable is mode choice
150 (urban railway, car, and motorbike).
100
96 The utility Unj that an individual n (n = 1, 2, …, N)
61 associates with a travel mode j (j = 1, 2, 3) as follows,
50 39
9 𝑈𝑛𝑗 = 𝛼𝑗 + 𝛽𝑗 𝑥nj + 𝜀𝑛𝑗 (1)
0
Never Seldom Sometime Often Almost Where αj is a constant term for travel mode j, βj is a vector
always of parameters associated with a vector of explanatory
variables xnj related to socio-demographic characteristics and
Figure 3. Influence of weather on mode choice behavior weather factors, and εnj is an error term that is Gumbel
distributed.
Fig. 3 shows the influence of weather on mode choice Conditional on ηnj, the probability that individual n
behavior. Less than one-fifth of respondents answered that chooses travel mode i follows the standard logit formulation:
their mode choice decisions were unaffected by weather. This
figure could be explained by having no other mode 𝑒 𝛽𝑖 𝑥ni
𝐿𝑛𝑖 (𝛽𝑗 |𝜂𝑛𝑗 ) = 𝛽 𝑥 (2)
alternatives or their preferences. Also, the degree of the ∑𝑗 𝑒 𝑗 nj
influence of weather conditions on individuals is different:
181 people (46.89%) thought to be influenced but not often; The unconditional probability is the integral of the
96 people (24.87%) confirmed that they were often affected conditional probability over all possible values of ηnj:
by weather factors; the remaining 39 people (10.1%) affirmed
471
𝑃𝑛𝑖 = ∫ 𝐿𝑛𝑖 (𝛽𝑗 |𝜂𝑛𝑗 ) 𝑓(𝜂𝑛𝑗 )𝑑𝜂𝑛𝑗 (3) Variables Estimate t-value
Age from 19 to 30 years old
The model estimation is done by using R with package (1: yes; 0: otherwise)
maxLik. It is hypothesized that people tend to use urban Cars -0.04 -0.05
railways and cars under extreme weather conditions. It should Urban railways 0.74 1.62
be noted that conventional factors related to attributes of mode Age from 31 to 50 years old
choice (i.e., travel cost, travel time, frequency, level of (1: yes; 0: otherwise)
service) are not used for the modeling because this study Cars -0.08 -0.10
focuses on the impact of adverse weather on travel mode Urban railways 0.99 2.11*
choice. Age over 50 years old
(1: yes; 0: otherwise)
IV. RESULTS AND DISCUSSIONS Cars 0.07 0.07
Urban railways 1.79 2.82**
The estimation results are shown in Table 1. The adverse
Individual monthly income
weather impacts are captured by extreme temperatures and (1: > 10 million VND; 0: otherwise)
heavy rain variables and noted that motorbikes are selected as Cars 0.54 2.59**
a reference choice among the three travel mode choices in the Urban railways -0.04 -0.24
model. Number of motorbikes per household
Cars -0.02 -0.17
Concerning the weather factors, almost all weather-related
Urban railways 0.09 1.31
variables in the model are statistically significant with
Receiving weather information
expected signs. Regular reception of weather information has (1: at least one per day; 0: otherwise)
a positive impact on using urban railways. Based on this Cars 0.01 0.06
finding, the local government may need to consider installing Urban railways 0.39 3.19**
electronic bulletin boards with weather information in public Extreme temperature
areas to increase getting the weather-related information to (1: very hot/cold; 0: otherwise)
citizens to raise urban railway usage. In addition, when the Cars 1.69 8.95***
outdoor temperature is unfavorable (above 35°C or below Urban railway 1.01 8.24***
10°C), Hanoi residents tend to use cars and the urban railway Heavy rain
instead of riding motorbikes. Having statistical evidence of the (1: heavy rain; 0: otherwise)
Cars 1.21 6.71***
effect of extreme temperature on an individual's travel
Urban railways 0.67 5.59***
behavior can help planners determine trends and predict the
Sample size N = 1544
travel demand in different seasons of a year, especially in
Initial log-likelihood -1696.26
summer and winter when the outdoor temperature is very high
Final log-likelihood -1384.29
or very low. Like the results for temperature variables, in the
McFadden's rho square 0.18
context of heavy rain, people also prefer to commute by car
Adjusted McFadden's Rho square 0.17
and urban railway than motorbikes on heavy rainy days. There
Notes: *** p < 0.001; ** p < 0.01; * p < 0.05
are two possible interpretations for this result: (1) riding a
motorbike in heavy rain is considered dangerous and risky; (2) V. CONCLUSION
as the underground drainage system in Hanoi is not good in
many places, local floods in those places often occur under the Commuting by motorbike is still dominant in Hanoi City
heavy rain weather, obstructing motorbike users most. In as well as other big cities in developing countries in Asia. Due
brief, both extreme weather variables have statistically to climate change, citizens living in the big cities suffer from
significant impacts on travel mode choice. Based on the adverse weather conditions such as extreme temperature,
hypothetical scenarios designed in this study, Hanoi citizens heavy rain, air mess, strong wind, and so forth. There is a
tend to use cars and urban railway under the weather of critical question of whether current citizens' travel mode
extreme temperature and rain. However, with the current choices will be changed under the adverse weather, resulting
public transport system in Hanoi City, especially having no in the importance of investigating the impacts of adverse
choice of urban railway, the citizens may tend to use cars weather on travel mode choice. This study contributes to the
instead of urban railways and motorbikes since the adverse growing literature on the influence of adverse weather on the
weather is becoming intense soon due to the climate change. mode choice behavior of commuters, especially in Hanoi City
Therefore, it is necessary to speed up the implementation of - a typical big city in a developing country suffering from
urban railway projects to create an urban railway network in climate change in recent years. Pointing out weather-related
Hanoi City as soon as possible. impacts on mode choices could contribute to policymakers
setting appropriate and necessary policies and strategies for
Table 1. Estimation result of the multinomial logit model decreasing private vehicles and increasing public transport
usage in Hanoi City in the future.
Variables Estimate t-value
This study aims to investigate potential factors affecting
Constants
the travel mode choice of people living in Hanoi City for
Cars -2.52 -2.94** commuting trips. A mode choice model is developed based on
-1.43 -2.88** a multinomial logit model with a focus on the adverse weather
Urban railways
Gender factors through surveys conducted from 386 respondents
(1: male; 0: otherwise) living in Hanoi City in 2018. This is initial evidence that the
0.13 0.70 mode choice behavior of Hanoi citizens is being affected by
Cars
adverse weather conditions.
Urban railways 0.00 0.02
472
The model estimation results showed that people who are [2] A. Singhal, C. Kamga, and A. Yazici, "Impact of Weather on
getting older prefer using urban railways to using motorbikes, Urban Transit Ridership", Transportation research part A:
implying that they are potential urban railway users. Also, the policy and practice, 2014, vol. 69, pp. 379-391.
high-income people tend to choose cars instead of motorbikes, [3] C. Liu, Y. O. Susilo, and A. Karlström, "The Influence of
which indicates a necessity of having policies to control the Weather Characteristics Variability on Individual’s Travel
increase in using cars in growing high-income people in Mode Choice in Different Seasons and Regions in Sweden",
Hanoi. In addition, people receiving daily weather Transport Policy, 2015, vol. 41, pp. 147-158.
[4] J. Kim, H. S. Mahmassani, and J. Dong, "Likelihood and
information prefer to use urban railways, indicating the
Duration of Flow Breakdown: Modeling the Effect of Weather"
importance of providing weather information in public areas
(in, Transportation research record, 2010, vol. 2188, (1), pp.
to increase urban railway ridership. Furthermore, extreme 19-28.
temperature and heavy rain had significantly positive impacts [5] M. Winters, M. C. Friesen, M. Koehoorn, and K. Teschke,
on using cars and urban railways. Based on the current context "Utilitarian Bicycling: A Multilevel Analysis of Climate and
of slow urban railway development, people living in Hanoi Personal Influences", American journal of preventive
would choose to commute by car instead of urban railways medicine, 2007, vol. 32, (1), pp. 52-58.
under intensely adverse weather. Therefore, it is necessary to [6] V. W. Stover, and E. D. McCormack, "The Impact of Weather
speed up the implementation of urban railway projects in on Bus Ridership in Pierce County, Washington", Journal of
Hanoi City. Public Transportation, 2012, vol. 15, (1), pp. 6.
[7] A. J. Kalkstein, M. Kuby, D. Gerrity, and J. J. Clancy, "An
It is important to note that this study has some limitations. Analysis of Air Mass Effects on Rail Ridership in Three Us
First, fixed and random effects coupled with heterogeneity Cities", Journal of transport geography, 2009, vol. 17, (3), pp.
among individuals may exist as the multinomial logit model 198-207.
was developed based on the SP data in which each respondent [8] M. Sabir, M. Koetse, and P. Rietveld, "The Impact of Weather
answered the four scenarios. Second, two conventional factors Conditions on Mode Choice: Empirical Evidence for the
- travel cost and travel time - were not addressed in the mode Netherlands": "Proceedings of the Bivec-Gibet Transport
choice model. Despite these limitations, a potential expansion Research Day 2007" (2007), pp. 512-527.
of this study is a comparison of the effects of adverse weather [9] T. K. N. Le, "Influenced Factors of Mode Choice Behavior on
factors between groups of car ownership and motorbike Intended Use of Bus Rapid Transit (Brt) System in Da Nang,
ownership. Vietnam" (in Vietnamese), University of Economics of Ho Chi
Minh City, 2013.
ACKNOWLEDGMENT [10] T. H. Nguyen, and Q. T. Ho, "Travel Behavior on Travel Mode
This research was conducted under the Project for Human Choice in Vietnam's Urban Area: A Case Study in Ho Chi Minh
and Can Tho" (in Vietnamese), Journal of Economics and
Resource Development Scholarship by Japanese Grant Aid
Development, 2015, vol. 17, (1), pp. 51-58.
(JDS) (former Japanese Grant Aid for Human Resource
Development Scholarship (JDS).
REFERENCES
[1] A. J. Khattak, and A. De Palma, "The Impact of Adverse
Weather Conditions on the Propensity to Change Travel
Decisions: A Survey of Brussels Commuters", Transportation
Research Part A: Policy and Practice, 1997, vol. 31, (3), pp.
181-203.
473
Travel Behavior on Ground Access Mode Choices

by Introducing a New City Air Terminal: A Case
Study of Vientiane Airport, Laos
Komack Keochampa Canh Do

Department of Civil Aviation Graduate School of Advanced Science and Engineering
Ministry of Public Works and Transport Hiroshima University
Vientiane, Lao PDR Higashihiroshima, Japan
okidkcp@gmail.com ORCID [0000-0003-1316-3012]
Akimasa Fujiwara Hong Nguyen

Graduate School of Advanced Science and Engineering Graduate School of Advanced Science and Engineering
Hiroshima University Hiroshima University
Higashihiroshima, Japan Higashihiroshima, Japan
ORCID [0000-0002-3354-0631] ORCID [0000-0001-8323-2082]
Abstract: Vientiane Airport - an official named Wattay I. INTRODUCTION

International Airport (WIA) - is Laos's major hub airport. It
Toward transforming the Lao People's Democratic
seems like a first gateway and is crucial to the country's Republic (Lao PDR or Laos) from a land-locked country to a
economic activities and tourism. The airport has handled land link that allows domestic and international firms to
three-fourths of annual air passengers in the whole country. access global and regional value chains, the Lao Government
Its annual air passengers tend to increase rapidly, and has had to undertake bold policy reforms. These reforms will
consequently, traffic congestion will become serious around be necessary to make the country more attractive as a new
the airport. In order to improve airport ground access to the investment destination and link it to major production and
WIA, this research investigates factors influencing airport consumption areas in the Association of Southeast Asian
mode choices in accessing the WIA by introducing the City Nations (ASEAN) and the world. Lao PDR could develop
Air Terminal in Vientiane (VCAT). This terminal will into a logistics hub with efficient logistics services, while
connect the WIA and Vientiane CBD by a non-stop targeted investments in agriculture and tourism could result
limousine bus service, making passengers can do the check- in new export opportunities. Among these reforms,
in process inside the CBD preliminary. It may contribute to improving transportation connectivity plays a vital role,
reducing the number of private vehicles accessing WIA and especially in the air transport sector, stimulating trading,
increasing the convenience of airport passengers. A revealed tourism, and economic growth.
and stated preference survey was exclusively conducted Recently, air passenger demands have vastly increased
with 600 respondents at the airport in October 2019. A according to the national economy and have gradually grown
multinomial logit (MNL) model was calibrated to to about 7% annually [1]. WIA is one of the largest airports
investigate access mode choice behavior on the new airport in Laos, located northwest of Vientiane – the capital of Laos.
service. The results indicate that airport access mode choice It seems like a first gateway and is crucial to the country's
is significantly affected by travel cost, in-vehicle travel time, economic activities and tourism. Moreover, WIA is a major
delay time, waiting time, and parking fee. The socio- hub airport in Laos, operating international and domestic
demographic and trip characteristics are also crucial flights. It has covered around three-fourths of the annual air
attributes. Especially, dependency on others and weather passengers in Laos, and the number of annual air passengers
conditions are significant factors affecting airport access tends to increase rapidly. Statistically, in 2018, WIA handled
mode choice to WIA. Nine scenarios were revealed to show almost 2.3 million air passengers, and it would reach up to
how improvements in travel time, in-vehicle travel cost, and 6.4 million air passengers in 2040 [2]. This increase in air
waiting time associated with public transportation reduce travel demand in WIA has led to escalated surface traffic
the modal share of private cars. The research findings volumes because air travelers have to reach the WIA with
various ground transportation modes. Consequently,
provide policymakers insight into improving the planning of
congestion on access routes, overcrowded car parking
airport access to WIA.
facilities, and overcrowded check-in counters would be the
main adverse affected objectives.
Keywords: airport ground access mode choices, Wattay
international airport, City Air Terminal, Vientiane CBD, To solve the dilemmas mentioned above, the City Air
value of time (VOT). Terminal in the Vientiane capital (VCAT) was introduced to
deal with the rapid increase of air passengers and reduce

