Part E - Proceedings Articles Theme 4 - Mechanical, Manufactuirng and Industrial Engineering - Final

7th International Conference on Engineering Research, Innovation and Education
January 12-14, 2023, Sylhet, Bangladesh
PAPER ID: 032
Study the impact of SINGLE MINUTE EXCHANGE OF DIE (SMED)

in a particular RMG factory
Mrinmoy Kanti Dey*, Syed Misbah Uddin
Industrial and Production Engineering Department

Shahjalal University of Science and Technology, Sylhet, Bangladesh.
Email:mrinmoydey181@gmail.com,misbah-ipe@sust.edu
*Corresponding author [Mrinmoy Kanti Dey]
Abstract: The basic idea behind lean manufacturing is to eliminate

Keywords:
waste, which does not add value to the final product. Lean manufacturing
● Lean manufacturing; is one of the activities that focus on cost reduction by eliminating non-
value-added activities. The core purpose behind this research work was
● SMED tool;
to study a lean manufacturing concept in an RMG factory in Bangladesh.
● Changeover time; In this concern, Single Minute Exchange of Die tools and techniques have
● Non-productive time; been studied to find out how much it can contribute in reducing
● Time-study; changeover time. The whole study was conducted in the sewing section
of a particular RMG factory. A total of four sewing lines were selected.
● Non-value added
In every line the total operation time before and after the application of
activities; the SMED tool was measured through time study. The related data were
collected both from direct observation and interviewing responsible
personnel. The whole analysis was conducted by MS EXCEL software.
The end result shows that SMED reduces non-productive time by
streamlining and standardizing the operations using simple techniques
and easy applications. In every line there were more than a fifth percent
(20%) of total changeover time was reduced. This signifies that a huge
amount of time is being wasted due to the changeover.
1. Introduction
Bangladesh is a fast-growing economy powered by the readymade garments (RMG) industry which has
promoted the country in the world through the motto ‘Made in Bangladesh. The overall impact of the
readymade garments exports is certainly one of the most significant social and economic developments in
contemporary Bangladesh. But to maintain this growth in a positive direction it is necessary to ensure proper
utilization of every resource. In addition to this, in today’s competitive world, the most important driver for
success is time; the company that delivers goods with a shorter lead time is the market winner.
Page | 1232
In order to survive in this type of market any industrial unit has to reduce costs and simultaneously produce
quality products with shorter downtime [R. Godina, E. M. G. Rodrigues, and J. C. O. Matias, 2018].
However, the companies intend to introduce improvements at the level of the change of format in order to
minimize downtime and thus achieve its objectives. One of the best manners to reduce downtime is through
Lean manufacturing [D. McCarthy and N. Rich ,2015].
In Bangladesh, there are different types of lean tools like VSM, Kaizen, 5S, kanban are used in RMG
factories for process improvement. But, during an ongoing production, one of the Lean manufacturing tools
which can be used for reducing downtime is SMED – an acronym for “Single Minute Exchange of Die”,
which can be translated as "quick tool change". In practice, SMED is a set of techniques belonging to lean
manufacturing that aim to reduce the setup time of a machine. When properly applied, it allows machines
to take less time to attach, giving more flexibility to the production line [M. D. la Vega-Rodríguez et al,
2018]. SMED provides greater flexibility in die changing, by reducing changeover time. This is achieved
by separation of internal and external activities. It involves analyzing all activities performed during
changeover so that non-value added activity can be eliminated or can be converted into external activity
[Change Over Time Reduction Using SMED, 2018]. Moreover, in order to meet this shorter downtime and
because of global competition, manufacturing systems have to adopt new production methods [Bicheno
John, (2000)]. The method itself is mainly focused on significant reduction of the time required for the
changeover of machinery. SMED involves three basic terms such as change over time, internal activity and
external activity. [ Deepak Maurya et al,2018].
Changeover time is the total elapsed time between the last units of good production of the previous run,
to the first unit of good production of the succeeding run.
Internal activity is an activity which is performed after the machine is stopped.
External activity is an activity which can be done without stopping the machine.
The aim of this research work is to study the influence of lean manufacturing concept SMED in a RMG
factory of Bangladesh. It brings the outcome of a close observation held at sewing floor of a RMG factory.
The end result shows that SMED has a significant effect in reducing changeover time.
2. Experimental
Garments manufacturing system is a technological process of making a complete part by arranging several
man and machine according to some sequential operations where are obliged by some terms and conditions
to maintain the customer satisfactory level and quality. When manufacturing large-scale products, some of
the internal and external elements and tool exchanges take place where changeover time always plays a
vital role for slow production. A rapid changeover is widely acknowledged as an essential prerequisite to
flexible and responsive production. The SMED system is a method that make possible to perform
equipment setup and change over operations within 10 minutes. The five steps of the SMED are as follows
[A. P. Dillon and S. Shingo]
2.1 Steps followed in SMED Implementation
 Step One – Identify Pilot Area
 Step Two – Identify Elements
 Step Three – Separate External Elements
 Step Four – Convert Internal Elements to External
 Step Five – Streamline Remaining Elements
Page | 1233
2.1 Identify pilot area
In this step the areas for SMED implementation is selected. Usually the area is the production floor.
Employees who are familiar with the equipment (operators, maintenance personnel, quality assurance, and
supervisors) must be engaged and motivated. In this research a sewing floor total four sewing lines were
selected for the observation.
Figure 2.1: SMED process steps [www.whatissixsigma.net]
2.2 Identify Elements

In SMED, changeover elements are divided in to two categories.
An external activity involves operations that can be done while the machine is running and before the
changeover process begins. Internal Elements are those that must take place when the equipment is stopped.
Table 2.1 External and Internal activities

External Elements Internal Elements
 Sample collection  Change machine
 Trim card arrangement  Folder adjustment
 Pre-production meeting  Input setting
 Prepare layout sheet  Clean machine
 Identify and analyze critical operation  Needle change
 Select exact folder  Thread change
 Prepare folder  Pressure feed change
 Arrange special machine  Set pattern
 Make pattern  Set measurement
2.3 Separate External Elements

In this step, elements of the changeover process that can be performed with little or no change while the
equipment is running are identified and moved “external” to the changeover (i.e. performed before or after
the changeover). It is not unusual for changeover times to be cut nearly in half with this step alone.
Page | 1234
2.3 Convert Internal Elements to External
In this step, the current changeover process is carefully examined, with the goal of converting as many
internal elements to external as possible. This list should be prioritized so the most promising elements are
acted on first. Once the list has been prioritized work can begin on making the necessary changes.
2.4 Streamline remaining elements

In this step, the remaining elements are reviewed with an eye towards streamlining and simplifying so they
can be completed in less time. First priority should be given to internal elements to support the primary
goal of shortening the changeover time.
3. Calculation
Data Analysis is the process of applying techniques to describe and evaluate data. At the beginning of the
data analysis the time of all external and internal activity were arranged in a table. The overall calculation
was carried out in MS EXCEL. Details of data analysis are discussed below.
3.1 Analysis of External activity
Time against all of the external activities were tabulated in a excel sheet. Each activity consists data from
four lines. The summation of data (in minute) in each line were calculated then
Table 3.1: Different types of External activity (in minute)
External activity line5 line 6 line 7 line 8
Sample collection 12 15 16 12
Trim card arrangement 8 17 15 14
Pre-production meeting 22 34 17 30
Prepare layout sheet 34 5 34 3
Oil change of machines 7 8 6 10
Identify & analysis critical operation 5 14 13 12
Input 0 22 0 25
Select exact folder 33 38 22 40
Prepare/Make folder 32 0 38 0
Arrange special machines 20 68 0 65
Make pattern 28 32 68 32
Total 201 253 229 243
3.2 Analysis of internal data

The procedure is almost as same as internal. Time against all of the external activities were tabulated in a
excel sheet. Each activity consists data from our lines. The summation (in minute) of data in each line were
calculated then.
Table 3.2: Different types of Internal Activity (in minute)
Internal Activity line5 line 6 line 7 line 8

Change machines 110 105 115 120
Folder adjustment 115 135 132 130
Input setting 28 28 34 30
Clean machines 50 42 44 40
Page | 1235
Needle change 43 41 36 35
Thread change 40 13 18 13
Pressure feed change 20 12 19 17
Set pattern 30 32 34 35
Set measurement 40 35 23 25
Technician teach to operator about process 90 105 95 90
Checking the parts 93 93 97 95
Operator run the process 98 98 125 130
checking output 5 7 5 6
Total 762 746 777 766
3.3 Overall time in four lines

Table 3.3: Total time (in hour) in both activity
IT(min.) ET(min.) TOTAL(hour)
line5 762 201 16.05
line6 746 253 16.65
line7 777 229 16.76
line8 766 243 16.81
4. Results and discussion

The result shows that before implementing SMED the total time in line5, line6, line7 and line8 are 16.05,
16.65, 16.76 and 16.81hour respectively. However, after practicing SMED the external tasks get finished
by the responsible departments. For example, the activities like sample collection, trim card arrangement
and pre-production meeting are completed by quality control department. As a result only the time of
internal activities will come into consideration after SMED implementation.
Table 4.1: Overall improvement
Before After Time Save Change

(hour) (hour) (hour) (%) 20
line 5 16.05 12.7 3.35 21% 15
line 6 16.65 12.43 4.21 25%
10
line 7 16.76 12.95 3.81 23%
line 8 16.81 12.76 4.05 24% 5
Total 66.27 50.84 15.42 93% 0

Average 16.57 12.71 3.86 23% line5 line6 line7 line8
Before After
The total time in performing internal activities in above mentioned lines are 12.7, 12.43, 12.95, 12.76 hour
respectively. The total time saved in each lines are 3.35, 4.21, 3.81 and 4.05 hours. Highest improvement
was 25% in line 6 and in line 5, line 7, line 8 it was 21%, 23% and 24% respectively.
Page | 1236
4.1 Discussion
This project work was focused on studying how the changeover time is reduced by practicing a lean
manufacturing tool name SMED. To conduct the study four lines in the sewing section of a particular RMG
factory was chosen. The SMED methodology which is followed in this study is based on the separation of
internal and external operations as well as the conversion of internal to external operations. After studying
the existing scenario it was found that activities were converted successfully from internal to external in a
definite time, a changeover board is used where the responsible personnel make it sure that all external
activities are finished before the start of assembly production. The whole analyses was done in MS EXCEL
(2013 version). Based on a series of data collected during the setup activities in a woven garments which
produces long sleeve of men it is found out that on an average 3.86 hour was being wasted during the
changeover.
5.0 Conclusions
Nowadays the flexible processes require a high changeover frequency. However, in order for the small lot
productions to be profitable, the tool changeover has to be the shortest possible. One way to reduce the
changeover time is through SMED. In this paper a particular RMG industry was selected to study the impact
of SMED. A comparison of results and achievements before and after the SMED implementation was made
to measure the effectiveness of SMED to reduce setup time. The aim was to study and observe how SMED
can significantly reduce machine downtime during operations. The result shows that a great amount of time
is being saved by applying SMED. This also signifies that a momentous amount of time is being wasted
due to changeover.
References
1. A. P. Dillon and S. Shingo, A Revolution in Manufacturing: The SMED System. CRC Press, 1985
2. Bicheno John, (2000). The Lean Toolbox: Picsie Books, Buckingham, England, PP-21-24.
3. Change Over Time Reduction Using SMED: An Industrial Case Study. Volume 9, Issue 3, March-2018
13 ISSN 2229-5518.
4. Deepak Maurya, Yagnesh Yadav, Deepak Pandey ,Prof Rajeshwar.S.Deshmukh “ Change Over Time
Reduction Using SMED: An Industrial Case Study”, International Journal of Scientific & Engineering
Research Volume 9, Issue 3, March-2018 13 ISSN 2229-5518
5. D. McCarthy and N. Rich, Lean TPM: A Blueprint for Change. Butterworth-Heinemann, 2015.
6. https://www.whatissixsigma.net/wp-content/uploads/2018/05/smed-process.jpg
7. M. D. la Vega-Rodríguez, Y. A. Baez-Lopez, D.-L. Flores, D. A. Tlapa, and A. Alvarado-Iniesta,
“Lean Manufacturing: A Strategy for Waste Reduction,” in New Perspectives on Applied Industrial Tools
and Techniques, Springer, Cham, 2018, pp. 153–174
8. R. Godina, E. M. G. Rodrigues, and J. C. O. Matias, “An Alternative Test of Normality for Improving
SPC in a Portuguese Automotive SME,” in Closing the Gap Between Practice and Research in Industrial
Engineering, Springer, Cham, 2018, pp. 277–285
Page | 1237
PAPER ID: 041
DESIGN AND CONSTRUCTION OF HYBRID COOKSTOVE AND

ITS COMPARATIVE PERFORMANCE EVALUATION WITH
TRADITIONAL STOVE
Riffat Jahan Juthi, Muhammad Rashed Al Mamun*, Gulam Kibria Robbani,

Shermin Aktar
Department of Farm Power and Machinery, Faculty of Agricultural Engineering and
Technology, Sylhet Agricultural University, Sylhet, Bangladesh
Email: riffatjahanjuthi@gmail.com; rashed.fpm@sau.ac.bd; kibria.info0@gmail.com
and shermin.fpm@sau.ac.bd
*Corresponding author: Muhammad Rashed Al Mamun
Abstract: One of the prime sources of indoor air pollution is household

Keywords: cookstoves. Traditional cookstoves made of clay are mostly used for cooking
in rural areas in Bangladesh. Biomass are used as fuel for this purpose which
● Hybrid Stove; emits huge smoke, resulting indoor air pollution and health problems. Besides,
● Traditional these stoves have a significant contribution to global climate change.
Unsustainable wood harvesting for cooking adds to deforestation, which
Cooking Stove;
reduces carbon uptake by forests. To prevent these harmful effects a greener
● Renewable energy; solution is required. In this study, a Hybrid Stove (FPM-HS) was designed,
fabricated, and tested to evaluate its performance. The stove was made of
● Solar;
concrete and operated in a dual fuel mode. The Hybrid Stove (FPM-HS) is
● Biogas. composed of two sections. These are- Solar Electric Section and Biogas
Section. In the availability of sunlight, the Solar Electric Section is used and
in the absence of sunlight, the Biogas Section is used. The thermal efficiency
of the biogas section of the manufactured Hybrid Cookstove was 30.87% and
the burning rate 12.16g/min. The efficiency of the solar electric section of the
Hybrid Stove was 72.07%. Cooking rice in the biogas section of the hybrid
stove takes less time than that in a household gas stove. The research suggested
that this technology could be beneficial for cooking in rural areas, lowering
environmental pollution, and solving the energy crisis.
1. Introduction
One of the most popular and traditional methods of cooking in rural areas of the majority of
developing nations is the use of biomass as fuel. However, burning this fuel has negative effects
on the environment and human health. Household air pollution (HAP) from cooking with solid
fuels has become a significant cause of death and disability in many developing countries,
including Bangladesh (Khan, 2017) . Approximately 3 billion of the world's poorest population
rely on biomasses burned on ineffective and highly polluting stoves for cooking. About 4.3 billion
deaths occurred due to this pollution in 2012 (WHO, 2014). To meet WHO’s recommendations
for indoor air quality, emission rates of stoves must be hundreds of times lower than those for
conventional stoves using solid fuels but the intervention with this stove rarely shows an
improvement in health outcomes (Weyant, et al., 2019). The use of biofuels significantly affects
Page | 1238
household budgets, women's time and activities, gender roles and relationships, safety and
hygiene, and the environment worldwide. According to estimates, 50% of the world's harvested
wood is utilized as fuel (Administration, 1995).
Gas stoves are also used in rural regions. The emission of indoor air pollutants in this stove is
much less but the drawback here is that the gas supply is inaccessible in all areas. People have to
purchase LPG gas cylinders for cooking but it may not be available in rural areas and even if it is
available, it may not be affordable to the people of those regions.
The use of electric stoves has increased in both urban and non-urbanized areas. Compared to gas
stoves, electric stoves cost less money at the time of purchase. These stoves are eco-friendly as
they don't emit any kind of gas. The supply of electricity and load shedding can be a matter of
concern for rural areas while using electric stoves. To solve this issue solar energy can be used to
provide electricity to electric stoves through solar panels.
This study was initiated to overcome the drawbacks of the available stoves in the rural market by
designing a hybrid cooking stove. It can run on dual fuel mode. It is designed and constructed by
utilization of biogas and solar energy. Its performance was tested and compared with that of
different types of available stoves in the market. It emits lesser indoor air pollutants than
traditional, LPG gas and electric cook stoves. The stove is more sustainable than traditional stoves
as renewable energy is being used here.
2. Methodology
2.1 Materials
Careful consideration is needed to select appropriate materials, and equipment for designing the
Hybrid Stove. The materials used to construct the stove are Portland cement, sand, gravel, mild
steel rod, gas burner, control knob, electric coil plate, gas flow pipe, and nozzle. The tools and
machines used are measuring tape, steel scale, screwdriver, welding machine, hammer, mortar
and pestle, wooden mold, pointing trowel, finishing trowel, color, and brush.
2.2 Design Consideration

Size, material cost, durability, portability, consumer price, marketability, approximate cooking
time, indoor air pollution, assembly, safety, repair & maintenance, availability of materials,
cooking power or overall efficiency, etc. characteristics were considered while designing the
stove.
2.3 Design Parameters

The stove was designed to meet the cooking energy requirements of a family of five. The Hybrid
Stove will be 10.5 inches long, 11.5 inches wide, and 6.5 inches high. The burner hole should be
5.5 inches in diameter. The gas burner should have a 4-inch diameter. It should have 45 holes in
it. The gas pipe should have a 1.5-inch diameter. The tray should be 10 inches long, 9.2 inches
wide, and 0.5 inches thick, correspondingly. The electric burner's diameter should be 7.5 inches,
and its height should be 0.5 inches. The electric coil should be 1000W.
2.4 Computer-Aided Design
The FPM-HS was designed according to the determined specified dimensions using AutoCAD
2020. The design of the hybrid stove is shown below:
Page | 1239
Figure1: Design of Hybrid Stove when tray is closed and opened
2.5 Method of Fabrication
At first, the dimensions were specified for all components. The components were fabricated
separately in a metal shop. Then, sand and cement were mixed with a pointing trowel. A wooden
mold was used to give shape to the concrete stove. A finishing trowel was used to provide a
smooth finish to the concrete structure. Then parts were assembled in the lab. A welding machine
was used for the attachment of the gas pipe. The attachment of the pipe and burner with a nozzle
was done with a wrench. Lastly, painting and other aesthetic tasks were completed.
2.6 Experimental Method
The equipment used to test the performance of the stove are: a K-Type Thermocouple, a
Stopwatch, and a Weight Balance. The thermal efficiency and burning rate of the biogas part of
the hybrid stove were determined by the Water Boiling Test (Rob Bailis, 2007). The test was
repeated five times with a mass of 2000g, 1750g, 1500g, 1250g, and 1000g of water, and then the
average of the values of the parameters was determined.
The performance test of the solar electric part of was done by boiling 1Litre of water on the stove.
The time taken to boil a given quantity of water was recorded. Efficiency of the stove was
calculated.
The functionality of the stove was also tested by cooking different masses of rice and measuring
their time of cooking. The temperature of the flame was also recorded.
2.7 Data Analysis
Recorded data from this study were coded and entered into a Microsoft Excel worksheet,
organized, and processed for further analysis.
3. Calculation
3.1 Calculation of Thermal Efficiency of Biogas part of Hybrid Stove
It was measured by the following equation:

𝑄ℎ𝑒𝑎𝑡𝑖𝑛𝑔
η=
𝑄𝑐𝑜𝑚𝑏𝑢𝑠𝑡𝑖𝑜𝑛
𝑐 × 𝑚 × 𝛥𝑇
= (1)
𝑞 × 𝑚՛
Page | 1240
3.2 Calculation of Thermal Efficiency of Solar Electric Section of Hybrid Stove
The efficiency was measured by Eq. (6):
Efficiency, η = 𝑂𝑢𝑡𝑝𝑢𝑡 𝐸𝑛𝑒𝑟𝑔𝑦

𝐼𝑛𝑝𝑢𝑡 𝐸𝑛𝑒𝑟𝑔𝑦
(𝑚𝑖−𝑚𝑓)×𝐶𝑣+(𝑇𝑓−𝑇𝑖)×𝑚𝑓𝐶𝑝
⸫η= (2)
𝑃𝑡
4. Result and Discussions
After accomplishing all the steps mentioned the FPM-HS was fabricated (Fig.2).
Figure 2: Hybrid Stove when tray is closed and opened
In the case of the electric section of the Hybrid Stove, solar energy was utilized to cook. When
there is availability of sunlight, the solar panel converts solar energy to electricity. When there is
no availability of sunlight, the tray is opened, so the electric burner comes out and the gas burner
appears. The nozzle is turned on so that the gas from the gas line pipe enters the burner.
4.1 Performance Test of Biogas Section of Hybrid Stove
40 16
Thermal Efficiency (%)
35
Burning Rate (g/min)
14
30
12
25
20 10
15 8
10
6
1000 1250 1500 1750 2000
1000 1250 1500 1750 2000
Weight of Water (g) Weight of Water (g)
(a) (b)
Figure 3: (a)Thermal Efficiency and (b)Burning Rate at different weight of water
The result of the analysis in Fig.3(a) is clearly showing that there is a strong positive relationship
between weight of water and thermal efficiency. Hence, it is clear that with the increase in weight
of water, the thermal efficiency increases. The average thermal efficiency is 30.87%. In Figure
3(b), it was demonstrated with the increase in weight of water, the burning rate decreases. Due to
Page | 1241
the continuous use of fuel, the pressure of the gas has decreased. As a result, the burning rate has
also decreased. The average burning rate 12.16g/min.
4.2 Performance Test of Solar Electric Section of Hybrid Stove
Thermal Efficiency was found 72.07 %
4.3 Comparative Study of Cooking between Hybrid and Traditional Stove
Temperature:
Figure 4 shows comparison of temperatures of flame while cooking rice of biogas section of hybrid
stove and household gas stove while cooking different masses of rice. It is also clear that all
temperature of flame of the Hybrid Stove was more than that of the household gas stove while
cooking 250g and 500g rice except 750g of rice
700
Temperature (℃)
640
580
520
460
400
250 500 750
Mass of Rice (g)
Household Gas Stove
Biogas Section of FPM-HS
Figure 4: Comparison of temperatures of flame while cooking rice
Time
Figure 6 shows the comparison between the time to cook rice in hybrid stove and household gas
stove. In both cases the time to cook was increased with the increase in mass of rice. It is also
clear that the time to cook rice in Hybrid Stove was less than that in the household gas stove.
21 45
Time to Cook Rice (min)
Time to cook (min)
35
16
25
11 15
5
6
250 350 450 550 650 750
250 500 750 Mass of Rice
Mass of Rice (g) Household electric Stove
Household Gas Stove
Biogas Section of Hybrid Stove electric Section of Hybrid Stove
Figure 5: Comparison on the time to cook rice
Page | 1242
5. Conclusion
This study demonstrates the solar-biogas hybrid system for cooking in dual fuel mode, which
lowers greenhouse gas emissions and ultimately averts global warming. The stove facilitates less
daily fuel usage by utilizing free solar energy, and it promotes season-independent cooking, as
biogas can be used when solar energy is insufficient. It was observed that it takes lesser time to
cook rice in the hybrid stove than that in a household gas stove. So, the study concluded, the
Hybrid Stove is an excellent alternative to a traditional cook stove. Not only is its performance
better than the traditional stoves it also uses renewable energy to cook.
Acknowledgment
This research was supported by the Ministry of Science and Technology, and Sylhet
Agricultural University. The authors would like to thank kibria for her cooperation during
sample collection and preparation.
References
Administration, E. I. (1995). Renewable Energy Annual 1995. Washington DC: Officeof Coal,
Nuclear, Electric and Alternation Fuels, US Department of Energy, Washington DC, 20585.
Khan, M. B. (2017). Household air pollution from cooking and risk of adverse health and birth
outcomes in Bangladesh: a nationwide population-based study.
Rob Bailis, V. B. (2007). Performance testing for monitoring improved biomass stove
interventions: experiences of the Household Energy and Health Project1 1This paper is one of
six describing work done as part of the Household Energy and Health (HEH) Project,
coordinated by the In. Science Direct, 57-70.
Weyant, C. L., Thompson, R., Lam , N. L., Upadhyay , B., Shrestha , P., Maharjan, S., . . .
Pokhrel , A. K. (2019). In-Field Emission Measurements from Biogas and Liquified Petroleum
Gas (LPG) Stoves. MDPI.
WHO. (2014, November 09). Air pollution: Indoor air pollution. Retrieved from World Health
Organization: https://www.who.int/news-room/questions-and-answers/item/air-pollution-
indoor-air-pollution
Page | 1243
International Conference on Engineering Research, Innovation and Education
PAPER ID: 050
Causes and Consequences of Bullwhip Effect on Boutique

Industry of Dhaka City
Rukhsana Ahmed Noshin1*, Choudhury Abul Anam Rashed1* and Minhazul Islam1
1
Department of Industrial And Production Engineering, Shahjalal University of Science &
Technology (SUST), Sylhet-3114, Bangladesh
rukhsanaahmed043@gmail.com, rashed-ipe@sust.edu, minhaz74692@gmail.com.
ABSTRCT: Bullwhip effect is a burning issue in today’s supply chain. Every echelon of the
supply chain faces the negative consequences of BWE somehow. So, it is important to find out
the causes behind the bullwhip effect to mitigate the consequences. Boutiques industry in
Bangladesh is a rapidly growing industrial sector. In this study, we focused on finding the causes
and consequences of bullwhip effect on boutique industry of Dhaka city. The main objectives of
this study are to find out the causes behind the bullwhip effect, to identify which causes are the
most significant from the perspective of Dhaka city, and the main consequences of bull whip
effect. Through studying previous literatures sixteen common causes have been identified behind
bullwhip effect. In this study survey-based method has been used and respondents are chosen
through clustered sampling. Through a structured questionnaire the necessary data have been
collected. In this study among all the 16 causes 6 causes are found as the most significant causes
from the perspective of retailers and wholesalers of Dhaka city. SPSS Version 26 has been used
for statistical analysis to make final decision. We also found ten consequences those are
commonly faced by these two echelons of boutiques’ supply chain because of bullwhip effect,
these are excessive inventory cost, manpower wastages and higher labor cost, higher
replenishment lead-time, higher transportation cost, tension in buyer supplier relationship,
product unavailability, loss of profit, poor customer service etc.
KEYWORDS: Supply chain Management; Bullwhip Effect; Boutique Industries; Causes;

Consequences.
1. INTRODUCTION
Bangladesh's boutique fashion business has risen swiftly and has become an important element
of the economy. There are many boutique stores in our country that provide a wide variety of
product kinds to meet the local demand for apparel. Boutique house manufacturing extends to a
variety of distinct product tiers combined into a single product line. Productions have
ramifications for both operations and marketing. The producer provides items to fulfill the needs
and demands of merchants and customers. As a result, the boutique fashion business has a wide
range of products, short product life cycles, volatile market demand, and stringent supply
methods. (Şen, 2008).Every boutique industries have a particular supply chain management of
their own. A supply chain consists of all parties involved, directly or indirectly in fulfilling the
customers’ requests. Not only the suppliers and customers, recent interest in supply chain
management centers around co-ordination among various members of supply chain comprising
manufacturers, distributors, wholesalers, retailers.
Page | 1244
One important mechanism for the coordination in a supply chain is the information flows among
the members of the supply chain. These information flows have a direct impact on the product
scheduling, inventory control and delivery plans of individual members of supply chain (Hau L.
Lee, 1997). This information flows sometimes have some distortions and these distortions can be
considered as a term called “bullwhip effect”. Bullwhip effect occurs when small order variability
at customers level amplifies the orders for upstream players, such as wholesalers and
manufacturers, as the orders moves up along the supply chain. The term, bullwhip effect, first
coined by the logistics executives of Procter and Gamble (Lee, 1997).
Consumers demand
200 Retailer's order to wholesaler
wholesaler's order to manufacturer
manufacturers' order to supplier
150
100
ORDER QUANTITY
50
0
1 2 3 4 5 T6IME 7 8 9 10 11 12 13 14 15 16 17 18
Figure 1: Order variability during times