congestion at WIA. VCAT has missions to share the burden results. The last section concludes the research presented in
of overcrowded air passengers and reduce road congestion the paper and future research work.
on access routes to WIA. VCAT was designed as a new
modern public transportation system, which serves air II. LITERATURE REVIEW
passengers to access and egress from WIA by using airport For more than four decades, numerous researchers have
limousine buses with a specific lane for traveling. VCAT attempted to understand the behaviors of air passengers
was expected to be located in the vicinity of Vientiane. related to airport access mode choices. A better
Moreover, VCAT would offer air passengers many benefits, understanding of airport access mode behavior is
such as a smooth and continuous flight check-in and considerable to airport administration and management [4].
immigration clearance process at VCAT's air terminal before
departure at WIA. Regarding factors that affected airport access mode, one
of the preliminary research studies in this area was feasibly
Basic concepts of the VCAT were expected as follows: studied by Ellis, Bennett and Rassam [5]. The author aimed
to examine air travelers' access mode choice by developing
• VCAT's airport limousine buses would offer faster,
statistical models at the Baltimore-Washington airport.
safer, punctual, and more comfortable services to
Results found that trip purpose, travel time, and travel cost
WIA from the Vientiane city areas. This would ease
were essential factors in which travel time and costs became
airport users' worries about delays for their departure
two main concerns of air passengers when choosing a travel
flights.
mode to access the airport. Another study by Harvey [6]
• VCAT's airport limousine buses would arrive near or investigated air passengers' access mode choice to San
in front of the airport departure gates, so air Francisco Bay Area airports. The author grouped air
passengers and those who have not traveled by air passengers into different groups of business and non-
before can easily access the departure areas or airline business passengers. The study showed that travel cost and
check-in counters. time are sensitive factors; in particular, business passengers
were more sensitive to travel time than non-business
• Airport users would be able to proceed with their passengers and vice versa. The result also stated that the
airline check-in and immigration process at the number of carrying luggage was one of the critical factors
VCAT, and they will be free from their heavy affecting airport access mode choice. Clark and Lam [7] and
luggage in traveling to the airport (Figure 1). Pels, Nijkamp and Rietveld [8] reported that trip purpose,
• Air passengers would easily access the air terminal of travel cost, origin, residential area, and party size
VCAT, which is conveniently located near or around significantly affect airport access mode choice. Keumi and
the city center of Vientiane. Murakami [9] exposed that travel time, travel cost, waiting
time, delay cost, and frequency of services affected airport
access mode.
On the other hand, the air passengers' primary objective
of ground travel was on-time arrival at the airport because
they care about reliable travel time to ensure their on-time
arrival and not lose their flights. Besides, comfort, ease of
accessibility, luxury, and safety are services that air
passengers prefer in their journey to the airport; as a result,
Figure 1. Concepts of the City Air Terminal in Vientiane, Laos. travel costs might be variance based on the amenities that the
Source: Adapted from the International Aviation Transport air traveler requires most. In addition, air passengers were
Association (IATA) website [3] more willing to pay in the morning than in the afternoon.
Mamdoohi, Saffarzadeh, Taherpour and Yazdanpanah [10]
Therefore, this study analyzed and treated the VCAT as a investigated airport access mode at Imam Khomeini
new modern transportation mode and effectively enhanced International Airport, Iran. They found significant travel
its use as the primary access and egress mode choice to WIA. time, trip purposes, monthly household income, and private
Understanding air passengers' travel behaviors of airport vehicle ownership. Moreover, business air travelers were
access mode choice at WIA would be valuable to effectively likely to use private cars rather than non-business air
design a link between the WIA and the city and better travelers to access the airport.
manage road traffic of airport access. The research findings
would contribute profound insights into air passengers' The safety margin became one of the fundamental
behavior in airport ground access mode choice by variables affecting airport access mode choice, particularly
introducing the VCAT to the government of Laos and the for air passengers with business trip purposes. Air passengers
WIA authority. This would help develop and improve the usually arrive earlier at the airport to guarantee no late arrival
accessibility of airport access mode to the WIA in the future. and missing their flight. Hence, the safety margin was
defined as the time difference between air travelers' preferred
This paper is divided into five sections. The following and expected arrival times [11]. A survey was performed to
section shows a literature review about airport ground access investigate the magnitude of the safety margin on airport
mode choice behaviors. The survey design and descriptive access mode choice at Hong Kong International Airport
analysis are described in the third section. The fourth section (HKIA). The result disclosed that various airport access
summarizes the ground access mode choice model and its modes had differences in safety margin measures, and that is
because multiple airport access modes have multiple travel
The Project for Human Resource Development Scholarship by Japanese time reliability. Then again, Koster, Kroes and Verhoef [12]
Grant Aid (JDS) financially supported this research. evaluated the variability in estimates of the cost of access
475
time for Dutch air passengers. They discovered that business significant variables across two domestic airports. The
and non-business travelers are similar in terms of costs research also found that air travelers' socio-economic status
allocated for access time. They also observed that the costs significantly affects access mode choice for non-business
of access travel time vary between 0% and 30% of total rather than business trips.
access travel costs for business travelers and non-business
travelers between 0% and 25%. Regarding travel luggage and party size factors when
traveling by air, Akar [18] investigated the airport access
Besides travel cost and travel time, several significant mode choices at Port Columbus International Airport, Ohio
factors were measured and examined to understand airport (USA). Research findings revealed that the number of
access mode for air passengers. The first example is a study baggage encourages air passengers to descent utilizing public
by Psaraki and Abacoumkin [13]. They grouped air transit access to the airport. Moreover, Tam, Lam and Lo
passengers into socio-economic and trip characteristics (e.g., [11] and Akar [18] mentioned the party size of traveling.
business and non-business trip purposes) to investigate They stated that the larger the traveling party size, the less
airport access mode choice at the Athens International likely they will choose public transit and use private
Airport in Greece. They aggregated that 52.3% and 45.9% of automobiles to access the airport. Besides that, Chang [19]
air passengers accessed the airport using private cars and examined factors affecting airport access mode choice for
taxis, which private vehicles were utilized to drop off more elderly air travelers at Taipei International Airport, Taiwan.
than parking at the airport. The results expressed that Athens The results showed that elderly air travelers requested their
residents with domestic business trips were likelier to drive family members to drive them to the airport. Moreover,
and park their cars at the airport. In contrast, air travelers safety, user-friendly, and convenience for storing luggage
with international non-business flights were likelier to be were most important in choosing mode access. In contrast,
dropped off at the airport or use taxi cabs. car ownership in a household and the number of luggage
were significant variables that increased the preference for
The second example is Jou, Hensher and Hsu [14] 's using private car access to the airport.
study. They performed a survey when a mass rapid transit
system was constructed to link the Taoyuan International With the intention to reduce the use of private cars to
Airport (TIA) to the Taipei train station and High-Speed Rail access the airport, public transit has been promoted and
station in Taiwan. The authors added three additional socio- examined as the primary mode of airport access [16, 20-22].
economic variables (i.e., job, education, and the number of Budd, Ison and Ryley [20] suggested replacing the number
motorcycles). The results showed that students with lower of private cars by promoting public transit with the assistance
monthly personal incomes tended to ask their families or of airport administrators. Gupta, Vovsha and Donnelly [16]
friends to drive them off at the airport. Besides that, summarized that, in New York City, the AirTrain was likely
accompanying travelers, airline category class, the day away, to link between rail stations and airports. Gokasar and Gunay
and the frequency of air travel were also included in the [21] stated that the number of public transit to access the
study. They revealed that most business passengers with a airport would increase if the journey origin is located near or
higher frequency of traveling by air were more likely to use around the influence zones of the public transport in
private cars to the airport because their trip length is not far. traveling to Atatürk International Airport in Turkey. Zaidan
On the other hand, the lower frequency of travel by air was and Abulibdeh [22] organized a survey by introducing Doha
more likely to be dropped off and picked up at the airport. Metro at the Hamad International Airport (HIA) in Qatar.
However, they showed that total and in-vehicle travel time The result revealed that airport users' socio-economic and
are essential variables influencing the access mode choice to trip characteristics significantly affected the current and
TIA. future access mode choice to HIA.
The third example is a study by Alhussein [15]. The In summary, much previous research has presented
author performed a survey at the King Khaled International crucial and consistent insight into the factors influencing air
Airport (KKIA) in Saudi Arabia. The study aggregated that travelers' airport access mode choice, mainly travel cost and
55.5% of air passengers used private cars, and taxi and travel time. Compared with the previous research, this study
limousine passengers dominated more than 40%. Only 2.3% emphasizes a situation where a new service is proposed and
of air passengers chose the Saudi Public Transit Company assumed in future scenarios during the survey. Psaraki and
(SAPTCO) to get to KKIA. Research findings showed that Abacoumkin [13] stated: "The access conditions observed at
income, luggage, travel access time, and nationality any airport are unique and depend on the nature and volume
significantly affected mode choice behavior in accessing of airport traffic; the location and geographical setting of the
KKIA. airport and the urban areas it serves, and the economic, social
and political structure of the country in which it is situated."
The fourth study by Gupta, Vovsha and Donnelly [16] Since no studies regarding airport access mode choice in
investigated airport access mode choice in New York City Vientiane have been performed previously, this research
and found the socio-economic characteristics that affected would be a valuable contribution to the Laos Government
the airport access mode choices, especially age, gender, and and the WIA authority to improve and develop airport mode
income level, were significant. accessibility to WIA. Notably, these research findings would
The fifth study by Choo, You and Lee [17] examined also provide a good insight into policy decision-making by
Gimpo Airport and Daegu Airport as two domestic airports airport management and operations and transport authorities
in South Korea. They analyzed three types of airport access in Vientiane, Laos.
mode choices (i.e., auto and public transit, auto and bus, and
auto and subway) and considered some explanatory variables III. SURVEY AND DATA
in the model. The results showed that travel time, travel A. Survey Design
distance, trip purpose, age, gender, job, and income were
476
This research investigates factors influencing airport The experimental attributes of VCAT have considered
mode choice in accessing the WIA by introducing the VCAT some travel mode attributes, such as the travel cost, in-
in the future. A paper-based survey was designed with vehicle travel time, delay time, access time, and waiting
revealed preference (RP) and stated preference (SP) parts time. The orthogonal fractional technique, a common
and conducted at WIA. The RP part asked air passengers approach, was used to construct the hypothetical scenarios in
about their current access mode choice. In contrast, the SP the stated preference (SP) approach to present to the survey
question examined the probability of using the VCAT as respondents. In order to avoid multicollinearity among
their primary mode in accessing the WIA if the service attributes, the SP situations were designed to ensure that the
launches in the future. The hypothetical situations of the SP defined attributes could express themselves virtually
used in the research based on research purposes, in which independently of each other. Table 1 presents travel mode
procedures of its experimental techniques comply with the attributes and levels for the SP choice experiment.
instruction of Sanko [23], stated guidelines for SP
experiment design. The questionnaire consisted of three B. Questionnaire Data and Descriptive Analysis
sections: (1) socio-demographic characteristics, including The questionnaire was designed in a sampling manner
nationality, age, gender, education, monthly household and translated into Lao, English, and Thai. The survey used a
income, and household car ownership; (2) trip random systematic survey to collect data from departing
characteristics, including the trip purpose, check-in luggage, passengers at the WIA. It was conducted over one month in
accompanying traveler, dependency on others, and weather 2019 (i.e., October 09 to November 15), including weekdays
conditions; and (3) the SP access mode choice questions and weekends at the departure area of the airport terminal
which were designed regarding accessing the WIA from the buildings (both international and domestic). Air passengers
Vientiane, consisted of two parts, proposed five alternatives, were interviewed based on a systematic approach: the
such as motorbike, private car, taxi, bus, and VCAT, in interviewer randomly selected and gave the questionnaire to
which VCAT mode is currently not available (only offered in the first person sitting in the first row or column of seats in
the future) at the WIA; therefore, air passengers faced these the boarding waiting zone. Then, the next interviewee was
alternatives as hypothetical situations. It should be noted that chosen, and so forth. The following person was asked if any
each respondent answered two scenarios in the SP part. air passenger denied participating in this survey. Finally, six
hundred questionnaires were completed; after a validating
Table 1. Attribute levels of travel modes in the SP part process, 521 completed surveys from international and
Alternatives Travel mode attributes Level of services domestic air passengers were utilized for further analyses.
Motorbikes Travel cost • 16,500 LAK (fixed)
Parking fee • 5,000 LAK (fixed) Table 2. Descriptive analysis of socio-demographic characteristics
In-vehicle travel time • 20 min (fixed)
Delay time • On-time arrival (fixed) Contents Sample Contents Sample
Weather conditions • Rainy size (%) size (%)
• Sunny Age Gender
Private cars Travel cost • 42,000 LAK (fixed) ≤ 20 9.60% Male 50.10%
Parking fee • 10,000 LAK (fixed) 21-30 42.80% Female 49.90%
31-40 28.60% Education
In-vehicle travel time • 25 min
41-50 7.68% High school or under 11.32%
Delay time • On-time arrival
51-60 5.76% College 17.47%
• 10 min delayed
> 60 5.57% Bachelor 46.26%
Weather conditions • Rainy Nationality Master/Doctor 24.95%
• Sunny Vientiane citizens 41.84% Monthly household income
Taxis Travel cost • 66,000 LAK (fixed) (LAK)
In-vehicle travel time • 25 min (fixed) Non-Vientiane citizens 12.67% < 1.3mil 4.80%
Delay time • On-time arrival Foreigners 45.49% [1.3mil – 2mil) 9.98%
• 10 min delayed Marital status [2mil – 4mil) 17.66%
Weather conditions • Rainy Single 54.89% [4mil – 7mil) 13.82%
• Sunny Married 45.11% [7mil – 10mil) 8.64%
Buses Travel cost • 19,000 LAK (fixed) Occupation [10mil – 15mil) 8.86%
In-vehicle travel time • 35 min (fixed) Government officers 27.26% [15mil – 20mil) 5.95%
Delay time • On-time arrival Businesspersons 7.29% > 20mil 30.33%
• 10 min delayed Private firm employees 22.46% Driver license
Waiting time • 20 min (fixed) Salespersons 3.45% Motorbike 15.16%
Access time • 10 min Self-employed persons 3.07% Car 36.47%
• 15 min Students/Pupils 19.96% Both 32.44%
Weather conditions • Rainy Professors/Teachers 7.87% No license 15.93%
• Sunny Housekeepers 0.38% Household car ownership
VCAT Travel cost • 29,000 LAK Doctors/Nurses 2.30% Yes 75.05%
• 34,000 LAK Retired/Unemployed 4.80% No 18.43%
In-vehicle travel time • 18 min Other 1.15% Share 6.53%
• 23 min
Delay time • On-time arrival (fixed) Table 2 provides an overview of the survey data, which
Waiting time • 15 min shows that the most significant percentage of air passengers
• 20 min were Lao passengers (54.51%), of which Vientiane residents
Access time • 10 min dominated 41.84%. Air passengers (71.40%) were mainly
• 20 min between 21 and 40 years old, 50.10% were males, and
Weather conditions • Rainy 45.11% were married. In terms of education, the most
• Sunny
477
significant percentage of them (46.26%) had a bachelor's check-in luggage, followed by two pieces (12.09%), more
degree, followed by a master's and doctor's degree (24.95%), than two pieces (6.25%), and only one piece of carry-on
a college graduate (11.66%), and a high school or under luggage (67.56%). The data showed that single air passenger
(11.32%). Most air passengers had a vehicle driver's license or one accompanying traveler (60.46%), sole air passenger or
(84.07%) and owned a car in a household (81.58%). The parties of two persons (64.10%), for an average of less than
sample showed that the most dominant occupation was one week of trip duration (40.50%), and frequency of once or
government officers (27.26%), followed by private firm twice per year (59.88%), as shown in Table 3.
employees (22.46%) and students (19.96%). Over 72% of air
passengers revealed a monthly household income of 7 IV. MODEL ESTIMATION
million LAK or over (i.e., 1 US dollar (USD) is equal to C. Multinomial Logit Model (MNL)
11,865 LAK [24]).
The discrete choice model is a favorite method widely
Table 3. Descriptive analysis of trip characteristics applied in behavioral analysis. It can be utilized to analyze
and predict the individuals' choice from a choice set with
Sample Sample various alternatives. In this case research, the individuals are
Contents Contents
size (%) size (%)
air passengers who would like to access the WIA. The
Airline category Trip origin alternatives applied in the choice set are motorbike, private
National airlines 32.44% Home 51.06% car, taxi, bus, and VCAT as given modes mentioned above.
International airlines 67.56% Hotel 36.28%
Ben-Akiva and Lerman [25] stated that well-established
Airline class Work/office 9.60% approaches when the utility maximization principle rules
Business 4.99% Other 3.07% travel mode choice. The probability of an air traveler
Economy 94.43% Trip duration (days) choosing a private mode to travel to the WIA relies on utility
Other 0.58% <3 8.64%
benefits or how much utility they have when traveling by the
mode. In other words, the alternative in the choice set, which
Accompanying travelers 3–7 31.86% proposes the highest utility, will be selected. It includes
0 27.45% 8–15 17.27% individual characteristics of air travelers, trip characteristic
1 33.01% 16–30 14.20% attributes, and alternative attributes. The equation of the
2 13.05% ≥ 30 28.02% utility Uin of the alternative i for an individual n as the
following form:
3 7.87% Travel frequency (per year)
≥4 18.62% 1 35.12%
Party size 2 24.76% ()
1 29.17% 3 11.52%
2 34.93% 4 6.72% where Vin is the observable component of utility estimated by
3 9.79% ≥5 21.88% the analyst; in is the unknown component of utility;β is the
vector of parameters to be estimated; Xin is an explanatory
≥4 26.10% Airport access mode
variable for the individual n from the alternative i.
Carry-on luggage Motorbike-drop off 2.88%
0 4.99% Motorbike-parked 2.11%
Aforesaid above, an individual n selects an alternative i
in the choice set solely when its utility is greater than the
1 67.56% Car-drop off 50.67%
utility of the other alternatives Uin as the following:
2 25.53% Car-parked 4.03%
3 1.54% Taxis 15.55%
≥4 0.38% Hotel buses 3.26% ()
Check-in luggage Organisation cars 5.37%
0 19.39% Tuk-tuk/Jumbo 5.95% Similarly, the probability of the alternative i have chosen
is:
1 62.00% Rental cars 0.77%
2 12.09% Tourist buses 4.80%
3 4.22% Airport buses 1.92% ()
≥4 2.03% Other 2.69%
The MNL is a model that belongs to the discrete choice
The survey data presents briefly the characteristics model family. The model is famous and widely utilized in
relating to the personal trip, which shows the most prominent modeling travel mode choice behaviors in several studies due
air passengers (67.56%) used international airlines, and over to its uncomplicated specification and ease of interpreting the
94% of air passengers flew economy class. The most model estimation results. The MNL is a discrete choice
significant percentage of air passengers (65.07%) arrived at model between two or more alternatives among dependents
the airport by a privately-owned vehicle. Over half of them and independent variables. The model provides the choice
began their journey to the airport from their homes, followed probability of each alternative in the choice set as a function
by hotels (36.28%) and workplaces (9.60%). of a systematic segment in utility. The error components
Holiday/tourism (46.83%) was reported as the most common across the alternatives and observations are assumed to be
trip purpose of air passengers, followed by education/training identically and independently distributed (IID) and follow
(26.49%) and business (13.63%). According to the air the Gumbel distribution.
passenger responses, most of them (62%) had one piece of
478
As mentioned above, the probability of choosing an The table implied that business passengers are less likely
alternative (or travel mode) in the choice set given by the to choose bus and VCAT; this may be because personal
MNL model is illustrated in the following equation: vehicles offer independence and flexibility rather than other
modes. Regarding check-in luggage, showed less tendency to
choose motorbike, taxi, and VCAT, whereas air passengers
() with accompanying travelers have less tendency to choose
motorbikes, taxis, buses, and VCAT in accessing the WIA.
where Prn(i) is the probability of individual n choosing Concerning dependency on others, air passengers tend to
alternative i; Vin is a utility of individual n from an alternative choose private cars and VCAT. Furthermore, they tend to
i; Vjn is a utility of individual n from alternative j; K is the choose private cars, buses, and VCAT to travel to the WIA in
number of means of transportation mode in the choice set. rainy conditions.
Air passenger behavior on airport access mode choice to Table 4. Estimation results of the multinomial logit model
the City Air Terminal in Vientiane is analyzed based on the Variables Multinomial Logit Model (MNL model)
assumption that an air passenger will select the most Private Motorbikes Taxis Buses VCAT
preferred means of transportation. In order to make it simple cars
for estimation, the alternative access modes are limited to Alternative- n/a -6.38* 2.09 -0.44 1.20**
five, consisting of the motorbike, private car, taxi, airport, specific
constants
and VCAT. Therefore, the MNL applied to the critical
assumption of the Independence of the Irrelevant Alternative Socio-demographic Characteristics
(IIA). Laos passengers n/a -0.64 -1.92*** -1.183*** -1.50***
D. Estimation Results and Discussions Male n/a 0.92* -0.37 -0.01 -0.03
Elderly n/a -11.89*** 0.05 -0.15 -0.59
The research focuses on factors influencing airport mode passengers
choice in accessing the WIA by introducing the VCAT. The
Bachelor's n/a -0.34 -0.79** -0.43* -0.26
purpose is to measure the preference of air passengers to
degree
choose this proposed transportation mode to the WIA. An air
Higher monthly n/a -0.27 0.17 -0.13 -0.85***
passenger will select the most preferred means of
household
transportation. In order to make it simple for estimation, the income
alternative access modes are limited to five choice modes,
Household car n/a -1.01** -0.78** -0.06 -0.23
and the choice set consists of the motorbike, private car, taxi, ownership
airport, and the VCAT. Therefore, the MNL applies to the
Trip Characteristics
critical assumption of the Independence of the Irrelevant
Alternative (IIA). Business trips n/a -0.48 0.42 -1.23** -1.01*
** *
Check-in n/a -1.12 -0.65 -0.40 -1.05**
The variables used in the model consist of socio- luggage
demographic characteristics (i.e., nationality, gender, age, Accompany n/a -1.48*** -0.88*** -1.07*** -1.61***
education, monthly household income, and household car travelers
ownership), trip characteristics (i.e., check-in luggage, Dependency on 0.49* 0.18 0.28 n/a 0.45*
accompanying travelers, trip purposes, dependency on other
others, and weather conditions), and alternative attributes Rainy condition 1.04** n/a 0.56 0.96* 1.08**
(i.e., travel cost, in-vehicle travel time, delay time, access Travel Mode Attributes
time, waiting time, and parking fees). The RStudio software
Travel cost -0.17*
and several R packages were used to estimate the model. The
model was calibrated to investigate the behavior of air In-vehicle travel time -0.22**
passengers accessing the WIA, and private cars were treated Delay time -0.06+
as the base mode reference. Therefore, the model had no Access time -0.07
alternative-specific constant for private car mode. Waiting time -0.20***
Table 4 presents the result of the MNL model. It is Parking fees -0.46***
observed that Laos passengers have less tendency to choose Interaction Variables
taxi, bus, and VCAT. Also, males tend to choose motorbikes; Male × business trip × travel cost -0.008
generally, men love traveling by motorbike rather than Laos × business trip × travel cost -0.010
women. In the age group variable, air passengers were
Elderly × business trip × travel cost -0.034+
grouped under 60-year-old represented young passengers,
and above 60-year-old represented elderly passengers. The Initial log-likelihood -1677.034
results showed that elderly passengers are less likely to Final log-likelihood -1332.851
choose motorbikes to access the airport. Air passengers with McFadden’s Rho-squared 0.205
a bachelor's degree are more willing to choose taxis and Adjusted McFadden's Rho-squared 0.171
buses to travel to the airport. At the same time, air passengers VOT of male business passenger (LAK/h) 75,252
with a higher monthly household income tend less to choose
VOT of Lao business passenger (LAK/h) 74,643
the service of VCAT. Household car ownership as another
explanatory variable implied that air passengers have a less VOT of elderly business passenger (LAK/h) 65,518
tendency to choose motorbikes and taxis in accessing the Number of observations 1042
WIA. Significant level ***: 0.1%; **: 1%; *: 5%; +: 10%;
n/a = Not applicate
479
In line with travel mode attributes, travel cost, in-vehicle transportation market is comparatively low owing to the
travel time, delay time, access time, waiting time, and slight increases in probability (1.32%) even though multiple
parking fees have negatively influenced air passengers' mode service attributes (including in-vehicle travel time and
choice. The results indicated that the offered modes with the waiting time) were already improved and lower compared
higher travel cost and in-vehicle travel time, the less with the other modes. Simultaneously, the probability of the
tendency to choose the modes. In addition, our model results VCAT increases by 1.02%, 0.89%, and 0.67%, respectively,
showed that the more significant the waiting time and access when improving 10% in travel cost, in-vehicle travel time,
time, the less the tendency to select buses or VCAT to access and waiting time. Additionally, an increase of 10% in
the WIA, even though the access time is insignificant. parking fees for the private car at the WIA also reduces the
Additionally, the higher the delay time, the fewer air modal share of private vehicles accessing the WIA by
passengers choose private cars, taxis, or buses. With 0.41%. The results suggest that even a slight improvement in
increasing parking fees at the airport, there is less tendency travel cost, in-vehicle travel time, and waiting time would
to choose private cars and motorbikes in accessing the WIA. promote the choice probability of the VCAT.
The value of time (VOT) is computed from the model. Table 6. Sensitivity analysis for changing public transport modes'
The VOT can be described as an air passenger's willingness attributes
to pay for travel time saving [26]. It can be calculated from
the estimated coefficients of travel cost and in-vehicle travel Private
Motorbike Taxi Bus VCAT
time. By generalizing the model, we can have considerable Scenarios car
(%) (%) (%) (%)
differences in VOT as the VOT of the male business (%)
passengers and Lao business passengers equal 75,252 LAK/h Scenario 1 -0.03% -0.10% -0.06% 0.40% -0.20%
and 74,643 LAK/h, respectively. In contrast, the VOT of the Scenario 2 -0.06% -0.21% -0.13% 0.83% -0.43%
elderly business passenger is equal to 65,518 LAK/h (Table Scenario 3 -0.03% -0.12% -0.08% 0.49% -0.26%
4). Scenario 4 -0.09% -0.34% -0.20% 1.32% -0.69%
The MNL model results presented that the behaviors' Scenario 5 -0.09% -0.31% -0.19% -0.42% 1.02%
airport access mode choice of air passengers to the WIA by Scenario 6 -0.07% -0.27% -0.17% -0.37% 0.89%
introducing the VCAT was significantly affected by socio- Scenario 7 -0.06% -0.21% -0.13% -0.28% 0.67%
demographic characteristics, trip characteristics, and travel Scenario 8 -0.13% -0.48% -0.29% -0.65% 1.56%
mode attributes. It is shown that McFadden's Rho-squared Scenario 9 0.03% -0.41% 0.06% 0.14% 0.19%
has a value of 0.205, which verifies the statistical test on the
overall goodness-of-fit of the model and shows the MNL V. CONCLUSION
model's appropriateness. This study has investigated air passengers' airport access
E. Sensitivity Analysis mode choice behavior by introducing the City Air Terminal
in Vientiane as a new high-quality public transportation
In order to comprehensively analyze airport access mode system. Questionnaires and survey implementation were
choice behavior by the introduction of VCAT and provide designed to obtain reliable and valid data for analysis. The
insight for policymakers to increase public transport usage in survey data was exclusively collected through an RP and SP
Vientiane, it is necessary to simulate the various probability survey questionnaire from departing passengers at the WIA.
of public transport modes. Based on the MNL model, nine The MNL model was calibrated to analyze the key factors
scenarios constructed in Table 5 shows an overview of the influencing the airport access mode choice.
effects of various proposed strategies to increase the usage of
airport buses and VCAT. The MNL model results, this research concludes the
influencing variables into three parts: Socio-demographic
Table 5. Definition of cases for public transport modes characteristics, trip characteristics, and travel mode
attributes. The results indicate that service attributes such as
Scenario Explanation
travel cost, in-vehicle travel time, delay time, waiting time,
Scenario 1 ▪ Airport buses: travel cost decreases by 10% and parking fee are significant variables influencing airport
Scenario 2 ▪ Airport buses: in-vehicle travel time decreases by 10% access mode choice to the WIA. Socio-demographic
Scenario 3 ▪ Airport buses: waiting time decreases by 10% characteristics and trip characteristic attributes (i.e., gender,
Scenario 4 ▪ Airport buses: in-vehicle travel time decreases by 10% nationality, age, education, monthly household income,
▪ Airport buses: waiting time decreases by 10% household car ownership, trip purpose, check-in luggage, and
Scenario 5 ▪ VCAT: travel cost decreases by 10% accompanying travelers) significantly influence access mode.
Scenario 6 ▪ VCAT: in-vehicle travel time decreases by 10%
Moreover, dependency on others and weather conditions are
two additional factors found vitally influencing the air
Scenario 7 ▪ VCAT: waiting time decreases by 10%
passengers' behaviors of airport access mode to the WIA.
Scenario 8 ▪ VCAT: in-vehicle travel time decreases by 10%
▪ VCAT: waiting time decreases by 10% The VOT was computed, and the differences in the VOT
Scenario 9 ▪ Private cars: parking fee increases by 10% were associated with various levels of air passengers. The
VOT can be described as an air passenger's willingness to
Regarding transportation services, airport buses and pay for their travel time saving [26]. Our study estimated the
VCAT are almost similar in accessing the WIA. Table 6 VOT based on gender, nationality, and age for specific trip
indicated that the modal share of airport buses would be purposes. The findings of VOT estimation are consistent
increased by 0.40%, 0.83%, and 0.49%, respectively, if with the model results. The sensitivity analysis for specific
travel cost, in-vehicle travel time, and waiting time decreased service attributes for public transportation modes was
by 10%. It shows that the competitiveness of the buses in the conducted to identify significant factors requiring
480
improvement to increase their utilization. The results [9] C. Keumi, and H. Murakami, "The Role of Schedule Delays
indicated the low competitiveness of the bus in accessing the on Passengers' Choice of Access Modes: A Case Study of
WIA. Even though multiple service attributes were Japan's International Hub Airports", Transportation Research
improved, the modal share of the bus still increased slightly. Part E: Logistics and Transportation Review, 2012, vol. 48,
To enhance the VCAT's usage, a suggestion for the airport (5), pp. 1023-1031.
operators to consider improving the specific service [10] A. R. Mamdoohi, M. Saffarzadeh, A. Taherpour, and M.
attributes such as travel cost, in-vehicle travel time, and Yazdanpanah, "Modeling Air Passengers' Ground Access
Mode Choice a Case Study of Ikia", International Journal of
waiting time in the design stage of the VCAT project, which
Modeling and Optimization, 2012, vol.
would increase the effectiveness of their operations.
[11] M. L. Tam, W. H. K. Lam, and H. P. Lo, "Modeling Air
Moreover, regarding airport management, the usage of Passenger Travel Behavior on Airport Ground Access Mode
private vehicles would be reduced by increasing the parking Choices", Transportmetrica, 2008, vol. 4, (2), pp. 135-153.
fees at the airport. The research findings provide [12] P. Koster, E. Kroes, and E. Verhoef, "Travel Time Variability
policymakers insight into improving the planning of airport and Airport Accessibility", Transportation Research Part B:
access to WIA. Methodological, 2011, vol. 45, (10), pp. 1545-1559.
Further improvement of research relating to airport [13] V. Psaraki, and C. Abacoumkin, "Access Mode Choice for
Relocated Airports: The New Athens International Airport", J
access mode choice by introducing the City Air Terminal in
Air Transp Manag, 2002, vol. 8, (2), pp. 89-98.
Vientiane is required. Firstly, future research should
[14] R.-C. Jou, D. A. Hensher, and T.-L. Hsu, "Airport Ground
concentrate on service attributes to better understand airport Access Mode Choice Behavior after the Introduction of a
access mode choice. Secondly, the different statistical New Mode: A Case Study of Taoyuan International Airport in
modeling techniques should be calibrated to analyze the Taiwan", Transportation Research Part E: Logistics and
survey data to confirm the model results. The VCAT system Transportation Review, 2011, vol. 47, (3), pp. 371-381.
was not in operation or under construction during the survey [15] S. N. Alhussein, "Analysis of Ground Access Modes Choice
duration; hence the attribute levels designed into the King Khaled International Airport, Riyadh, Saudi Arabia",
scenarios could be inconsistent with the current situation. So, Journal of Transport Geography, 2011, vol. 19, (6), pp. 1361-
the information related to the VCAT project in future 1367.
research is required to keep updated in order to obtain [16] S. Gupta, P. Vovsha, and R. Donnelly, "Air Passenger
effective forecasting. Thirdly, the sample data comprises Preferences for Choice of Airport and Ground Access Mode
only departing passengers. Therefore, an additional survey in the New York City Metropolitan Region", Transportation
on arriving passengers should be conducted in future Research Record, 2008, vol. 2042, (1), pp. 3-11.
research. Analyzing such complete sample data would allow [17] S. Choo, S. You, and H. Lee, "Exploring Characteristics of
us to understand better the airport access/egress mode choice Airport Access Mode Choice: A Case Study of Korea",
behavior of air passengers at WIA. Finally, a mixed logit Transportation Planning and Technology, 2013, vol. 36, (4),
model can be developed instead of the multinomial logit pp. 335-351.
[18] G. Akar, "Ground Access to Airports, Case Study: Port
model to capture fixed and random effects coupled with
Columbus International Airport", J Air Transp Manag, 2013,
heterogeneity among individuals.
vol. 30, pp. 25-31.
ACKNOWLEDGMENT [19] Y.-C. Chang, "Factors Affecting Airport Access Mode Choice
for Elderly Air Passengers", Transportation Research Part E:
The authors would like to thank the support of the Logistics and Transportation Review, 2013, vol. 57, pp. 105-
Department of Civil Aviation of Laos for data collecting. 112.
[20] T. Budd, S. Ison, and T. Ryley, "Airport Surface Access in the
REFERENCES Uk: A Management Perspective", Research in Transportation
[1] Ministry of Planning and Investment, "The 8th Five-Year Business & Management, 2011, vol. 1, (1), pp. 109-117.
National Socio-Economic Development Plan (2016–2020)". [21] I. Gokasar, and G. Gunay, "Mode Choice Behavior Modeling
PDR UNDP Lao, 2016. of Ground Access to Airports: A Case Study in Istanbul,
[2] "Final Report: The Preparatory Survey for the Vientiane Turkey", J Air Transp Manag, 2017, vol. 59, pp. 1-7.
International Airport Terminal Expansion Project in the Lao [22] E. Zaidan, and A. Abulibdeh, "Modeling Ground Access
People's Democratic Republic". (JICA) Japan International Mode Choice Behavior for Hamad International Airport in the
Cooperation Agency, 2013. 2022 Fifa World Cup City, Doha, Qatar", J Air Transp
[3] https://www.iata.org/en/services/consulting/airport-pax- Manag, 2018, vol. 73, pp. 32-45.
security/level-of-service/, accessed April 6, 2022. [23] N. Sanko, "Guidelines for Stated Preference Experiment
[4] "Airport Ground Access Mode Choice Models". National Design", Master of Business Adminstration diss., Ecole
Academies of Sciences Engineering, and Medicine, 2008. Nationale des Ponts et Chaussées. s, 2001, vol.
[5] R. H. Ellis, J. C. Bennett, and P. R. Rassam, "Approaches for [24] Wise Payments, https://wise.com/gb/currency-converter/usd-
Improving Airport Access", Transportation Engineering to-lak-rate?amount=1, accessed April 9, 2022.
Journal of ASCE, 1974, vol. 100, (3), pp. 661-673. [25] M. E. Ben-Akiva, and S. R. Lerman, "Discrete Choice
[6] G. Harvey, "Study of Airport Access Mode Choice", Journal Analysis: Theory and Application to Travel Demand", (MIT
of Transportation Engineering, 1986, vol. 112, (5), pp. 525- Press, 1985).
545. [26] R. Paleti, P. Vovsha, D. Givon, and Y. Birotker, "Impact of
[7] M. C. Clark, and W. H. K. Lam, "Airport Ground Access Individual Daily Travel Pattern on Value of Time",
Considerations", Airports into the 21st Century. Hong Kong Transportation, 2015, vol. 42, (6), pp. 1003-1017.
Institution of Engineers, Hong Kong, 1990, vol., pp. 5-7.
[8] E. Pels, P. Nijkamp, and P. Rietveld, "Access to and
Competition between Airports: A Case Study for the San
Francisco Bay Area", Transportation Research Part A: Policy
and Practice, 2003, vol. 37, (1), pp. 71-83.
481
A Study of the Tensile Strength for the Mixing