The objectives of the study are
1. To identify the causes of bullwhip effect.
2. To analyze which causes are more influential on bullwhip effect from the perspective of
selected cluster of boutique industry in Dhaka.
3. To identify the consequences of bullwhip effect in boutique industry.
1.1. CAUSES OF BULLWHIP EFFECT
The causes behind the BWE can be classified into two categories. They are (1) operational causes
and (2) behavioral causes.
Operational causes include the following cause:
1. Demand forecasting, 2. Order Batching, 3. Price fluctuation, 4. Shortage gaming and rationing,
5. Replenishment policy, 6. Inventory policy, 7. Lead time. 8. Lack of transparency, 9. Improper
control system, 10. Lack of synchronization, 11. Misperception of feedback, 12. Number of
echelons, 13. Multiplier Effect, 14. Capacity limits, 15. Company processes
According to Lee HL, Padmanabhan V, Whang S (1997) the interaction between two rational echelon
in a supply chain is also a factor behind bullwhip effect. Behavioral causes includes fear of empty stocks
(Hau L. Lee, 1997).
Page | 1245
1.2. CONSEQUENCES OF BULLWHIP EFFECT:
In several research we found some consequence due to bullwhip effect. The most common
consequences are given bellow:
1. Excessive inventory cost (Hau L. Lee, 1997), (Sunil C, 2007).
2. Higher Manufacturing Cost (Hau L. Lee, 1997), (Leedy & Ormrod, 2001)
3. Insufficient capacity (Sun HX, 2005)
4. Product unavailability (Heydari Jafar, 2009), (Sunil C, 2007)
5. Man power wastage and higher labor cost (Sunil C, 2007)
6. Poor customer service & loss of goodwill (Hau L. Lee, 1997)
7. Tense buyer supplier relationship (Blinder, 1982), (Sunil C, 2007)
8. Higher transportation cost (Sunil C, 2007)
9. Higher replenishment lead time (Jr, 1965), (Jakšič M, 2008)
10. Loss of profit
2. RESEARCH METHODOLOGY
Among the several approach in quantitative research like content analysis, experiment, use of
secondary data and survey method, a structured questionnaire-based survey approach is employed
in this study
3. DATA ANALYSIS & FINDINGS
Cronbach's alpha is a metric for determining the internal consistency or reliability of a set of scale
or test item.
Table 1: Reliability statistics of retailers’ opinion
Cronbach’s Alpha No of Item
.661 15
Table 2: Reliability statistics of wholesales’ opinion
Cronbach’s Alpha No of Item
.655 16
In table 1 the alpha coefficient for the fifteen variable is .661, suggesting that the variables have
relatively good internal consistency. In table 4.2 the alpha coefficient for the sixteen variable is
.655. This value is also suggesting that the variables have relatively goof internal consistency.
Page | 1246
ASSESSMENT OF RESPONSIBILITIES OF IDENTIFIED CAUSES ( ON THE BASIS OF THE
RETAILER’S POINT OF VIEW) BEHIND THE BULLWHIP EFFECT (SAMPLE SIZE N=100
90
80
70
60
50
40
30
20
10
0
Highly Responsible Responsible Neutral Less Responsible Not Responsible
Figure 2: The responsibilities of the identified causes behind bullwhip

effect
3.1. CROSS TABULATION
Cross tabulation is performed to check dependencies between two variables. If the P value of
Pearson Chi-Square’ is greater than 0.05 then there will be dependency between those variables.
Here we checked the dependencies between each cause and the bullwhip effect (Qualtricks, 2021).
Cross tabulation between the causes and bullwhip effect has been analyzed to find the dependency
among these.
3.2. CORRELATON AMONG THE VARIABLES:
Pearson correlation is performed to identify the interrelationship between the variables (causes).
As we know that the value of Pearson correlation should be within +1 to -1. A positive correlation
arises when one variable increases at the same time as the other, i.e. when the high numerical
values of one variable correspond to the high numerical values of the other. Negative correlation
arises when one variable drops as the other grows, i.e. when one variable's high numerical values
correspond to the low numerical values of the other.
Page | 1247
3.3. FACTOR ANALYSIS
The Kaiser-Meyer-Olkin (KMO) test is a sampling adequacy assessment. The value should be
greater than or equals to 0.5. here this value is .556 which is greater than 0.5 (standard value for
factor analysis).
Table 3: KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy. 0.556
Approx. Chi-Square 118.679
Bartlett's Test of Sphericity Df 78
Sig. 0.002
3.4. FACTOR ANALYSIS DATA REDUCTION FOR CAUSES
Table 4: Communalities for factor analysis

Communalities
Initial Extraction
Demand Forecasting 1.000 0.632
Order Batching 1.000 0.622
Price Fluctuation 1.000 0.612
Inventory Policy 1.000 0.776
Replenishment Policy 1.000 0.760
Improper Control Processes 1.000 0.714
Lack of Transparency 1.000 0.691
Number of Echelon 1.000 0.589
Multiplier Effect 1.000 0.590
Misperception of feedback 1.000 0.562
Company Process 1.000 0.545
Capacity Limit 1.000 0.712
Fear of Empty Stock 1.000 0.562
Extraction Method: Principal Component Analysis.
It shows that each factor has loading greater than .5, which suggest their accuracy valid for factor
analysis. According to thumb rule variables having factor loading in communalities less than .5
explain less than 50% of total variance. So, these variable can be consider as least significant.
Here, lead time, rationing and shortage gaming & lack of synchronization are eliminated as their
factor loadings are less than 0.5. Principle components analysis with varimax orthogonal rotation
is used to reduce total 16 causes into 13 causes which are relatively more responsible for bullwhip
effect.
Page | 1248
3.4.1. PRINCIPLE COMPONENT ANALYSIS
The goal of principal component analysis is to reduce a big collection of variables to a small set
that retains the majority of the information in the large set. The principle component analysis
approach allows us to construct and use a smaller collection of variables known as main factors.
In this technique “Total Variance Explained Table” has been used to develop several components
as the main causes in which all 13 causes are grouped those obtained from communalities value
previously.
3.4.2. ROTATED COMPONENT MATRIX
Table 5: Rotated Component Matrix

Rotated Component Matrixa
Component
1 2 3 4 5 6
Demand Forecast 0.737
Price Fluctuation 0.602
Order Batching -0.508
Lack of Transparency 0.761
Fear of Empty Stock 0.635
Misperception of Feedback 0.628
Capacity Limit .838
Number of Echelon .539
Improper Control Process 0.807
Company Process 0.556
Replenishment Policy 0.831
Inventory Policy 0.809.
Multiplier Effect 0.641
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.a
a. Rotation converged in 17 iterations.
Now the obtained components have to be named based on which variable or causes load high
with which factor. And variables with a high load will provide a good idea of what that factor is
measuring or suggesting. We read those variables word by word, observing what they're
measuring and what they have in common. That is the proper naming for the component.
Followings are the named six components, which are responsible the bullwhip effect according
to the retailers’ point of view.
Based on the name of the causes or variables associated with the components, the naming are
described below. The numeric value of the respective variables describes how much the variables
are contributing to the formed components.
1. Prediction of Ordering Quantities: Demand forecasting (.737), price fluctuation (.602) and
order batching (-.508) can be termed as prediction of ordering quantities. Among these elements
demand forecasting and price fluctuation explains the largest correlation with this factor, as the
demand of customer or how amount of product customer will order can be changed according to
the fluctuation of the price of that product, so these causes explains highest contribution to this
factor.
Page | 1249
1. Information Lacking: Lack of transparency (.761), fear of empty stock (.635),
misperception of feedback (.628) can be termed as information lacking. The lack of
transparency shows the highest contribution to form this factor. It can be told that there
should be transparency among the echelon. Insufficient transparency shows the highest
contributing element to information lacking. This is named as information lacking
because the causes grouped into this component are related to the flow of information.
2. Limitation of Echelon: Capacity limit (.838) and number of echelons (.539) are termed
as limitation of echelon. Capacity limit is the most contributing element to this component
explaining the largest correlation with this component.
3. Company’s Strategy: Improper control process (.809) and company process (.556) can
be termed as company’s process. Here improper control process are the most contributing
element to this factor explaining the largest correlation with this factor. This factor is
named as company strategy because all the element in this factor related to the strategies
that a company follows.
4. Ordering Policy: Replenishment policy (.831) and inventory policy (.809) can be termed
as the ordering policy. These two causes or elements are almost equally contributing to
this factor. Both element indicate how and in which way the company order product from
their supplier, that’s why this component named as ordering policy.
5. Multiplier Effect: As in this factor there is single element only which is multiplier factor
(.641). So, it is named as same as the element.
3.4.3. SCREE PLOT
Figure 3: Scree Plot
From figure it is noticed that the 1st elbow occurs on point 3 in the curve. So it can be concluded
that 1st 2 components that are explained before are the most influencing factor in bullwhip effect.
From table we found demand forecasting, price fluctuation and order batching grouped in 1 st
components as the name of ordering pattern and fear of empty stock, lack of transparency and
misperception of feedback are grouped in 2nd component with the name information lacking.
Therefore, it can be concluded these three causes are the most influencing in bullwhip effect.
Page | 1250
3.5. REGRESSION ANALYSIS
Regression analysis is the process of estimating the dependency of on variable (dependent

variable) on one or more other variables (independent variable). Regression analysis is performed
to see how much the bullwhip effect is affected by the six groups or components. Considering
overall bullwhip effect as the dependent variable and prediction of ordering quantities,
information lacking, limitation of echelons, company’s strategy, ordering policy and multiplier
effect as the independent variable the regression model can be generated as follows:
Y= β0 + β1 X1+ β2 X2 +…..+ βp Xp +Ɛ
Where Xj are the independent variables
Y is the dependent variable
β0 is the regression constant
βj are the regression coefficients of independent variable
Ɛ is the error term
3.5.1. MODEL SUMMARY
Table 6: Model Summary

Model R R Square Adjusted R Square Std. Error of the Estimate
1 .794a 0.630 .622 0.646743
a. Predictors: (Constant), prediction of ordering quantities, information lacking
b. Dependent Variable: Bullwhip Effect
This represents the model summary of regression analysis. In this table R is the multiple
correlation coefficients. The value of R is 0.794. R2 indicates that how close the data are
fitted to the regression line. The value of R2 is 0.630 which means 63.0% of the total
variation in bullwhip effect has been explained by the predictor variables.
3.5.2. ANOVA
Table 7: Anova
ANOVAa
Sum of Mean
Model Df F Sig.
Squares Square
Regression 69.067 2 34.533 82.561 .000b
1 Residual 40.572 97 0.4182
Total 109.64 99
a. Dependent Variable: Bullwhip Effect
a. Predictors: (Constant), prediction of ordering quantities, information lacking
The F-ratio in the ANOVA table (table 4.28) tests whether the overall regression
model is good fit for the data or not. The table represents that independent
Page | 1251
variables significantly predict the dependent variable (2, 97) = 82.561, P<0.05 so
the model is a good fit to the data
3.5.3. COEFFICIENT
The significant value (P-value) of regression coefficients suggests the apart from
the “information lacking” all other variables have significant influence on
bullwhip effect.
Table 8: Coefficient
Unstandardized Standardized
Model Coefficients Coefficients
Std.
B Beta t Sig.
Error
Constant 0.033 0.319 .103 .918
Prediction of 0.480 0.066 0.466 7.301 0.000
Ordering
Quantities
Information 0.600 0.071 0.537 8.420 0.000
Lacking
Unstandardized coefficient are the ordinary least squares estimates of regression model
parameters and indicates how much the average value of dependent variable varies with a
unit change in the independent variable when all other independent variables are held
constant. Now the estimated regression line can be written as:
Y= .0333-.480* Prediction of Ordering Quantities* +.600* `Information Lacking + error
term
Y= Bullwhip Effect
4. CONSEQUENCES OF BULLWHIP EFFECT
150
100
50
0
Figure 4: Consequences those boutique industries face due to bullwhip

effect
The
frequency of each consequence is shown in this chart. The main consequences of bullwhip effect
are excessive inventory cost, manufacturing cost, insufficient capacity, product unavailability,
poor customer service and loss of good will, higher transportation cost, tense in buyer supplier
relationship higher replenishment lead time, loss of profit.
Page | 1252
5. DISCUSSION & CONCLUSION
Form analysis the statistical tool and quality control tool both has common six causes which are
mostly responsible for bullwhip effect.
Table 9: Significant Causes after analysis

Retailer’s Opinion
Statistical Analysis Demand Forecast, Price Fluctuation, Order Batching, Lack of
Transparency, Fear of Empty Stock, Misperception of Feedback,
Capacity Limit, Number of Echelon, Improper Control Process,
Company Process, Replenishment Policy, Inventory Policy, Multiplier
Effect
Wholesaler’s Opinion
Statistical Analysis Demand Forecasting, Multiplier Effect, Number Of Echelon, Price
Fluctuation, Order Batching, Rationing And Shortage Gaming,
Misperception Of Feedback, Fear Of Empty Stock, Lack Of
Transparency, Inventory Policy, Replenishment Policy, Improper,
Control Processes, Lack Of Synchronization, Capacity Limit,
Company Process, Lead Time
The purpose of this study was to identify which causes are mostly significant for bullwhip effect
on boutique industry of Dhaka city. After reviewing several literatures, sixteen causes are found
which are responsible for bullwhip effect. Among them few common causes are found which are
identified as highly responsible for bullwhip effect in case of both retailers and wholesalers’
responses such as demand forecasting, price fluctuation, order batching, fear of empty stock, lack
of transparency and misperception of feedback. This study also showed what types of
consequences they face due to bullwhip effect such as excessive inventory cost, higher
manufacturing cost, insufficient capacity, product unavailability, poor customer service, higher
labor cost and manpower wastage, higher transportation cost, tense in buyer supplier relationship,
higher replenishment lead time and loss of profit.
Page | 1253
6. References
Ben-Daya M, R. A. (1994). Inventory models involving lead time as a decision variable.

Journal of the Operational Research Society, 45(5), 579–582.
Blinder, A. S. (1982, June). The American Economic Review, 72(3), 334-348.
Creswell, J. W. (2003). Researh Design: Quantative, Qualitative and Mixed Method Approach
(1st ed.).
EH, B. (1963). Consistency and optimality in managerial decision making. Management

Science, 9(2), 310–321.
Geary, Disney, S. &., Towill, S. &., & D.R. (2006). On bullwhip in supply chains--
historical review, present practice and expected future impact. International
Journal of Production Economics, 101(1), 2-18.
Handfield BR, N. L. (2002). Supply chain redesign: transforming supply chains into
integrated value systems. Financial Times Prentice Hall, 371.
Hau L. Lee, V. P. (1997, 4 1). Information Distortion in a Supply Chain: The Bullwhip Effect.
Management Science, 43(4), 546-558.
Heydari Jafar, K. R. (2009). A study of lead time variation impact on supply chain performance.
Int J Adv Manuf Technol, 40(11-12), 1206–1215.
Jakšič M, R. B. (2008). The effect of replenishment policies on the bullwhip effect: a transfer
function approach. European Journal of Operational Research, 184(3), 946–961.
JL, B. (1983, July 25-29). Automated production control with a simulation capability.
IFIP Conference on System Modeling and Optimization, (pp. 1-14). Copenhagen.
Jr, V. A. (1965). Optimal policy for a multi-product, dynamic, nonstationary inventory problem.
Management Science, 12(3), 206–222.
JW, F. (1958). Industrial dynamics—a major breakthrough for decision makers.

Harvard Business Review, 36(4), 37-66.
Khan, J. A. (1987, Septemer). The American Economic Review, 77(4), 667-679.
Lee, H. P. (1997). The bullwhip effect in supply chains. Sloan Management Review, 38(2),
93-102.
Leedy, P. D., & Ormrod, J. E. (2001). Practical Research: Planning and Design (7th ed.).
Prentice Hall and SAGE Publications.
Liang W-Y, H. C.-C. (2006). Agent-based demand forecast in multi-echelon supply chain.
Decision support systems, 42(1), 390–407.
Page | 1254
Liao C-J, S. C.-H. (1993). An analytical determination of lead time with normal demand.
International Journal of Production Research, 11(9), 72–78.
Moyaux, T., Chaib-draa, B., & D'Amours, S. (2007). Information Sharing as a Coordination
Mechanism for Reducing the Bullwhip Effect in a Supply Chain. IEEE Transactions
on Systems, Man, and Cybernetics, Part C (Applications and Reviews), 37(3), 396-
409.
OJ, B. (1983). The production and inventory behavior of the American Automobile Industry.
Journal of Political Economy, 91(3), 365-400.
Paik, S.-K., & Bagchi, P. K. (2007). Understanding the causes of the bullwhip effect in a
supplychain. International Journal of Retail & Distribution Management, 35(4), 308-
324.
Price: Economics Onliine. (2021). Retrieved September 08, 2021, from Economics Online
Web site: https://www.economicsonline.co.uk/Definitions/Price.html
Qualtricks. (2021, March). Retrieved from cross-tabulation:

https://www.qualtrics.com/experience-management/research/cross-tabulation/
S., B. A. (1986). Can the Production Smoothing Model of Inventory Behavior be Saved?
Journal of Economics, 101(3), 431-454.
Şen, A. (2008). The US fashion industry: A supply chain review. International Journal
of Production Economics, 114(2), 571-593.
Sterman, J. D. (1989). Modeling Managerial Behavior: Misperceptions of Feedback in

a Dynamic Decision Making Experiment. Management Science, 35(3), 321–
339.
Sun HX, R. Y. (2005, September 11-13). The Impact of forecasting methods on bullwhip effect
in supply chain management. IEEE International Engineering Management
Conference, 215-219.
Sunil C, M. P. (2007). Supply chain management: strategy, planning and operation (3rd
ed.). india: Pearson Education.
SY, S., & M, L. (2008). The Effect of Forecasting and Information Sharing in SCM for Multi-
Generation Products. European Journal of Operational Research, 186(1), 276-287.
The Analysis Factor. (2021, March). Retrieved from https://www.theanalysisfactor.com/factor-

analysis-how-many-factors/
Warburton RDH, D. S. (2007). Order and inventory variance amplification: the equivalence of
discrete and continuous time analyses. International journal of production economics,
110(1-2), 128–137.
Page | 1255
PAPER ID: 055
Automated Braking Control System for Automobiles

Tanvir Ahmed1*, Choudhury Abul Anam Rashed1*
Department of Industrial and Production Engineering, Shahjalal University of Science &
Technology (SUST), Sylhet-3114, Bangladesh
tanvirahmed.sustipe@gmail.com , rashed-ipe@sust.edu
Abstract
Hilarious growth in automobile industry needs to take care of automobile user’s safety with its
dynamic advancement in the technology. Though numerous safety mechanisms are available, still
few automobiles meet with accident due to mechanical failure and human errors such as non-
application of the brake when necessary, leads to fatal accidents. Currently, vehicles are often
equipped with active safety systems to reduce the risk of accidents, many of which occur in the
urban environments. The most popular safety systems include Antilock Braking Systems (ABS),
Traction Control and Stability Control. All these systems employ different types of sensors to
constantly monitor the conditions of the vehicle, and respond in an emergency. An Ultrasonic
setup is placed in front of vehicles and that setup consists of an emitter and a receiver. Ultrasonic
emitter always emits the Ultrasonic waves, whenever an obstacle is detected then wave gets
reflected and receiver receives the signal. Reflected wave sends the signal to the Arduino Nano
from that based upon distance of object it actuates the brakes. The resulting system can achieve
measurements with high accuracy and improved short distance measurement. This distance
measurement is used to control smart braking system for safety applications.
Key Words: Automobiles, Automatic Braking System (ABS), Traction Control, Stability
Control, Ultrasonic sensor, Arduino Nano, Safety.
1. Introduction
Braking systems of commercial vehicles were always given the highest importance concerning
safety issues and in particular active safety. Once the vehicle is started, it must be
stopped some place. Brakes are connected on the wheels to stop the vehicle. Before applying the
brakes, speeding up is cut off to stop the fuel supply framework in this manner the motor
builds up no more energy to run the vehicle, and after that Clutch is additionally withdrawn which
associates the motor from transmission framework (Hasan Oktem et al., 2018). In this manner,
when the vehicle is standing, the motor is as yet running at lingering. Make enough deceleration
to stop the vehicle as fast as the driver wishes, without surpassing the drivers comfort level as to
pedal exertion or pedal travel. The stopping mechanism coverts the dynamic vitality into warm
Brakes are the most vital segments of a vehicle as they are the essential source to convey the
vehicle to an end (Kavatkar et al.,2017). Inability to brake can bring about a catastrophe. Such a
complex task imposed to the control of braking system cannot be based on the driver abilities and
need to be done independently of the driver. An improved Intelligent Braking System (IBS)
braking forces management would certainly enable to reach the given task. The advanced strategy
Page | 1256
for the braking force management, proposed by (sidek and Salami, 2000), is based on intelligent
controlling of the braking forces distribution between the front and rear axle of power-driven
vehicle. The system when integrated with other subsystems like automatic traction control system,
intelligent throttle system, and auto cruise system etc. will result in smart vehicle maneuver (sidek
and Salami, 2000).
An intelligent mechatronic system includes an Ultrasonic wave emitter provided on the front
portion of a car producing and emitting Ultrasonic waves frontward in a pre-determined distance.
An Ultrasonic receiver is also placed on the front portion of the car operatively receiving a
reflective Ultrasonic wave signal. The reflected wave (detected pulse) gives the distance between
the obstacle and the vehicle. Then a microcontroller is used to control the speed of the vehicle
based on the detection pulse information to push the brake pedal and apply brake to the car
instantaneously for safety purpose (Sairam et al., 2013).
S. N. Sidek et al (2010) found that intelligent braking system has a lot of potential applications
especially in developed countries where research on smart vehicle and intelligent highway are
receiving ample attention. The system when integrated with other subsystems like automatic
traction control system, intelligent throttle system, and auto cruise system etc. will result in smart
vehicle maneuver. The impact of such design and development will cater for the need of
contemporary society that aspires quality drive as well as to accommodate the advancement of
technology especially in the area of smart sensor and actuator. The emergence of digital signal
processor enhances the capacity and features of universal microcontroller (Sidek, Salami and
M.J.E, 2001). Aleksendric, D.et al. (2012) reported that intelligent modeling, prediction and
control of the braking process are not an easy task if using classical modeling techniques,
regarding its complexity. They proposed a new approach has for easy and effective monitoring,
modeling, prediction, and control of the braking process i.e. the brake performance during a
braking cycle.
Currently, vehicles are often equipped with active safety systems to reduce the risk of accidents,
many of which occur in the urban environments. The most popular include Automatic Braking
System (ABS), Traction Control and Stability Control. All these systems employ different types
of sensors to constantly monitor the conditions of the vehicle, and respond in an emergency
situation (Perez et al., 2010).
1.1 Objectives
The main objectives of this project are:
i. To develop an Intelligent Braking System for Automobiles.

ii. To find the braking distance at various vehicle speeds.
Page | 1257
Working Function
Vehicle Start
Ultrasonic
Sensor
Obstacle Detected
No Yes
Normal Drive Reflect Wave
Sensor Receiver
Braking Circuit
Vehicle Brake
Figure 1: Flow diagram of Automatic Braking Car System.
2. DESIGN AND SELECTION

2.1 Design
Based on research methodology two primary design concept has been generated. Those are:
I. In the front end, one ultrasonic sensor is employed in conjunction with a servo motor.
Whose function is to recognize objects or vehicles that are inside the range of the
ultrasonic sensor. The servo motor's job here is to spin the sensor 180 degrees in order
to track the item or vehicles in three directions.
II. Three ultrasonic sensors are installed on the car's three sides. One is in the car's front
end. The other two are on the car's left and right sides, respectively. The sensors can
identify an object or a vehicle in three directions at the same time.
Page | 1258
Front View Top View Left Hand Side View
Figure 2: 3D printed model of the concept 1
Front View Left Hand Side View Top View
Figure 3: 3D printed model of the concpet 2
2.2 Selection
The second model is chosen from the two available options. The second model offers a number
of benefits over the first. In comparison to the first, the second model does not require the use of
a servo motor. In the second model, three ultrasonic sensors were utilized to simultaneously
identify an item and a vehicle in three directions. The second model, which can detect in three
directions, will be far more useful in reducing accidents than the first. The first model is good for
avoiding front-end collisions with other vehicles, whereas the second model can prevent not only
front-end collisions but also left and right-side collisions. There is a risk of an accident if random
passers-by or animals try to cross the road in a highway, but if the automobile is equipped with
three ultrasonic sensors, there is less risk of an accident since the car would stop as soon as any
item enters the sensors' measuring area. It assists in avoiding any type of collision when parking
an automobile. In addition, the second model will minimize the likelihood of a collision in a
reckless overtaking situation, which is a major cause of accident on Bangladeshi highways. In
the given scenario, the second model will be superior than the first, thus we will choose the second
model.
Page | 1259
Table 1: Concept Screening
Selection Concept no. 1 Reference Concept no. 2

Criteria “0”
Ease of Use 0 0 0
Durability - 0 +
Efficiency - 0 +
Cost efficient - 0 0
Maintenance 0 0 0
Sum +’s 0 0 2
Sum 0’s 2 5 3
Sum –‘s 3 0 0
Net Score -3 0 2
Rank 3 2 1
Continue? No No Yes
Page | 1260
Table 2: Concept Scoring
Selection Weight Concept no. 1 Reference “0” Concept no. 2

Criteria
% Rating Weighted Rating Weighted Rating Weighted
Rating
Rating Rating
Ease of Use 10% 2 .2 0 0 3 .3
Durability 20% 1 .2 0 0 4 .8
Efficiency 50% 1 .5 0 0 5 2.5
Cost 15% 2 .3 0 0 2 .3
efficient
5% 3 .15 0 0 3 .15
Maintenance
Total 1.35 0 4.05

Score
Rank 3 2 1
Continue? No No Develop
2.3 Components
Followings are the components in this intelligent braking system:

(1) Ultrasonic Sensor, (2) Arduino UNO, (3) Wheels, (4) Rechargeable Battery, (5) Gear Motor,
(6) Resistor, (7) Transistor, (8) Jumper Wire, (9) Car Chassis, (10) Motor Driver IC
3. FABRICATION
Fabrication Steps are-
I. Take a car chassis as a frame. Cut it along as calculated. Take measurement by
calculations according to the wheels.
II. Rubber stopper can be the blocks of the wood as we are just creating a suspension to the
front wheels. Just hole in them with drill and try to make a 90-degree L suspension for
wheels. Make hole equivalent to the Bolt in the car chassis. Fix the wheel and Rubber
stoppers to the chassis. Add a hole for the motor of slight larger equivalent to the neck of
the motor in the center. Fix the motor with gear hub. Gear head to plane by cutting of
extra part for smooth steering. Add two more small holes on the edges of the stoppers to
fox the plastic strips and get along the Gear for movement. Calculate the length of plastic
strips accordingly. Fix the small hole with screw or thin bolts as needed.
III. Attach two gear motors at the backside of the plywood.
Page | 1261
IV. Place two wheels at the rear of the vehicle. Two Smart Robot Car Tyres Wheels For
Arduino TT Gear Motor Chassis has used here. Two wheels attach at the front side of the
car.
V. Now Arduino UNO has mounted at the middle of the base and screw has been used to
tighten them up. Wires has connected with the Arduino.
VI. Sensor holder placed at the top front side of the vehicle and Ultrasonic sensor device is
placed upon this. Sensor wire is connected with microcontroller.
VII. Wire ends are connected with Arduino in the following manner:
Positive to +5v
GND to GND
Trig to A0
Echo to A1.
VIII. Two 9v rechargeable battery mounted at the backside of the car and the wire that are
attached with the battery are made connection with microcontroller.
IX. Program code to run the car inputted in the micro controller device and now the car can
run through with or without help of the auxiliary device and work as a plug and play
device.
3.1 Final Construction
Figure 4: Top view of the developed Model
Page | 1262
3.2 Calculations
The braking distance is the main factor considered in this system. Braking distance for a particular
speed is the distance between the point of application of the brakes and the point at which the
vehicle comes to a complete stop from the present speed (J.V. Sai Ram et al., 2017). It is calculated
by using following formula.
Braking Distance = V / 2μg (meter)
Where
V= Velocity of the vehicle (m/s)
μ = Coefficient of friction of road = 0.8
g = Acceleration due to gravity = 9.81(m/s2)
Now, for velocity 10 km/hr.
Braking Distance = (10*1000/3600)/ (2*0.8*9.81)
= 0.18 m
For velocity 50 km/hr.
= 0.88 m
= 1.59 m
= 2.30 m
= 2.48 m
3.3 Applications
I. Controlling Automobiles
This system is best suited for controlling the braking system of automobiles. This
system can be used in all types of vehicles like Car, Motorcycles, Heavy and Light
vehicle etc.
II. Automated guided vehicles in Industries
Heavy Industries like Steel Manufacturing Company, Billet Manufacturing Company
used computer operated automatic cranes to transfer scrap, hot material and finished
product from one place to another place. If both cranes come too close to each other’s or
overlapped than a serious accident can be occurred. To avoid these types of unwanted
Page | 1263
accidents a sensor is must to detect the obstacles or cranes between its ranges to brake
the system automatically.
In production line, both manual operator and automated system operated
simultaneously. For work pressure, stress factor or tiredness it is not possible for
workers to hold their concentration all the time and this lack of concentration can lead
to a serious accident. If industries use the system than the automated vehicles like
overhead cranes can detect the worker easily from a far distance and can avoid
accidents during manufacturing in Production Lines.
III. Military Application
This system can be used in military application such as spy robots to collect news from
the opposition while making a brake-seeing enemy is from the far distance.
4. DISCUSSION
In the present work, a prototype of an ultrasonic braking distance for stationary obstacle is
measured and controlling the speed of vehicle accordingly to predetermined distance is shown.
An ultrasonic sensor is cheaper and less demanding of hardware than other types of sensors
presently used in automobiles, such as the sensors based on computer vision or radar which is
used to measure the distance between vehicle and the obstacle. The relative braking distance of
the vehicle with respect to the obstacle is estimated using consecutive samples of the vehicle’s
velocity. These two quantities are used by the control system which is Arduino to take the actions
on both the accelerator and also the brake, thus to adjust the speed in order to maintain a safe
distance to prevent accidents. As ultrasonic sensors can detect obstacles of different types, this
system can also prevent collision of the vehicle with pedestrians, or can at least reduce the
probability of accidents. Coupled with the fact of lower cost of ultrasonic sensors compared with
other kinds of sensors, could facilitate the application and mounting of the system in many low-
end vehicles, helping to improve the safety.
6. CONCLUSION
Safe Driving is very important as it concerns with the human lives. A prototype of the automated
braking control system is developed. The obtained results showed that if the distance between the
object and the vehicle is less than 0.5m and the velocity of the vehicle is 10km/hr-20km/hr the
system can stop the vehicle smoothly. When object distance is in between 0.5 to 0.88m with the
vehicle speed ranging between 30km/hr to 50km/hr collision can be avoided.
Page | 1264
References
Aleksendrić, D., Jakovljević, Ž., & Ćirović, V. (2012). Intelligent control of braking
process. Expert Systems with Applications, 39(14), 11758-11765.
Chen, M., Liu, F., Ren, C., & Gao, Z. (2012, August). A Control System of Vehicle Rear-end
Anti-collision. In 2012 4th International Conference on Intelligent Human-Machine
Systems and Cybernetics (Vol. 2, pp. 46-48). IEEE.
Di, Z., & He, D. (2016, June). Forward Collision Warning system based on vehicle detection
and tracking. In 2016 International Conference on Optoelectronics and Image
Processing (ICOIP)(pp. 10-14). IEEE.
Gokulakrishnan, P., Ragavan, M. N. V., & Suresh, M. K. (2018). SIBS: Smart Intelligent
Braking System to Eliminate Human Perception and Reaction Time using
VANET. International Journal of Pure and Applied Mathematics, 119(12), 16555-
16562.
J.V.Sai Ram, K.M.S.V. Manikanta, G. Pavanth, B. Jagadeep, Dr. B.Raghu Kumar (2017).
Automatic Braking System Using Ultrasonic Sensor. In 2017
International Journal of Innovative Science and Research Technology (Vol. 2, pp. 398-
404). International Journal of Innovative Science and Research Technology (IJISRT)
Kavatkar, M. T., Salve, H, & Rah ate, M. (2017). Design and Analysis of Intelligent Braking
System, International Journal of Engineering Development and Research
(IJEDR1701019) 5 (1), 119-131.
Mahtab, A., Singh, M., Sharma, V. K., & Kumar, A. (2017, April). Anti-collision vehicle with
voice recognition. In 2017 International conference of Electronics, Communication and
Aerospace Technology (ICECA) (Vol. 2, pp. 126-129). IEEE.
Pérez, J., Seco, F., Milanés, V., Jiménez, A., Díaz, J. C., & De Pedro, T. (2010). An RFID-based
intelligent vehicle speed controller using active traffic signals. Sensors, 10(6), 5872-
5887.
Sairam, G. V., Suresh, B., Hemanth, C. S., & Sai, K. K. (2013). Intelligent mechatronic braking
system. International Journal of Emerging Technology and Advanced Engineering,
ISSN, 2250-2459.
Sanjana, T., Fuad, K. A. A., Habib, M. M., & Rumel, A. A. (2017, February). Automated anti-
collision system for automobiles. In 2017 International Conference on Electrical,
Computer and Communication Engineering (ECCE) (pp. 866-870). IEEE.
Page | 1265
PAPER ID: 070
Multivariate Time Series Analysis and Forecasting on Air

Quality Index with Industrial Statistics: A Case Study of
Bangladesh
MD Shafikul Islam1*, Azmine Tousik Wasi2*

1
Department of Industrial and Production Engineering, SUST, Sylhet,3114
2
Department of Industrial and Production Engineering, SUST, Sylhet,3114
Email:
1
shafikul37@student.sust.edu
2
azmine32@student.sust.edu
*Corresponding author [MD Shafikul Islam, Azmine Tousik Wasi]
Abstract: The Air Quality Index (AQI) analyzes all the observed particles
Keywords:
and pollutants; converts them to a single scale number to represent how
● Air Quality Index
fresh or polluted the air is in a particular location. AQI values correlated
with many factors including environment, manufacturing,
● Multivariate industrialization, transportation, and weather. In this research, a
Analysis multivariate time series analysis of the Air Quality Index was conducted
● Industrial Intensity
with other statistics such as the Industrial Intensity Index, Manufacturing
Value Added per capita, Motor Vehicles and CO2 Emission. These
Index variables are analyzed concerning time to identify their relationship with
● Motor Vehicle AQI by the Pearson correlation method. According to our findings,
Registered Manufacturing Value Added and Total Motor Vehicles Registered in a
Year are highly correlated with AQI, with correlation values of 0.84 and
● Facebook Prophet 0.85, respectively; defining AQI is primarily affected by pollution from
transportation and manufacturing. The Prophet time series model was
also used to learn trends and forecast future AQI values. According to
the model, the AQI value rises from 2016 to 2022 but falls sharply in
2020; the average AQI value is predicted to be 178.89 in 2023, which is
higher than any previous year's value.
1. Introduction
The Air Quality Index (AQI) measures the quality of the ambient air in relation to the applicable
international air quality standard. This is a thorough index that considers multiple pollutants at
once and can reflect the overall status of the ambient air quality of the experiment area. (Baldwin
Page | 1266
R, et al. (2007)). Furthermore, the Air Quality Index (AQI) concentrates on the health effects of
air pollution that may appear hours or days after inhaling polluted air.
Air quality was identified as the top environmental concern for human health worldwide in the
Environmental Health Policy Objectives of the 2022 Environmental Performance Index (EPI)
(Wendling, et al. (2022)). According to the World Air Quality Report 2020, Bangladesh has the
worst air quality in the world, with Dhaka ranking second among the most polluted cities (Ahmed
H.U. (2020)). Not only Bangladesh, but the entire world, is suffering from the effects of air
pollution. Approximately 93% of children also breathe toxic air (Anik S.S.B (2020)).
The UNIDO competitive Industrial Performance (CIP) Index, developed primarily for developing
countries, benchmarks the country's competitive industrial activity against the backdrop of
liberalization and globalization. Manufacturing value added and the Industrial intensity index is
two important components of it. MVA per capita is a proxy for a country's level of
industrialization and is deflated by population to account for the size of the economy. The
intensity of industrialization is calculated by taking the simple average of the share of MVA in
GDP and the share of medium and high-technology (MHT) activities in MVA (Komninos N.
(2005)). Besides CO2 emissions are the primary cause of global climate change. It is widely
acknowledged that to avoid the worst effects of climate change, the world must urgently reduce
emissions. (Ritchie H, et al. (2020)).
Several studies have been conducted to demonstrate the significance of the Air Quality Index.
According to Islam and Salam, gaseous pollutants contribute significantly to Dhaka's particulate
pollution (2000 and 2003). Ahmed and Hossain investigated particulate pollution in Dhaka and
Chittagong, Bangladesh's two largest cities (2008). With time series modeling, Islam and Sharmin
predicted the air quality of the Sylhet and Dhaka divisions (2020). Azad, Rahman, and
collaborators conducted a study in Bangladesh to examine the changes in air quality and COVID-
19 transmission in various lockdown scenarios (2021). However, the relationship between the Air
Quality Index and other industrial and environmental parameters is not well understood. Most
importantly, making decisions requires forecasting the air quality using other parameters because
the impact of COVID-19 has almost been neutralized.
Given the significance of the Air Quality Index, the goal of this study is to identify the overall
pattern of the Air Quality Index concerning other industrialization statistics in Bangladesh, such
as the Industrial Intensity Index, Manufacturing Value Added, and Motor Vehicle Registered. We
have also considered one more Environmental parameter, CO2 emission to find the co-relation
with the Air Quality Index.
2. Methodology
Basic Overview of the method

The purpose of our research is to determine the relationship between the Air Quality Index and
other factors such as industrial, environmental, and transportation statistics in Bangladesh. To
accomplish this goal, data was gathered from various sources. AQI values are analyzed to see
monthly, and yearly trends and the COVID effect. Then, Facebook prophet time series modeling
was used to forecast future AQI values.
Page | 1267
All other factors along with AQI values are scaled to mean 0 and standard deviation 1, for better
processing in the next steps; also, the p-value threshold for statistical significance at two-tailed
tests was fixed at p ≤ 0.05. Multivariate time series analysis was conducted on scaled data to find
correlations between them using the Pearson method. Finally, some statistical tests were
performed to verify if the co-relations are statistically significant and responsible.
Figure 1: Methodology Overview: An Overview Diagram Illustrating the Flow of Analysis
Data Description
AQI data are obtained from (https://waqi.info/) on October 1, 2022. Data on Bangladesh's
Manufacturing Value Added and CO2 emissions are accessed
from (https://datacatalog.worldbank.org). To retrieve data for the Industrial Intensity Index and
Motor Vehicle Registration, we created a web scraping algorithm using the Python package
Beautifulsoup (Richardson, L. (2007)).
Pearson Correlation
The Pearson correlation method is the most used approach for numerical variables. It assigns a
value between -1 and 1 for presenting the correlation with two variables where 1 representing
maximum positive correlation and -1 representing maximum negative correlation. A positive
correlation indicates that if variable A increases, variable B will also increase, while a negative
correlation indicates that if A increases, B decreases (Boslaugh & Watters, 2008).
Facebook Prophet Model
Facebook Prophet is a method for forecasting time series data based on an additive model with
yearly, weekly, and daily seasonality, as well as holiday effects. Prophet is resistant to missing
data and trend shifts, and it typically manages outliers well (Letham,2017). At the core of this
modeling, the output function Y(t) is the addition of growth (g(t)), seasonality s(t), holiday (h(t))
and error e(t).
Page | 1268
𝒀(𝒕) = 𝒈(𝒕) + 𝒔(𝒕) + 𝒉(𝒕) + 𝒆(𝒕)
3. Result and Data Visualization

AQI: Time Series Analysis of AQI and Forecasting with Facebook Prophet
The Air Quality Index (AQI) is a measure of concentrations of different air pollutants. It assigns
numbers on a scale of 0 to 500 showing current air quality. It also can be used to predict when air
quality will be unhealthy by time series analysis with enough data. From the time series analysis
of AQI, it can be noticed that there was always a steady rise of AQI value in starting and ending
part of every year.
Figure 2: AQI – Original Series and One Year Moving Average

As per figure 2, the AQI values (Original Series) follows a trend of higher AQI values at the start
and end of the year and stable in the middle. And, in the One Year Moving Average (Moving
Average (Yearly)) graph, we can notice a subtle decrease in AQI in 2020, pointing to COVID-19
timeline.
Figure 3: AQI - Future Forecasting with Facebook Prophet
Page | 1269
Figure 3 demonstrates that, in 2023, it is expected to have some high outliers at the beginning of
2023 with an average AQI of 178.89. Also, the average AQI value is rising regularly since 2020.
It is predicted to rise by 2.22% and 9.07% in 2022 and 2023 respectively from previous years,
raising a huge concern for Bangladesh.
Multivariate Time Series Analysis

In Figure 4, a visualization was prepared for the 4 factors along with AQI in a line chart. In visual
inspection, Manufacturing Value Added per Capita (MVA) and Total Motor Vehicle Registered
(MVR) demonstrated close and positive co-relation. Industrial Intensity Index (III) also followed
a similar trend except in 2019.
Figure 4: Multivariate Time Series Analysis of AQI and Other Factors
Pearson Correlation and Statistical Significance

The Pearson method was used to determine the correlation of all the factors. The results are shown
in Table 1 and visualized as a heatmap in Figure 5, where darker regions denote higher co-
relation.
- AQI CO2 MVA III MVR

AQI 1.000 0.387 0.847 0.272 0.856
CO2 0.387 1.000 0.571 -0.249 -0.142
MVA 0.847 0.571 1.000 0.221 0.617
III 0.272 -0.249 0.221 1.000 0.428
MVR 0.856 -0.142 0.617 0.428 1.000
Table 1: Correlation Matrix of Factors and AQI Figure 5: Correlation Heatmap
Page | 1270
The figure shows that Manufacturing Value Added per Capita (MVA) and Total Motor Vehicle
Registered (MVR) illustrate very higher correlation with AQI; 84.7% and 85.6% respectively
verifying the visual inspection. On the other hand, CO2 Emission and Industrial Intensity Index
(III) show weak but positive correlation; 38.7% and 27.2% respectively. In a two-tailed Pearson
significance test, both the correlations of Manufacturing Value Added per Capita (MVA) and
Total Motor Vehicle Registered (MVR) to AQI is recognized as statistically significant
maintaining p ≤ 0.05.
4. Discussion and Conclusion
This study correlated the air quality index with other statistics to determine the actual concerning
area in order to reduce the increasing trend of the air quality index. The Pearson correlation test
revealed that the relationship between Manufacturing Value Added and Motor Vehicle Registered
is statistically significant. This research will help to develop an integrated approach to recovering
from the negative effects of the environment and will lead the way for relevant recommendations
to improve the country's air quality. However, the correlation between the Industrial Intensity
Index and the Air Quality Index is not statistically significant. Despite rapid industrialization,
Bangladesh has a good reputation for establishing its Green Industry (Star Business Report,
2019).
Correlation of AQI with other statistics

Correlation P value
Statistics Type Significance
value (Two-tailed test)
Industrial Intensity Index 0.27 0.601 Not significant
Manufacturing Value Added 0.84 0.033 Significant
Motor Vehicle Registered 0.85 0.029 Significant
CO2 Emission 0.38 0.44 Not significant
Table 2: Correlation, P value, and statistical significance of different statistics and AQI index
Air quality forecasting was done using the Facebook Prophet model, which is a useful technique
as it takes seasonality, holidays, and other errors into account. The ability to predict future air
quality with 95% confidence intervals shows how accurately the models measure their
parameters. However, for more precise forecasting, some unpredicted variables (rainfall,
temperature, etc.) could be incorporated into the model. According to the study's findings, air
quality in the country remains unhealthy or extremely unhealthy during the dry season
(December-March). The severity of most construction and urbanization grows, and the level of
concern during this time grows more serious. As our model predicts that contamination will
increase in the future, additional research involving both the public and private sectors will be an
excellent way to address this critical issue.
Page | 1271
References
Ahmed H.U., ‘Stuck in air pollution and traffic congestion, The Financial Express, Dhaka, (2020)
Anik S.S.B, ‘Experts: Too many private cars to fix traffic in Dhaka’, Dhaka tribune, Dhaka,
(2020)
Azad, M. A. K., et al., How air quality and COVID-19 transmission change under different
lockdown scenarios? A case from Dhaka city, Bangladesh. Science of the Total
Environment, 762, 143161. 2021
Baldwin R, Calkins D, (World Bank Group, Washington, D.C), 2007, ‘Bangladesh urban air
quality management: an institutional assessment, Available in: https://documents.worldbank.org/
Boslaugh, S., & Watters, P. A. 2008. Statistics in a Nutshell: A Desktop Quick Reference (In a
Nutshell (O’Reilly)).
Islam M, Chemical speciation of particulate matter pollution in urban Dhaka City.2000,
Bangladesh Environ 2000:51–58
Islam, M.M., Sharmin, M. & Ahmed, F. Predicting air quality of Dhaka and Sylhet
divisions in Bangladesh: a time series modeling approach. Air Qual Atmos Health 13,
607–615 (2020)
Komninos N. (URENIO),2005, “The CPI Index”, Available in:
https://www.urenio.org/2005/08/26/the-cip-index/, accessed 1 October 2022
Letham, B. Forecasting at scale. PeerJ Preprints. 2017, Available in:

https://peerj.com/preprints/3190v2/
Richardson, L. Beautiful Soup Documentation — Beautiful Soup 4.9.0 documentation. (n.d.).,
Available in: https://www.crummy.com/software/BeautifulSoup/bs4/doc/ (accessed 1 October
2022)
Ritchie, H., Roser, M., et al (Our World in Data),2020." CO₂ and Greenhouse Gas Emissions".
Available in: https://ourworldindata.org/co2-and-other-greenhouse-gas-emissions (accessed 1
October 2022)
Salam A, Bauer H, Kassin K et al., Aerosol chemical characteristics of a mega-city in Southeast

Asia (Dhaka–Bangladesh). 2003. Atmos Environ 37:2517–2528.
Star Business Report, Bangladesh leads world in green RMG production. The Daily Star. 2019
Wolf, M. J., Emerson, J. W., Esty, D. C., de Sherbinin, A., Wendling, Z. A., et al. (Yale Center
for Environmental Law & Policy) 2022.Environmental Performance Index. Available in:
https://epi.yale.edu/ (accessed 1 October 2022)
Page | 1272
PAPER ID: 081
Application of CTS (Computer to Screen) Machine in Printing

Industries for Process Improvement & Material Optimization
Tarequl Islam 1, *
1, *
Shahjalal University of Science and Technology, Kumargaon, Sylhet-3114, Bangladesh
Email: 1, * tareq.ipe.sust@gmail.com
Abstract: The printing and labeling industries are struggling to meet the need
Keywords:
for more complex and dynamic design requirements coming from the
customers. It is now crucial to implement technological advancements to
● Process manage workflow, productivity, process optimization, and continual
Improvement; improvement. There has never been a time when the imagery and
embellishments of apparel has been more commercially viable as it is now.
● Material Images and text are fused directly to fabric by heat transfer printing and
Consumption labeling. For screen development which is required for heat transfer label mass
Optimization; production, many industries are still using the conventional method of screen
development process. A CTS (computer-to-screen) innovates the printing and
● CTS (Computer to labeling industries by enhancing workflow, lowering consumable
Screen) Machine; consumptions and chemical usage, speeding up setup, guaranteeing flawless
design, and raising the print quality of the producing screens. The study's
● Heat Transfer objective is to assess how CTS machines are used and how they affect existing
Labels; heat transfer screen development processes in one of Bangladesh's leading
printing and labeling companies. The study's primary goal is to highlight and
● Printing Industries; analyze how the use of CTS machines reduces material and operational costs
by optimizing the process. Costs for CapEx and OpEx are computed and
compared for using CTS technology before and after adoption. Savings data
such as material, consumable, and operating cost savings versus depreciation
and machine payback period analysis were taken into consideration. It is clear
from this study that CTS machines in the printing and labeling industries can
guarantee profitability on top of Capital Expenditures.
1. Introduction
Competitiveness has been the key factor for the survival of the companies. The economic crisis
that marked the beginning of this millennium forced the total readjustment of processes and
operations which, in some cases, gave origin to deep changes in the organizations (1). The
transformation trend towards digital technology to achieve sustainability targets and meet legal
regulations has been visible in many industries. The printing sector has already been increasingly
boosting sustainability performance through digitalization to automate workflows of processes
(2). Also, the price of raw materials is increasing and manufacturers face cascading challenges
through the supply chains (3). Thus the study will assess how adoption of CTS technology in a
Page | 1273
printing industry make the process optimized & remove a significant amount of materials
consumption focusing on the Heat Transfer Screen development process.
Heat Transfer Printing & Labeling is one of the most important product lines in digital printing
industries based on business contribution because it is very widely used in the apparel industry.
Heat transfer labels convey the same information as their fabric equivalents – branding, washing
information, size details etc. (4). A heat transfer is a method of taking a printed image and fixing
the image directly to the garment in order to copy the image onto the garment itself. The design
image is printed onto special transfer paper or synthetic film. This substrate has a special coating
known as a release layer. This aids the transfer of the image onto the garment (5). When printed
paper or synthetic film is pressed with a certain temperature & pressure, it provides tag-free strong
brand story & information. Heat Transfer technologies follow a reference printed screen to be
printed for large scale production. Each reference screen can be used for a certain amount of
production run. To develop the reference screen as a prepress tool it is required to use artwork,
design plate printing machine, silk screen, certain film, chemicals, exposure machine etc. & this
is the conventional process of screen development. Whereas A CTS, or Computer-To-Screen
printer, brings innovation to the screen printing industry. CTS machines do not use film; therefore,
the cost of screen printing film is eliminated. Go straight from the computer to the emulsion-
coated screen. By removing the need for film, image quality improves, and exposure times are
dramatically reduced (6). T. Schweizer studied that using CTS technology in printing plate
production helps to eliminate costly film material and expensive and environmentally harmful
chemicals needed for developing the film (7).
The technology has been adopted by one of the major printing industries in Bangladesh. After
successful launch of the project it is evident that CTS machine can improve screen development
process and reduce screen development lead time ensuring quality & consistent design. The aim
of the study is to understand how application of CTS machines have made a difference in the
existing screen development process and make the process efficient ensuring optimized process
& lowering material consumption.
2. Methodology
Adoption of this technology was one of the automation projects of Xyz printing & packaging ltd
which is a market leader in the printing & labeling industry. After sufficient research and
feasibility study, they decided to invest in a CTS machine in October, 2021 to replace the
conventional process which was about to become obsolete. This was one of their global
automation budgets. Background data were collected by collaborating directly with the project
team. The existing system studied thoroughly & the impact of CTS technology also analyzed to
make the project economically viable. Firstly, based on incoming customer demand the number
of design plates required per day to produce a certain amount of reference screen to feed the
machine was studied. Secondly, historical consumption of chemical & materials to fulfill the daily
demand of reference screens were analyzed on a specific project timeline which were about to be
consumed zero after CTS implementation. Required chemicals & materials cost had been
analyzed & also impact on labor cost calculated as a savings amount. Machine purchasing cost,
installation & training cost, freight cost & site preparation cost had been considered into
investment amounts & also the electricity & depreciation cost after purchasing the machine had
also been calculated. Based on the cost & savings analysis, financial data had been prepared about
the IRR, NPV payback periods.
Page | 1274
After a financial feasibility study, it was found that the project had a small payback period with a
significant IRR while ensuring a significant amount of savings.
2.1 Current process of screen development

For large scale production of heat transfer screens, every machine follows a prepressed & printed
reference screen. Based on customer requirements the design has been created using any design
tool such as Adobe Photoshop, Illustrator etc. Design file is loaded to the KATANA Imagesetter
machine which requires time to develop & print out the positive plate. The plate is then mounted
with a chemical coated mash or reference screen attached with red laser film (23’’) so that the
printing of a positive plate is exposed to the mash. Developer replenisher & fixer is used to attach
the laser film with the screen & positive plate. Exposure machine is used to develop the mash
through exposing, hardening & rinsing out by water. The heat transfer printer can print on areas
that are not hardened so that ink can pass through and adhere to a surface. The reference or "mash"
screen, which will be used to print screens at bulk level, is finally created.
They were making an average 108 positives plates for 180 screens per day and the positives
making technology is in an obsolete stage and its maintenance cost is very high. Recently they
had a breakdown with the KATANA Imagesetter machine and the fixing cost is about 7.5 K USD
and machine parts are rare to find. For this they needed to outsource the positive and that costs
about 250 USD daily. With the introduction of CTS Machine, they are being able to eliminate the
positive development process. The investment will return within 3 years ensuring more than 70K
USD saving per year.
Fig1: Current screen development process

2.2 Screen Development on CTS Machine
CTS is a helpful device for imaging and exposing screens simultaneously needed to feed the
printing machine for a specific impression sheet. The CTS machine does not require the chemical,
laser film, positive plate development & exposure machine. The design file is uploaded to the
CTS machine using RIP software. During the first pass over the screen frame, CTS uses specially
formulated water-based UV blocking ink, CTS inkjet, and industrial pinheads. The emulsion-
producing screen that can be transferred directly to washout is exposed on the return pass by the
Page | 1275
built-in high output scanning UV led light sources (8). The artwork is printed onto the screen by
industrial print heads & simultaneously after printing the screen is already coated by UV led light.
At the same time, the printed & exposed screen is hardened & rinsed out with water. Areas that
the heat transfer screen printer wants to print are not hardened, so ink can pass through it which
will be applied to the fabric surface through heat & pressure. The use of the Red Laser Film 23"
roll, Developer Replenisher & Fixer is eliminated because of the adoption of CTS.
Fig2: Screen development on CTS machine
3. Historical consumption data

Below tables shows how chemical & laser film required for reference screen development per
month before the technology adoption where unit costs for laser film is $252/roll & $4/Ltr for
both replenisher & fixer-
Table1: Historical usage of chemical & material
Month’21 Film 23" Developer Fixer
(Roll) Replenisher (Ltr) (Ltr)
January 16 60 60
February 12 60 40
March 24 80 40
April 23 70 45
May 12 60 40
June 24 80 60
July 12 60 5
August 30 90 55
September 36 100 60
Total 189 660 405
Total Price $47,690 $2,620 $1,608
Annual Usage $63,587 $3,494 $2,144
Page | 1276
4. Project investment summary
For successful CTS launching, the company had to invest a significant amount including direct
& indirect costs. The investment cost includes the machine purchasing cost, installation & training
cost, air freight cost & site preparation cost. Other two costs were analyzed & included in the total
investment cost & these are depreciation cost of five fiscal years & day to day operating electricity
cost. Below tables includes the cost breakdown of these particular projects:
Table2, 3,4: Machine investment cost, depreciation cost & electricity cost
(2) (3)
Project Investments of CTS Machine Depreciation Cost
Machine Cost($) $130,000 Number Of Machine Unit 1
Installation & Training Cost($) $30,000 Unit Price of CTS $190,000
Freight Cost($) $5,000 Total Cost $190,000
Site Preparation $25,000 Depreciation Year 5
Total Cost($) $190,000 Total Depreciation Cost $38,000
(4)
Electricity Cost
Total unit requirement (KW) 4
Average Run Hour 20
Efficiency 70%
Electricity rate $0.11
Per Day Cost $6
Total Electricity Cost $1,785
5. Savings calculator
Below table includes the key deliverables & savings highlights from the project:
Table5: Key deliverables after adopting CTS machine
Items Current Future Cost Saving Deliverables Highlights
Cost (KUSD/Yr) (KUSD
(KUSD/Yr ) /Yr)
Material 63.5 0 63.5 $ 63.5 K/Year Eliminating laser film
Chemical 5.6 0 5.6 $ 5.6 K/ Year Eliminating developer
replenisher & fixer.
Head count 1.6 0 1.6 $ 1.6 K/ Year Eliminating 0.5 HC from
positive screen development.
Breakdown of material, chemical & headcount savings data plotted in below tables:
Page | 1277
Table6: Materials savings value (Yearly) Table7: Head count savings value (Yearly)
Material Savings HC Savings
Red Laser Film 23" roll (Monthly) $21 Annual Average CTC $3,200
Cost/ Roll $252 Present HC 1
Monthly Savings $5,299 Future HC 0.5
Total Annual SAVE $63,587 Savings $1,600
Total Annual Savings $1,600
Table8: Chemicals savings value (Yearly)

Chemical Savings
Avg Monthly Developer Replenisher (Ltr) 73 Avg monthly use of Fixer (Ltr) 45
Cost / Ltr $4 Cost / Ltr $4
Monthly Savings $291 Monthly Savings $179
Yearly savings $3,494 Yearly savings $2,144
Total Annual Savings in Chemical ($) $5,637
5.1 Financial analysis summary