Ratio of Eva in PA6/EVA Blends
Minh-Nhat Nguyen Quy-Long Nguyen Ngoc-Thien Tran
HCMC University of Technology HCMC University of Technology and HCMC University of Technology and
and Education Education Education
18144272@student.hcmute.edu.vn 18144261@student.hcmute.edu.vn thientn@hcmute.edu.vn
Vinh-Tien Nguyen Minh The-Uyen Tran

Faculty of Chemical and Food Technology Faculty of Mechanical Engineering
HCMC University of Technology and Education HCMC University of Technology and Education
Ho Chi Minh City, Viet Nam Ho Chi Minh City, Viet Nam
tiennv@hcmute.edu.vn uyentmt@hcmute.edu.vn
Abstract: The mixing of plastics to create a material that (VA), the higher the ratio of VA [8-12], the more
meets the requirements of tensile strength and mechanical flexibility. Based on the ratio of VA in EVA, it can be
properties of the product has been of great interest in recent divided into 3 types: high VA rate (>60%) called Ethylene
years. The properties of Polyamide 6 (PA6), which is a plastic Vinyl Acetate rubber; moderate rate of VA (4-30%) is a
material with characteristics such as hard, durable, good heat copolymer with some rubber-like properties, called a
resistance, and corrosion resistance, but its ductility is quite thermoplastic elastomer material; low VA ratio (<4%) can
low. To compensate for that weakness, mixing PA6 with be called vinyl acetate modified polyethylene. In this study,
Ethylene Vinyl Acetate (EVA) - with the properties of a the materials used the average VA.
'rubber' with high ductility and flexibility at low temperatures Like other materials, the purpose of mixing different
- is one solution. Therefore, in this study, mixtures of EVA materials is to increase their mechanical properties and
and PA6 at 5%-30% (wt) were extruded to make tested produce the desired material with different use values [13-
samples. The results of the EVA/PA6 mixed extruded sample 15]. With the properties of PA6, which is a plastic material
test will include such indicators as Tensile strength, Young's with characteristics such as hard, durable, good heat
modulus, and Elongation. SEM micrographs of fractured resistance, and corrosion resistance [16], but high hardness
surfaces show the particles-dispersed phase morphology of materials like PA6, its ductility is quite low. To
PA6/EVA blends. The indicators show that tensile strength compensate for that weakness, mixing EVA with the
decreases with increasing of EVA ratio but has a higher properties of a 'rubber' with high ductility and flexibility at
elongation at break, and reach maximum when the ratio of low temperatures, has the ability to combine well with
EVA in the blend was 15%. The photomicrographs also show other plastics [17-20]. When mixed with a certain ratio, it
the spheres particles can be observed when the EVA percent is possible to change the properties of the material to create
reaches 15% and more obvious when the percentage of EVA such as hardness, and elasticity by increasing or decreasing
reaches 20-30%, the more spherical particles appear in the the ratio of EVA.
blends, the more adhesion ability of the two materials is
reduced. With the desire to create a material that is durable, can
be used in a wide temperature range, has good abrasion
Keywords: PA6/EVA blends, tensile strength, phase resistance but still keeps a beautiful surface, PA6 is fittable
morphology, mixing polymers to be the main material and EVA as the additive material.
Gradually increase the percentage of EVA in the mixture
I. INTRODUCTION from 5-10-15-20-30%, the rest is PA6.
Polyamide 6 (PA6) or Nylon 6 is a type of plastic The aim of the study was to find the appropriate mixing
widely used in industries, especially the automotive and ratio by weight to improve the tensile properties of
electronic industries, based on its outstanding features EVA/PA6 blends.
such as high strength and stiffness at high temperatures,
and resistance to high-temperature resistance [1-6]. Good II. EXPERIMENTAL MATERIAL AND METHODS
fatigue and wear resistance, resistance to chemical A. Material
corrosion and UV rays, and non-toxic to humans [7].
The blending ratio by weight of PA6/EVA plastic was
Ethylene Vinyl Acetate (EVA) is a copolymer polymer used in this experiment as in Table 1. The experimental
of Ethylene and Vinyl Acetate, which can change plastics were mixed in the mixing system of the Haitian
properties based on changing the ratio of Vinyl Acetate MA1200III plastic injection molding machine (Fig. 2)

Table 1. PA6/EVA blending ratio (%wt) (27±1oC), procedure at an extension rate of

Sample code EVA PA6 50mm/min. The results of related data are recorded on
E5 5 95 the computer: Maximum force at break, elongation,
E10 10 90
E15 15 85 Young's modulus... In each case will test at least 5
E20 20 80 samples and take the average value.
E30 30 70
4) SEM photographing: Photomicrograph at the
B. Process and Methods fractured surface by using the SEM method. Preparing
The sample in this study was created and tested follow step by specimens: use samples that have been broken on the
step as the Fig. 1 below. Testometric 5kN M350-10CT machine, cutting off a
small piece at the breakpoint. After that, fix it in the
Testing SEM Hitachi TM4000Plus electron microscope (Fig. 4).
Injection
Mixing measure- photo-
molding
ments graphing
Figure 1. The flow of the experiment
1) Mixing: Manually dry-mixed according to the pre-

divided ratio as shown in Table 1.
2) Injection molding: Using the extrusion method.
The plastic granules, after manual mixing, will be put
into the Haitian MA1200lll plastic injection molding
machine (Fig. 2) at a temperature of 230oC and a screw
speed of 80 rpm. Test specimens are pressed according Figure 4. The Hitachi TM4000Plus electron microscope
to ASTM D638, type 4, as seen in Fig. 3.
A. Tensile Property
Fig. 5 is the stress-strain curve of the EVA/PA6 blends
after tensile testing. The results taken in Table 2 show that
the elongation increases gradually with a rising ratio of
EVA in the mixture. Based on the graph, the elongation is
longest when EVA is 15%. However, when reaching the
ratio of 30%, stress at peak and stress at break showed
signs of a strong decrease, which can be seen in the data
table (Table 2) and the chart in Fig. 5.
Table 2. Mechanical properties of EVA/PA6 blends
Strain
Figure 2. Haitian MA1200III plastic injection molding machine Stress Stress at Tensile
Sample Elongation at
at peak break modulus
code (mm) Break
(MPa) (MPa) (MPa)
(%)
E5 126.77 38 58 518 195
E10 148.29 37 65 444 228
E15 175.60 33 68 423 270
E20 165.32 32 64 256 254
E30 152.72 27 49 202 235
a)
b)
Figure 3. (a). Specimens according to ASTM D638, (b). An
injected mold specimen
3) Testing measurements: Measuring tensile
properties according to the ASTM D638 standard: on
the Testometric 5kN M350-10CT machine. The test
specimens were measured at room temperature Figure 5. Stress-Strain curve of EVA/PA6 blends
483
Table 2 and Fig. 5 show that the stress at break, stress

at peak, and tensile modulus of the polymer blends drop
steadily as the percentage EVA increases. The stress at the
break of the groups tends to decrease but is not uniform.
At EVA 5% the stress at break is the lowest, then increases
to the percentage of EVA 15% and then gradually
decreases can be concluded that the rate of EVA 5% does
not affect the tensile strength of materials. Tensile strength
decreased as EVA content increased from 15-30%.
Similarly, when the EVA content was varied from 15- d) Diagram of elongation and weight % EVA.
30%, the material's tensile modulus was also decreased Figure 6. The chart of mechanical properties of EVA/PA6 blends
(Fig. 5).
B. Phase Morphology
The inclusion of rubber-like EVA molecules that
function as stress concentrations, resulting in lower overall
composite stresses than raw PA6, might explain the drop
in material's tensile strength of PA6/EVA mixes. The
blend's tensile modulus has dropped can be explained by
the copolymer EVA softening effect, as the EVA’s tensile
modulus is significantly lower than PA6. Because EVA is
a low-modulus polymer, its presence in the structure of
PA6 reduces the total tensile modulus of the composite. In a) The fractured surfaces of the E5 sample
addition, this is also influenced by the poor surface
adherence of the two components.
b) The fractured surfaces of the E10 sample
a) Diagram of stress at break and weight % EVA

c) The fractured surfaces of the E15 sample
d) The fractured surfaces of the E20 sample
b) Diagram of stress at peak and weight % EVA
e) The fractured surfaces of the E30 sample
Figure 7. Micrograph of the fractured surfaces using SEM

This part of the experiment uses SEM to observe and
analyze the microstructure and studies the phase
morphology of PA6/EVA blends with the percentage of
EVA change from 5 to 30%. The micrographs scanning
through fractured surfaces of each test subject with
c) Diagram of tensile modulus and weight % EVA
different percentages of EVA after the tensile test were
484
shown in Figure 7. SEM analysis of the specimens tested [5] Bhattacharyya, A. R., Maiti, S. N., Misra, A., Mechanical
shows that the particles- dispersed phase morphology can properties and morphology of PA6/EVA blends, Journal of
be observed in the blends with EVA percentage of about Applied Polymer Science, 85 (2002) 1593 – 1606.
5-30%, shown in Figure 7 (a- e). [6] Shang-Han Wu, Feng-Yih Wang, Chen-Chi.M. Ma, Wen-Chi
Chang, Chun-Ting Kuo, Hsu-Chiang Kuan, Wei-Jen Chen.
Through the SEM micrographs of fractured surfaces, Mechanical, thermal and morphological properties of glass
there were fiber structures and spheres particles scattered fiber and carbon fiber reinforced polyamide-6 and polyamide-
around. The specimens containing 5-10% EVA had short 6/clay nanocomposites, Materials Letters, 2001.
fiber with high density but thin and low coalescence [7] P. Maiti. Polyamide nanocomposites, Thomas, S., Zaikov, G.
leading to the lower tensile strength. On the other hand, the E. (Eds.), Polymer nanocomposites research advances, New
length and thickness of the plastic fibers in the EVA York, 2008, pp. 175 – 227.
samples by 15-30% increased with the proportion of EVA, [8] M Palabiyik, S Bahadur. Mechanical and tribological
coalescence between fibers led to the increase in tensile properties of polyamide 6 and high density polyethylene
strength. In the central part, spheres particles can be polyblends with and without compatibilizer, 2000.
observed when the EVA percent reaches 15% and more [9] Karsli, Nevin Gamze, Aytac, Ayse. Tensile and
obvious when the percentage of EVA reaches 20-30%, the thermomechanical properties of short carbon fiber reinforced
more spherical particles appear in the blends, the more polyamide 6 composites, Composites Part B: Engineering,
2013.
adhesion ability of the two materials is reduced. At the
higher ratio of EVA, more holes will appear in the [10] Kuan-Hua Su, Jia-Horng Lin, Chih-Ching Lin., Influence of
PA6/EVA blends, increasing the elongation ability of the reprocessing on the mechanical properties and structure of
polyamide 6, Journal of Materials Processing Technology,
blends. It can be seen by the average elongation at the 2007.
break of the test specimens was increased as the percentage
[11] Izaro Goitisolo, José I. Eguiazábal, Jon Nazábal., Effects of
of EVA increased. reprocessing on the structure and properties of polyamide 6
IV. CONCLUSION nanocomposites, Polymer Degradation and Stability, 2008.
[12] C.G.Martins, N.M.Larocca, D.R.Paul, L.A.Pessan.
There are two main conclusions of this study as follows: Nanocomposites formed from polypropylene/EVA blends.
1. In PA6/EVA blends, tensile properties decrease Polymer. 2009.
with the addition of percent EVA, but has a higher [13] Jamaliah Sharif, Sharifah Hanisah Syed Abdul Aziz,
elongation at break, and reach maximum when the Kamaruddin Hashim. Radiation effects on LDPE/EVA
blends. Radiation Physics and Chemistry, 2000.
ratio of EVA in the blend is 15%.
[14] Arup R. Bhattacharyya, Anup K. Ghosh, Ashok Misra.
2. The micrographs of fractured surfaces observed by Ionomer compatibilised PA6/EVA blends: mechanical
SEM show that at higher ratio of EVA, EVA properties and morphological characterization. Polymer,
molecules linkage and high surface tension between 2003.
EVA and PA6 lead to an increase in the elongation [15] Ramesh Babu Adusumalli, S. Aparna, D. Purnima. Review on
ability but lower in tensile strength due to the EVA Various Compatibilizers and its Effect on Mechanical
particles were incompatibility with PA6. Properties of Compatibilized Nylon Blends. Polymer-Plastics
Technology and Engineering, 2017.
Furthermore, to obtain higher results, the blends of [16] Fabrício Bondan, Juliano Roberto Ernzen, Javier Amalvy,
PA6/EVA need to be compatibilized and should be studied Ana Vera Machado, Johnny De Nardi Martins, Otávio
more in a different article. Bianchi. Influence of dynamic crosslinking on the
morphology, crystallization, and dynamic mechanical
ACKNOWLEDGMENT properties of PA6,12/EVA blends. Journal of Applied
This study is supported by the Ho Chi Minh City Polymer Science, 2016.
University of Technology and Education (HCMUTE). [17] M. Kryszewski, Ezio Martuscelli, Rosario Palumbo, Marian
Kryszewski. Polymer Blends: Processing, Morphology, and
Properties, 2012.
REFERENCES
[18] Isabel S. Miles, Andrew Zurek. Preparation, Structure and
[1] J. Avery. Injection molding alternatives: a guide for designers properties of two-phase co-continuous polymer blends.
and product engineers, Munich: Carl Hanser Verlag, 1998. Polymer Engineering & Science, 1988.
[2] Whelan, T. Polymer technology dictionary, first ed., Chapman [19] J. F. Mano, D. Koniarova, R. L. Reis. Thermal properties of
& Hill, London, 1994. thermoplastic starch/synthetic polymer blends with potential
[3] Gaymans, R. J. In Rubber Toughened Engineering Plastics; biomedical applicability, 2003.
Collyer, A. A., Ed., Chapman and Hall: London, 1994. [20] D. Heikens, N. Hoen, W.M. Barentsen, P. Piet, H. Ladan.
[4] M. Tasdemir, H. Yildirim. Achieving Compatibility in Blends Mechanical properties and morphology of copolymer
of Low-Density Polyethylene/Polyamide-6 with Addition of modified polymer blends. Journal of Polymer Science.
Ethylene Vinyl Acetate, Istanbul, Turkey, 2001. Polymer Symposia, 1978.
485
Cytotoxicity of Isoxazole/Pyrazole Curcuminoids