When the business case was raised, the finance team analyzed all the background data & found
that the project has an internal rate of return of around 30% considering the best, worst & most
likely scenario. The project had a payback period of around three years which indicates that
adoption of this project can ensure profitability.
Table9: Financial analysis summary
Particulars Most Likely Scenario Worst Case Scenario Best Case Scenario
Project IRR 30.3% 30.2% 30.4%
Net Present Value $82,654 $81,977 $83,129
Project Payback (yrs) 3.06 3.06 3.05
6. Conclusion
Since October 2021, the company has significantly decreased the time and work needed to make
the screens by adopting CTS technology. CTS screens are found superior to conventional screen
development processes and they are offering more details and transition to meet customer
preferences. With fast screen exposures, simple designs, consistent quality, and high speed, CTS
has already reduced process and enhanced productivity in the Heat Transfer Label (HTL)
department of the company. The space and maintenance needed for all the screens used every day
has already decreased. Using CTS, they quickly print the reference screens and support the bulk
production based on the needs of the production. Because customer preferences vary frequently
and there are many variable types of information on the screens incoming these days, it is clear
that adopting CTS made the process more efficient than the traditional screen development
approach while ensuring profitability. It is recommended that any screen printing industries can
adopt it as a part of their sustainable business operation based on their experience.
Page | 1278
Acknowledgement
I am grateful to the whole team of Xyz printing & packaging ltd who were directly or indirectly
involved in the launching of this project & supported me with all the relevant information &
guidance.
References
1. Moreira, A., Silva, F., Correia, A., Pereira, T., Ferreira, L., & de Almeida, F.
(2018). Cost reduction and quality improvements in the printing industry. Procedia
Manufacturing, 17, 623-630. https://doi.org/10.1016/j.promfg.2018.10.107
2. Gladysz, B., Krystosiak, K., Ejsmont, K., Kluczek, A., & Buczacki, A. (2021).
Sustainable Printing 4.0—Insights from a Polish Survey. Sustainability, 13(19), 10916.
https://doi.org/10.3390/su131910916
3. Bragagni, Maurizio & Xhaferraj, Lorenc. (2021). RAW MATERIAL PRICES
INCREASE. https://doi.org/10.13140/RG.2.2.10320.43526.
4. How to Print Labels Using Heat Transfers. Blog.focuslabel.com. (2016). Retrieved 19
August 2022, from https://blog.focuslabel.com/print-labels-using- heat-transfers.
5. Lee, D. (2019). What Is A Heat Transfer Label/Tag?. Blog.focuslabel.com.Retrieved 11
September 2022, from https://blog.focuslabel.com/what-is-a-heat- transfer-label.
6. Computer-To-Screen Machines | CTS Imaging. Lawson Screen & Digital Products.
(2022). Retrieved 9 August 2022, from https://www.lawsonsp.com/screen-printing-
equipment/auxiliary-equipment/cts-film-printers.
7. Schweizer, T. (2010). Computer-to-screen system handles large-scale applications.
Https://Www.Researchgate.Net/, 32, 33-34. Retrieved 12 September2022.
8. M&R i-Image XE Direct to Screen Machine - Screen Print World. Screen PrintWorld.
(2022). Retrieved 10 September 2022, from https://screenprintworld.co.uk/product/mr-
i-image-xe-direct-to-screen-machine.
Page | 1279
PAPER ID: 156
LASER ENGRAVING OF NON-METALLIC MATERIALS-PMMA
Mst. Nasima Bagum1*, Mustofa Tanbir Kuhel2*, Md. Ahsan Habib3

nasima-ipe@sust.edu; tanbirkuhel@gmail.com; ahsan43@student.sust.edu
1,2,3
Department of Industrial and Production Engineering, Shahjalal University
of Science and Technology, Sylhet-3114, Bangladesh
*Corresponding author [Mst. Nasima Bagum; Mustofa Tanbir Kuhel]
Abstract: Laser machining plays a vital role in the manufacturing industry

Keywords: in producing innovative products and creating designs that were impossible by
the traditional manufacturing process. Due to its consistency in achieving
 Laser machining; rapid cutting speed, high-quality cut surfaces, high precision, small heat-
 Non-metallic affected zone (HAZ), and minor work damage, laser machining has been used
in various applications over the last two decades. Because of these advantages,
materials; laser machining is ideal for a wide range of thin nonmetallic materials,
 Engraving depth; including PMMA sheets, PVC sheets, and wood. However, Identifying the
ranges of process parameters for a specific material and optimizing those
 Taguchi orthogonal parameters is challenging. Taguchi L25 orthogonal array is designed to
array; conduct the experimental run, where scanning speed and power are selected
as controllable factors, and engraving depth is considered a response factor. A
 S/N Ratio; 40 mm microchannel is imprinted on PMMA using the parameters (Power
 Optimization; =13W-50W, Scanning Speed =100mm/s-1000mm/s), and the depth of all 25
experiments is measured by microscope. The effects of process parameters are
observed by plotting all experimental results. It is observed that engraving
depth increases with an increase in power and decreases with an increase in
scanning speed. A maximum engraving depth of 0.717mm is obtained at
50Watt power and 100mm/s scanning speed.
1. Introduction
Laser machining has been used in various applications over the last two decades due to its
consistency in achieving rapid cutting speed, high-quality cut surfaces, high precision, small heat-
affected zone (HAZ), and minor work damage. Because of these advantages, laser machining is
ideal for a wide range of thin nonmetallic materials, including PMMA sheets, PVC sheets, and
wood. Regardless of the diversity of laser applications, the complexity of photons interacting with
matter and the Multiphysics nature of the phenomena (thermal, fluidic, optical, mechanics,
etc.) involved during laser processing make it challenging to identify the ablation threshold or
materials’ thermal effect. Besides, the rate of material removal, the heat-affected zone (HAZ),
surface roughness, etc., are also affected by input variables such as air pressure, laser power, and
scanning speed. The acronym for the laser is ‘Light Amplification by Stimulated Emission of
Radiation.’ Here, solid bulk is ablated in a predetermined pattern using a laser beam (Kaldos et
el., 2004).
Page | 1280
Microchannels are primarily used in biomedical and microfluidic applications (Prakash et al.,
2015). Because of its low cost and relatively more straightforward fabrication processes, PMMA
has emerged as a solution to many microfluidic devices in recent years. PMMA is entirely
transparent and is used as a replacement for glass. When treated using laser processing techniques,
it generates less heat-affected zone (HAZ) than other polymer materials such as polycarbonate
(PC) and polypropylene (PP) (Muck et al., 2004). Microchannel fabrication has always been a
tricky process utilizing traditional production technology. Several fabrication processes for
various purposes and materials have been developed and used throughout the years. Laser
machining is chosen for creating microchannels to produce microfluidic devices more quickly
and cheaply, as well as to establish sustainable development. Regarding process complexity,
laser-assisted microchannel manufacturing techniques are the most basic, capable of producing
nearly any form of microstructure. Kasman and Saklakoglu (2011) investigated the effect of deep
engraving on the surface roughness and engraving depth of WC (tungsten carbide) and AISI H13
and showed that ‘scanning time’ and ‘speed’ had a considerable impact; as scanning speed
increases, surface roughness increases. The rest of the parameters had minimal impact on the
desired output. Experimental investigation (Agalianos et al.,2018) of laser engraving on steel
showed that the ablated layer thickness (Dz) is increased by raising the average power (P) and
reducing the scanning speed (V). Furthermore, raising the average power (P), scanning speed (V),
and repetition rate (F) increases MRR (material removal rate). Research conducted by Haron et
al. (2019) on the effect of laser engraving parameters on titanium showed that increased frequency
results in a better surface finish. Better surface quality is obtained when the scanning speed is set
to 200mm/s. Tosun and Ozler (2004), in their experiment of hot turning on the manganese steel
workpiece, described the application of the Taguchi experimental design. Based on S/N values,
the parametric effects were evaluated. The outcomes of this exploratory study demonstrated that
parameters cutting speed and feed rate significantly impacted cutting performance. Gadallah et
al. (2015) worked on the Modelling and Optimization of Laser Cutting operations. Because of
their excellent thermal and chemical properties and low-cost non-metallic material, are replacing
metallic materials over the past decades. Moreover, low Sublimation points and thermal
conductivity make laser cutting easier for PMMA.
2. Methodology
2.1 Experimental Details
2.1.1 Laser Cutting Machine and work materials
BCL1006X laser engraving machine was used to create a micro blind channel on the surface of
PMMA workpieces of dimension 40×10×2 mm of PMMA with proper protection. The
specification of the laser system and properties and chemical composition of PMMA are given in
Tables 2.1 and 2.2, respectively.
Page | 1281
Table 2. 1: Specification of the laser system Table 2. 2: Properties of PMMA
Characteristics Value/unit. Property Variable Value Unit

Model No. BCL 1006X. Melting Point T 433.15 K
Laser Type CO2 Sealed Laser Tube Heat capacity cp 1470 Jkg-1K-1
Cooling Mode Water Cooling at constant
Engraving speed 0-60,000 mm/min pressure
(mm/min) Density ρ 1190 Kgm-3
Drive System High-precision 3-phase Thermal k 0.18 Wm-1K-1
Stepper Motor Coefficient
Spot Diameter 55 μm Coefficient of α 7.0e-5 K-1
Cutting area (mm) 1000mm x 600mm. thermal
(23.675" x 39 3/8") expansion
Cutting Speed (mm 0-40,000 mm/min Young’s E 3.2e9 Pa
per min) Modulus
Max power 80 W Poisson’s ratio σ 0.35 -
Source type CO2 sealed LASER
Wavelength 10.6 µm
Focal length 167 mm
LASER frequency 50-60 Hz
Power Supply AC110V/220V+10% V
2.2 Parameter selection
The melting temperature of PMMA is 433K; hence when the surface temperature rises above the
melting temperature, material removal is possible using a laser. Finite element analysis using
Multiphysics software shows that 13W is the least amount of laser power needed for engraving
depth. Also, in the lab experiment, at 10W laser power, no engraving depth was found. Hence the
minimum power level was chosen as 13 watts. Experiment was conducted with a combination of
five-speed levels (100, 300,500,700,1000 mm/s) and five power levels (13,20,30,40,50W).
2.3 Experimental procedure
In this research, a straight line was drawn, and the required scanning speed and power were
defined in RD WORKS V8. The software data was transferred to the laser engraving machine
by connecting the PC using a WLAN network. At first, the workpiece was cleaned using a glass
cleaner, and the workpiece was set under a laser beam. In a laser engraving machine, power is
shown as a percentage of total power. The laser beam travels on the workpiece and generates the
desired design by starting the laser engraving machine. The detailed setup is given in Figure 2.1.
In a single workpiece, five microchannels are imposed by varying laser power in figure 2.2. A
total of 5 workpieces comprise all 25 experiments. After the experiment, the workpiece was
cleaned, and the microscopic image was captured using an optical microscope, ‘Motic BA310’.
The image is opened in the software’s graphics window, and depth is calculated. Figure 2.3 shows
the 10X magnified view of the groove produced at 50 watts at 100 mm/s speed and depth
measured in software Motic Image Plus 3.0’.
Page | 1282
Figure 2.2: Engraved Surface at
Figure 2.1: BCL laser engraving machine with the
100 mm/s scanning speed at
workpiece.
different power
.
Figure 2.3: Microscopic view and measurement for 50W 100mm/s speed
3. Result, Analysis, and Discussions
According to Finite Element Analysis, 13 watts is the recommended minimum power to achieve
engraving depth. Results revealed that no engraving depth was found for both FEA and
experiment at 10 watts. The heat generated for the stipulated period at the machining surface is
insufficient to generate the required depth.
Taguchi robust design is a technique for determining the optimum control factor values to reduce
sensitivity to uncontrollable noise sources and make the system robust (Relia Soft Publishing,
2011). There are two input parameters (scanning speed and power) and one response parameter
(engraving depth). Each of the input parameters has five-level. Hence an L25 orthogonal Taguchi
design is chosen to conduct the experiment. Table 3.1 represents the measured depth of engraving
found from 25 experimental runs on PMMA in a single pass using a Laser Engraving Machine.
To estimate the effect of all parameters on the engraving depth, the mean signal-to-noise noise
ratio S/N has been calculated using Minitab 18, considering the larger, the better categories using
equation (i). Table 3.2 shows the signal-to-noise ratio and mean for each set of parameters.
1 1
𝑆 = −10𝑙𝑜𝑔 [ ∑𝑛 ] .............. (i)
𝑁 10 𝑛 𝑖=1 𝑦𝑖2
Where n is the number of experiments and y is the experimental value of engraving depth.
Page | 1283
Table 3. 1: Engraving depth from experimental run
Exp. No. Scanning Power (W) Power as % in Laser Engraving Depth

Speed(mm/s) Machine (mm)
1 100 13 16.25% 0.528
2 100 20 25% 0.532
3 100 30 37.50% 0.616
4 100 40 50.00% 0.643
5 100 50 62.50% 0.717
6 300 13 16.25% 0.487
7 300 20 25% 0.518
8 300 30 37.50% 0.598
9 300 40 50% 0.626
10 300 50 62.50% 0.645
11 500 13 16.25% 0.486
12 500 20 25% 0.504
13 500 30 37.50% 0.511
14 500 40 50% 0.602
15 500 50 62.50% 0.618
16 700 13 16.25% 0.411
17 700 20 25% 0.439
18 700 30 37.50% 0.458
19 700 40 50% 0.584
20 700 50 62.50% 0.604
21 1000 13 16.25% 0.384
22 1000 20 25% 0.417
23 1000 30 37.50% 0.428
24 1000 40 50% 0.554
25 1000 50 62.50% 0.569
To analyze and explain the effects of the process parameters on engraving depth, main effect plots
for S/N ratios and means are shown in Figure 3.1. The figure illustrates the effects of the process
parameters on engraving depth, with an increase in scanning speed, the S/N ratios drop. Thus, it
can be said that the relationship between scanning speed and engraving depth is inverse. On the
contrary, it has a positive relationship with power.
Page | 1284
Table 3.2: The signal-to-noise ratio and mean for each set of parameters.
Exp. No Power (W) Speed(mm/s) Engraving depth (mm) S/N Ratio Mean
1 13 100 0.528 -5.54732 0.528
2 13 300 0.487 -6.24942 0.487
3 13 500 0.486 -6.26727 0.486
4 13 700 0.411 -7.72316 0.411
5 13 1000 0.384 -8.31338 0.384
6 20 100 0.532 -5.48177 0.532
7 20 300 0.518 -5.71340 0.518
8 20 500 0.504 -5.95139 0.504
9 20 700 0.439 -7.15071 0.439
10 20 1000 0.417 -7.59728 0.417
11 30 100 0.616 -4.20839 0.616
12 30 300 0.598 -4.46598 0.598
13 30 500 0.511 -5.83158 0.511
14 30 700 0.458 -6.78269 0.458
15 30 1000 0.428 -7.37112 0.428
16 40 100 0.643 -3.83578 0.643
17 40 300 0.626 -4.06851 0.626
18 40 500 0.602 -4.40807 0.602
19 40 700 0.584 -4.67174 0.584
20 40 1000 0.554 -5.12980 0.554
21 50 100 0.717 -2.88962 0.717
22 50 300 0.645 -3.80881 0.645
23 50 500 0.618 -4.18023 0.618
24 50 700 0.604 -4.37926 0.604
25 50 1000 0.569 -4.89775 0.569
Figure 3.1 Main Effects Plot for Signal to Noise Ratios and Means
Analysis of Variance (ANOVA) was performed on Minitab software 18 to investigate the
significant process parameter. The stepwise regression method automatically removes
insignificant model terms and selects ‘Linear model’ to present the response and exhibits its
significant terms. Table 3.3 shows that this model's R-squared and adjusted R-squared values are
very high and close to each other. Here, the numeric value of ‘Model-F’ and ‘Model-P’ indicates
Page | 1285
that the chosen model is significant. Here P-value is less than 0.05; then the model is regarded as
statistically significant at a 95% confidence interval (Zulkali et al., 2006). The percentage
obtained shows the highest variability in the laser power data set. Then comes the other parameter,
scanning speed, whose contribution is about 34.76%. So, LASER power can be regarded as the
most critical factor in the engraving process. A similar change in the 3D surface plot of engraving
depth vs power vs speed is shown in Figure 3.2.
Table 3.3: Analysis of variance for S/N ratios (Linear Model)
Source DF Seq SS Adj SS Adj MS F P Contribution
Power 4 29.421 29.421 7.3554 43.64 0.000 59.76%

Speed 4 17.115 17.115 4.2787 25.39 0.000 34.76%
Error 16 2.697 2.697 0.1685 5.48%
Total 24 49.233 100%
R-Sq 94.52%
R-Sq(adj) 91.78%
Figure 3.2: 3D surface plot of engraving depth

The optimal operating parameters for obtaining maximum engraving depth are done based on
the maximum value of mean S/N shown in Table 3.4.
Table 3.4: Response (mean) of S/N
Symbol Level_1 Level_2 Level_3 Level_4 Level_5
A -6.820 -6.379 -5.732 -4.423 -4.031
B -4.393 -4.861 -5.328 -6.142 -6.662
Here, A stands for - Power, B stands for – Scanning speed
Hence, the optimum parameters are power at level 5 and scanning speed at level 1. Here, power
at level 5 is 50W, and scanning speed at level 1 is 100mm/s. The microscopic view of maximum
engraving depth at an optimal parametric level of 50Watt power and 100mm/s scanning speed is
shown in Figure 3.3.
Page | 1286
0.717mm
Figure 3.3: Engraving depth at 50W,100mm/s
4. Conclusions and Recommendations

This study investigates the ranges of selected process parameters, their effects on the engraving
process, and their optimization based on maximizing engraving depth. For laser engraving of
PMMA, the following conclusion can be drawn
 10W laser power should be avoided as it is too low to ablate PMMA material.
 Increasing the power increases the ablation rate at a constant scanning speed, which
means the higher the power density, the higher the engraving depth. On the contrary, at
constant power, the ablation rate decreases with an increase in scanning speed
 Laser power cause more variability in the data set, so more attention should be given
while selecting and changing laser power.
 The optimum parametric combination is 50watt power and 100 mm/s scanning speed,
achieving 0.717 mm depth.
 Using the multiple numbers of passes could give a finely engraved surface.
5. Reference:
1. Agalianos, F., Patelis, S., Kyratsis, P., Maravelakis, E., Vasarmidis, E., &
Antoniadis, A. (2011). Industrial applications of laser engraving: Influence of the
process parameters on machined surface quality. World Academy of Science,
Engineering and Technology, 59, 1242-1245.
2. Gadallah, M.H. and Abdu, H.M., 2015. Modeling and optimization of laser cutting
operations. Manufacturing Review, 2, p.20.
3. Haron, M. N., and Romlay, F. R. (2019). Parametric study of laser engraving process
of AISI 304 stainless steel by utilizing fiber laser system. IOP Conference Series:
Materials Science and Engineering, 469, 012124. https://doi.org/10.1088/1757-
899x/469/1/012124
4. Kaldos, A., Pieper, H. J., Wolf, E., and Krause, M. (2004). Laser machining in die
making—a modern rapid tooling process. Journal of Materials Processing
Technology, 155, 18151820.
Page | 1287
5. Muck, A., Wang, J., Jacobs, M., Chen, G., Chatrathi, M. P., Jurka, V., ... and
Schöning, M. J. (2004). Fabrication of poly (methyl methacrylate) microfluidic chips
by atmospheric molding. Analytical Chemistry, 76(8), 2290-2297.
6. Prakash, S., and Kumar, S. (2015). Fabrication of microchannels: A review.
Proceedings of the Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture, 229(8),1273-1288.
7. ReliaSoft Publishing. (2011). ReliaSoft's reliability HotWire: Issue 122. Reliability
Engineering, Reliability Theory and Reliability Data Analysis and Modeling
Resources for Reliability Engineers.
https://www.weibull.com/hotwire/issue122/index.htm
8. Kasman, Ş., & Saklakoğlu, İ. E. (2011, January). A Statistical Study for Optimum
Surface Roughness and Engraving Depth. In AIP Conference Proceedings (Vol.
1315, No. 1, pp. 1559-1564). American Institute of Physics.
9. Tosun, N., and Ozler, L. (2004). Optimization for hot turning operations with
multiple performance characteristics. The International Journal of Advanced
Manufacturing
10. Zulkali, M. M. D., Ahmad, A. L., & Norulakmal, N. H. (2006). Oryza sativa L. husk
as heavy metal adsorbent: optimization with lead as model solution. Bioresource
technology, 97(1), 21-25. https://doi.org/10.1016/j.biortech.2005.02.007
Page | 1288
PAPER ID: 158
Prediction and experimental validation of laser parameters for fabrications of

a microchannel.
Md. Ahsan Habib1*, Mst. Nasima Bagum 2*, Md. Mehedi Hasan Kibria3, Maaroof-Ul-
Alam4
ahsan43@student.sust.edu,nasima-ipe@sust.edu,kibria-ipe@sust.edu,
maaroof17@student.sust.edu;
1,2,3,4
Department of Industrial and Production Engineering, Shahjalal University of Science and
Technology, Sylhet-3114, Bangladesh.
* Corresponding author
Abstract
Laser beam machining gets a widely popular and effective technology in recent years due to its variety of
applications such as in industry, metallurgy, medical, and electronics. A laser beam is used for engraving
the workpiece; many different variables and parameters can influence how well it works. Laser beam
machining removes material from work material through the Photochemical (chemical ablation) process,
the Melting and blowing process, and the Evaporation (sublimation) process. However, it’s a big challenge
to identify the process parameters to create the desired channel. In this research, our main target is to predict
the depth and width of the engraving as well as threshold energy and compare it with the experimental
result. For doing this, ‘Finite Element Analysis’ is done using COMSOL Multiphysics software, and the
results show that the PMMA sheet starts to melt when the power and the scanning speed ratio is 0.013. But
in the experiment, the ratio is found to be 0.0156 hence the amount of heat loss is almost 32.67%. Based
on the threshold power, depth and width of the microchannel were calculated using a theoretical model. A
microchannel was imprinted on 11 workpieces using laser, and the depth and width were measured using
Celestron Microscope. The effects of process parameters are analyzed by observing the microscopic image
of the workpiece and by plotting all experimental results, and FEA outputs in graphs and the data are
compared. It shows the depth and width increase with the power increase and the scanning speed decrease.
The material removal mechanism is susceptible to the pulse beam mode. Moreover, observation shows that
78W and 15mm/s scanning speed gives accurate engraving. The experimental data indicate that higher
power, lower scanning speed, and continuous wave are preferable for a smooth and consistent channel.
Keywords: Threshold energy, Finite element analysis, pulse beam mode, microchannel.
1. INTRODUCTION
LASER (Light Amplification by the Stimulated Emission of Radiation) is a device that causes atoms or
molecules in a substance to release energy in the form of light at certain wavelengths and then amplifies
that energy. It emits a very narrow beam of radiation. The material is melted and vaporized to make the cut
in the laser cutting process. When the laser beam is focused on the material, a portion of the laser beam
energy is absorbed by the surface of the workpiece because electromagnetic radiation and electrons in the
workpiece interact. The laser energy is then turned into heat, raising the workpiece's surface temperature
above the melting point (Ivan Kubovský, 2020). The workpiece's surface melts, creating an erosion/cutting
Page | 1289
front (a molten layer). A pressurized gas jet blows away the eroded/cut substrate front. Evaporation requires
six times as much energy as Melt-and-Blow. This method is commonly used to cut acrylic, thermoset
polymers, rubber, specific thermoplastic polymers, wood, paper, leather, and several ceramics. Metal
vaporization cutting is only viable if the relative contribution of the molten state is kept to a minimum.
Laser beam cutting has grown in popularity in recent decades due to its numerous uses in various
manufacturing processes in industries and its unrivaled benefits. High cut quality, consistency in achieving
rapid cutting speeds, low waste, high precision, excellent end product, good surface finish, less work
damage, tiny heat-affected zone (HAZ), and many other benefits have aided laser machining to gain
increased interest in metal forming and manufacturing industries (Agalianos, 2011). Laser machining is
used to cut various metallic and non-metallic materials, especially non-metallic materials such as PMMA
sheets, PVC sheets, glass, and wood.
Antoniadis et al. (2018) investigated laser engraving on steel and discovered that increasing the average
power (P) and decreasing the scanning speed increases the ablated layer thickness (Dz). In addition,
increasing the average power (P), scanning speed (V), and repetition rate (F) increases the MRR (material
removal rate). Manninen et al. (2015) experimented to determine the impact of laser ‘pulse length’ on
nanosecond pulsed laser engraving performance. With stronger pulses, the material removal rate and
processing temperature increase. The quality of the etched surface degraded when the pulse length was
increased. When the pulse duration is less than 20ns, the temperature rises slowly; however, when the pulse
duration is greater, the temperature rises quickly.
Kubovsky et al. (2020) experimented with a study of CO2 laser engraving on wood. To calculate the surface
free energy, researchers looked at discoloration, surface morphology, and surface wetting with a typical
liquid. The free surface energy reduced as the raster density grew at 4% laser intensity, owing to a reduction
in the 'polar component' of this energy. At 8 percent laser power, the differences in surface free energy were
negligible. S. Genna et al. (2015) studied the surface mechanism formation during the laser milling on poly-
methyl-methacrylate (PMMA). This experiment aims to study the influence of the process parameters on
the removed volume, depth, and roughness. They used a 30W CO2 laser source in continuous and pulsed
wave mode. The results clearly show that the mechanisms of interaction change considerably regarding the
independence of the processing parameters and the independence of laser modes. Due to the
delicate measurements and dynamic properties of laser machining, it is tough to monitor the material
removal process, HAZ, and surface finishing by just watching an experiment. Simulation is an excellent
technique to learn more about what's going on within a workpiece. Finite Element Analysis (FEA) can
provide a better idea of what's happening inside the workpiece. FEA tools like COMSOL, ANSYS, and
ABAQUS to obtain our desired visualization of a workpiece cut by a laser beam. Several studies are
available on the width and depth of the groove created by laser beam cutting (LBC) and their interaction
with the parameters (Laser power, scanning speed, spot diameter, compressed air pressure, the position of
the focus point, workpiece thickness, and material thermal characteristics) of LBC in earlier studies and
articles published. For a laser beam machining process, the first and most crucial step in every laser
operation is to figure out how much material needs to be removed in each layer, which is measured offline
using a trial-and-error technique, and what should be the appropriate combination of this process.
This research focuses on the appropriate parametric combination of laser parameters to make a proper
groove in a PMMA material. The threshold energy is estimated using FEA with COMSOL Multiphysics
software. Based on the relationship between the numerical and experimental data, the cutting parameter for
PMMA is determined. Based on the threshold power, depth and width of the microchannel were calculated
using a theoretical model. The sensitivity of laser mode (continuous or pulsed beam mode) is also
determined.
Page | 1290
2. EXPERIMENTAL DETAILS
2.1 Equipment
The experiment is done by a 120 W CO2 pulsed laser beam machine. A computer controls the full cutting
process for various cutting parameters. According to the Design of Experiment, different average laser
power (Pa) and Scanning speed (Ss) are operated on 11 pieces of PMMA sheet having dimensions 20×15×5
mm (LxWxT). Laser machine specifications are shown in table-1.
2.2 Material
PMMA was selected for the experimental analysis because of its lower heat capacity, degradation
temperature, heat conductivity, and transition temperature. These characteristics of PMMA make it rapidly
increase the temperature and absorb radiation. The properties of PMMA are given in table-2.
Table 1: Laser cutting machine’s specification Table 2: Material properties of PMMA

Features Name/value Properties Value
Melting Point(T) 433K (160ºC).
Brand Name Bodor Density(ρ) 1180kgm-3
Model No BCL1309XM Thermal Coefficient(k) 0.19Wm-1K-1.
Cooling mode Water cooling Heat Capacity at Constant 1470Jkg-1K-1
Laser 120W CO2 Laser Tube Pressure(cp)
Beam Diameter 0.4 mm Poisson’s Ratio(σ) 0.35
Beam type Pulsed beam Young’s Modulus(E) 3.2E9
Focal Length 2cm
2.3 Methods
Pre-processing: Identify the dimension of design in RD Works V8.