against Human Oral Epidermal Carcinoma-KB Cell
Line
Hoang Minh Hao* Ho Dung Manh Vo Thi Nga
Department of Chemical Department of Pharmacy Department of Chemical
Technology, Faculty of Chemical Lac Hong University Technology, Faculty of Chemical
and Food Technology Bien Hoa City, Dong Nai Province, and Food Technology
Ho Chi Minh City University of Vietnam Ho Chi Minh City University of
Technology and Education manhhodung@gmail.com Technology and Education
haohm@hcmute.edu.vn ngavt@hcmute.edu.vn
Abstract: Curcumin, an attractive compound with broad- the synthesis and structure–activity relationship (SAR) studies
spectrum activities found in the rhizome of Curcuma longa of small molecules containing fluorine have been developed
L. is chemically composed of two ortho-methoxy phenols to interpret the characteristic effects of fluorine on biological
connected symmetrically by two α,β- activities.
unsaturated carbonyl groups. Structure-activity relationship Oral cancer is a serious disease with more than 350,000
(SAR) studies on curcumin revealed that the substituents in cases worldwide and about half of the patients died from it
aromatic ring systems and 1,3-dicarbonyl moieties are [13]. The conventional treatments for oral cancer such as
responsible for biological activities. Therefore, structural primary surgery followed by radiotherapy and/or
modifications of these functional groups can be envisioned as chemotherapy gave undesirable side effects. In recent years,
a strategy to screen curcumin analogs with promising natural products extracted from plants showing selective,
activities. In previous work, our group targeted the synthesis potent and relatively less toxic properties have received much
of curcuminoids containing isoxazole/pyrazole rings. As part attention from researchers [14].
of our ongoing research on exploring anticancer inhibitory Human oral cancer cell line (KB-Keratin-forming tumor
curcumin-based agents, we now report the cytotoxicities of cell line HeLa) has been commonly used to evaluate the in
isoxazole/pyrazole curcuminoids bearing various vitro anticancer activity of potential anticancer agents. In our
substituents in aromatic moieties. Fourteen present work, we aimed at analyzing the effects of
isoxazole/pyrazole curcuminoids (O1-7 and P1-7) were isoxazole/pyrazole ring and substituents in aromatic moieties
evaluated for their in vitro anticancer activities against human in isoxazole and pyrazole curcuminoids on the in vitro
oral epidermal carcinoma-KB cell line. Structures O1 (IC50 = anticancer activities against KB cancer cell line. The
13.851.17 M) and P1 (IC50 = 3.400.22 M) displayed observations were compared to the results of their
good cytotoxicity profiles compared with curcumin (IC50 = corresponding mother structure, i.e., curcumin.
33.352.66 M), and were thus identified as potential hit
II. MATERIAL AND METHOD
scaffolds for further studies.
A. Isoxazole and pyrazole curcuminoids
Keywords: curcumin, isoxazole/pyrazole curcuminoid,
The synthetic procedures of isoxazole (O1-7) and pyrazole
anticancer activity, KB cancer cell line, SAR (P1-7) curcuminoids (Figure 1) have been reported in our
previous work [4,15].
I. INTRODUCTION
From a biological point of view, curcumin (Cur), a main
component of turmeric (Curcuma longa) displays a diversity
of interesting properties including anti-inflammatory [1],
antioxidant [2] and anticancer [3] activities. Possessing these
characteristics, curcumin has been considered as a potential
lead structure for drug development. Among curcumin-based
analogs, isoxazole and pyrazole curcuminoids, which are
prepared by a heterocyclization of free 1,3-dicarbonyl in
curcumin structure with hydroxylamine or hydrazine
exhibited increased anticancer activities [4-10].
Due to special properties of fluorine (F) such as small size
and hydrogen bonding acceptor, fluorine substituents
displayed interesting biological responses [11,12]. However, Figure 1. Curcumin (Cur), isoxazole (O1-7) and pyrazole (P1-7)
the occurrence of organo-fluorine bioactive compounds curcuminoids were evaluated for their cytotoxicity against KB
originating from natural sources is virtually absent. Therefore, cancer cell line.