 Choose a straight line to engrave onto the workpiece using RD Works V8 and connect
the PC to the machine using a cable.
 Match the image’s dimensions with the material to be cut.
Processing: Clean the workpiece and

 Place the material to be cut on a perforated table inside the laser cabinet.
 Clean the lens of the machine gently using cotton wool during each cutting process.
 Now that the material to be cut is in the correct position.
 Choose the type of cutting by specifying the scanning speed and laser power.
 Start the engraving and make a straight line onto the workpiece.
 Return the laser nozzle to its initial position.
 Remove the engraved part from the laser cabinet.
Post-processing: Assemble the Celestron Microscope and install Celestron Micro Capture Pro software
V2.
 Use the calibration scale to identify the magnification of the Celestron microscope.
 Set the workpiece under the microscope and fix it using the clips.
 Take a picture of each sample of the workpiece.
 Measure the width and in the same manner, use tool maker microscope to measure depth.
Page | 1291
2.4 Finite Element Analysis
Finite Element Analysis was done using COMSOL Multiphysics (version 5.5.0.359). In COMSOL, the
boundary conditions are first constructed after creating a 3D model of the physical domain of the workpiece
having a dimension of 20mm×0.4mm×5 using regular mesh (0.09mm-0.5mm). The laser beam passes
through the x-axis of the top surface, as shown in Figure 1. Heat penetrates vertically to the workpiece when
a laser beam falls on it. The physics of heat transmission and its boundary conditions are provided as such
if the right-hand side of the equation is set to zero and the moving heat flux is utilized as the boundary
condition along the defined route, the laser beam can be characterized as a moving heat source throughout
the physical domain. The shape of an engraving groove is primarily affected by temperature. The material's
density changes when a laser heats it to the point of vaporization with high intensity. The moving mesh
mode offers an alternative method, in which the meshes are bent and moved in line with a change in
temperature or any other user-defined function.
3. RESULTS ANALYSIS AND DISCUSSION

3.1 Threshold Energy
To obtain the threshold energy, it is necessary to gain heat above the melting temperature of work materials.
FEA shows that when the power is 13Watt, and the Scanning speed is 1000mm/s, the average surface
temperature is 440 K which is above the melting point of PMMA; that means the material yields to start
losing the material as at this combination of power and scanning speed, while in the experiment at the same
scanning speed minimum power required to remove material is found to be 15.6 watts. The results of FEA
at the stated combination of power and speed are given in Table 3, and surface temperature rise is shown
in Figure 1.
Figure 1: Temperature from FEA (a) P=13 W, Ss=1000mm/s and (b) P=15.6 W, Ss=1000mm/s.
Table 3: Temperature and depth obtained at the threshold energy
Laser Power Scanning Laser Power and Temperature Average Depth

(W) Speed Scanning speed ratio Range Temperature (mm)
(mm/s) (P/Ss) (K) (K)
13 1000 0.013 430-450 440 0.074
15.6 1000 0.0156 480-510 495 0.091
Page | 1292
Hence it is clear that although at ratio P/Ss, 0.013 surface temperature is above the melting temperature, the
material removal process initiates when the ratio is 0.0156 instead of 0.013. The density of PMMA is
1180kg/m3 (Raita G, 2015), and the specific heat is 1466 J/kg/K. the Depth of heat penetration obtained in
the Finite Element Analysis at 13W is 0.074mm. And at 15.6 W is 0.091mm, respectively. The amount of
heat absorbed can be calculated using equation (i)
Q1 = mass × specific heat × temperature changes (i)
-7
So, the mass at 13W is, mass1 = 6.98×10 Kg
Heat absorbed, Q1 = mass × specific heat × temperature changes
= 6.98×10-7 Kg × 1466 J/Kg/K × (440-293) K = 0.171 J
The depth obtained in the Finite Element Analysis at 15.6W is 0.091mm.
So, the mass at 15.6W is, mass2= 8.59×10-7 Kg
Heat absorbed, Q2 = mass × specific heat × temperature changes
=8.59×10-7kg × 1466 J/kg/K × (495-293) K=0.254 J
The amount of heat loss, h= 0.254−0.171×100% = 32.67%
0.254
i.e., there is a 32.7% heat loss while performing machining operations, which might be due to radiation and
convection (Raita G, 2015).
3.2 Engraving model
Figure 2 shows the schematic view of the groove. To predict the depth (D) of the engraving, it is assumed
that the relation between the quantity of heat (Q) and the amount of material needed to evaporate to make
a particular depth (D) is linear. The energy (Qin) absorbed from the laser is greater than the threshold heat
(Qth). Hence depth of engraving is estimated using equation (ii) (S. Genna, 2015), where Here k stands for
the proportionality constant.
D=k∆Q=k(Qin-Qth) (ii)
Figure 2 Schematic view of groove.

𝑥2
2𝛼 (− ) 𝑃𝑎
From equation (ii), D=k[( . 𝑟2. 𝑒 𝑅2 . ) − 𝑄𝑡ℎ] (iii)
√𝜋 𝑅2 𝑆𝑠
Here α is the CO2 laser absorption coefficient of PMMA (0.92), R is the spot radius and r is the beam course
length (wavelength) and x is any distance of the laser beam along the x-axis. Here, k and Qth is unknown.
To find the terms, it was imposed that the depth of the profiles was obtained in the first two test phases.
From FEA it is ascertained that the combination of 15.6W laser power and 1000mm/s laser speed reach the
threshold heat (Qth), where D and x are zero. Here α=0.92, d=0.4mm, R=0.2mm, when x=0 then r=0.2mm
Page | 1293
Then, Qth=0.0161 [J]. Finding is that for 25W of laser power and 750mm/s scanning speed, the depth is
0.11mm, then, k=0.594 [mm/J].
Using different combinations of power and scanning speed, an experimental run was carried out and both
the starting and ending depths of the channel were measured using a digital microscope. Also, theoretical
depth was calculated by adopting the values of Threshold energy (Qth) and experimental coefficient (k).
Table 4 represents both the depth data found from the experiment and theory.
Again, the engraving width (W) depends on the process parameters. If the laser has a spot diameter d,
average power Pa and scanning speed Ss. Then width can be estimated by equation (iv) (S. Genna, 2015)
𝑃𝑎⁄ −(𝑃𝑎⁄ )
𝑆𝑠 𝑆𝑠
W= d(1 − 𝑒𝑥𝑝 (− 𝐶3
𝑡ℎ)) (iv)
Where, Path is the power and Ssth is the scanning speed of laser required to reach the threshold heat (Q th)
and C3 stands for the experimental constant. The correlation of theoretical and experimental depth (C3)
is=0.993 was found by using Microsoft Excel. Table 4 represents the theoretical and experimental width
value obtained at different power and speed ratio.
Table 4: Experimental and Theoretical depth and width value
Ex Avera Scannin Pa/Ss Theoreti Experimen Deviati Theoreti Experimen Deviati

p ge g Speed ratio cal tal Depth on of cal tal Width on of
No Laser (mm/mi (J/m Depth (mm) Depth Width (mm) width
. power n) m) (mm) (mm)
(W)
1 78 15 5.2 3.14 2.95 0.19 0.795 0.785 0.01
2 78 50 1.56 0.86 0.94 0.02 0.631 0.625 0.006
3 78 75 1.04 0.63 0.72 0.09 0.514 0.491 0.023
4 78 100 0.78 0.47 0.69 0.22 0.429 0.405 0.024
5 78 200 0.39 0.23 0.46 0.23 0.251 0.267 0.016
6 78 250 0.312 0.18 0.37 0.19 0.206 0.227 0.021
7 84 50 1.68 1.02 0.98 0.04 0.650 0.62 0.03
8 72 50 1.44 0.88 0.85 0.03 0.609 0.605 0.004
9 60 50 1.2 0.73 0.73 0 0.557 0.545 0.012
10 48 50 0.96 0.58 0.62 0.04 0.490 0.465 0.025
11 36 50 0.72 0.44 0.52 0.08 0.406 0.387 0.019
Threshold Pa/Ss ratio = 0.0156 and C3=0.993
In figure No. 3, it is clearly understood that there is a slight mismatch between theoretical and actual data
when the average laser beam intensity is high, and the scanning speed is too low or too high. The engraving
depth is more than projected when Pa/Ss ratio falls. Starting from Pa/Ss ratio 0.96 to 1.68, predicted and
actual values are all almost nearby. Higher power and moderate speed are thus acceptable to accurately
predict engraving.
Page | 1294
Figure 3: Variation of depth and width concerning the ratio of the parameters.
. Figure 4: Depth and width variation at constant power and scanning speed
Figure 5: Depth and Width of the grooves at different ratios.
Page | 1295
Figure 6: Channel width at continuous and pulsed beams.
From the figure-4, it is visible that increasing average power with moderate settings of other parameters
leads to increased depth and width due to a linear increase in laser beam energy. Here extra energy melts
and evaporates more material from the micro-grove zone creating more deviation (Sen, 2015). It is evident
that engraving depth substantially decreases with increased scanning speed which means at a low Ps/ss
ratio. When scanning speed rises, the time the beam interacts with the work surface decreases, leading to
less material layer removal from the work surface and a shallower engraving (Tosun, 2004). Figure 5 shows
the surface texture of the work material (PMMA), the design and distance between the edges of the channel,
and the shape of the groove of the channel at different power and scanning speed ratios. The surface is
smooth; the edges are parallel to each other. The shape of the track is like ‘U,’ an English alphabet at a
higher ratio of power and speed. At pa/ss ratio 5.2 gives more accurate engraving. Also, the surface is not
so smooth; the edges are rough, uneven, and not that parallel to each other. The shape of the groove is more
like a ‘V’ at a lower ratio. The channel of the continuous beam produced a more similar and smooth
engraving than the pulsed beam mode, as shown in Figure 6. So, a smooth and uniform engraving
continuous beam, mode laser is preferable for fabrication. Whereas the roughness in pulsed beam mode is
higher. So, for fabricating a smooth and uniform engraving, continuous beam mode laser is preferable.
4. CONCLUSIONS
This investigation clarifies that some heat is lost during the laser-cutting process. Higher laser power
intensity and slower scanning speeds can provide a consistent groove with a decreased roughness rate.
Theoretical model can predict the width of the channel more accurately and when Power to scanning speed
ratio 0.96 to 1.68, predicted and actual values of depth are nearby. A pulsed beam is more sensitive to
producing roughness than a continuous beam. Obtained results shows that 78W and 15mm/s scanning speed
gives accurate engraving. The experimental data indicate that higher power, lower scanning speed, and
continuous wave are preferable for a smooth and consistent channel.
5. REFERENCES
Agalianos, F. P. (2011). Industrial applications of laser engraving: Influence of the process parameters on
machined surface quality. World Academy of Science, Engineering and Technology, 1242-1245.
Allmen, M. V. (2013). Laser-beam interactions with materials: Physical principles and applications.
Springer Science and Business Media.
Ivan Kubovský, D. K. (2020). Structural Changes of Oak Wood Main Components Caused by Thermal
Modification. Polymers 2020, 481-485. doi:doi:10.3390/polym12020485
Manninen, M. H. (2015). Effect of pulse length on engraving efficiency in nanosecond pulsed laser
engraving of stainless steel. Metallurgical and Materials Transactions B, 2129-2136.
doi:https://doi.org/10.1007/s11663-015-0415-x
Page | 1296
Milosˇ Madic´, S. M. (2020). Laser cutting optimization model with constraints: Maximization of material
removal rate in CO2 laser cutting of mild steel. J Engineering Manufacture, 1-20. doi:doi:
10.1177/09544054209
Nyon, K. Y.-R. (2012). Finite element analysis of laser inert gas cutting on Inconel 718. The International
Journal of Advanced Manufacturing Technology, 995-1007.
Raita G, T. I. (2015). Encycloipedia of Polymeric Nanomaterials. Prog Polym Sci, 189-276. doi:doi:
10.1016/0079-6700(88)900004-4
S. Genna, C. L. (2015). An Experimental Study on the Surface Mechanism Formation During the Laser
Milling of PMMA. CIRTIBS Research Centre, 242-245. doi: DOI 10.1002/pc.234
Sen, A. D. (2015). Fiber laser micro-machining of ti-6al-4V. Lasers Based Manufacturing, 255-281.
doi:https://doi.org/10.1007/978-81-322-2352-8_15
Tosun, N. a. (2004). Optimization for hot turning operations with multiple performance characteristics. The
International Journal of Advanced Manufacturing Technology, 11-12.
doi:https://doi.org/10.1007/s00170-003-1672-4
Page | 1297
PAPER ID: 165
AN APPROACH TO REDUCE PROCESS WASTAGES OF AN

APPAREL MANUFACTURING ORGANIZATION USING VALUE
STREAM MAPPING
Qausar Rahman1*, MD Farhad Hossain2, Saurav Paul3, Abdullah Al Rahi4

1
Data Analyst Quality and Logistics,Carhartt WIP. 731 Monipur, Mirpur-02, Dhaka-1216.
2
Logistics Officer, Carhartt WIP. House No 14, Road No 14, Nikunjo, Dhaka.
3
Lecturer, European University Of Bangladesh.
4
Supply Production Leader, Decathlon. Uttara Sector 14 road 13 house 36, Dhaka
*Qausar Rahman
Abstract: Waste is a key issue for every manufacturing company, and in the
Keywords: garment production system, where the majority of processes are manual and
labor-intensive, the quantity of waste is substantial. Due to product variation,
● Lean most tasks in garment manufacturing cannot be automated. Anything that does
Manufacturing not bring value to the process is considered waste. Waste elimination should
become one of the most important and long-term objectives for any
System;
manufacturing firm. In the RMG industry, the waste of retention time,
● Value Stream inventory of work-in-progress, and the transportation time between operations
are significant issues. Due to incorrect expansion, the optimal plan for
Mapping;
manufacturing is frequently disregarded, resulting in additional waste. Due to
● Productivity waiting time (retention time, work in process inventory, transportation time,
Improvement; etc.), the production lead time grows, which in turn affects the manufacturing
plant's total capacity, causes production schedule interruptions, etc. Value
Stream Mapping is utilized because it facilitates the scientific representation
of material and information flow. The current state of the VSM map was
created by mapping the information and product flow of a Dhaka, Bangladesh-
based multinational RMG company. This project's objective is to identify
wastes through Value Stream Mapping (VSM), highlight improvement
opportunities for the overall process, and develop a future state map to reduce
these wastes.
1. INTRODUCTION
Apparel manufacturing units which are called the heart of our economy plays a key role in the
development of our country. This sector is one of the biggest contributors to our gross domestic
product (GDP). Although RMG sector’s contribution to our GDP has been declining for the last
consecutive five years, it doesn’t mean that this sector is declining. Only the other sectors are also
growing which is needed for a stable economy of a country. But the sudden hike in dollar price
and the American Apparel and Footwear Association (AAFA) promoting Bangladesh’s apparel
manufacturing units to produce more orders in the next 2 years is a huge opportunity for this
Page | 1298
sector of our country. According to the Export Promotion Bureau (EPB) the garments imported
into Germany registered a 27.74 percent year-on-year growth. Garment shipment to European
Union (EU) increased by 33.87 percent year-on-year in the last fiscal year. Export to the UK
increased by 30.56 percent, Canada by 33.22 percent and apparel shipment to non-traditional
markets grew by 25.40 percent [1]. So after the recovery from the COVID-19 pandemic, the
business in this sector is getting more attention than before. Also the interest of the US top brands
to move their orders from China, Indonesia and Vietnam to Bangladesh will also be a big boon
for this sector’s growth. Also the empowerment and employment of our country’s women are
directly or indirectly is influenced by this sector. Almost 2 million workers in 4000 plus factories,
which is about one-fourth of the number of employees engaged in the manufacturing sector,
constitute the real backbone of the country’s economy [2]. As the business is getting competitive,
the very important driver for both manufacturers and employers is “time”; the manufacturing unit
that delivers goods within a short time with conferring to the quality of the goods mentioned will
be the market winner. But do to improper strategy, plan and expansion the financial growth of the
apparel manufacturing units isn’t as per expected. The financial growth of any company depends
upon productivity improvement and waste minimization. While products differ in each factory,
the typical wastes found in the manufacturing environments are quite similar. After years of work
to eliminate waste, Taiichi Ono, a former executive at Toyota, identified seven categories of
waste. These wastes are [3]: Waste from overproduction, waste from waiting time, waste from
transportation, waste related to useless and excess inventories, waste in the production process,
waste for useless motion, and waste from scrap and defects.
The success of any manufacturing organization highly depends on production lead time, quality
of product, production cost, and overall productivity level. Among these factors, this study
focuses on the production lead time, which is strongly associated with value-added and non-
value added time. Identifying the value added and non-value added steps will give a guideline to
remove the non-value added time to reduce production lead time for faster response to markets.
The focus of this research will be to reduce the non-value which can also be named retention
time to improve the production lead time. There can be many causes of retention time but here
the focus will be the work-in-process transportation time. To find these retention times and
visualize them in an organized way the Value Stream Mapping method will be applied. The data
are collected from a reputed MNC based in Dhaka. For privacy reasons we cannot give the
exact name of the company, so we will simply refer to it as "Company H" throughout this paper.
This research has been conducted with some specific objects, which are :
a) Identification of retention times applying Value Stream Mapping (VSM)
b) Calculate the production lead time from the Value Stream Mapping and assess
the retention times.
c) Reduce retention times and propose a future state VSM to reduce the
production lead time.
2 SIGNIFICANCE OF THE PROJECT:

In concern organization will be able to-
 Visualize all kinds of waiting time and information delays at a glance.
Page | 1299
 Understand current stage of value stream map and concern about lead time.
 Revise & will draw the future stage of value stream map.
 Will be able to develop implementation strategy based on future stage map.
3 BRIEF COMPANY PROFILE:

“Company H” was established in 1992, we have been providing quality fashion lingerie and
swimwear to major global retailers for over two decades. At this moment with over 33,000
employees and 13 manufacturing facilities across 3 countries, with centralized pre-production
office and logistics center have 3 goals to maintain for it’s customers which are price, quality,
and speed. The considered unit of “Company H” has a production of 3.5 Million units with 170
production lines.
4 METHODOLOGIES:
In this project, Value Stream Mapping is used for lean implementation. Value stream mapping
(VSM) is a method for illustrating and analyzing the logic of a production process. The
terminology stems from the metaphor of the production process as a steady stream of products
where value is added for each step that the products take downstream. This metaphor and the
terminology also strengthen the notion of continuous flow as the ultimate form of production – at
least in terms of efficiency.[4] As the manufacturing unit only produces undergarments it has only
two types of product categories top and bottom. So this study will be done by averaging the
process time for the top category only as only one floor is dedicated to the bottom units.
Product Category Machine Required Average SMV Pitch Time

Top 24 12 Minutes .5
Bottom 12 6 Minutes .5
4.1 PROCESS STEPS

A value stream map gives a graphical overview of the flow of material and information in a
production process. This is a good foundation for understanding how activities and operations are
connected and form a basis for analyses of the process which can help management, Industrial
engineers, production associates, planners, suppliers, and customers recognize wastes related to
materials and information flow that affects lead time, space and cost. This project includes the
below steps:
1. Creating the current state map using Microsoft Visio.
Page | 1300
a) Create an outline of the
process: this first step, we
create the ‘backbone’ of
the VSM. We identify the
operations and draw them
in a straight line. Then we
add the external sources;
customers to the upper
right and suppliers to the
upper left.
b) Draw the flow of
information and Figure 1.4.1: VSM Methodology
materials: In the second
step, we add information about how materials and information flow through the value
stream.
c) Add process data: When adding process data, it is important to recognize what is
useful for the given situation and purpose.
d) Add timeline and calculations: In this final step, we need to calculate the takt time,
the process times and waiting times (inventory lead times) and add these to our VSM.
These will be used for estimating the total lead time, process time and process
efficiency.
2. Analyze the current step map.
a) Analyze process capacity.
b) Identify the bottlenecks/ retention times.
3. Look for wastes.
4. Creating the future state map using Microsoft Visio.
a) Move towards continuous flow: One of the most important principles of Lean and
the key to an efficient value stream is the goal of continuous flow.
“Flow where you can, pull where you must.”
5. Rebalancing the process.
4.2 DATA COLLECTION WITH THE HELP FOR VSM LOG SHEET.
Here SIPOC diagram [5] (suppliers, inputs, process, and output, customers) to document the process
steps. Starting with the customers then work forward. Going by this process when completed, there
should be enough data to understand the value stream and begin creating a current-state value stream
amp with a visual representation including a more detailed depiction of the value stream. Data is
collected from every sections of the production plant. The current state of the value stream map is
drawn based on the analysis of these section-wise data.
4.3 ANALYSIS OF DATA
After collecting and organizing the data simple analytical technique is used for data sorting and
analysis. Microsoft Excel program is used for some calculations. The value stream map is drawn
with the required data using Microsoft Visio VSM module.
Page | 1301
Here the total lead time minus the value added time is considered as non-value added time. The toal
lead time includes the lead time of every process of the
facility.
After making the whole process Value Stream Map the

Total Lead time comes to 70.03 Days. The Value added
time for a unit is 1128.8 Sec, the total Process/Cycle
Time is 1264.71 sec and the Non-Value Added Time for
a unit is 6049411.88 sec. So the Value Added % is
.0186% for a unit. The distance traveled by a unit is
615.52 meters.
Figure 2: Current State Map (Value Stream Map)

The lead time for major sections of the current state map
is 20.02 days for Raw Material, 10.21 days for Cutting, 4
days for sewing, .45 days for AQL, and 35.35 days for
Packing.
Table 1: Current State Map VSM Metrics
Current State Map VSM Metrics

Value Added Time 1128.8 Sec
6049411.88
Non Value Added Time
Sec
Value Added % 0.0186%
Total Process Time 1264.71 Sec
Total Inventory Points 15
Total Distance Travelled By 615.52
a Unit Meters
Total Lead Time 70.03 Days
Raw Material Lead Time 20.02 Days
Cutting Lead Time 10.21 Days
Sewing Lead Time 4 Days
AQL Lead Time .45 Days
Packing Lead Time 35.35 Days
The overall scenario of the studied areas is finally

portrayed in a value stream map (Fig-2.), which
portrays the existing material and information flows
along with Cycle/Process time, capacity, and Lead times.
Page | 1302
4.4 REVISE AND DRAWING THE FUTURE STATE MAP
After analyzing the current state map, the lead time for the packing is the most time-consuming
for the facility. It is due to the facility layout of the manufacturing unit. The facility is 10 storey
tall building. The facility warehouse is on the 1st floor (Ground Floor), from 2nd to 6th floor are
sewing floors, the 7th floor is for canteen, 8th floor is for
Cutting, 9th floor for packing and 10th floor is for
Management Office. On the figure below the green
arrows indicate the smooth flow of the product and
the red line shows us the excess/waste in
transportation due to poor facility planning. Here not
only the packing section increases the product’s total
distance traveled and non-value added time but also
the lead time. One of the major goal of VSM is to
smooth the flow of material throughout the
production facility. To reduce the lead time for
packing the change of the floor arrangement is
needed. If we move the packing section in the flow
of product processing flow then the lead time for the
packing section will improve tremendously. The
revised facility layout is suggested by re-locating the
packing section to 2nd floor. For this the cutting
section need to be moved to 9th floor and the 2nd floor
sewing section will be moved to 8th Floor. By taking
this measurement the smooth flow of the production Figure 4: Floor Layout and
process is ensured. Product Flowing Path.
5S is a method to reduce waste and optimize Figure 5: Revised Floor Layout
productivity through maintaining an orderly workplace and Product Flowing Path.
and using visual cues to achieve more consistent operational results. This tool is a systematic
method for organizing and standardizing the workplace. By applying 5s in the store, cutting,
sewing, and finishing section, the organization can organize the workplace. It is found that the
hand tape, sticker, gum, poly, stapler and some hand tools are not placed in a particular location.
This causes a problem to the users to find them at right time. A significant amount of time is
wasted just for this. By applying 5s tool, non-value added time can be reduced notably. Work-
in-process (WIP) or Process inventory is found a notable problem in the organization that adds a
significant amount of non-value added time. Minimizing WIP with the application of JIT
philosophy, the organization could save waiting time for materials in the workstations.
The key findings and strategies that are used for future state implementation:
Page | 1303
1. Need to Improve “Supplier’s material
Quality to reduce the cost of quality.
2. Total Cutting process improvement
from input to output based on the “Pull”
system.
3. Develop standard production methods
and layouts for Bra & Briefs.
4. Improve “Quality” throughout the
processes by improving the current
Figure 3: Future State Map (Value Stream Map)

state of 11 quality gates.
5. Improve “Style changeover”
performance using QCO.
6. Organize Kaizen Blitz and 5s in every
section to improve efficiency and
production capacity.
7. Eliminating the raw material
quarantine process as it isn’t
required anymore.
After adopting all of these techniques and
lean tools on the manufacturing plant, the
production of sewing (one of the most
important operations to increase the pull
system along the manufacturing process)
was boosted to 118,000 units with the same
amount of resources. The elimination of the
quarantine process for raw materials and
this rise in the pull system decreased the
lead time of our future system from 70.03 to
59.27 days. As a result, the manufacturing
process' value-added percentage grew
to.0220% from.0186% previously. The
major future state map metrics are shown in
Table 2.
Page | 1304
1.5 CONCLUSIONS
Table 2: Future State Map VSM Metrics
This project depicts the entire situation,
from the procurement of raw materials Future State Map VSM Metrics
through the shipment of finished goods. Value Added Time 1128.8 Sec
The value added % is derived from the Non-Value Added
5119799.20 Sec
entire order lead time. Extremely Time
concerning is the fact that only a small Value Added % 0.0220%
proportion of an order's overall processing Total Process Time 1264.71 Sec
time is spent adding value. In an era where Total Inventory 14
the fashion industry is attempting to Points
shorten production lead times with smaller Total Distance 580.68 Meters
order quantities and frequent style Travelled By a Unit
changes, the reduction of material Total Lead Time 59.27 Days
inventories and mapping of the value chain Raw Material Lead 14.49 Days
have become crucial for all manufacturing Time
Cutting Lead Time 9.64 Days
units. As observed in this project, if
Sewing Lead Time 3.72 Days
"Company H" follows the practice
AQL Lead Time .45 Days
introduced in the future state map, the lead
Packing Lead Time 32.92 Days
time will be drastically reduced from 70.03
to 59.27 days. These strategies will also result in more efficient space utilization. Even
after standardizing all of these tactics and lean tools, "Company H" can examine its VSM
periodically to attain the average aim of 50 days lead time by progressively improving
the lead time. This project reveals that the retention time (work-in-process) is a significant
concern for the analyzed "Company H," resulting in decreased productivity, inefficient
space usage, and increased production lead time. Every garment manufacturing plant may
have these challenges; therefore, they should employ the VSM approach to identify and
eliminate them in order to boost their production and decrease their lead time, thereby
improving their market competitiveness.
1.6 REFERENCES:
1. https://www.thedailystar.net/business/news/apparel-shipment-us-rises-51-
3066406
2. www.acronymfinder.com/Bangladesh-Garment-Manufacturers.
3. Liker, Jeffery K., „The Toyota Way – 14 Management Principles from the World‟s
Greatest Manufacturer‟, Page 27, McGraw Hill, 2004
4. Langstrand. (2016). An introduction to value stream mapping and analysis.
5. SIPOC Diagram: Definition and Steps - Tallyfy. (2017, November 13). Retrieved
September 5, 2022, from https://tallyfy.com/sipoc-diagram/
Page | 1305
PAPER ID: 166
QUALITY MANAGEMENT FOR RMG’s: A QUALITY GATE