B. Cytotoxicity assay against KB cancer cell line 1H), 6.80 (d, H6, 3J (H,H) = 16.5 Hz, 1H), 6.90 (d,
Curcuminoids containing isoxazole and pyrazole rings H3,3,5,5, 3J (H,H) = 8.0 Hz, 4H), 6.97 (d, H2, 3J (H,H) =
were assayed on the human oral cancer cell line-KB using the 17.0 Hz, 1H), 7.10 (d, H7, 3J (H,H) = 16.5 Hz, 1H), 7.28 (d,
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium H1, 3J (H,H) = 16.0 Hz, 1H), 7.45 (d, H2,2,6,6, 3J (H,H) =
bromide (MTT) method. The experimental setup can be found 9.0 Hz, 4H); 13C-NMR (125 MHz, CDCl3, ppm): 55.3
in the literature [4]. The in vitro anticancer activities of 14 (OCH3), 55.4 (OCH3), 97.7 (C4), 111.0 (C6), 114.0 (C2),
curcumin analogs were assessed against KB cancer cell line 114.3 (C3,5, 2C), 114.4 (C3,5, 2C), 128.3 (C2,6, 2C),
by determining their half-maximal inhibitory concentration 128.4 (C1), 128.5 (C2,6, 2C), 128.7 (C1), 134.4 (C1),
(IC50, μM). In our experiment, we utilized “inactive-NA” to 135.2 (C7), 160.2 (C4), 160.5 (C4), 162.2 (C5), 168.6 (C3).
show structures exhibiting no activities against KB (i.e., IC50 ESI-MS m/z calc for [M+H]+: 334.15; found: 333.90.
> 128 g/mL).
3,5-bis((E)-3-methoxystyryl)isoxazole (O4): Yield 56%
III. RESULTS AND DISCUSSION (111.9 mg), yellow solid, C21H21NO3 [333.14 g/mol]; Rf =
0.45 (HEX/EA = 9/1); m.p. 102.8 oC; 1H-NMR (500 MHz,
C. NMR and MS data CDCl3, ppm): 3.86 (s, OCH3, 6H), 6.95-6.97 (dd, H4,4, 3J
(1E,4Z,6E)-5-hydroxy-1,7-bis(4-hydroxy-3- (H,H) = 8,0 Hz, 4J (H,H) = 2.5 Hz, 2H), 7.13 (s, H2,2, 2H),
methoxyphenyl)hepta-1,4,6-trien-3-one (Cur): Yield 53% 7.22 (d, H6,6, 3J (H,H) = 7.5 Hz, 2H), 7.31 (t, H5,5, 3J
(1.95 g), red-orange solid, C21H20O6 [368.13 g/mol]; Rf = 0.31 (H,H) = 8.0 Hz, 2H), 7.38 (d, H2,6, 3J (H,H) = 15.5 Hz, 2H),
(n-hexane (HEX)/ethyl acetate (EA) = 3/2); m.p. 182.3 oC; 1H- 7.73 (d, H1,7, 3J (H,H) = 15.5 Hz, 2H). 13C-NMR (125 MHz,
NMR (500 MHz, CDCl3): δ = 3.94 (s, OCH3, 6H), 5.79 (s, CDCl3, ppm): 55.4 (OCH3), 113.5 (C2,2), 116.4 (H4,4),
H4, 1H), 6.57 (d, H1,7, 3J (H,H) = 16.0 Hz, 2H); 6.92 (d, 120.1 (C2,6), 121.3 H6,6), 129.9 (C5,5), 143.4 (C1,1),
H5,5, 3J (H,H) = 8.0 Hz, 2H), 7.04 (d, H2,2, 3J (H,H) = 1.5 143.4 (C1,7), 160.0 (C3,5). ESI-MS m/z calc for [M+Cl]-:
Hz, 2H), 7.10 (dd, H6,6, 3J (H,H) = 8.0 Hz, 3J (H,H) = 2.0 368.64; found: 368.80.
Hz, 2H). 13C-NMR (125 MHz, CDCl3): δ = 56.0 (OCH3),
3,5-bis((E)-4-fluorostyryl)isoxazole (O5): Yield 41%
101.1 (C4), 109.7 (C5,5), 114.8 (C2,2), 121.8 (C6,6),
(76.0 mg), yellow solid, C19H13F2NO [309.10 g/mol]; Rf =
127.7 (C1,1), 140.5 (C1,7), 146.8 (C3,3), 147.9 (C4,4), 0.46 (HEX/EA = 9/1); m.p. 158.6 oC; 1H-NMR (500 MHz,
183.3 (C3,5) ppm. ESI-MS m/z calc for [M+H]+: 369.14;
CDCl3, ppm): 5.81 (s, H4, 1H), 7.07 (d, H3,3, 3J (H,H) = 9.0
found: 368.90.
Hz, 2H), 7.09 (d, H5,5, 3J (H,H) = 8.5 Hz, 2H), 7.53 (d,
4,4'-((1E,1'E)-isoxazole-3,5-diylbis(ethene-2,1- C2,2, 3J (H,H) = 9.0 Hz, 2H), 7.54 (d, H6,6, 3J (H,H) = 8.5
diyl))bis(2-methoxyphenol) (O1): Yield 44% (96.4 mg), Hz, 2H), 7.61 (d, H1,7, 3J (H,H) = 16.0 Hz, 2H), 7.63 (d, H2,6,
white solid, C21H19NO5 [365.13 g/mol]; Rf = 0.34 (HEX/EA = 3
J (H,H) = 15.5 Hz, 2H). 13C-NMR (125 MHz, CDCl3, ppm):
1:1); m.p. 163.6 oC ; 1H-NMR (500 MHz, CDCl3, ppm): δ = 101.7 (C4), 116.0 (C3,3), 116.1 (C5,5), 123.7 (C2,6),
3.95 (s, OCH3, 3H), 3.96 (s, OCH3, 3H), 5.76 (s, Ar-OH, 1H), 129.9 (C2,2), 129.9 (C6,6), 131.2 (C1,1), 139.3 (C1,7),
5.79 (s, Ar-OH, 1H), 6.41 (s, H4, 1H), 6.79 (d, H6, 3J (H,H) = 162.8 (C4,4), 164.8 (C3,5). ESI-MS m/z calc for [M-H]-:
16.0 Hz, 1H), 6.91 (d, H5'', 3J (H,H) = 8.0 Hz, 1H), 6.92 (d, 308.09; found: 307.90.
H5, 3J (H,H) = 8.0 Hz, 1H), 6.96 (d, H2, 3J (H,H) = 16.5 Hz,
1H), 7.01 (d, H2'', 4J (H,H) = 2.0 Hz, 1H), 7.02 (d, H2, 4J 3,5-bis((E)-3-fluorostyryl)isoxazole (O6); Yield 57%
(H,H) = 2.0 Hz, 1H), 7.06-7.08 (dd, H6'',6', 3J (H,H) = 8.0 Hz, (105.7 mg), yellow solid, C19H13F2NO [309.10 g/mol]; Rf =
4
J (H,H) = 2.0 Hz, 2H), 7.08 (d, H7, 3J (H,H) = 16.5 Hz, 1H), 0.26 (HEX/EA = 9:1); m.p. 157.8 oC; 1H-NMR (500 MHz,
7.26 (d, H1, 3J (H,H) = 16.5 Hz, 1H); 13C-NMR (125 MHz, CDCl3, ppm): 6.50 (s, H4, 1H), 6.94 (d, H6, 3J(H,H) = 16.5
CDCl3, ppm): δ = 55.9 (OCH3), 55.9 (OCH3), 97.6 (C4), Hz, 1H), 7.01 (m, H4,4, 2H), 7.13 (d, H2, 3J(H,H) = 16.5
108.2 (C2), 108.8 (C2), 110.9 (C5), 113.8 (C5), 114.6 Hz, 1H), 7.14 (d, H7, 3J(H,H) = 16.5 Hz, 1H), 7.21 (d, H2,
4
(C6), 114.8 (C6), 121.5 (C2), 121.6 (C6), 128.2 (C1), 128.5 J (H,H) = 1.5 Hz, 1H), 7.23 (d, H2, 4J (H,H) = 1.5 Hz, 1H),
(C1), 134.8 (C1), 135.6 (C7), 146.7-147.0 (C4,4, C3,3, 7.28 (d, H6,6, 3J (H,H) = 8.0 Hz, 2H), 7.31 (d, H1, 3J(H,H)
4C), 162.2 (C5), 168.5 (C3). ESI-MS m/z calc for [M+H]+: = 17.0 Hz, 1H), 7.33 (m, H5,5, 2H); 13C-NMR (125 MHz,
366.14; found: 366.00. CDCl3, ppm): 99.1 (C4), 113.3 – 113.5 (C2,2, 2C), 114.1
(C6), 115.7 – 116.1 (C4,4, 2C), 117.4 (C2), 122.8 – 123.2
3,5-di((E)-styryl)isoxazole (O2): Yield 56% (91.7 mg),
(C6,6, 2C), 130.3 – 130.4 (C5,5, 2C), 133.8 (C7), 134.6
white solid, C19H15NO [273.12 g/mol]; Rf = 0.43 (HEX/EA =
(C1), 138.1 (C1,1, 2C), 161.7 (C3), 162.2 (C3), 164.1
9:1); m.p. 168.3 oC; 1H-NMR (500 MHz, CDCl3, ppm): 7.03
(C5), 167.9 (C3). ESI-MS m/z calc for [M+H]+: 310.11;
(s, H4, 1H), 7.23 (d, H6, 3J (H,H) = 16.5 Hz, 1H), 7.26 (d, H2,
3 found: 309.90.
J (H,H) = 16.5 Hz, 2H), 7.30 – 7.38 (m, H2,2, 2H), 7.38 –
7.46 (m, H1,7,6,6,4,4, 6H), 7.67 (dd, H3,3,5,5, 3J 3,5-bis((E)-2-fluorostyryl)isoxazole (O7): Yield 61%
(H,H) = 7.5 Hz, 4H); 13C-NMR (125 MHz, CDCl3, ppm): 99.2 (113.1 mg), yellow solid, C19H13F2NO [309.10 g/mol]; Rf =
(C4), 113.3 (C6), 115.6 (C2), 127.1 (C2,6, 2C), 127.2 0.59 (HEX/EA = 9/1); m.p. 120.1 oC; 1H-NMR (500 MHz,
(C2,6, 2C), 128.8 (C4), 128.8 (C4), 128.9 (C3,5, 2C), CDCl3, ppm): 6.57 (s, H4, 1H), 7.07 (d, H6, 3J (H,H) = 16.5
129.1 (C3,5, 2C), 134.4 (C1), 135.3 (C1), 135.6 (C1), Hz, 1H), 7.07-7.12 (m, H3,3, 2H), 7.15-7.19 (td, H5,5, 3J
136.2 (C7), 161.9 (C5), 168.0 (C3). ESI-MS m/z calc for (H,H) = 7.5 Hz, 4J (H,H) = 1.0 Hz, 2H), 7.20 (d, H2, 3J (H,H)
[M+H]+: 274.13; found: 273.90. = 16.5 Hz, 1H); 7.28-7.32 (m, H6,6, 2H), 7.33 (d, H7, 3J
(H,H) = 16.5 Hz, 1H); 7.46 (d, H1, 3J (H,H) = 17.0 Hz, 1H);
3,5-bis((E)-4-methoxystyryl)isoxazole (O3): Yield 48% 7.53-7.63 (td, H4,4, 3J (H,H) = 7.5 Hz, 4J (H,H) = 1.5 Hz,
(95.9 mg), yellow solid, C21H19NO3 [333.14 g/mol]; Rf = 0.28 2H); 13C-NMR (125 MHz, CDCl3, ppm): 98.9 (C4), 115.5
(HEX/EA = 4:1); m.p. 172.2 oC; 1H-NMR (500 MHz, CDCl3,
(C6), 115.9 (C3), 116.0 (C3), 118.3 (C2), 123.5 (C1, 123.7
ppm): 3.83 (s, OCH3, 3H), 3.84 (s, OCH3, 3H), 6.41 (s, H4,
487
(C1), 124.4 (C5,5), 127.4 (C4), 127.8 (C1), 128.1 (C7), 132.6 (C1,1), 147.0 (C3,5), 161.6 (C4), 163.6 (C4) ppm;
128.2 (C4) , 130.2 (C6), 130.4 (C6), 159.6-161.9 (C2,2, ESI-MS m/z calc for [M+H]+: 309.12; found: 308.90.
2C), 162.1 (C5), 168.3 (C3). ESI-MS m/z calc for [M+H]+:
3,5-bis((E)-3-fluorostyryl)-1H-pyrazole (P6): Yield 76%
310.11; found: 309.90.
(0.14 g), colorless solid, C19H14F2N2 [308.11 g/mol]; Rf = 0.22
4,4'-((1E,1'E)-(1H-pyrazole-3,5-diyl)bis(ethene-2,1- (HEX/EA = 7/3); m.p. 187.3 oC; 1H-NMR (500 MHz, CDCl3):
diyl))bis(2-methoxyphenol) (P1): Yield 73% (0.16 g), red- δ 6.65 (s, H4, 1H), 6.96 (td, 3J (H,H; H-F) = 8.0 Hz, 3J (H,H)
orange powder, C21H20N2O4 [364.14 g/mol]; Rf = 0.46 = 2.5 Hz, H4,4, 2H), 7.03 (d, 3J (H,H) = 16.5 Hz, H2,6, 2H),
(HEX/EA = 3/2); m.p. 215.3 oC; 1H-NMR (500 MHz, CDCl3- 7.11 (d, 3J (H,H) = 16.5 Hz, H1,7, 2H), 7.15 (t, 3J (H,H) = 2.0
MeOD): δ = 3.93 (s, OCH3, 6H), 6.55 (s, H4, 1H), 6.85 (d, 3J Hz, H2, 2H), 7.22 (d, 3J (H,H) = 7.5 Hz, H6,6, 2H), 7.29 (m,
(H,H) = 16.0 Hz, H2,6, 2H), 6.88 (d, 3J (H,H) = 8.0 Hz, H5,5, 2H) ppm; 13C-NMR (125 MHz, CDCl3): δ 101.0 (C4),
H5,5, 2H), 6.99 (d, 3J (H,H) = 8.5 Hz, H6,6, 2H), 7.03 (s, 112.8-113.0 (C2,2), 114.9-115.1 (C4,4), 118.4 (C2,6),
H2,2, 2H), 7.04 (d, 3J (H,H) = 16.5 Hz, H1,7, 2H) ppm; 13C- 122.5 (C6,6), 130.2 (C5,5), 130.3 (C1,7), 138.7 (C1,1),
NMR (125 MHz, CDCl3-MeOD): δ = 55.8 (OCH3), 99.3 (C4), 162.2 (C3), 164.1 (C3) ppm; ESI-MS m/z calc for [M+H]+:
108.6 (C2,2), 114.8 (C5,5), 114.9 (C2,6), 120.6 (C6,6), 309.12; found: 308.90.
129.0 (C1,1), 131.2 (C1,7), 146.2 (C4,4), 147.1 (C3,3)
ppm; ESI-MS m/z calc for [M+H]+: 365.15; found: 364.9. 3,5-bis((E)-2-fluorostyryl)-1H-pyrazole (P7): Yield 81%
3,5-di((E)-styryl)-1H-pyrazole (P2): Yield 79% (0.13 g), (HEX/EA = 3/2); m.p. 160.2 oC; 1H-NMR (500 MHz, CDCl3):
yellow solid, C19H16N2 [272.13 g/mol]; Rf = 0.27 (HEX/EA = δ 6.72 (s, H4, 1H), 7.05 (m, H3,3, 2H), 7.09 (m, H5,5, 2H),
9/1); m.p. 161.2 oC; 1H-NMR (500 MHz, CDCl3): δ 6.66 (s, 7.12 (d, 3J (H,H) = 16.5 Hz, H2,6, 2H), 7.22 (m, H6,6, 2H),
H4, 1H), 7.05 (d, 3J (H,H) = 16.0 Hz, H2,6, 2H), 7.16 (d, 3J 7.27 (d, 3J (H,H) = 17.0 Hz, H1,7, 2H), 7.53 (m, H5,5, 2H)
(H,H) = 16.0 Hz, H1,7, 2H), 7.27 (t, 3J (H,H) = 7.5 Hz, H44, ppm; 13C-NMR (125 MHz, CDCl3): δ 100.6 (C4), 115.7-
2H), 7.31 (t, 3J (H,H) = 7.0 Hz, H3,3,5,5, 4H), 7.47 (d, 3J 115.9 (C3,3), 119.6 (C2,6), 123.8-123.8 (C1,1), 124.2
(H,H) = 7.5 Hz, H2,2,6,6, 4H) ppm; 13C-NMR (125 MHz, (C5,5), 124.3-124.4 (C1,1), 127.3 (C4,4), 129.3 (C6,6),
CDCl3): δ 100.4 (C4), 116.9 (C2,6), 126.6 (C2,2,6,6), 147.1 (C3,5), 159.5 (C2), 161.5 (C2) ppm; ESI-MS m/z calc
128.2 (H4,4), 128.7 (C3,3,5,5), 131.9 (C1,7), 136.3 for [M+H]+ 309.12; found: 308.90.
(C1,1), 147.0 (C3,5) ppm; ESI-MS m/z calc for [M+H]+:
273.14; found: 272.90. D. Cytotoxicity against KB cancer cell line
As summarized in Table 1, phenolic compounds bearing
3,5-bis((E)-4-methoxystyryl)-1H-pyrazole (P3): Yield
an isoxazole (O1, IC50 = 13.851.17 M) or a pyrazole (P1,
69% (0.14 g), yellow solid, C21H20N2O2 [332.15 g/mol]; Rf =
IC50 = 3.400.22 M) ring exhibited higher activities than
0.42 (HEX/EA = 3/2); m.p. 176.1 oC; 1H-NMR (500 MHz,
CDCl3): δ = 3.81 (s, OCH3, 6H), 6.57 (s, H4, 1H), 6.86 (d, 3J curcumin (IC50 = 3.400.22 M). The presence of isoxazole
and pyrazole moiety led to a 2- to 10-fold greater potency
(H,H) = 8.5 Hz, H3,3,5,5, 4H), 6.88 (d, 3J (H,H) = 16.5 Hz,
toward KB cancer cell line. The absence of substituents (-OH
H2,6, 2H), 7.06 (d, 3J (H,H) = 16.5 Hz, H1,7, 2H), 7.39 (d, 3J
and -OCH3) or the presence of para-OCH3 group in phenyl
(H,H) = 8.5 Hz, H2,2,6,6, 4H) ppm; 13C-NMR (125 MHz,
rings showed no anticancer activities. It should be noted that
CDCl3): δ = 55.3 (OCH3), 99.6 (C4), 114.2 (C3,3,5,5), the presence of pyrazole scaffold in P2 structure exhibited
115.3 (C2), 113.9 (C6), 127.8 (C1), 129.3 (C1), 129.3 better anticancer activity with an IC50 of 24.06 μM regardless
(C2,2,6,6), 130.8 (C1,7), 159.7 (C4,4) ppm; ESI-MS m/z of the removal of -OH and -OCH3 groups. Compound P3
calc for [M+H]+: 333.16; found: 333.00. bearing -OCH3 at para-position showed poor anticancer
3,5-bis((E)-3-methoxystyryl)-1H-pyrazole (P4): Yield activity when compared with curcumin. Structures containing
71% (0.14 g), colorless solid, C21H20N2O2 [332.15 g/mol]; Rf isoxazole/pyrazole ring and meta-substituted -OCH3 group
= 0.32 (HEX/EA = 7/3); m.p. 106.5 oC; 1H-NMR (500 MHz, displayed close (O4, IC50 = 34.732.76 M) or lower (P4,
CDCl3): δ = 3.79 (s, OCH3, 6H), 6.62 (s, H4, 1H), 6.76 (dd, 3J IC50 = 51.964.09 M) inhibitory activities respecting to
(H,H) = 8.5 Hz, dd, 3J (H,H) = 2.5 Hz, H4,4, 2H), 6.98 (t, 4J curcumin. These observations indicated the effects of
(H,H) = 1.5, H2,2, 2H), 7.01 (d, 3J (H,H) = 16.5 Hz, H2,6, isoxazole/pyrazole ring and substituents (nature and position)
2H), 7.01-7.06 (m, 2H), 7.09 (d, 3J (H,H) = 16.5 Hz, H1,7, in aromatic moieties on the inhibitory activities of isoxazole
2H), 7.22 (t, 3J (H,H) = 8.0 Hz, H5,5, 2H) ppm; 13C-NMR and pyrazole curcuminoids.
(125 MHz, CDCl3): δ = 55.2 (OCH3), 100.4 (C4), 111.7 Table 1. Cytotoxicity against KB of curcumin (Cur), isoxazole
(C2,2), 113.8 (C4,4), 117.8 (C2,6), 119.2 (C6,6), 129.6 (O1-7) and pyrazole (P1-7) curcuminoids. MTT viability assay
(C5,5), 131.0 (C1,7), 137.9 (C1,1), 147.09 (C3,5), 159.9 after 72 h, n=3, mean  standard deviation (SD), Cpd: compound.
(C3,3) ppm; ESI-MS m/z calc for [M+3H]+: 335.18; found: CPD IC50SD (M) CPD IC50SD (M)
335.00.
Cur 33.352.66
3,5-bis((E)-4-fluorostyryl)-1H-pyrazole (P5): Yield 68% O1 13.851.17 P1 3.400.22
(HEX/EA = 7/3); m.p. 229.1 oC; 1H-NMR (500 MHz, O2 NA P2 24.069.00
CDCl3): δ 6.61 (s, H4, 1H), 6.95 (d, 3J (H,H) = 16.5 Hz, H2,6, O3 NA P3 252.8920.23
2H), 6.99 (t, 3J (H,H) = 8.5 Hz, H5,5, 4H), 7.09 (d, 3J (H,H) O4 34.732.76 P4 51.964.09
= 16.5 Hz, H1,7, 2H), 7.40 (d, 3J (H,H) = 9.0 Hz, H2,2, 2H), O5 207.0516.17 P5 NA
7.40 (d, 3J (H,H) = 9.0 Hz, H6,6, 2H) ppm; 13C-NMR (125
O6 NA P6 245.4918.98
MHz, CDCl3): δ 100.3 (C4), 115.6 (C3,3), 115.8 (C5,5),
O7 NA P7 NA
117.2 (C2,6), 128.0 (C2,2), 128.1 (C6,6),), 130.0 (C1,7),
488
As reported, the free radicals are involved in cell damage [2] N. Sreejayan and M. N. Rao, “Curcuminoids as potent
by attacking DNA strands [16]. Considering chemical inhibitors of lipid peroxidation”, J. Pharm. Pharmacol., vol. 46,
pp. 1013–1016, Dec. 1994.
structure of a para-substituted phenolic motif, the anticancer
[3] K. Mehta, P. Pantazis, T. McQueen, and B. B. Aggarwal,
capacity of curcumin can be attributed to the favorable “Antiproliferative effect of curcumin (diferuloylmethane)
formation of phenoxyl radical, which is stabilized by a against human breast tumor cell lines”, Anticancer. Drugs, vol.
conjugated -system [17,18]. In addition, curcumin was found 8, pp. 470–481, Jun. 1997.
to generate reactive oxygen species (ROS), which are [4] V. T. B. Pham, T. V. Nguyen, H. V. Nguyen, T. T. Nguyen,
responsible for cell apoptosis [18,19] and H. M. Hoang, “Curcuminoids versus Pyrazole‐Modified
Analogues: Synthesis and Cytotoxicity against HepG2 Cancer
The ability of curcumin to induce cell apoptosis in cancer Cell Line”, ChemistrySelect, vol. 5, pp. 11681–11684, Oct.
cells can be explained through the detoxification mechanism. 2020.
Curcumin is classified as an , -unsaturated ketone, which [5] A. Theppawong et al., “Synthesis of Non ‐ Symmetrical
Nitrogen‐Containing Curcuminoids in the Pursuit of New
prone to Michael addition with nucleophiles such as -SH Anticancer Candidates”, ChemistryOpen, vol. 8, pp. 236–247,
groups of glutathione S-transferases (GSTs). In other words, Feb. 2019.
curcumin is an inhibitor of detoxification enzymes GSTs [6] B. C. Jordan, B. Kumar, R. Thilagavathi, A. Yadhav, P. Kumar,
through Michael addition reaction [20-22]. and C. Selvam, “Synthesis, evaluation of cytotoxic properties
of promising curcumin analogues and investigation of possible
Note that isoxazole/pyrazole cyclization leads to a more molecular mechanisms”, Chem. Biol. Drug Des., vol. 91, pp.
rigid structure and elimination of keto-enol tautomerism. In 332–337, Jan. 2018.
other words, curcumin sensitizes to Michael donors while the [7] J. Sup Shim et al., “Hydrazinocurcumin, a novel synthetic
isoxazole/pyrazole analogs are not taken the detoxification curcumin derivative, Is a potent inhibitor of endothelial cell
mechanism into account. However, the replacement of - proliferation”, Bioorg. Med. Chem., vol. 10, pp. 2439–2444,
Aug. 2002.
diketone group in curcumin by isoxazole/pyrazole ring (O1,
[8] J. Das, S. Pany, S. Panchal, A. Majhi, and G. M. Rahman,
P1) resulted in significant anticancer activities against KB “Binding of isoxazole and pyrazole derivatives of curcumin
cancer cell line. These observations are similar to the reported with the activator binding domain of novel protein kinase C,”
results, in which the anticancer activities of isoxazole and Bioorg. Med. Chem., vol. 19, pp. 6196–6202, Nov. 2011.
pyrazole curcuminoids were increased due to their improved [9] D. Simoni et al., “Antitumor effects of curcumin and
solubility and stability when compared to curcumin [4,5,7,23- structurally β-diketone modified analogs on multidrug resistant
25]. The better anticancer activities of these analogs might cancer cells,” Bioorg. Med. Chem. Lett., vol. 18, pp. 845–849,
Jan. 2008.
rely on other mechanisms rather than detoxification, but this
[10] M. W. Amolins, L. B. Peterson, and B. S. J. Blagg, “Synthesis
remained unclear. and evaluation of electron-rich curcumin analogues”, Bioorg.
The isoxazole and pyrazole curcuminoids (O5-O7, P5-P7) Med. Chem., vol. 17, pp. 360–367, Jan. 2009.
displayed poor/inactive anticancer activities toward KB [11] B. E. Smart, “Fluorine substituent effects (on bioactivity)”, J.
Fluor. Chem., vol. 109, pp. 3–11, Jun. 2001.
cancer cell line, indicating that fluorine (-F) group in phenyl
ring gave a negative effect on the inhibitory activities. [12] M. Schlosser and D. Michel, “About the ‘physiological size’ of
fluorine substituents: Comparison of sensorially active
Although the special effects of fluorine atom on the biological compounds with fluorine and methyl substituted analogues”,
activities were reported, the mechanism has been difficult to Tetrahedron, vol. 52, pp. 99–108, Jan. 1996.
divine [26]. [13] F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L. A. Torre,
and A. Jemal, “Global cancer statistics 2018: GLOBOCAN
IV. CONCLUSION estimates of incidence and mortality worldwide for 36 cancers
in 185 countries”, CA. Cancer J. Clin., vol. 68, pp. 394–424,
Structure-activity relationship analysis suggested that i) Nov. 2018.
the isoxazole/pyrazole phenolic motif (O1, P1) is responsible [14] C.-P. Wu, S. Ohnuma, and S. V. Ambudkar, “Discovering
for better anticancer activity against KB cancer cell line, natural product modulators to overcome multidrug resistance
whereas the fluoro substituent in phenyl ring gives a negative in cancer chemotherapy”, Curr. Pharm. Biotechnol., vol. 12,
effect on the inhibitory activity. ii) The scaffold of meta-OCH3 pp. 609–620, Apr. 2011.
isoxazole curcuminoid (O4) demonstrated close cytotoxicity [15] H. M. Hoang, N. T. Vo, and T. H. Le, “Synthesis of New
against KB. In general, we pointed out a similar trend to the Isoxazole Analogs of Curcuminoid”, vol. 24, pp. 2141–2145,
Dec. 2021.
previous reports and confirmed that curcumin-based isoxazole
[16] P. A. Cerutti, “Oxidant stress and carcinogenesis”, Eur. J. Clin.
and pyrazole analogs exerted stronger anticancer effects than Invest., vol. 21, pp. 1–5, Feb. 1991.
curcumin. In addition, we demonstrated for the first time the [17] T. Chuprajob, C. Changtam, R. Chokchaisiri, W. Chunglok, N.
anticancer activity of curcuminoids bearing both fluorine- Sornkaew, and A. Suksamrarn, “Synthesis, cytotoxicity against
substituted aromatic group and isoxazole/pyrazole ring. human oral cancer KB cells and structure–activity relationship
studies of trienone analogues of curcuminoids”, Bioorg. Med.
ACKNOWLEDGMENT Chem. Lett., vol. 24, pp. 2839–2844, Jul. 2014.
We gladly acknowledge the use of the Ho Chi Minh City [18] S. Mishra et al., “Differential apoptotic and redox regulatory
activities of curcumin and its derivatives”, Free Radic. Biol.
University of Technology and Education facilities to complete Med., vol. 38, pp. 1353–1360, May 2005.
our work. [19] M. Sökmen and M. Akram Khan, “The antioxidant activity of
some curcuminoids and chalcones”, Inflammopharmacology,
REFERENCES vol. 24, pp. 81–86, Jun. 2016.
[20] M. Labbozzetta et al., “Lack of nucleophilic addition in the
[1] Y. X. Xu, K. R. Pindolia, N. Janakiraman, R. A. Chapman, and isoxazole and pyrazole diketone modified analogs of curcumin;
S. C. Gautam, “Curcumin inhibits IL1 alpha and TNF-alpha implications for their antitumor and chemosensitizing
induction of AP-1 and NF-kB DNA-binding activity in bone activities”, Chem. Biol. Interact., vol. 181, pp. 29–36, Sep.
marrow stromal cells”, Hematopathol. Mol. Hematol., vol. 11, 2009.
pp. 49–62, 1997.
489
[21] S. Awasthi et al., “Curcumin–glutathione interactions and the [24] M. Ahmed, M. A. Qadir, A. Hameed, M. Imran, and M.
role of human glutathione S-transferase P1-1”, Chem. Biol. Muddassar, “Screening of curcumin-derived isoxazole,
Interact., vol. 128, pp. 19–38, Aug. 2000. pyrazoles, and pyrimidines for their anti-inflammatory,
[22] A. T. Dinkova-Kostova and P. Talalay, “Relation of structure antinociceptive, and cyclooxygenase-2 inhibition”, Chem.
of curcumin analogs to their potencies as inducers of Phase 2 Biol. Drug Des., vol. 91, pp. 338–343, Jan. 2018.
detoxification enzymes”, Carcinogenesis, vol. 20, pp. 911–914, [25] R. Sribalan et al., “Synthesis of a Water-Soluble Pyrazole
May 1999. Curcumin Derivative: In Vitro and In Vivo AGE Inhibitory
[23] S. Mishra, S. Patel, and C. G. Halpani, “Recent Updates in Activity and Its Mechanism”, ChemistrySelect, vol. 2, pp.
Curcumin Pyrazole and Isoxazole Derivatives: Synthesis and 1122–1128, Jan. 2017.
Biological Application”, Chem. Biodivers., vol. 16, 47 pages, [26] C. Isanbor and D. O’Hagan, “Fluorine in medicinal chemistry:
Feb. 2019. A review of anti-cancer agents”, J. Fluor. Chem., vol. 127, pp.
303–319, Mar. 2006.
490
Modeling of Flow Mixed with Polymers in Open