CONCEPT
Qausar Rahman1*, MD Farhad Hossain2, Abdullah Al Rahi3

1
Data Analyst Quality and Logistics,Carhartt WIP. 731 Monipur, Mirpur-02, Dhaka-1216.
2
Logistics Officer, Carhartt WIP. House No 14, Road No 14, Nikunjo, Dhaka.
3
Supply Production Leader, Decathlon. Uttara Sector 14 road 13 house 36, Dhaka
*Qausar Rahman
Abstract: In recent times, as the value of the dollar has risen and our reserves
Keywords: are decreasing at an alarming rate, the responsibility of reserving dollars
through our Ready-Made Garments has become crucial. In the next two years,
● Lean Bangladesh will receive more orders from the United States than Vietnam,
Manufacturing Indonesia, and China, according to the USFIA 2022. Therefore, the objective
of our RMGs should be to increase quality in order to make the business more
System; sustainable, given that China and Indonesia are superior in this regard. In order
● Quality to reduce cost and improve the quality of the garments a quality management
concept is used to establish quality gates in a lingerie manufacturing unit with
Management;
a capacity of producing 4 million products per month. A well-structured
● Quality Gates; procedure is suggested for identifying and handling fluctuations in the quality
of work-in-process products, leading to a reduction in defective products by
● Sustainability;
detecting deviations at early process stages and, additionally, offering the
possibility for process control and feedback. By tracing prior inspections, a list
of significant, concurrent, and recurrent defects was compiled. By establishing
internal decision points (Quality Gates), steps of checking can be aggregated,
minimizing effort for quality control and summarizing information on
pertinent quality parameters of work-in-process products. Finally, a quality
gate is recommended for the organization for enhanced process control.
1. Introduction
Many manufacturing companies are exposed to international competition in which product

quality is becoming increasingly important for long-term business success in the trade-off
between productivity and customer satisfaction [1,2]. In this context, quality is characterized by
high product quality along with high process reliability and transparency within the
manufacturing process, low scrap and rework rates, high delivery capability as well as on-time
delivery [3]. The increase in demand for garments in apparel manufacturing industries is placing
a huge burden on the manufacturing units’ capacity for production. Increasing capacity is a time-
consuming and costly process, so the only way to increase our capacity is to reduce rework by
improving the quality of production through quality management. Although the terms ‘‘quality
management’’ (QM) and ‘‘quality assurance’’ (QA) are used extensively, there is little common
agreement on their exact meaning, and there are many misconceptions about QM/QA systems,
procedures, and tools. The key to a successful QM/QA is a relentless focus on processes and on
Page | 1306
giving people involved in organizations’ processes the tools necessary, and also the responsibility,
to improve quality. Quality management attempts to ensure a high level of quality in all areas that
have a direct or indirect impact on customer satisfaction. With the specification of guidelines,
QM forms the basis for a minimum of rejects and maximum customer satisfaction. It is important
to document and analyze all processes from production to delivery. These are then optimized (if
possible). The customer requirements for the product should be exceeded. Thus, QM deals with
the task of improving the quality of products through organizational tasks. However, QM is in a
constant state of change and must be continuously adapted. In quality assurance, only
potentially negative influences on the quality of the end product are analyzed and then at best
eliminated. This ensures that the specifications of the quality management are observed and
correctly implemented [4]. So if the goal is to reduce the rework/re-process of the produced
products to reduce production time, quality management is the way to do it. There are many tools
that are used in quality management among which include failure mode and effects analysis
(FMEA), fault tree analysis (FTA), flowcharting, process mapping, and quality gates. Among
these tools FMEA, FTA is used intensively to correct a single defect by going to the root of the
cause and possibly eliminating it. Here the tools process mapping and quality gates are used to
identify the possible inspection points and their defects list to focus only on the required defects
for the particular quality gates thus reducing inspection time and quality of the inspection. This
will also help to further analyze the defects using FMEA and FTA methods later.
Consequently, quality inspection planning is an integral part of the manufacturing planning

process. It is often derived from a weak information basis and therefore relies on the experience
of the production planner. Currently, quality inspection test already produced parts and therefore
have a retrospective character. Although there is an increasing availability of data from the
manufacturing phase as well as data analytics tools that can turn this data into valuable
information, there is a lack of revision of inspection activities as part of quality control.
To overcome these challenges, it is necessary to consider the entire manufacturing system rather
than individual and isolated processes. Thereby, early detection of deviations and the analysis of
cross-process interactions are enabled to complement the classic measures for error prevention
and avoidance. It is therefore important to understand the dynamics and interactions between the
individual elements (e.g. products and processes) within the manufacturing system in order to
derive targeted and case-specific quality management strategies. [5–7]
Page | 1307
This paper will discuss the process mapping through value stream mapping (VSM) and then
establish the quality gates to focus only on the defects that are detectable on the particular quality
gate thus saving time and cost of the processes.
1.1 BRIEF COMPANY PROFILE:
“Company H” was established in 1992, we have been providing quality fashion lingerie and
swimwear to major global retailers for over two decades. At this moment with over 33,000
employees and 13 manufacturing facilities across 3 countries, with centralized pre-production
office and logistics center have 3 goals to maintain for its customers which are price, quality, and
speed. The considered unit of “Company H” has a production of 3.5 Million units with 170
production lines. After the pandemic the increase in orders
2. THEORETICAL BACKGROUND
2.1 Quality Inspection and Quality Gate Concept
Quality gates, as presented in the stage-gate model [14], are nowadays a widespread quality
management tool in product development and project management, i.e. in the automotive sector
[15]. They have been shown to be applicable to a large variety of processes and process chains,
enabling a comprehensive control of quality-relevant features [16,17]. Although initially
proposed for use in product and software development, an adapted stage-gate model has been
shown to be applicable to support quality management for production systems, especially in the
context of zero-defect manufacturing [18]. While in strategical planning and development
processes, a strong focus is on the consequent application of quality gates for control of process
Page | 1308
results, quality gates in production and assembly processes focus on control of process parameters
[19].
Within a manufacturing process, raw materials are transformed into final products by value-
adding to meet customers’ requirements. The value-adding process consists of the physical
transformation of the products (e.g. transforming geometry, state of matter, and chemical
composition). Consequently, product states with their respective features change with every
process step. Product quality characteristics or their variances are accumulated. The final product
quality is thus influenced by all transformation processes within the process chain [8,9]. Against
this background, an in-depth understanding of the relationship between process, intermediate
product and final product is required in order to achieve the best possible quality of the end
product. [10]
In this context, quality inspections are important elements of manufacturing systems to ensure the
production of high-quality products. They are often performed offline, after an intermediate
product is finished at the end of a process chain. This can be achieved by measuring certain quality
attributes defined within the quality requirements of products (product quality) or process
requirements (process quality) [11]. The idea of quality gates (QGs) is to systematically divide a
defined process chain into different quality-relevant decision points.
Along these QGs, the perceived quality of the object is inspected, securing that defined quality
attributes are met before allowing any further progress [12]. Figure 1 (a) gives an overview of the
QG concept for manufacturing systems.
At each QG, the current product states are measured with specific technologies such as vision
inspection. An intermediate product state (IPS) characterizes an intermediate product at a certain
observation point along the manufacturing process chain by a combination of intermediate
product features (IPFs) [12]. An IPF is defined as a quantitative (e.g. temperature) or qualitative
(e.g. material composition) description of the product itself as well as definable and deterministic
measurements [7,10]. At a specific QG, the IPS with its measured IPF is checked against
previously defined target values and classified, e.g. as “good” or “not good” products. Moreover,
the process quality can be determined by analyzing certain process parameters (PPs). In this
context, the process capability can be assessed, or a certain PP can be evaluated regarding its
position within a predefined process window.
Within a manufacturing system, different relations, or dependencies between PPs and IPFs can
occur. Bi-directional relations between IPFs and PPs may exist. Additionally, a forward relation
can appear between a PP and an IPF at successive observation points. Forward relations can also
occur between two following IPFs or between random IPFs. Due to the direct influence of PPs
on IPFs and vice versa, the consideration of the process quality in addition to the product quality
is of great importance for the overall quality [5,6].
2.2 Concept of Product Phenotypes in Manufacturing Systems
To analyze the behavior and interactions of different intermediate products within a

manufacturing system, a comprehensive (cross-)QG modelling and analysis of IPS is required.
Therefore, IPFs can be clustered to define intermediate product classes (IPCs) at defined
observations (e.g. QGs). These IPCs can then be used to e.g. analyze the behavior of different
intermediate products at certain observation points to later derive certain, e.g. IPC-specific,
inspection strategies.
From a system perspective, the propagation of product class affiliations can be examined in more
detail. In order to examine these propagation paths, it is necessary to know the probabilities, in
Page | 1309
which IPC at QG n+1 the intermediate product will be and if it is assigned to a certain IPC at QG
n. Considering QGs with the associated IPC, individual propagation pathways with high
propagation probabilities can be identified. In the following, these propagation paths of IPC shall
be referred to as "product phenotypes". In the context of manufacturing systems, a phenotype
describes the characteristic of a certain product within the manufacturing system. The
characteristic is defined by the aggregation of the IPF represented by the assignment to IPCs.
Figure 1 (b) gives an overview of intermediate product classes and product phenotypes.
2.3 Inspection Strategies in Manufacturing Systems
Quality inspections are initially planned during the development phase of a manufacturing system
and thus before the start of production. This phase is characterized by high uncertainties, various
assumptions, and only limited availability of data. In contrast, quality inspection strategies should
be implemented as part of quality control. Figure 2 provides an overview of the current state of
research on inspection strategies in manufacturing systems [13–25]. On the horizontal axis,
several identified publications are plotted. Different evaluation criteria were defined on the
vertical axis. These criteria are used to evaluate individual publications. The color red indicates
that the criterion was not met, yellow partially met, and green fully met.
Currently, there is a lack of inspection strategies that include product individual requirements
based on their specific intermediate features. Individualization of inspections should be achieved
in an adaptive way and could be supported by a variation of inspection intensity. As specific
inspection strategies for each individual product lead to high planning complexity, products must
Page | 1310
be aggregated. Moreover, inspection strategies must consider the manufacturing system instead
of isolated processes. No current approach addresses these requirements.
In the planning of inspection strategies, various scopes can be identified. Therefore, they are
classified into either process-specific or product-specific inspection strategies. Process-specific
strategies are characterized by a specific consideration of individual quality gates whereas
products are viewed uniformly. They are applied when inspection equipment has significant
influences on product properties or inspection errors are an issue. In this regard, Schmitt et al.
introduced an approach for virtual quality gates that calls for zero type I (or type II) inspection
errors combined with a physical inspection of the conforming (or non-conforming) products [13].
Product-specific inspection strategies employ a systemic perspective on the manufacturing
system. However, a granular consideration of products is necessary. This requires the generic
description of products. The product state concept characterizes intermediate products at certain
observation points (i.e. quality gates) based on a combination of intermediate product features
(IPFs) [7]. Similar intermediate product states (IPSs) at a given quality gate can be aggregated as
intermediate product classes (IPCs) using clustering algorithms [5]. This work was not continued
with regard to the definition of inspection strategies.
3. QUALITY GATE ESTABLISH METHODOLOGY

First to identify all the processes and sub-processes the help of value stream mapping is taken.
VSM is applied to get the current state map and identify all the processes and sub-processes.
After identifying all the processes and sub-processes to get the quality process map and quality
gates 5W’s and 1H of quality gates is used, also the previous quality defects data is analyzed to
get the probability of the defect in
various gates. The intermediate product
classes (IPC’s) with similar
intermediate product states (IPS’s) can
be inspected through the same quality
gates classification. In manufacturing
systems, the end-of-line inspection
generally classifies products as either
“good” or “not good”. Consequently,
only two IPCs representing these
classifications exist for the final QG.
This characteristic can be used to Fig 3: IPC and IPF Relation.
determine the scrap probability of a
phenotype at any position within the
process chain: To identify the scrap probability at process step n, all IPC observations of processes
1 to n-1 are taken into consideration. The scrap probability then equals the probability that the
product lands in the scrap-IPC at the end of line inspection. The identification of high scrap
probabilities allows for the anticipatory scrapping of presumably defective products. This implies
that, based on data analytics, an inspection strategy may call for removing a product phenotype
at a point within the process chain where it is still conforming.
Page | 1311
Who?
• Process Managers.
What?
• Checking qaulity criteria.
When?
• Whole lifecycle - from initiation to closing.
Where?
• Departments and processes.
Why?
• Gain control over quality development and status.
How?
• Define and assess quality criteria and gates.
Also, there are some practices that are followed by project managers when applying the quality
gate concept in a manufacturing unit:
Fig 4 : 5W’s and 1H of Quality Gates.
I. Pay special attention to the positioning of deliverables, and gates

The correct positioning of quality criteria especially in the development process is key
for successful quality gate application.
II. Apply E2E quality gate architecture
Ensure that quality gates cover the whole project lifecycle and do not only cover a
particular part.
III. Add actual names to the quality gate checklist and not only a department.
Put the responsible people, e.g. the department’s manager (John Doe, Morten
Menigmand, Erika Mustermann) into the column for deliverable responsible instead of
the department (HR, IT, R&D). Otherwise, afterward, a discussion can come up when a
target is not met and who was responsible for it.
IV. The number of quality gates should be as low as possible for a unit to get the best out of
this.
However always assess the number of quality gates needed and required according to the
product lifecycle, its characteristics, scope, and the company’s general procedure and
quality standard.
V. You cannot manage what you cannot measure
It is desirable to have measurable criteria, however, this is not mandatory,. It can also be
in the Boolean form of yes or no. Nevertheless, all sub-criteria must be met before passing
on quality gates.
VI. One quality gate for each transition from one project phase to another.[20]
Page | 1312
The quality process map and quality gates are standardized for each phenotype, but the inspection
method for each phenotype will vary based on its intermediate product features (IPF) and
intermediate product states (IPS).
Taking into account the IPCs and IPFs, a quality process map with 11 quality gates from G1 to
G11 and 6 categories from A to F were generated after the completion of all steps. Analysis of
historical data reveals the major defects and their occurrence percentages and is included in the
quality process map. Microsoft Visio is used for the visualization of the quality process map [Fig.
5]. The swim lane model is used to represent the quality gates to show the subsequent processes
along the departments.
Page | 1313
Page | 1314
Fig. 5: Quality Process Map & Quality Gates.
4. APPLICATION OF THE QUALITY GATES
In a prototype case study, the proposed quality gate will be utilized. According to the quality
process map, "Company H's" manufacturing system will include eleven quality gates from the
receiving of raw materials through final product inspection. At the inspection gates, the products
will be categorized as certain product phenotypes based on their IPFs, and certain methods will
be implemented based on the inspection gate in order to reduce rework and save time. As the
gates will work as a cross-functional inspection system, the same IPCs will be inspected numerous
times to lower the percentage or chance of defects on the final products, hence enhancing the
conformance to customers' requirements and increasing customer satisfaction. The quality gate
gives a target of defects to base the inspection strategy for identifying the most defects from the
raw materials/products inspected. This method in result will reduce cost by saving time and
reworking process. After implementing the quality gates the defects data from the quality gates
will be further analyzed to get all possible defects with their probabilities to further strengthen the
inspection strategies. "Company H" records defects using defect codes. Currently, defects are
recorded on paper and entered into an Excel file for analysis, but in the near future, this system
will be replaced with a Quality Management System (QMS) to obtain real-time data.
5. CONCLUSION
The higher demands require a higher capacity of production but due to budgetary issues and time
restrictions, it is not always feasible to raise the capacity of a manufacturing plant. The possible
way to tackle this situation is to reduce the time of re-processing through quality management
and use that saved time to produce more to increase the capacity. This will also result in
conformance to customers’ demand and get higher customer satisfaction. The quality gate concept
doesn’t only establish quality ensuring points in the manufacturing system but also as an
information hub to collect data for implementing the best inspection strategies for a certain
phenotype. The quality gates are focused on product phenotypes so their custom strategy for the
specific phenotype is optimized to identify it’s defects in the earlier stages so that the re-process
can be easier and Type I errors can be avoided. This will ensure that the defective intermediate
product isn’t passed onto the next process. The key challenges for this method is the strategies
developed for the phenotypes and the training quality inspectors on those strategies. Also, the
collection of real-time data is also a challenge as “Company H’ is still working without a QMS.
Integrating the quality gate concept with a quality management system and the most effective
phenotypic tactics allows for additional development. In addition to phenotypes, the notion can
be expanded to include intermediate product classes (IPC). Although the procedure will be more
complicated, its precision will increase.
Page | 1315
References:
1. Haas, M.A., 2009. Quality Networking.
2. Westkämper, E., Decker, M., 2006. Einführung in die Organisation der Produktion.
Springer-Verlag Berlin Heidelberg, Berlin, Heidelberg.
3. Lieber, D., 2018. Data Mining in der Qualitätslenkung am Beispiel der
Stabstahlproduktion. Dissertation. Shaker, Aachen.
4. Topp, M. (2020, April 14). 4 Differences Quality Management and Assurance - sentin.
Retrieved September 14, 2022, from https://sentin.ai/en/differences-qm-qa/
5. Filz, M.-A., Gellrich, S., Herrmann, C., Thiede, S., 2020. Data-driven Analysis of
Product State Propagation in Manufacturing Systems Using Visual Analytics and
Machine Learning. Procedia CIRP 93, 449–454.
6. Filz, M.-A., Gellrich, S., Turetskyy, A., Wessel, J., Herrmann, C., Thiede, S., 2020.
Virtual Quality Gates in Manufacturing Systems: Framework, Implementation and
Potential. JMMP 4 (4), 106.
7. Wuest, T., 2015. Identifying product and process state drivers in manufacturing systems
using supervised machine learning. Dissertation. Springer, Cham.
8. Du, S., Xu, R., Huang, D., Yao, X., 2015. Markov modeling and analysis of multi-stage
manufacturing systems with remote quality information feedback. Computers &
Industrial Engineering 88 (88), 13–25.
9. Du, S., Xu, R., Li, L., 2018. Modeling and Analysis of Multiproduct Multistage
Manufacturing System for Quality Improvement. IEEE Transactions on Systems, Man,
and Cybernetics: Systems 48 (5), 801–820.
10. Wuest, T., Irgens, C., Thoben, K.-D., 2014. An approach to monitoring quality in
manufacturing using supervised machine learning on product state data. J Intell Manuf
25 (5), 1167–1180.
11. Wirtz, A., Gächter, C., Wipf, D., 1993. From Unambiguously Defined Geometry to the
Perfect Quality Control Loop. CIRP Annals 42 (1), 615–618.
12. Wuest, T., Liu, A., Lu, S.C.-Y., Thoben, K.-D., 2014. Application of the Stage Gate
Model in Production Supporting Quality Management. Procedia CIRP 17, 32–37.
13. Schmitt, J., Bönig, J., Borggräfe, T., Beitinger, G., Deuse, J., 2020. Predictive model-
based quality inspection using Machine Learning and Edge Cloud Computing. Advanced
Engineering Informatics 45.
14. Cooper R G. Stage-gate systems: a new tool for managing new products. Business
horizons 33 (1990) 3, 44-54.
15. Braunsperger M. Designing for quality—an integrated approach for simultaneous quality
engineering. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture 210 (1996) 1, 1-10.
16. O'Connor, P. R.: Implementing a stage-gate process. Journal of Product Innovation
Management (1994); 11:183-200.
17. Lucassen M. Quality gates: empirische Analyse und Gestaltungsempfehlungen. 2009.
München, Techn. Univ., Diss., 2009.
18. Wuest T, Liu A, Lu S C-Y, Thoben K-D. Application of the stage gate model in
production supporting quality management. Procedia CIRP 17 (2014), 32-37.
Page | 1316
19. Hawlitzky N. Integriertes Qualitätscontrolling von Unternehmensprozessen:
methodische Gestaltung eines Quality-Gate-Konzeptes zur Planung, Messung und
Steuerung der Prozessqualität: TCW, Transfer Centrum für Produkt.
20. Quality Gates in Project Management - apppm. (n.d.). Retrieved September 14, 2022,
from http://wiki.doing-projects.org/index.php/Quality_Gates_in_Project_Management
Page | 1317
PAPER ID: 184
Payload Estimation and Factor Analysis of a Lighter-than-Air

System as Aerostat for Day-Long Environmental Data
Collection at A Specific Altitude
Md. Tasnim Rana1*, Md. Shahidul Islam 2

Student (Department of Mechanical Engineering, Khulna University of Engineering
and Technology, Khulna, Bangladesh; email address:rana.me.kuet@gmail.com)
Professor (Department of Mechanical Engineering, Khulna University of Engineering
and Technology, Khulna, Bangladesh; email address: shahidulbitk@me.kuet.ac.bd
*Corresponding author [Md. Tasnim Rana]
Abstract: The Experts employ special drones or conventional weather

Keywords: balloons to monitor the weather or collect data from a specific altitude.
However, they take a lot of time and are highly expensive. Due to their limited
● Lighter than air; flight time, they can only provide data for relatively brief periods. LTA
● Weather data ; technology can bring a new dimension to data collection in the context of
weather monitoring. Using lighter-than-air (LTA) technology, weather data
● Payload estimation; may be collected at a particular altitude. The study presented in this writing is
● Environment primarily concerned with an LTA-based environment monitoring system with
its payload estimation in a different case study and the factors that will play a
monitoring; key role while designing the system. Through the efforts of LTA, a stable
● High altitude; system for long-term weather monitoring at a specific altitude has been
established by considering the relevant factors with the desired payload. The
● Surveillance. designed system can be utilized not just for weather monitoring but also for
other data collection or high-altitude video surveillance. The system includes
some useful alternative approaches and can be utilized to establish remote
communication networking systems.
1. Introduction
Long-time flying is one of the most critical and challenging research areas in the aerial research
field. Drones or other aerial robotics can be used as surveying or data collection medium at high
altitudes. But, battery storage limitations or high costs of technical support can make that process
difficult. Solar-powered fix wing drones are recently used for long-time flights but they are much
more expensive and also they are not able to cover a fixed altitude as they have to move at a
certain speed for generating lift. On the other hand, some researchers also use weather balloons
as a data collection system but it does not set stable at a fixed point, so continuous data collection
at a certain altitude could not be achieved with weather balloons. Tethered aerostat can make the
difference with an additional aerodynamic structure which will result in a stable system for
Page | 1318
surviving in air at a high altitude position. But, no clear data were found regarding payload
estimation and factors that will affect the flight of tethered aerostats with spherical structures.
According to the authors' understanding, Kruchinin at Moscow University conducted the first
investigation on the oscillations of a tethered sphere in a fluid flow (1977) [1]. The lift and drag
forces of ascending spherical weather balloons were described by Scoggins (1967) [2]. Other
studies in the literature focus on how surface waves affect a buoyant spherical construction that
is anchored. These include the works of Ogihara, Shi-Igai, and Kono (1969), as well as Harleman
and Shapiro from 1961. (1980)[3]. Cornell University is the home of the first team to
systematically study the transverse oscillations of a tethered sphere in a fluid flow (Govardhan
and Williamson, 1997); (Williamson and Govardhan, 1997)[4]. A "lock-in" phenomenon of the
primary vortex shedding as observed on a fixed sphere and the body motion was used by Jarvis
et al. (2001) to explain the sphere oscillations [5]. They also found that there was a mode of
oscillation at much higher flow rates that could not be explained by the conventional "lock-in"
theory because the fixed sphere's vortex shedding in that flow regime would not have any
frequency content close to the tethered sphere's natural pendulum frequency. This unexpected
phenomenon was explained by Govardhan and Williamson in 2005. They attributed the tethered
sphere's oscillations to "movement-induced vibration," a theory first put out by Naudasher and
Rockwell (1994), in which the motion of the sphere creates self-sustaining vortex forces [6].
Bearman (1984) and Anagnostopoulos' study in this area is among other sources (2002) [7].
As, there was no clear work on factors affecting the tethered aerostat case of the sphere shape
system, working with the affecting factors would add a good impact on aerostat finding. Though
the overall conception is from basic aerodynamic laws, applying the law in a point-by-point
approach is missing in previous work. So, this paperwork will express the facts in an organized
manner.
2. Experimental
2.1 Material
The basic structure for the experiment was done with carbon fiber shaped 1m in length with cross-
hole connections with each other. For the envelop material and balancer one of the most coveted
qualities for the aerostat application is the high strength-to-weight of the synthetic fibers.
Polyesters, polyamides, aramids, liquid crystal polymers, poly-benzo-xazole, and ultra-high
molecular weight polyethylene are examples of synthetic fibers with a high modulus that is readily
accessible on the market. It is also possible to use high-performancers like Spectra, Vectran, M5,
Zylon, and Dyneema.But, for the experiment Polyethylene (PE) is used with heat sealing
envelope.
2.2 Fabrication and Data transfer system:
The data for the experiment was collected from a setup of the weather monitoring system, which
is a program with an IoT-based data transmission system network.at every minute the setup with
the IoT system would send a data row where we can find temperature, altitude, and humidity at a
certain altitude range.
Page | 1319
Fig 2: Final Fabrication output of the Envelope, Sensor, and data transmission setup
Fig 1: Flow diagram of the process
As this paper will show the payload estimation part. Payload estimation will lead to the weight
capacity that someone would use to give external load on the aerostat system. Most of the cases
it seems simple when to sum up all the components of weight to find out the system weight. But
there are other factors that will affect like wind velocity, control line weight, or drag force created
by not only the envelope but also the payload itself and the control line.
3. Theory/calculation
The estimated highest weight of the system= Weight of the envelope material +Weight of the
LTA gas + Weight of the control line + Weight of additional system structure for rigidity of the
envelope + Dag force component(Control line +envelope + payload)
Page | 1320
Fig 3: Components of weight for an ideal aerostat Fig 4: On-air flight condition
Where, Weight of the envelope (W) = volume of the material× Density of the material
=𝜋𝑎𝑏 × 𝜌m [𝜌m = Material Density]
Weight of the gas Lighter than Air (LTA) =4/3 𝜋𝑎𝑏𝑐 × 𝜌g [Density of the gas]
Weight of the control line (g) =s×p
Component of Drag force acting as body weight [Fd=1/2CD 𝜌Av2 Sin (Knot Angle)]
So the total estimated weight of the system will be
W= (𝜋𝑎𝑏 × 𝜌m ) +( 4/3 𝜋𝑎𝑏𝑐 × 𝜌g [Density of the gas])+( s×p)+{ 1/2CD 𝜌Av2 Sin(Knot
Angle)}+Other structural weight that will be taken using weight cell at standard atmospheric
pressure(STP)................................................................................................................. (Eq 01)
The total payload capacity would be the net force received by eliminating system weight from
the total lift force. So, Payload Capacity= (Total Lift force received from the aerostat itself + lift
force received from the buoyancy effect from LTA gas) – Total estimated weight of the system
Total lift force (Fb) = {(𝜌a-𝜌He) (𝜋/6 − 𝑉m/d3) d3-(Ma +f𝜋𝑑2 + Mp ) }g

So the estimated highest payload
=Fb-W
={(𝜌a-𝜌He)(𝜋/6 − 𝑉m/d3)d3-(Ma+ f𝜋𝑑2+ Mp )}g−{(𝜋𝑎𝑏 × 𝜌m )+( 4/3 𝜋𝑎𝑏𝑐 × 𝜌g [Density
1 2
of the gas])+(s×p)+{ /2CD𝜌Av Sin(Knot Angle)} +Other structural weight that will be
taken using weight cell at atmospheric pressure(STP) ..................................... (Eq 02)
And this estimated payload is one of the key system factors that affect the stability of the
aerostat system. The question comes with the CD, How we can find out the value of CD?
Page | 1321
Fig 5: A sphere's drag coefficient CD as a function of the Reynolds number Re. CD and Re are both
dimensionless numbers. Keep in mind the logarithmic scales. The abrupt decline in CD at Re 2.5105, known
as the drag crisis, is connected with the transition of the sphere's surface boundary layer from laminar to
turbulent. [8].
The following graph[Fig: 5] is used for finding out the Cd for spherical objects if the
surface condition is well, sometimes for better aerodynamic stability in air, some kite
shape additional structures would add additional drag force components which will
eventually affect the overall weight and payload capacity of the aerostat system.
Where additional
1 Fd1 would depend 1on CD
Cd= CD0+ C 2 =1.28 sin (a) + [2𝜋𝑎 ]2 ……….(Eq 03)
1 /{1 + 2𝜋𝑎/
(𝜋𝐴𝑅)}
.7πAR .7πAR
As
C1=2𝜋𝑎/ (Downwash effect)
{1 + 2𝜋𝑎/(𝜋𝐴𝑅)}
CD0=1.28 sin (a)
AR = S21 . For a spherical aerostat, the S would be the diameter and the area will be
A
circular area as per the experiment.