Channel Flow: Application on the Blumenau River
in Brazil
Walid Bouchenafa Airton Hoenicke Bruna Luiza Cunico
SNF SA AXCHEM Brasil Indústria AXCHEM Brasil Indústria
Rue Adrienne Bolland, 42163, Química. R. Frederico Jensen, 200 - Química. R. Frederico Jensen, 200 -
Andrézieux, France Itoupavazinha, Blumenau - SC, Itoupavazinha, Blumenau - SC,
wbouchenafa@snf.com 89066-300, Brazil 89066-300, Brazil
airton@axchem-brasil.com.br bruna@axchem-brasil.com.br
Huyen Xuan Dang-Vu Trong Dang-Vu

Ho Chi Minh City University of Technology (HCMUT) SNF HC
Vietnam National University Ho Chi Minh City One chemical Plant Road, Riceboro, GA 31323, USA
Ho Chi Minh City, Vietnam Tvu2@snf.com
xhuyen@hcmut.edu.vn
Abstract: Extreme weather with heavy and prolonged rain adding polymers can be considered an effective new
along with rapid melting snow is one of the main causes of technique to reinforce the measures already deployed in the
river flooding. Many techniques have been developed for river flood risk management protection strategy since it
flood control and while structural methods like sea walls, reduces the flow up to 18% of its initial height [5] and
dikes, and barriers are expensive and long processing, drag therefore avoids the overflow of rivers in case of flooding [6].
reducing polymer is a less expensive and easy-to-deploy non- The approach presented in this article aims to minimize
structural technique. In this study, as a part of the BLUMPOL the impacts of river flooding in urban areas. The present work
project (BLUMENAU POLYMER), a case study on the is based on the results of research works describing the
modeling of a watercourse containing a flow without and phenomenon of drag reduction by the addition of polymers in
with drag reducing polymers is presented. The objective is to a free surface flow on a watercourse under highly controlled
improve knowledge of the drag reduction effect on open hydraulic conditions [7]. During this work, it was found that
channel flows. The modeling results show that the addition of the addition of polymers, for free surface flows, with a low
polymers in the flows leads to a marked drag reduction quantity of 20 PPM (particles per million), leads to a
resulting in decreasing the water depth up to 18% of its initial significant reduction in drag.
depth. Indeed, the interactions between the polymer and the
turbulent structures of the flow tend to increase the thickness
Keywords: flood risk management, open channel flow, of the boundary layer and modify the velocity profile near the
drag reduction, polymer, Mike Hydro rivers wall by the establishment of an elastic sub-layer. between the
I. INTRODUCTION viscous layer and the logarithmic zone of the boundary layer
[8], which subsequently causes a decrease in the water depth.
Floods are the most common natural disasters often
resulting in considerable economic losses and human and Following this interpretation, the authors hypothesized
social tragedies [1]. Extreme events in India and the that similar phenomena also occur on the Blumenau River
Philippines in 2018, Vietnam in 2021, and Brazil in 2022 located in Brazil and prone to flooding.
show how devastating the effects of floods can be. To do this, first, a hydraulic simulation model faithfully
Additionally, some megacities such as Bangkok, Mumbai, describing the flow dynamics of the Blumenau River is
Jakarta, Rio De Janeiro, Shanghai, Hochiminh City, Dhaka, developed in this study. The quasi-two-dimensional
London, Paris, and Rotterdam are extremely vulnerable to hydrodynamic software Mike Hydro River (MHR) is used to
riverine flooding due to increased risk from urbanization and simulate the flows in our case study. Two models are
the effects of climate change [2]. developed and then compared with each other, the first
An effective river flood risk protection system can reference model is dedicated to a flow without polymers and
significantly improve public safety and reduce social validated using data from measurements of water heights and
damages and economic losses related to floods. Therefore, discharge. The second model is devoted to the same
flood risk management is still a huge challenge for Blumenau River considering the presence of a quantity of 20
government and local authorities [3], especially when the ppm of polymers in the flow. Mignot et al [8] measured the
adaptation of populations bordering rivers in the face of the variation in water depth by injecting different concentrations
risk of flooding is in perpetual evolution. of polymers flows in channels at the laboratory scale. Their
work showed that from 20 ppm of PAM the backwater curves
The preferred strategy for flood control is currently the become asymptomatic with a stable decrease in water depth
protection technique [4]. The method of reducing drag by

of 20% whatever the concentration above 20 ppm of make it possible to characterize the topography of a studied
polymers. system, such as, for example, satellite images that can be
obtained from national or international organizations
II. APPLICATION CASES involved in the collection and storage of satellite images such
The hydraulic modeling is carried out on a river called as the National Aeronautics and Space Administration NASA
Itajai, which crosses the city of Blumenau located in the and the Indian Institute of Remote Sensing IIRS. Some of this
southeastern part of Brazil at an altitude of 21 m. A length of data is freely available. In addition, the most common
91 km was selected from the Indaial department to the estuary technique for updating bathymetric data is called laser remote
at Fazenda to better evaluate the drag reduction processes by sensing or LIDAR (Light Detection And Ranging), it is a
the addition of polymers. The water depth during the 2016 useful technique for simultaneously measuring topography
flood varies along the river from 8.5 to 10 meters. The width and bathymetry [9]. The mapping of the study area can be
of the river varies between 250 to 330 meters, the slope is low reproduced using a digital terrain model (DTM) which
in the order of S0 ∼10-4. The bed of the Itajai River contains provides precise morphological and topographical data. The
pebbles, gravel, and alluvial materials. The flow regime of bathymetry of our study area is prepared from a digital terrain
Itajai is fluvial, the discharge during the 2016 flood was 2368 model and provided as input data for the model.
m3/s. The site selected for the study is a meandering Note that areas with little or no data face particular
watercourse, crossed by twelve bridges. challenges. Remote sensing and the use of the Geographic
Information System (GIS) technique are particularly useful
solutions. These techniques can also be used in areas where
physical accessibility is an issue. Although satellite imagery,
aerial photography, and LIDAR technology can generate real-
time data, the cost of data acquisition is always a concern: the
purchase of expensive data and technologies such as LIDAR
should be compared to the benefits of getting more accurate
modeling results.
Figure 1. Itajai River in Blumenau a) in 2019 and b) during the
flood in 2016
Figure 3. Bathymetry of the Seine river in Paris using the LIDAR

technique (www.norbit.com)
2. The flow resistance coefficient is estimated from the
Figure 2. Location of the study site and definition of measurement medium surface condition. It depends on the type of soil and
points characterizes the linear head losses. Generally, it is
The choice of the river for the modeling is justified by the interpreted by the Manning coefficient K intervening in the
possibility of carrying out in-situ trials within the framework Manning-Strickler formula which gives the expression of the
of the BLUEPOL project to better assess the phenomenon of average velocity for a flow in a uniform regime [10]:
2 1
drag reduction by the injection of polymers at the scale of a
real watercourse. 𝑣⃗ = 𝐾𝑅ℎ3 𝑆02
III. STATE OF THE ART ON DATA REQUIRED FOR Where 𝑅ℎ [m] is the hydraulic radius, 𝑆0 is the energy
1
MODELING slope, equal to that of the channel bottom, 𝐾 = where K
η
Three input databases are required to launch the modeling [m1/2s-1] and η [m1/2s-1] are the Strickler and Manning
regardless of the hydraulic model selected: 1) a map of the coefficients respectively.
study area to locate the watercourse and define its cross- 3. The initial conditions are described by hydraulic data
sections as well as the slopes and its considered length, 2) a such as the discharges at any point of the river for the start
flow resistance coefficient defining the structure of the of the simulation. In our case study, the hydraulic data on the
riverbed and its lateral banks, and finally 3) initial and discharge as well as the measurements of the water depth
boundary conditions characterizing the flow. The data entry along Itajai are used as input data for the modeling. The
method and its qualities define the final modeling results. downstream boundary condition is a known relationship
1. The cross-sections determine the relationship between between discharge and water depth. A DTM model is used to
water depth and surface. They are typically obtained based define the river morphology and its cross-sections. Itajai flow
on topographic surveys orthogonal to the bed of the is simulated during the September 2016 flood period. The
watercourse. Indeed, the topography is another important measurement data is used to validate the modeling results and
aspect for the evaluation of flow propagation within a stretch correct the flow model without polymers.
of a finite-length river. Traditionally, topographic and
bathymetric data are obtained from land surveying and
bathymetric surveys are usually collected from local
organizations. Nevertheless, currently, several technologies
492
IV. FLOW HYDRODYNAMIC MODELING laboratory conditions, that adding polymers to a large-scale
WITHOUT AND WITH POLYMERS watercourse reduces water depth. As the polymer arrived at a
cross section, water depth decreased quite linearly with time
The hydrodynamic processes of river flows could be
and eventually reached a lower plateau of maximum drag
represented by different simplified approaches [9], and the
reduction with a constant reduced water depth.
choice of a modeling approach must be consistent with the
study context. There are several main hydrodynamic modeling
software for open channel flows, such as HEC-RAS,
LISFLOOD-FR, TELEMAC, and MIKE HYDO. In the
context of modeling flows mixed with polymers, the
representation of the flow dynamics remains quite crucial to
better assess the effectiveness of the drag reduction technique.
The Mike Hydro River (MHR) hydrodynamic model is
considered best suited for modeling this type of flow. This
model makes it possible to explicitly describe and represent Figure 6. Simulated water depth over 30 km and the difference
the flow dynamics using the Barré de Saint Venant equations between water depth on the cross-section before and after the
describing the movement of open channel flows in shallow polymer injection
water [11]. The model calculates the water depth and flows The model of flow mixed with a concentration of 20 ppm
velocity at any point along the river. The principle of the MHR of polymers shows a significant reduction in the Itajai flow
model is discussed in [12] and [13]. resistance coefficient. Figure 6 shows the water depth for a
flow with polymers over a first distance of 30 km from the
injection point and the cross-section shows the difference in
water depth when injecting the polymers. Two hydraulic
phenomena are observed in the modeled system: a decrease
in water depth along the watercourse up to 18% of its initial
depth and in return, an increase in flow capacity up to 30%.
The relative water depth reduction decreased almost
linearly with distance from the injection point. and vanished
at about 25 km downstream, at which the water depth was
hardly affected by the polymer addition.
Indeed, the phenomenon of drag reduction caused by the
Figure 4. Visualization of Itajai flows modeled by Mike Hydro injection of polymers into a watercourse can reduce the local
River water depth while increasing the flow velocity and
In their studies [8], the authors showed the impact of the contributing subsequently to protection more effectively
polymer on the change in Strickler's coefficient. In the against the effects of flooding in the vulnerable area [7]. This
modeling of flows mixed with polymers, the approach used phenomenon makes it possible to improve river flood risk
to introduce the polymers into the model consists in varying management strategies.
the Manning coefficient for each profile across the modeled
section [14].
V. RESULTS
The modeled water depth of the Itajai river with polymer
injection is shown in the Fig. 5.
Figure 7. Water depth at Itajai before and after polymer injections
VI. CONCLUSION
The addition of polymers strongly decreases the drag force
that the walls apply on the flow, leading to highly increased
mean velocities and highly decreased water depths [7]. The
Figure 5. Modeled water depth with 20 ppm polymer results in the laboratory showed that the flow resistance
coefficient decreases to 50% when injecting 20 ppm of
In this work, the Mike Urban Hydro was used to model polymers. The Manning coefficient has been modified in our
Itajai flows. A comparison of the model results before model to introduce the polymers effect. This drag reduction
polymer injection, with the measurement data, showed a method has already proven to be effective for reducing the
remarkable capacity of the modeling system set up allowing water depth in real water course [7].
to reproduce effectively the flow dynamics of Itajai. The
measurement data made it possible to better correct the As part of the BLUM POL project which aims to carry
developed model. Fig. 6 confirms previous experiments in out a real trials in the Blumenau river. A modeling study was
carried out to show the simulations results. The model
493
developed was calibrated on the results obtained for another [5] W. Bouchenafa, A. Lefevre, and B. Quillien, “Augmenter le
river and it requires a final validation step during our next coefficient de Strickler par la technique de réduction de la
traînée dans les écoulements à surface libre”, International
trial in Blumenau river. Journal of Innovation and Applied Studies, ISSN 2028-9324
Numerical modeling of river flows can be used to assess vol. 24, No. 1 Aug. 2018, pp. 1-8.
the impact of river flood risk. Modeling can also be used to [6] R. Pich, “Method for limiting freshet levels and controlling
floods “, Patent# WO 2014199037 A1., 2014.
test the effectiveness of different flood risk solutions in more
[7] W.Bouchenafa, B. Dewals, A. Lefevre, E.Mignot, “Water
detail [15]. Any type of modeling can be a powerful tool to Soluble Polymers as a Means to Increase Flow Capacity: Field
help decision-makers understand flood risk. In our case study, Experiment of Drag Reduction by Polymer Additives in an
the hydraulic modeling showed interest in the drag reduction Irrigation Canal”. Journal of Hydraulic Engineering. vol. 147,
method to avoid the overflow of the banks in a major flood No. 8 (August 2021).
event. However, model results need to be validated against https://ascelibrary.org/doi/abs/10.1061/%28ASCE%29HY.19
43-7900.0001904
measured data. A measurement campaign on Itajai is in
[8] E. Mignot, N. Rivière, A. Lefevre and B Quillien “Smoother
preparation to better assess the effectiveness of the drag Than Smooth: Increasing the Flow Conveyance of an Open-
reduction technique by adding polymers and thus validate the Channel Flow by Using Drag Reduction Methods”. J. Hydraul.
numerical model. Eng., 2019, 145(4): 04019011, p. 8.
[9] W.J. Lillycrop, L.E., Parson and J.L Irish, “Development and
ACKNOWLEDGMENT operation of the SHOALS airborne LIDAR hydrographic
survey system”, Proceedings of SPIE 2694 CIS Selected
This project will mobilize more than ten people during the Papers: Laser Remote Sensing of Natural Waters: From Theory
next measurement campaigns. The modeling work was done to Practice, St Petersburg, November 01, 1996.26 (1996): 26 -
in partnership with the company AXCHEM Brazil. The 37.
authors thank AXCHEM represented by Armin, for the data [10] M. Soutter, A. Mermoud, A. Musy “ Infénierie des eaux et du
needed for the Blumenau river. sol. Processus et aménagements”, Presses polytechniques et
universitaires romandes. 2007, p. 294.
REFERENCES [11] B. J. Dewals, S. Detrembleur, P. Archambeau, S. Erpicum, J.
Ernst, M. Pirotton “Caractérisation micro-échelle du risque
[1] G. Merkuryeva, M. Kornevs, “Water Flow Forecasting and d’inondation: modélisation hydraulique détaillée et
River Simulation for Flood Risk Analysis”, Information quantification des impacts socio-économiques”, La Houille
Technology and Management Science, doi: 10.2478/itms- Blanche, N°2n 2011, pp. 28-34.
2013-0006, pp. 42-46, 2013. [12] Mike Powered by DHI “MIKE 1D DHI Simulation Engine for
[2] Z. W. Kundzewicz, D. L. T. Hegger, P. Matczak, and P. P. J. 1D river and urban modelling, Référence Manual”, p. 334,
Driessen, “Water Flow Forecasting and River Simulation for [13] S. Patro, C. Chatterjee, S. Mohanty, R. Singh, N. S.
Flood Risk Analysis”, National Academy of Sciences, vol. 115, Raghuwanshi, “Flood Inundation Modeling using MIKE
No. 49, 12321-12325, December 2018. FLOOD and Remote Sensing Data” J. Indian Soc. Remote
[3] Peter P. J. Driessen, Dries L. T. Hegger, Zbigniew W. Sens. March 2009. 37:107–118.
Kundzewicz, Helena F. M. W. van Rijswick, Ann Crabbé, [14] W. Bouchenafa, A. Lefevre, “Modélisation des écoulements
Corinne Larrue, Piotr Matczak, Maria Pettersson, Sally Priest, mélangés avec des polymères dans une rivière pour la gestion
Cathy Suykens, Gerrit Thomas Raadgever and Mark Wiering des risques d’inondation t”. International Journal of Innovation
“Governance Strategies for Improving Flood Resilience in the and Applied Studies. ISSN 2028-9324 vol. 28 No. 1 Dec. 2019,
Face of Climate Change”, Water MDPI, vol. 10, No. 1595, pp. 10-17
2018.
[15] L. Timbre, P. Willems, J. Berlamont, “Validation of a quasi-2D
[4] P. Driessen, D. Hegger, M. H. N. Bakker, H. Rijswick and Z. hydrodynamic river flood model using historical and ERSSAR
Kundzewicz,, “Toward more resilient flood risk governance”, derived flood maps”, XXXI IAHR CONGRESS, September
Ecology and Society, 21(4): 53, 2016. 11~16, 2005, Seoul, Korea, pp. 137-147.
494
Proceedings of
th
2022 6 International Conference
on Green Technology and Sustainable Development
(GTSD)
July 29-30, 2022 - Nha Trang City, Vietnam
HCMUTE-Vietnam, NTU-Vietnam, LHU-Vietnam, HCMC DOST-Vietnam,

RMUTL-Thailand, NPU-Taiwan, KSU-Taiwan
VIETNAM NATIONAL UNIVERSITY – HO CHI MINH CITY PRESS

E-mail: vnuhp@vnuhcm.edu.vn Website: www.vnuhcmpress.edu.vn
Headquarters: Representative office:
Room 501, VNU-HCM Headquarter, Linh Block K, University of Social Sciences and
Trung Ward, Thu Duc District, HCMC Humanities, 10-12 Dinh Tien Hoang Street,
Phone: 028 62726361 Ben Nghe Ward, District 1, HCMC
Phone: 028 62726390
Publication Director
Dr DO VAN BIEN
Editor
SIN KE DUYEN
Proofreader
NHU NGOC
Book Cover Designer

CHAU NGOC THIN, HCMUTE
Associate Partner of Organising Manuscript and Responsible for Copyright

HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION (HCMUTE)
1st Edition. Quantity: 200 copies, Size: 19 x 27 cm. Registration Number of Publication Plan: 194-
2023/CXBIPH/1-02/ĐHQGTPHCM. Publication Decision Number 25/QĐ-NXB by VNU-HCM Press issued
on on March 15, 2023. Printed at: Hung Phu Printing & Packaging Co., Ltd. Address: 162A/1, KP1A, An Phu
ward, Thuan An city, Binh Duong province. Legal deposit in 2022. ISBN: 978-604-73-9622-1.
Copyright of works has been protected by Publication Law and Vietnam Intellectual Property Law.
All forms of publication, copying and distribution of content are strictly prohibited
without the consent of the author and the Publisher.
PROTECT COPYRIGHT TOGETHER FOR GOOD BOOK

Proceedingsof20226thInternationalConference
onGreenTechnologyandSustainableDevelopment
(GTSD)
Co-organized by:
In cooperation with:
Sponsored by:
ISBN: 978-604-73-9622-1
NXB ĐHQG-HCM
Ho Chi Minh City University of Technology and Education (HCMUTE)

Add: 01 Vo Van Ngan Str., Linh Chieu Ward, Thu Duc City, Ho Chi Minh City, Vietnam
Website: www.hcmute.edu.vn Email: oia@hcmute.edu.vn - khcn@hcmute.edu.vn 9 786047 396221

Full Proceedings GTSD2022-IsBN 978-604!73!9622-1-VNUHCM Press

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Full Proceedings GTSD2022-IsBN 978-604!73!9622-1-VNUHCM Press

Uploaded by

Copyright:

Available Formats

PROCEEDINGS OF

GTSD2022 Organizing Committee......................................................................................................................... ix

Evaluating Green Marketing Trending Determinants by a Text Mining Approach ...................................... 5

A Simple Synthesis of Antibacterial and Antifungal ZnO Nanorods ............................................................... 8

An Assessment of Complementary Energy of HPFRCs under Tension ......................................................... 15

Digital Competence of University Students: A Comparative Study at Three Universities in Vietnam....... 67

Unique Competitive Advantages of Vietnam’s Garment Industry in the Sustainable Development

Impacts of Green Training on Green Competencies of Employees: Empirical Case of Industrial

Oxygen-LPG Torch for Thermal Spraying..................................................................................................... 152

Pharmacognostic Assessment of Polyscias Fruticosa Leaves in Vietnam ..................................................... 180

A Study on Mechanical Properties of Ca-Alginate Hydrogels ...................................................................... 204

Mechanical Characteristics of PBT Based Blend ........................................................................................... 208

A Simple Design Method for Piled Raft Foundations .................................................................................... 239

Fire Resistance Properties and Geopolymer Coating .................................................................................... 254

Transportation Infrastructure Strategy for Sustainable Development: A Case Study of Vietnam

Building Students’ Self-Reliant on Problem-Based Learning by Embedding Mind of Engineering

Modeling and Comprehensive Assessment of Construction Risks: A Perspective of PPP

Fatigue Life of Accelerated Corroded Steel Plate .......................................................................................... 377

A Study on Ultrasonic Shear Horizontal Waves in Composite Structures .................................................. 381

Designing and Modeling Pipe Welding Machine ............................................................................................ 385

Investigation of the Mechanical Properties of Lightweight Geopolymer Concrete Using Keramzite as

Environmental Sustainability: Exploring Managers’ Attitudes and Behaviours at High-End

Cytotoxicity of Isoxazole/Pyrazole Curcuminoids against Human Oral Epidermal Carcinoma-KB Cell

On behalf of GTSD2022 Organizing Committee

Assoc.Prof. Hoang An Quoc

GTSD2022 Organizing Committee

General Co-Chairs Program Chairs

Program and Steering Committee

Publication Chairs Local Arrangement Chairs

A Novel Mechanism Design following Augmented

Copyright © authors. This work is licensed under a Creative Commons 1

N li  p , x  always exists in any system, especially in suspension. The

Where, n is the order of the mechanism belonging to the

It is remarked that the input stiffness is defined as

Using Eq.(10), the criterion for design AVSM is

Evaluating Green Marketing Trending

The contributions of this study are as follows: A. Dataset description

Copyright © authors. This work is licensed under a Creative Commons 5

Figure 1. The word-cloud

Figure 4. Results for cluster 3 (blue)

Product is the element that appears the most with 1837

A Simple Synthesis of Antibacterial and Antifungal

Copyright © authors. This work is licensed under a Creative Commons 8

C. Antifungal activity D. Morphology and elemental analysis of microbial cells

IV. CONCLUSION [11] N. A. Ibrahim, A. A. Nada, A. G. Hassabo, B. M. Eid, A. M.

An Assessment of Complementary Energy of

understanding mechanical properties of HPFRCs, including

I. INTRODUCTION Figure 1. Illustration of complementary energy of HPFRCs

Copyright © authors. This work is licensed under a Creative Commons 15

Gauge length of 100 mm

Figure 2. Photo of the twisted fiber type in this research

Cement (Type III) 0.8

Silica fume 0.07

Table 2. properties of six fiber types

b) 1.0 vol.% T30/0.3

c) 1.5 vol.% T30/0.3 d) 2.0 vol.% T30/0.3

Figure 4. Tensile responses of HPFRCs with different fiber contents

Table 3. Tensile parameters with different fiber contents

b) Crack tip toughness ( CT )

[1] D. L. Nguyen, J. Song, C. Manathamsombat, D. J. Kim,

A Study of the Scavenging Process in a Two-stroke

Dao Huu Huy Huynh Van Loc Truong Hoa Hiep

Ngo Duc Huy Vo Bao Toan