As the experiment was done in the weather of Dhaka, the capital of Bangladesh, the air
density, average air velocity, temperature, and humidity were taken from that location
as reference. But, from the experiment we have the following data given in fig: 6 from
the Barometric pressure vs. Altitude graph, we have a relative view haw the pressure
drops with respect to the altitude. As the temperature is also changed due to altitude
variation, the final effect will happen on the density. Atmospheric density is the key
factor for calculating the payload capacity of the aerostat system as buoyancy force
mainly depends on the density. From the experimental setup, the temperature and
Page | 1322
pressure are known. From that data, we can find out the density which is a factor for
calculating the payload capacity of an aerostat system.
Barometric Pressure(pa) Density(%)

101309
100 99.80325
102000 100.5 99.6065
101069
Barometric Pressure(Pa)
100 99.213
101000 100349
99.5
Density(%)
99389
99 98.426
100000 100829 98.0325
98.5 97.639
98429 98.8195
99000 99869 97469 98 97.2455
98909 97.5 96.852
98000 97949
96989
97 96.4585
96.065
97000 96.5
96
96509
96000 95.5
0 100 200 300 400 500 600 0 100 200 300 400 500 600
Altitude(m) Altitude(m)
Fig 6: Barometric Pressure vs Altitude and Density vs Altitude graph
As the payload capacity depends on various components but the varying components are
the drag force component and the buoyancy forces. For the drag force component,
FD=1/2CD𝜌Av2 the main equation is also density (𝜌) dependent. This dependency must
affect the payload capacity.
The data-driven graph shows that the Barometric pressure will gradually decrease with
respect to altitude variation. The higher the aerostat can go the barometric pressure will
be declined. That is also correct from the temperature graph. More the altitude, the less
the temperature. As the density of air mainly depends on air temperature and pressure,
the density can be found. From the calculated data, a relative view from air density vs
Altitude can be found. It shows that the density of air will decrease with respect to
altitude. And, as the altitude increases the density of air would decrease. This varying
density will affect the drag force and also the buoyancy force both of which are the
component for payload estimation. The more the altitude less the buoyancy force and less
the payload capacity.
5. Conclusion
The key point of the work is to find out the payload estimation factors where total weight
and buoyancy force were summed up with a net lifting capacity which will be balanced
with the payload. With the equation and data analysis the novel information from this
paper is that, with the varying altitude, as the altitude increases, the density effect of air
will reduce the payload capacity during the flight over time (increasing altitude). So,
while calculating the payload, one of the components is the varying density during flight.
Though the accuracy of the collected data was modified with references and calibrated
sensors are used for collecting the data, some errors can be found with the data due to
data transfer problems or miss interpretation of encoder signals from IoT devices.
Moreover, other aerodynamic factors can also affect payload estimation which is not
included in the equation. In fact, dependency on aerodynamic stability can be a factor for
payload estimation which can be verified by future work.
Page | 1323
References
1. Coulombe-Pontbriand, P., Simulation and Experimental Validation of Tethered
Spherical. Aerostat Model. McGill University, Montreal-2005
2. . James R. Scoggins, Sphere Behavior and The Measurement Of Wind Profiles. Nasa
Technical Note 1967, George C. Marshdl1 Space Flight Cen Hiuntsuille, Ala., NASA
TN D-3994
3. C . H . K . W Illiamson , R . G Ovardhan, Dynamics and Forcing Of a Tethered Sphere
In A Fluid Flow. Journal of Fluids and Structures. 1997, 11 ,pp 293 – 305
4. M. Negri, D.Mirauda, S. Malavasi , An analysis method of the vortex-induced vibrations
of a tethered sphere. Journal of Meccanica.2020 , 55(3):1-26
5. W. Chenglei, H Tang, S. Yu, Control of vortex-induced vibration using a pair of
synthetic jets: Influence of active lock-on. American Institute of Physics . 2017,
http://dx.doi.org/10.1063/1.4996231
6. G. Göbel, M. Gebhardt, M. Deutscher, W. Metz , Numerical Simulation Of Flow-Induced
Vibration On Gates With Underflow. E-proceedings of the 38th IAHR World
Congress.2019 , doi:10.3850/38WC092019-0739
7. K. Mahmood, N. A. Ismail Tethered aerostat envelope design and applications: A
review. AIP Conference Proceedings 2226, (2020), https://doi.org/10.1063/5.0002358
8. P. Timmerman,J.P. Van , On the rise and fall of a ball with linear or quadratic drag.
American Journal of Physics.1999,https://doi.org/10.1119/1.19320
Page | 1324
PAPER ID: 185
A study on the Measures taken to sustain the flexibility of a

global supply chain during pandemic
Abdullah Al Rahi 1*, Qausar Rahman2, Eshayat Hasan Chowdhury3
1. Supply Production Leader, Decathlon. Uttara Sector 14 road 13 house 36, Dhaka. 2. Data Analyst
Quality and Logistics,Carhartt WIP. 731 Monipur, Mirpur-02, Dhaka-1216. 3. Student, Shahjalal
University of Sceince & Technology Sylhet.
Email: abdullahalrahi018@gmail.com, contact@qausarrahman.com,chowdhuryangon@gmail.com
* Abdullah Al Rahi
Keywords:
Abstract: During the pandemic era, one of the most impacted sectors
● Supply chain was the supply chain. The pandemic has been a prominent
phenomenon and it will take a while to get back to normal life, the
● Pandemic demand for daily needs and necessities never dropped. This study
aims to present how pandemics impacted a global supply chain by
● Flexible
restraining the flexibility of the supply chain. The data were collected
● Global from a reputed Global retail supplier the supplied goods are largely
textile-based finished goods. The study presents what were the main
● Sustainability areas where pandemics hit the most and what steps and modifications
were applied to find a solution to lessen the damage and survive
during the pandemics.The data were collected directly from the
company. And due to some privacy issues and license criteria, the
company name can’t be revealed in the study. SPSS, G-suite and
Microsoft Office was used during the study to analyze and present the
data.
1. Introduction
During the pandemic beginning, it was a common and shocking scenario to see Billion dollar
industries falling down within a very short time. The factors were many but the major difficulty
was the disrupted supply chain. Bangladesh being one of the giants in the global garments
industry, has experienced several difficulties during the Covid-19 pandemic. Since the
worldwide outbreak of COVID-19, the Bangladesh RMG supply chain has become prone to
disruption due to the raw material shortage from China , Buyers financial crisis, Sudden
cancellation of orders, and unprecedented decrease in demand. If the whole chain is looked at a
glance, it is visible that during the pandemic the priorities for clothes and textile based goods
Page | 1325
have been less for the mass and thus the global textile brands had to repilot their business,
decrease the orders and in some cases they had to close their operations in different countries
due to bankruptcy.
The RMG industry contributed USD34.13 billion to Bangladesh’s export earnings in 2018–19,
representing 84.21% of the total exports and demonstrating the sector’s importance for the
country’s economic survivability.[1] However, due to the current COVID-19 pandemic, export
growth declined by 18.12% in 2019–20 compared to the previous year, and further decreased by
1.2% in the first four months (July–October) in 2020–21 compared to the same period in 2019–
20 (BGMEA 2020a).[2] Different Buyers order cancellation has increased since the pandemic
started. By September 2020, USD 3.18 billion in exports were canceled, which affected the
livelihood of 2.28 million workers (BGMEA 2020b).[3] In this study, considering the impact of
this global context, one global chain retailer has radically changed their supply chain and
management strategy to avoid severe financial loss and at the same time it was able to uphold
the customer satisfaction. [4]
This study is based on the year long data from a Top Global Retailer. The global retailer work
methodology and different terms are well explained in the next chapters. The next section
presents the backdrop of the study, design process of the procedures that were taken for survival
during the pandemic, data collection and analysis. Section 3 presents the data calculation, and
Section 4 discusses the results of the study. Finally, Section 5 summarizes the study with the
limitations that were faced and future aspects of this study
2. Methodology
For the purpose of this study, real life data were taken from 5 RMG factories which have a
contribution for 200 million USD business in the global retail supplier. The global retail supplier
has 10 years of business relationship in Bangladesh. And had steady growth from 2010 until
2020. The sudden outbreak of COVID-19 completely changed the business trend within a very
short period of time. To tackle this unwanted scenario, the Global retailer took several
innovative decisions.
Several terms are very specific, in order for better understanding these terms will be explained.
And an overall data analysis will be presented to justify the action that was taken.
The Global retail supplier works based on the below flowchart, here the flowchart shows the
order cycle from the need being put in the system and the products being delivered to the
customers
- The lines here shows the physical flow
- The dotted lines show the information flow
The sales data that are extracted from the day to day purchase of different products through
shops and e-commerce websites are integrated with AI and previous years sales data trends.
And finally a Retail supplier concern analyzes the data and puts the need through orders in
different RMG factories. For complete understanding some of the terms are explained below
CHD is the short form of contractual handover date, it is the agreed date between the supplier
and the Global retailer, this date indicates the date of handing over the goods from bangladesh.
The finished goods are dispatched from CTG sea port to Europe. The main transport mode is via
Ship. Here another term is introduced and that is called CDD. CDD implicates the Contractual
delivery date. The CDD is the date when the orders reach the continental warehouse. From the
continental warehouse the orders are distributed to regional warehouses. This whole flow
Page | 1326
initiates from the analyzed data that are collected from real time purchase data from the end
customers. The AI integrates the real time data with previous years sales history.
Figure 2.1- Supply chain basic information and Physical goods flow
2.1 Data Collection
The data were collected from the main database of the global retail supplier. SAP is used to
collect and communicate through different mediums. The finished goods are mostly textile
based, the data that are collected are from March 2020 to March 2021. This year's long data
analysis helps to show the effectiveness of the steps that were taken to sustain the supply chain.
Initially these data were taken for 100 models, here the model defines different styles of finished
goods. The textile finished goods are T-shirts, Jackets and leggings. For privacy purposes details
of the finished goods are not discussed here.
2.2 Working Procedure

The Global retail supplier ensures the ontime handover of finished goods by giving earlier raw
material commitments to the RMG factory. By this way the RMG factories can ensure the
different raw materials, such as: fabric, care label, threads, RFID tags etc. Especially the raw
materials which are produced and sourced from China. In this raw materials are bought in
advance to ensure the smooth flow of the supply chain. However, the outbreak of the pandemic
created a massive shift in customers' needs as people were in a situation where the need of
Page | 1327
medicines and basic needs were more than ever. Thus the already placed orders were in the face
of cancellation and at the same time the materials were in different stages and different
warehouses. Below chart shares a clarity on the situation, the chart denotes the flowchart while
indicating the two path, one, if the firm order is canceled, another if the order is not canceled.
Figure 2.2: The flowchart shows the cascading effect of demand decrease
2.3 Data Collection and Analysis

The database related to orders and different RMG factory data, all are stored in the SAP. From
SAP data were collected and analyzed via the microsoft office. The data that were taken are the
primary and raw data that are shared between the retail and factories.
Here the data collections and analysis were different for the MTO (make to order) and
MTS(make to stock models). The MTO models(finished goods) are the models that have a
regular demand and are not selected as crucial business for the company whereas the MTS
Page | 1328
models are the models that are crucial for the company and at the same time they bring a lot
more revenue. [5]
For 80 MTO models the data analysis are given below, here the models are divided by product
type, the model description could not be shared for company security purpose,
Figure 2.3: Figure shows the decreasing demand for 80 models(MTO)
Figure 2.4: The QTY differences as per product type(MTO)

If these 2 graphs are studied it clearly shows how different the Demand decreased over time for
the MTO models it can be seen in the figure 2.3 some models started to do great business
Page | 1329
beyond the committed QTY; however it got drastically decreased after pandemic shutdown, the
scenario is same for one MTO models with a business over 330K Pcs.
Figure 2.5: The QTY differences as per product type(MTS)

The idea of make to stock is to produce the goods before the firm orders are placed in the
system, the MTS model is a trouser and there is a size wise QTY that is planned to produce
every week.
3. Calculation
For the MTO models the fabrics are not fully bought in advance, mostly they are bought 50%.
Dyed fabric is bought and shipped 50% and rest 50% is taken as per demand, in this way the
Global retail supplier doesn't have to take the full liability of the fabric. If the fabric need is
calculated product type wise and import/local fabric supplier wise-
For Import fabric-
Product type SUM of Fabric bought in KG SUM of Fabric bought in advanced 50%(KG)
Leggings 37902.955 18951.4775

Sweat-shirt 117904.758 58952.379
T-Shirt 155368.432 77684.216
Trousers 234883.232 117441.616
Grand Total 546059.377 273029.6885
For Local Fabric-
Page | 1330
Product type SUM of Fabric bought in KG SUM of Fabric bought in advanced 50%(KG)
T-shirt 90726.128 45363.064

Grand Total 90726.128 45363.064
For MTS models as the production is done based on the forecast not fully on the firm order the
fabric has to be bought in full QTY, that is why the risk prevails the most for this MTS.
But in this case , the production planning has been dynamically shifted to meet the decreased
demand scenario, the production planning week wise for week 01 to week 32 is given below-
Figure 3.1: The chart shows the production was decreased and then stopped and finally
increased to meet the fluctuating demand

The global retail supplier works on real time forecasts and AI data management. To tackle such
unprecedented decrease in demand, below drastic steps were taken. The main raw material is
fabric. Fabric suppliers are of two types local and Import. Import fabric supplier is China. 17%
of the total fabric is from Bangladesh. For the raw material production the main instructions
during the pandemic were
- Stop production immediately as communicated
- Hold the fabric in the warehouse Retail supplier will pay the warehouse holding cost
- For the most important fabric several machine should be running in case of sudden
demand rise
- for local fabric as the lead time is short and no risk of high stock as fabric are not
produced in advanced they are produced in JIT(just in time), most of the productions
were in hold
Page | 1331
- For the MTS model as the fabrics are already inhoued, the production planning had to
be shifted to meet the decreased demand.
- And also the business forecast was re-piloted so that the raw materials and future
plannings are minimized.
The main strong point that played a pivotal role is not to depend totally on firm orders. Unlike
other big buyers, this global retail supplier works on forecast and firm orders mixed. That is
why even when the pandemic shutdown occurred it could survive as not all the QTY were tied
to orders. The full study establishes how the demand dropped and also the company strategies to
cope with the situation.
It is inevitable that some firm orders had to be canceled from the retail supplier side, but the
percentage is 22% of the total order, also for this order the retail supplier paid the RMG
factories full liabilities. Thus the global retail supplier could retain their business and survive
during the disrupted supply chain period.
5. Conclusion
The study could have been expanded with more models and more months data, however due to
security and privacy issues more data could not be collected. next further studies can be made
with the dependency of forecast data vs the realized data. And during the pandemic outbreak the
detailed analysis of inventory management can also be shown.
References
1. BGMEA (2020), The Parel Story: January-March 2020, Pathways from Promise.
Bangladesh Garments Manufacturers and Exporters Association, available at:
https://www.bgmea.com.bd/media/newsletters/App_Story_Jan_March_2020/App_Story
_Jan_March_2020.pdf (accessed 24 April 2020).
2. Ali, M., Rahman, S.M. and Frederico, G.F. (2021), "Capability components of supply
chain resilience for readymade garments (RMG) sector in Bangladesh during COVID-
19", Modern Supply Chain Research and Applications, Vol. 3 No. 2, pp. 127-144.
https://doi.org/10.1108/MSCRA-06-2020-0015
3. Ahmed, N. (2009), “Sustaining ready-made garment exports from Bangladesh”, Journal

of Contemporary Asia, Vol. 39 No. 4, pp. 597-618, doi: 10.1080/00472330903076891.
4. Ashraf, H. and Prentice, R. (2019), “Beyond factory safety: labor unions, militant
protest, and the accelerated ambitions of Bangladesh’s export garment industry”,
Dialectical Anthropology, Vol. 43 No. 1, pp. 93-107, doi: 10.1007/s10624-018-9539-0.
5. Halawa, F., Lee, I.G., Shen, W., Khan, M.E. and Nagarur, N., 2017. The Implementation
of Hybrid MTS\MTO as a Promoter to Lean-Agile: A Simulation Case Study for Miba
Sinter Slovakia. In IIE Annual Conference. Proceedings (pp. 1006-1011). Institute of
Industrial and Systems Engineers (IISE).
Page | 1332
PAPER ID: 204
Analysis of Factors that Diminishing the Productivity in Jute

Sector, Bangladesh
Md. Sumon Rahman*, Rakesh Roy, Shoyed Muhammad, Md. Mahafuj Anam
Murad, Md. Masum Billah, Md. Sheikh Sadi, Al Imran Tashref
Department of Industrial and Production Engineering, Jashore University of Science and
Technology, Jashore-7408, Bangladesh.
Email: sumon.just16@gmail.com; rakeshroy996@gmail.com; sdiipgazi@gmail.com;

mahafujmurad.info@gmail.com; md.masumbillah1675@gmail.com, sadi.just@gmail.com,
tashref.ipe@gmail.com
*Corresponding author: sumon.just16@gmail.com
Abstract: The jute production sector is one of the oldest traditional

Keywords: production sectors of Bangladesh. Once upon a time jute was called the
Golden fiber of Bangladesh. In 2015, this sector earned 4.5 percent of the
● Jute sectors;
country’s total export, at present this sector accounts for only 3.9 percent of
● Less Productivity; the country’s total export. This sector is losing its productivity day by day.
There are several factors behind this less productivity. The objective of this
● Factors:
study was to investigate the causes of less productivity in the jute sector. A
● Multiple Linear survey was conducted on 250 jute mills workers from three jute industries at
Khulna, Bangladesh. A structured questionnaire was used to collect the data.
Regression
Multiple linear regressions were used to find out the relation between the
dependent and independent variables. The statistical package for social
science (SPSS) version 25 was used to analyze the data. The results showed
that environmental factors (p=0.04), machine factors (p=0.002) and
administrative or management factors (p=0.003) are statistically significant
on less productivity. On the other hand, personal factor (p=0.854) and
method (p=0.573) both are statistically insignificant on less productivity.
Finally, the authors made several recommendations to minimize less
productivity by including all people from workers to the top management of
the industries.
1. Introduction
Jute industry in Bangladesh is a prominent sector to earn foreign currency (Sadekin, et al.,
2015). In 2015, this sector earned 4.5 percent of the country’s total export (Uddin, et al., 2014).
At present this sector currently accounts for a more 3.9 percent of the country’s total export, this
sector loses their productivity day by day. Many factors can affect the productivity level of the
jute factories in Bangladesh. However, the factors causing less productivity in the jute factories
in Bangladesh are often overlooked by the management and the stakeholders of the vary
industry. Even though the jute sector contributes to the national, GDP and the country's total
export a lot. The machineries used in production are mainly overhauled old machineries from
the colonial era and the management is not an exception of that, usually traditional and not
improvement oriented. In the eve of fourth industrial revolution, productivity obviously is the
major concern in any industry. Several studies have been carried out over time regarding the
Page | 1333
jute sectors of Bangladesh. Uddin et al., 2014, studied the current scenario of Bangladeshi
jute sector from the economic and industrial viewpoint. Islam and Ali, 2017, figured out the
challenge in overall jute production system from cultivation to marketing. Their identified
challenges were inadequate government policy, lack of jute production, sudden factory
shutdown, improper cultivation system, lack of management skill, falling demand in global
market and so on. Whereas Rahman et al. assessed the future prospects of Bangladeshi jute
sector by observing its growth performance, profitability and performance of production
through a time series data over a period 1973-2013 (Rahman, et al., 2017).
Improving productivity is obviously the most important concern to any business organization to
convert its raw material to the consumable finished goods in the most effective and efficient
fashion(Wilcox, et al., 200). Mcguire and Mclaren concluded that working environment has
significant impact on the employees’ job satisfaction, well-being, integrated working approach
and innovation (McGuire and McLaren, 2007 ) Mathews and Khann 2016, investigated impact
of the workplace environment on the productivity. Machuca et al., 2020, discuss influence of
personal and human factor on the productivity and their associative prospects. Attar et al., 2012,
also outlined the relationships of equipment factors and method of working with the
productivity. Management skills and techniques also have considerable influence on the
productivity (Rodgers, and Hunter,1991). Researchers (Adebowale and Agumba, 2021, Hasan,
et al., 2018) also identified the influential factors for construction labor productivity.
As from the stated literature, factors that affect the productivity were discussed in several
sectors including construction industry, sugar industry and corporate office. Therefore, the
objective of this study was to investigate the causes of behind the less productivity in the jute
sector of Bangladesh.
2. Methodology
2.1 Data Collection
This study was conducted among the workers of 250 jute mills in the Khulna region of
Bangladesh. In order to collect the data, authors made a structural questionnaire form. The
questionnaire had two parts. First part included demographic data such as gender, age,
education level, marital status smoking consumption history. Second part contained different
independent variables (environmental, personal/human, machine, administration, and
methodological factors) to evaluate their impact upon less productivity. All questions were two
categories such as five point liker scale and open-end (yes/no). The author collected the
productivity data from the authorities of the jute industries.
2.2 Data analysis
The statistical data analysis consist of analysis of the participants characteristics including
working hour, proportion of self-reporting of subjective that were classified by genders and
analyzed using descriptive statistics including tabulated frequency, percentage, mean,
minimum, and maximum values. The multiple linear regressions was used to purpose of this
measure the relative impact on independent variable and dependent variable (less productivity).
The SPSS (statistical package for the social science) version 25 was used to analysis data.
The following regression model was used in analysis the data.
Y(Less Productivity in Jute sector) = b0+b1x1+b2x2+b3x3+b4x4+b5x5+e, Where, b0 = Constant,X1=
Environment Factors, X2= Personal Factors, X3= Machine Factors, X4= Administrative Factors
X5= Method, e = Error term.
Page | 1334
The relative significance of each of the independent variable on the dependent variable can be
measured from the associated coefficient. The authors also tested the hypothesis based on the
significance level below 0.05 where H0 is the null hypothesis and H1 is the alternative
hypothesis, the hypothesis breakdown is given below:
H0 : H1, H2, H3, H4 and H5 do not have any impact on less productivity in the jute sector
H1 : Environment, H2 : Personal factors, H3 : Machine factors, H4 : Administrative factors and
H5 : Method have a strong impact on less productivity in the jute sector.
3. Results and Discussion

3.1 General information of workers
Table-1: General information of Workers (N=250)
A comprehensive summary of general information of workers is shown in Table 1. Two hundred

fifty jute mill workers were interviewed during the study; 213 (85.2%) were men, while only 37
(14.8%) women of the participants. The number of women employed in Bangladesh's jute mills
is significantly lower than that of males since the workplace conditions are unfavorable for
women, and several barriers are preventing them from performing strenuous physical Labour.
34.4% of the workers in this research who participated in the survey were uneducated,
compared to 44% had only completed elementary school and only around one-fifth who had
completed secondary school. Although survey respondents ranged in age from 25 to 57, and
there were no workers under the age of 20 or over 60. Nevertheless, an earnest effort was made
throughout the survey to maintain the same number of employees across all skill levels.
3.2 Influential factor analysis

Regarding environmental factors, more than 60% of the participants strongly recognized that
being too hot hinders their performance, and more than 38% believe that being too hot at work
negatively affects productivity as presented at Figure 1. Similar patterns may be seen in survey
Page | 1335
results, with 64% of participants strongly agreeing that noise had a negative impact on their
performance and 32% merely agreeing when there was no disagreement. More than 97% of
participants agreed, with 72% strongly agreeing, that dust has an impact on productivity.
Participants were unconcerned with light and ventilation systems because they disagreed with
this claim or remained neutral about its impact on their productivity.
Figure 1: Environment factors associated with participant respondent
80% of participants said they were indifferent, and more than 19% disagreed, but none said that
a lack of ventilation systems hurt the productivity level. Similarly, 41.2 % of participants
rejected, and more than 40 % strongly disagreed with the effect of unregulated and suboptimal
light on productivity.
Figure 2 shows the response to the survey's question on Personal Factors. The majority of
participants, more than 51.2% agreed, and 25.2% strongly agreed that only a poor working
position reduces productivity.
Figure 2: Personal factors associated with participant respondent
Page | 1336
Concerning the remaining three components of the Personal Factors dimension, the participants
expressed either no opinion or a considerable disagreement. 36.8% of participants strongly
disagreed with the notion that a lack of skills reduces participants' productivity, whereas the
majorities were neutral on the subject. In a similar vein, over 50% of participants remained
mute when asked about the detrimental effects of high workplace stress on productivity, while
40% just refuted these effects. Unexpectedly, 63.6% of participants disagreed, and 28.8%t
strongly disagreed that a lack of PPE may have an impact on productivity.
Figure 3: Administrative factors associated with participant respondent

Figure 3 illustrates how survey respondents viewed all administrative component sub factors as
obstacles to high production, with the exception of training and facilities. Over one-third of
participants agreed that ineffective management, a lack of employee-employer cooperation, and
inadequate financial and non-financial benefits are obstacles to productivity, while
approximately one-fourth of participants had no opinion on each of these situations. However,
just 17.2 % strongly disagreed, and 62.3% disagreed that a lack of facilities and training
hampers production.
Figure 4: Machine factors associated with participant respondent
Page | 1337
According to figure 4, 99.60% of participants said old machines harmed their productivity,
while 0% had unfavorable views. 57.60 % of the participants said the breakdown machine
played a negative role in their performance level, while none disagreed. 62% of respondents
said that machine vibration has an impact on their productivity. Excessive Force as an
influencer of low productivity declined by above 42% of the respondents, while 38% were
neutral.
Figure 5: Method factors associated with participant respondent

It is pretty conspicuous in figure 5, that outdated systems significantly hinder productivity, as
90% of respondents strongly agreed, whereas Bad Working Layout was overlooked by the
participants as a factor for less productivity because 35.6% just refused the idea and 18.8%
firmly denied.
3.3 Regression Analysis
The regression model Summary of this study is shown in Table 2. The independent factors
(organizational culture, leadership, and communication) produced R-values of 0.230, showing a
perfect link between the dependent variable and independent variables.. An R2 score of 0.63
indicates that the independent variables (Environmental Factors, Personal Factors, Machine
Factors, Administrative Factors and Method) can explain the associated variance in Less
Productivity, showing that the model adequately fitted the study data.
Table 2: Model Summary
Model R R2 Adjusted R Std. Error of the Durbin-

Square Estimate Watson
1 0.230* 0.63 0.34 1.602 0.134
Page | 1338
3.3.1 Analysis of Variance (ANOVA)
Table 3: Analysis of Variance (ANOVA)
Model Sum of Squares Df Mean Square F Significance

(p-value)
Regression 35.0505 5 7.010 2.730 0.020*
Residual 626.550 244 2.568
Total 661.600 249

1
P*<0.05
The total variance (661.600) in Table 3 can be explained by the independent variable (model),
the variance, and the F-test (F= 2.730, P=0.020<0.05), which indicates that the test is
statistically significant.
3.3.2 Regression Coefficients of determination
Table 4: Coefficients
Model Unstandardized Standardized t Significance

Coefficients Coefficients (P-value)
B Std.Error Beta
1 Constant - 2.729 -162 0.17
4.419
Environment 188 440 0.227 3.438 0.04*
Factor
Personal Factor 057 311 0.012 0.184 0.854
Machine Factor 592 422 0.189 4.540 0.002*
Administrative 837 282 0.190 2.970 0.003*
Factor
Method 154 273 0.036 0.564 0.573
Table 4 shows that, the impacts of the constant components would be 4.419. The administrative
factors (p=0.003), machine factors (p=0.002), and environmental factors (p=0.04) were
statically significant. Whereas, Personal factors (p=.854) and method (p=.573) were
insignificant. The relative coefficient of independent variables conveys how significant it is to
contribute less productively to the jute industries. Less productivity in jute sector (Y)
=4.419+188*Environmental factor+057*personal factor+592*machine
factor+837*administrative factor+154*method
3.4 Testing of Hypothesis

Here, the results indicated that Environment factor (p=0.04), Personal factor (p=0.854),
Machine factor (p=0.002), Administrative factor (p=0.003), Method (p=0.573) impact on less
productivity in the jute sectors of Bangladesh. From these results it can be validate the
following hypothesis: H1: Environment factors, H2: Machine factors and H3: Administrative
Page | 1339
factors have a strong impact on less productivity in the Jute sector. Others hand, H 4: Human
factors and H5: Method factors have no strong impact on less productivity in the Jute sector.
4. Conclusion
From the above study the authors have arrived conclusion that environmental factor (p=0.04),
Machine factor (p=0.002), Administrative or management factor (p=0.003) are statistically
significant on less productivity. The reject value showed that personal factor (p=0.854), method
(p=0.573) both are statistically insignificant on less productivity. Therefore, at the end of this
study, the author has suggested some recommendation for both management and workers to
eliminate or minimize the jute industries workers such as: provide adequate tanning and
education, safety aide facilities, Personal Protective Equipment (PPE), regular monitoring etc..
References
Adebowale, O.J. and Agumba, J.N., 2021. A meta-analysis of factors affecting labour
productivity of construction SMEs in developing countries. Journal of Engineering,
Design and Technology.
Attar, A.A., Gupta, A.K. and Desai, D.B., 2012. A study of various factors affecting labour
productivity and methods to improve it. IOSR Journal of Mechanical and Civil
Engineering (IOSR-JMCE), 1(3), pp.11-14.
Hasan, A., Baroudi, B., Elmualim, A. and Rameezdeen, R., 2018. Factors affecting construction
productivity: a 30 year systematic review. Engineering, Construction and Architectural
Management.
Islam, M. M., & Ali, M. S., 2017. Economic importance of jute in Bangladesh: production,
research achievements and diversification. International Journal of Economic Theory
and Application, 4(6), 45-57.
Machuca-Villegas, L., Gasca-Hurtado, G.P., Restrepo Tamayo, L.M. and Morillo Puente, S.,
2020, September. Social and human factor classification of influence in productivity in
software development teams. In European Conference on Software Process
Improvement Springer, Cham, (pp. 717-729).
Mathews, C., and Khann, I. K., 2016. Impact of Work Environment on Performance of
Employees in Manufacturing Sector in India: Literature Review. International journal
of science and research (IJSR), 5(4), 852-855.
McGuire, D. and McLaren, L., 2007.The Impact of Physical Environment on Employee
Commitment in Call Centres: The Mediating Role of Employee Well-Being. Team
Performance Management, Vol. 14, No. 5/6.
Rahman, S., Kazal, M. M. H., Begum, I. A., & Alam, M. J., 2017. Exploring the future potential
of jute in Bangladesh. Agriculture, 7(12), 96.
Rodgers, R., and Hunter, J. E. (1991). Impact of management by objectives on organizational
productivity. Journal of Applied Psychology, 76(2), 322.
Sadekin, M., Muzib, M., & Al Abbasi, A. A., 2015. Contemporary Situation of FDI and its
Determinants: Bangladesh Scenario. American Journal of Trade and Policy, 2(2), 121-
124.
Uddin, M., Hossain, J. J., & Hoque, M. A., 2014. Present conditions of jute sector in
Bangladesh. Banglavision Research Journal, 14(1), 68-79.
Wilcox S, Stringfellow B, Harris R, Martin B., 2000. Management and productivity.
Washington (DC): Transportation Research Board, Committee on Management and
Productivity.
Page | 1340
PAPER ID: 206
Experimental Study on Aerodynamic Characteristics of NACA

4412 Aerofoil with Three Different Planforms
Mohummad Shariful Islam Mohammad Ali Md. Quamrul Islam

Ex Associate Professor, Mechanical, Professor, Mechanical, Professor, Mechanical
MIST, Bangladesh BUET, Bangladesh MIST, Bangladesh
Email: sharif6016@gmail.com Email:mali@me.buet.bd Email:quamrul@me.mist.ac.bd
Abstract
The aim of this research paper is to explore aerodynamic characteristics from three planforms (wooden-
plastic) models. In this research, NACA 4412 aerofoil planforms are used to analyze the performance of
lift and drag properties. Models are prepared to keep the overall surface area alike with changing the
aspect ratio (AR) of wings. All the models are tested at an airspeed of 79.20 kph (0.06 Mach) i.e. at
Reynolds Number 1.74 x 105 in the closed circuit wind tunnel. Lift and drag coefficient were calculated
and obtained results were plotted against the angle of attack(AOA). It has been observed that the critical
AOA of all the three planforms remain around 16˚ beyond which the stall occurs. So, to obtain maximum
lift from NACA 4412 aerofoil, the wing needs to be positioned at around 16 degrees with respect to the
flight path whereas slanted trailing edge planforms of NACA 4412 shows the better result. Therefore, this
type of model may be considered to design a wing of aircraft.
Keywords: NACA 4412 aerofoil, wing planform, lift and drag coefficient, angle of attack, stall.
1. Introduction
The wing is a very essential part for an aircraft to generate lift force to make the flight possible
in the air. This force is generated by a special wing cross-section which is called aerofoil and the
silhouette of the wing when viewed from above or below is known as wing planform. The
effects of wing size are crucial to aerodynamic characteristics on which the efficiency and
performance of aircraft depend. The geometry of the wing can be varied span wise to search for
better performance.
When the wing is generating lift, it has reduced pressure on the upper surface and increased
pressure on the lower surface. If there is a wing of infinite aspect ratio(span/ average chord), the
air flows over the wing surface without any inward or outward deflection, and therefore no
wing-tip vortices, no induced drag. But such a thing is impossible in practical flight, and for this
reason, the effective aspect ratio is increased as large as is practicable by changing cord and
length. The extension of wingspan permits the lowering of drag but this comes at a cost of
increasing the strength of the wing and hence its weight. So the researchers in the world are
always trying to search the max possible lift with minimum possible drag on different
wing/geometries to overcome the problem. In this backdrop, this research made an endeavor to
find out the improved aerodynamic characteristics from three planforms (NACA 4412 aerofoil)
of the wing with a view to reducing the drag for efficient flying.
In aerodynamics, the main source of the airplane drag is related with the wing [1]. Around two-
thirds of the total drags of typical transport aircraft at cruise, conditions are produced by the
wing [2]. Hossain et al. [3] conducted an experimental analysis for the aerodynamic
characteristics of a rectangular wing with and without bird feather like winglets for different
Reynolds Number which shows 25~30% reduction in drag coefficient and 10~20% increase in
lift coefficient by using bird feather like winglet at 8° angle of attack(AOA). Ghods et al. [4]
conducted an experimental analysis of NACA 2415 wing in wind tunnel where he shows lift
increases as the angle of attack increases between -5 and +17 degrees and at +17 degrees
maximum lift is generated. Kandwal et al. [5] study is done on airflow over a two-dimensional
Page | 1341
NACA 4412 Airfoil using ANSYS FLUENT. The CFD simulation results show close
agreement with those of the experiments, thus suggesting a reliable alternative to the
experimental method in determining drag and lift.
Mashud [6] introduced a flow separation on NACA 4215 airfoil profile and found that the lift
increased about 14% and drag reduced about 23% at 8˚AOA. Nazmul [7] studied with three
different planforms like the wooden model of NACA 4412 aerofoil in equal length (span) and
surface area. It is found that the critical AOA of three planforms remains around 16° beyond
which stall occurs. National Aerofoil Data for NACA 4412 [8] aerofoil shows the critical AOA
for rectangular planform around 16° AOA.
A typical aerofoil section is shown in Figure - 2, where several geometric parameters are
illustrated [9, 10] and the few definitions are given below:
Figure-1: Geometric Features of an Aerofoil [9,10].

The straight line connecting the leading and trailing edges is the chord line and the precise
distance from the leading to the trailing edge is called the chord. A line on a cross section of a
wing of an aircraft that is equidistant from the upper and lower surfaces of the wing. The
camber is the maximum distance between the mean camber line and the chord line. The angle
between the chord line and the direction of air flow is called the angle of attack (AOA). An
aerodynamic condition in which the angle of attack increases beyond a certain point, and lift
begins to decrease.
The formulas used in this experiment are mentioned below:
a. The 32 pressure taps provide pressure values determined from the manometer as:
b. The pressure coefficient (Cp) is a dimensionless number which describes the

relative pressures throughout a flow field in fluid dynamics:
Where ½ρU∞² is the free stream dynamic pressure head.
c. When n is a number of panels, the equations can be converted to:
Page | 1342
Figure - 2: Resultant Aerodynamic Force and its Components [7,11].
d. Lift and drag coefficient can be obtained from:
e. By using following formulas lift and drag force are calculated:
Where, L and D are the lift force and drag force respectively, free stream dynamic
pressure as q∞ =½ρU∞², s be a reference area and ρ is the density of air.
2. Test Models
Three wooden -plastic models of wings (using NACA 4412 aerofoil) are prepared which
having the equal surface area (31115 mm2) as shown in Figure-4. Each model is provided with
32 pressure tapings along the span and chord (16 at the upper surface and 16 at the lower
surface). Along the span, the wings are divided into four equal segments (A, B, C and D) as per
the figure. Four pressure tapping points at the upper surface and four pressure tapping points at
the lower surface are made at 20%, 40%, 60% and 80% of the average chord length of each
segment of all the wing models. Dimensions are mentioned in the figure-4 of experimental wing
models:
Model-1: Rectangular Planform (Reference). Model-2: Slanted Leading Edge

Planform.
Model-3: Slanted Trailing Edge Planform.
Figure-3: Experimental Wing Models.
Page | 1343
3. Experimental Setup
All the experimental data are taken at room temperature of 29˚C and at an air speed of 22 m/s
(79.20 kph) and the air flow is considered incompressible throughout the experiment. The
Reynold number and Mach number are 1.74 x 105 and 0.06 respectively. The specific density of
both air and water corresponding to room temperature is assumed to be 1.15 kg/m3 and 996
kg/m3 respectively.
Figure-4: Photograph of Experimental Set-up.
The pressure coefficients of both upper and lower surfaces of different planforms were
measured through the wind tunnel testing to analyze aerodynamic characteristics. The pressure
distributions of both upper and lower surfaces along the chord length of four segments
(Segment- A, B, C and D) of five experimental wing planforms for -40, 00, 40, 80, 120, 160, 200
and 240 AOA are plotted and analyzed. It is observed that the slanted-edge wing ( Model- 2
and 3) planforms can produce more lift due to the increased surface area near the root of the
wings.
After analyzing, it is found (as per figure-6) that the lift increases with an increase in AOA to a
maximum value and thereby decreases with further increase in AOA. The lift coefficient curve
rises from -4˚ to 16˚ AOA for all the planforms and then falls rapidly beyond 16˚ AOA. The
critical AOA of all the three planforms remains around 16˚ beyond which the stall occurs.
Slanted trailing edge models are having better lift coefficient.
Figure- 5: Variation of Lift Coefficient with AOA .Figure-6: Variation of Drag Coefficient with AOA.
Page | 1344
As per figure-7, the drag coefficient increases slowly with an increase in AOA to a certain value
and then it increases rapidly with further increase in AOA. It is observed that the values of the
drag coefficient for Model planform is much lower than that of the Model-1 & 2 planforms. It is
seen that after the peak occurs at 16° and the coefficient of drag starts increasing at an
exponential rate. This exponential rate of increase demonstrates that if the AOA is increased any
further the drag will dominate lift and stall will occur.
The values of lift to drag ratio are plotted for various AOA in figure-8. It is found that Model-3
is having higher L/D than Model-1 & 2 planforms for all AOA. Drag is the lowest for the
slanted trailing edge (Model-3) planforms among the three experimental wings. Here it is
mentioned that those statistics show a similar nature to Ghods analysis [4], Nazmul analysis [7]
and National Aerofoil Data NACA 4412 [8].
Figure-7: Variation of Lift to Drag Ratio with AOA.
5. Conclusions
The total surface area of the wings remains the same. For a given wing area, a wing of high
aspect ratio (Span/Average Chord) will produce less induced drag than a wing of low aspect
ratio because there is less air disturbance at the tip of a longer and thinner wing. By changing
the airfoil section near the wingtips allows more lift to be generated nearer the wing root and
less towards the wingtip. In this research, the slanted-edge wing planforms can produce more
lift due to the increased surface area towards the root of the wings. The critical AOA for all
wing planforms remains around 16˚. Stalling occurs after 16˚ AOA for all three wing planforms.
The slanted trailing edge (Model-2 & 3) planforms exhibit the best result like the lift coefficient
is better, the drag coefficient is lower and L/D is better. As a final comment, the slanted trailing
edge planform of NACA 4412 model may be considered by the aircraft designers and engineers
to design a wing model with a view to reducing the drag for efficient flying.
Acknowledgment
The authors gratefully acknowledge the Faculty of Mechanical Engineering, BUET, Dhaka due
to allowing to use wind tunnel for this experimental investigation.
Page | 1345
References
[1] Anderson, J.D., “Fundamentals of Aerodynamics”, McGraw-Hill Series in Aeronautical

and Aerospace Engineering, 3rd Edition, pp. 15-22, 2010.
[2] Lynch, F.T., “Commercial Transports- Aerodynamic Design for Cruise Performance
Efficiency.” Progress in Astronautics and Aeronautics, Vol. 81, pp. 81-144, 1982.
[3] Hossain, A., Rahman, A., Iqbal, A.K.M.P., Ariffin, M., and Mazian, M., “Drag Analysis
of an Aircraft Wing Model with and without Bird Feather like Winglet”, Int. J. of AE
and Mech Engg, Vol. 6, pp. 8-13, 2012.
[4] Ghods, M., Theory of Wings and Wind Tunnel Testing of a NACA 2415 Airfoil, M.Sc.
Engg. Thesis, Department of Mechanical Engineering, The University of British
Colombia, 2001.
[5] Kandwal, S., and Singh, S., “Computational Fluid Dynamics Study of Fluid Flow and
Aerodynamic Forces on an Airfoil.” International Journal of Engineering and
Technology, Vol. 1, pp. 1-8, 2012.
[6] Mashud, M., and Hossain, M. F., “Experimental Study of Flow Separation Control of an
Airfoil by Suction and Injection”, Proceedings of the 13th Asian Congress of Fluid
Mechanics, Dhaka, Bangladesh, pp. 166-169, 2010.
[7] Nazmul M. and Ali M. “An Experimental Investigation on the Aerodynamic
Characteristics of NACA 4412 Aerofoil with Curved-Edge Planform”, M.Sc. Thesis,
Mechanical Engineering, BUET, Dhaka-1000, Bangladesh, 2015.
[8] National Aerofoil Data NACA 4412, Online data,
http://airfoiltools.com/airfoil/details?airfoil=NACA 4412-il.
[9] Kermode, A.C., “Flight without Formula”, Pitman, 8th edition 1970.
[10] Clancy, L.J., “Aerodynamics”, John Wiley, New York, 1975.
[11] Alam, G. M. J., “Interference Effect and Flow Pattern of Four Biplane Configurations
using NACA 0024 Profile”, Proceedings of The International Conference on
Mechanical Engineering, Dhaka Bangladesh, 2011.
Page | 1346
PAPER ID: 212
Implementation of Cyber-physical Systems in Banking Sector
Tamjid1,*, S., Rashed2,*, C.A.A. and Bagum3, M.N.

1,2,3
Department of Industrial and Production Engineering, Shahjalal University of
Science and Technology, Sylhet, Bangladesh
1
sayemtamjid71@gmail.com
2
rashed-ipe@sust.edu
3
nasima-ipe@sust.edu
Abstract: The banking industry is one of the most critical segments of the
Keywords: global economy. The banking sector must embrace cyber-physical systems to
keep pace with international business. An integrated set of physical and
● Banking Sector; computational components, such as an embedded system, network, software,
● Internet Banking; and physical systems, is called a cyber-physical system. The study aims to
determine the cyber-physical system’s existing state from Bangladesh’s
● ATM;
perspective. Multiple case studies were conducted for this study. 20 public and
● Cyber-physical private banks participated in the survey using a semi-structured questionnaire.
The approach for gathering data was face-to-face interviews. Data were
System;
collected from bank managers, IT industry field specialists, professionals, and
● Cybersecurity. employees. The questionnaire was verified through cooperation with
academics and industry specialists in the banking sector. The data was
examined using Microsoft Excel 2019 software. Based on the results, the
components required to build the cyber-physical system and the technologies
accessible in the banking industry have been identified. The critical
components of a cyber-physical system are cybersecurity, technical
proficiency, accessibility, and interconnection. Banks offer automated teller
machines (ATMs), online banking, remote financial transfers, utility services,
and other technologies. Banks appear to be 16% ready to establish a cyber-
physical system based on the adoption level. The study’s findings suggest that
banks are striving to develop a cyber-physical system but are not yet prepared.
1. Introduction
The banking industry is critical to the global economy. Its principal concerns are protecting
depositors’ investments and issuing loans to people and enterprises. Any modern financial system
promotes economic progress and enhances living standards through technology by offering
numerous services to the economy. Information technologies allow storing, gathering, and
analyzing data, whereas operational technologies permit the creation of tangibly valuable goods
and services. The two combine to create the much desired “cyber-physical systems,” made
feasible by a centralized system that oversees and regulates activity at all levels. In the framework
of the digitized sector, connectivity and interoperability have taken on a significant role in
integrating humans, machines, services, and the environment into a single intelligent financial
system. By combining them with computer technology, the banking industry has reached
Page | 1347
previously unheard-of levels of customer satisfaction. Therefore, banks must embrace cutting-
edge IT solutions such as cyber-physical systems for an efficient and successful banking system.
Cyber-physical systems (CPS) are intelligent systems with collaboratively built interacting
networks of computational and physical components. These highly linked systems provide new
capabilities to enhance the quality of life and support technological advancements in required
fields. Cyber-physical systems may encompass numerous activities, from detecting a small
surgical needle to accomplishing billion-dollar financial transactions. Information and
communication technology (ICT) improvements have altered every aspect of the business. As a
result, both the general public and business professionals today urgently require a robust digital
platform and operational flexibility. The banking industry is also attempting to use rapidly
developing technology to provide online client services while maintaining competitiveness and
customer satisfaction. Therefore, it is necessary to investigate whether the banking industry,
whether public or private, is prepared for cyber-physical system implementation.
In recent years, the banking sector has transformed how it processes information and provides
services to its clients. The government requests that the collaborators develop strategies to create
and adopt new technologies. Financial inclusion is being embraced more widely as online banking
gains popularity. As a result, data security for the banking industry has become increasingly
important, and the bank management must make sure that the risks are correctly detected and
controlled. Banks safeguard customers’ data from illegal access, alteration, exposure, and
extinction. And for any organization to be sustainable, its information assets must be protected
and maintained. An institution like a bank that successfully implements a cyber-physical system
may address these difficulties and start excellent governance, increasing productivity. Therefore,
the purpose of this study is to analyze the adoption level of a cyber-physical system (CPS) in
Bangladesh’s perspective.
An increasing number of operations are being computerized due to the growth of Internet
technology. Information systems are developed for businesses’ integrated and effective
procedures, with corporate networks serving as their primary constituent. According
to (Srivastava, 2016), information and communication technologies (ICT), which are driving
automation, provide several benefits to the industrial and service sectors worldwide. The
components of Industry 4.0, such as the Internet of Things (IoT), cyber-physical systems (CPS),
cloud computing, and virtual reality, have altered production and service delivery throughout their
entire supply chains.
(Afrizal, et al., 2020) stated that CPS is a management style for industry 4.0 that integrates cyber-
physical systems with online human beings. The CPS Industry 4.0 design must take the human
factor into account. Humans play a significant role in the industry of the future. Human resources
must be trained to be flexible in light of the industry’s potential future growth.
(Fatimah, et al., 2020) developed an intelligent waste collection system model that combines real
and virtual environments to calculate, regulate, and interact with all waste management elements,
including equipment, data, sensors, actuators, and networking. Their suggested method takes a
new technique by integrating computing with the physical garbage collecting system through
cyber-physical systems. The garbage is easily identifiable and traceable, the waste collecting
procedure is automated, and the information is accurate.
(Chandran K.R. and Tholath, 2022) highlighted the significant change in the banking paradigm
with the introduction of fintech and mobile banking services. Online banking services increase
Page | 1348
options and competitive pressure in financial services. Therefore, online banking services must
develop more significant moral standards, protective measures, and awareness.
(Pogăciaș and Dovleac, 2021) stated that the technology and competencies of banks to rapidly
combat the new techniques of analyzing ever-increasing volumes of data would be the basis for
all future improvements. The financial structure must be regulated concurrently with the
digitalization process to prevent or reduce risks.
(Broby, 2021) stated that technological advancements would improve banking service delivery
and lower the cost of it while also forcing the future bank to deal with competitive concerns.
Additionally, it emphasizes the significance of the Open Banking trend and how it makes banking
as a service more manageable. (Curran and Meuter, 2005) provided a conceptual model of how
the degree of innovation adoption in the banking industry fluctuates depending on the pertinent
technological aspects.
Human civilization is becoming increasingly dependent on the efficient integration of digital and
physical systems to carry out complex automation and control tasks. Organizations should focus
on innovation to survive in a world with the rapid advancement of technology and scarcity of
resources. Cyber-physical systems’ efficiency, security, and dependability must be engineered, a
field of crucial relevance. Due to a lack of skilled personnel, inadequate technology assistance,
and inadequate infrastructural facilities, the banking industry is experiencing significant
challenges. Therefore, the existing state of cyber-physical systems has to be given more priority
(CPSs). The objective of the current study is to determine the level of CPS adoption and the
technologies currently used in the banking industry.
The research aims to implement cyber-physical systems in the banking sector. Therefore, the main
focus of the study is
 to assess the existing situation regarding the adoption level of cyber-physical systems;
2. Methodology
The study is a multiple case study-based investigation in which the goal was to gather baseline
information on the technologies that certain banks were utilizing or to what extent they had
accepted them. We made an effort to pinpoint the technologies needed to construct cyber-physical
systems effectively. In the beginning, a pilot study was conducted in two banks. A semi-structured
questionnaire was used to collect data. The questionnaire was finalized on the advice of academic
specialists and industry professionals to obtain data. In-person interviews were carried out with
field experts in the IT industry, including bank branch managers who may also be members of
the central branch of those specific banks. The data analysis was performed using Microsoft Excel
2019. The obtained results were evaluated, and the researched institutions’ preparedness for a
cyber-physical system was acknowledged.
3. Demographic Information
For this study, the respondents were interviewed from 20 branches of public and commercial
banks across Dhaka and Sylhet divisions in Bangladesh. The selection of the respondents was
based on how well they addressed challenges connected to the banking system’s use of technology
and innovation. Among the responders, there were diverse age groups and years of experience
represented by 30% of women and 70% of men. As the research is concerned with cyber-physical
systems from a bank’s standpoint, all these bank managers and IT specialists in Bangladesh could
Page | 1349
comprise the sample framework as a population. Of the 20 bank branches (sample size N = 20),
four were state-owned commercial banks (SOCBs), twelve were private commercial banks
(PCBs), two were Islami shariah-based PCBs, one was a foreign commercial bank (FCB), and
one was a specialized bank (SDB).
4. Result and Findings
4.1 Available Technologies in Banking Sector
The respondents were questioned about the banking technologies they now use. The technology
that banks have access to has been determined from the interviews and is discussed below. The
following technologies need to be looked into: automated teller machines (ATMs), debit/credit
(cardless banking), point of sale, loan repayment, insurance, utility (bill pay/tuition fee), online
banking, and remote fund transfer. The state of these technologies in the current banking
environment is shown in Table 1 as either at infancy, developing, or fully operational.
Table 1: Available technologies in studied banks
Response (Percentage)
Technology
At Infancy Evolving Fully Functional
Automated Teller Machines (ATMs) 45% 35% 20%
Smartcard Banking (Debit/ Credit) 25% 50% 25%
Point of Sale 70% 20% 10%
Loan Repayment 90% 10% –
Insurance 85% 15% –
Utility (Bill Pay/Tuition Fee) 85% 10% 5%
Online Banking 15% 60% 25%
Remote Fund Transfer 25% 55% 20%
The most prevalent features, according to Table 1, are internet banking, smartcard banking,
ATMs, and remote financial transfers. These available technologies are a component of the slowly
developing prerequisites for effective banking. These technologies are a part of cyber-physical
systems in overall context, but they would need to be completely operational for a successful
cyber-physical system implementation.
The majority of technologies—55%—are still in their infancy, according to all available
viewpoints on technology. This indicates the use of technology in the context of the present and
places an emphasis on long-term strategy development. However, every bank that was examined
worked to find solutions, touching an average of 32% developing technology and 13% completely
operational operations.
Page | 1350
4.2 Scenario of the Adoption Level for CPS in the Banking Sector
The respondents were asked for feedback on the financial services they now use. The data
gathered has been accessed by reviewing a number of impacting factors from the interviews.
Table 2: Elements required for CPS in banks
Response (Percentage)
Factors
Operative Laggard
Technical Expertise 20% 80%
Infrastructure 15% 85%
Digital Transformation 20% 80%
Accessibility 15% 85%
Ease of Use 10% 90%
Marketing and Promotion 5% 95%
Interconnectedness 15% 85%
Cybersecurity 25% 75%
The following elements should be prioritized for the effective deployment of cyber-physical
systems: i) Technical competence; ii) Infrastructure; iii) Digital transformation; v) Accessibility;
v) Ease of use; vii) Interconnectedness; and viii) Cyber-security. Table 2 shows whether these
factors are operational or trailing in the current banking environment.
An average of 16% of the components have been covered for the deployment of cyber-physical
systems. As a result, it may be inferred that while the banking industry is not yet prepared for
CPS implementation, they are at least beginning to operate.
4.3 The Demand for Self Service Technology (FinTech)
Following the COVID-19 outbreak, it has been seen that customers are using self-service banking
technology more frequently. Additionally, during conversations with banking staff members, the
demand for self-service banking technology became apparent. Because all bank branches are
connected centrally, the core IT team can address issues that could occur at any branch. But the
individual department must communicate with the main office to exchange problems and wait for
a repair. Returning to the previous condition takes time, and customers become upset with delayed
services. Common difficulties include ATM cards trapped in the machines, bad network
connectivity; power outages; password-related issues, etc.
According to the study, 16 banks share the same software platform. Eight banks make use of the
“PC Bank,” “Flora,” and “BEXI Bank” technology. The other four employ EBBS, IBBS, PIBS,
and Kernel banking systems. From the study, it can be concluded that bank branches require
assistance to resolve problems on their own under the direction of the central office. Alternatively,
the divisional head office may be given the authority to do maintenance and address issues, which
Page | 1351
would benefit customers by providing timely service. The hiring of IT specialists at divisional
offices, the introduction of a user-friendly software interface, and the quicker resolution of
customer complaints may all be listed as hurdles.
5. Conclusion
From the study, it can be concluded that–
 The banking sector is 16% ready for the implementation of cyber-physical systems.
 From the technological requirements, banks are 13% fully operational to embrace cyber-
physical systems.
Acknowledgment
There are no acknowledgements.
References
Afrizal, A., Mulyanti, B., Widiaty, I., Development of Cyber-Physical System (CPS)
implementation in industry 4.0, IOP Conference Series: Materials Science and Engineering. 2020,
https://doi.org/10.1088/1757-899X/830/4/042090
Broby, D., Financial technology and the future of banking, Financial Innovation. 2021,
https://doi.org/10.1186/s40854-021-00264-y
Chandran K.R., P., Tholath, D.I., Digital Revolution in Banking Services and Ethical Aspects: A
Review, Journal of Positive School Psychology. 2022, 6(5); 2364 – 2370.
Curran, F.M., Meuter M.L., Self-Service technology adoption: comparing three technologies,
Journal of Services Marketing. 2005, 19(2); 103–113.
Fatimah, Y.A., Widianto, A., Hanafi, M., Cyber-physical System Enabled in Sustainable Waste
Management 4.0: A Smart Waste Collection System for Indonesian Semi-Urban Cities, Elsevier.
2020, https://doi.org/10.1016/j.promfg.2020.02.169
Pogăciaș, C., Dovleac, R., Implementation and impact of Industry 4.0 and Quality 4.0 in the
banking sector, MATEC Web of Conferences. 2021,
https://doi.org/10.1051/matecconf/202134209008
Srivastava, S.K., Industry 4.0, Lucknow: BHU Engineer’s Alumni. 2016.
Page | 1352
PAPER ID: 242
Identification of Causes of Defects and their Solution

Processes in the Sewing Section of Selected Apparel
Industries in Bangladesh
Jahid Hasan1*, Syeda Kumrun Nahar2 and Himel Kishor Barua3

1,2
Assistant Professor, Department of Industrial and Production Engineering, Shahjalal
University of Science & Technology, Sylhet 3114, Bangladesh; Email:
1
j.hasan.ipe@gmail.com, 2aurjoma@yahoo.com
3
Technology, Sylhet 3114, Bangladesh: Email: himel1569@gmail.com
*Corresponding author [1Jahid Hasan]
Keywords: Abstract: Due to the fast changes in the economy, the global business climate has
grown increasingly competitive. Bangladesh has a sizable number of garment
● Quality factories; therefore, quality improvement may be important for boosting
productivity and economic growth. The DMAIC methodology of the Six Sigma
Improvement;
approach, which includes five stages called Define, Measure, Analysis,
● DMAIC; Improvement, and Control, is used in this study to give a detailed analysis of the
quality improvement of chosen garment factories in Bangladesh by minimizing the
● Six sigma;
defects rate in the sewing section. Pareto analysis was conducted to determine the
● Garment; main defects that cause additional defects. The two most common defects in the
chosen apparel companies following data analysis and defect rate computation are
● Sewing
Skip Stich and Broken Stich. A cause-effect diagram was used to find out the root
Section; causes of those major defects. Some potential solutions are suggested to overcome
● Defects. those causes. After the implementation of the solutions in one garment company,
the outcome is very significant. The defects percentage has been reduced from
11.09% to 9.01% and the sigma level has been improved from 2.72 to 2.84.
1 Introduction
RMG is the largest exporting industry in Bangladesh, which experienced phenomenal growth
during the last 30 years. According to the statement on the export of RMG of BKMEA, in the
financial year 2018-2019, 84.21% of Bangladesh’s export earnings came from the garment
industry, the total export of RMG 34133.27 in million US$, and its contribution to the GDP was
8.76% (BGMEA, 2019). Although being the largest exporting, rejected garments after shipment
are very usual in the garment manufacturing industries (M. M. Islam et al., 2013). Rejection
causes waste and decreases resource efficiency. The defect rate has a direct impact on the
product's profit margin and raises the cost of quality throughout production (Kolarik & William,
1999). It also hampers the organization’s reputation (Kumar & Naidu, 2011). This situation shifts
the buyers to other competitive countries like India, Pakistan, Indonesia, Cambodia, Vietnam, etc.
The requirement for rework and order rejection must be reduced if garment manufacturers are to
enhance the quality of their products in today's world of increased demand for lower prices (Bala
Page | 1353
& Gupta, 2009). For minimizing the rejection rate, it is very important to follow a particular
methodology in their quality management system. To achieve the overall objective of minimizing
defects and rejection of finished products, it is needed to establish documents and maintain a
system capable of ensuring that products conform in total to standard specifications (Dean &
Bowen, 1994).
According to a study, the final assessment of garments reveals more than 75% of them have
sewing-related flaws (Kayaalp & Erdoğan, 2008). To maximize the consistency and accuracy of
process outcomes of sewing area, this research focuses on identifying the defects occurring
frequently and their root causes in the sewing section of selected garment factories in Bangladesh
and finding solutions to minimize the percentage of the defects by applying the appropriate
methodology.
2 Research Methodology
For this study, survey base case study research approach was used. The study will be conducted
at a total of five garment manufacturers that are dispersed around Bangladesh. A list of defects
(Garments Merchandising, 2016) is identified after detailed observation of the manufacturing
process. Data on sewing line flaws were collected manually. Finally, a questionnaire was formed
to identify the root causes contributing to serious defects in each company's sewing department.
Six Sigma employs a well-structured continuous improvement methodology to tackle process
variability and drive out waste rigorous manner. DMAIC (Define, Measure, Analyze, Improve,
Control) methodology of Six Sigma resolves issues of defects or failures, deviation from a target,
excess cost or time, and deterioration. (Bañuelas & Antony, 2003). Several tools of DMAIC
methodology such as; SIPOC, Process Sigma Calculation, Pareto Chart, Cause and Effect
Diagram, Five Whys, Brainstorming; were used in this study. After examining the sewing-related
flaws in garment factories, this study not only identifies the main sources of these flaws but also
offers remedies that are finally implemented in one particular garment factory.
3 Calculation
For this study, primary data was collected through manual inspection for different garments. This
one type of data was collected for analysis from five different garment factories.
Table 1: Combined defective and cumulative defective data for selected garment factories.
Types of Skip Broken Uneven Raw-edge Open Joint Missing Off Over Needle Fabric Poor
Others Tension Damages
Defects Stitch Stitch Stich Out Stitch Stitch Stich Placement Stich Cut Fault Shape
Defective Garments 3,488 2,912 1,994 1,737 1,684 1,568 1,239 1,216 994 607 469 367 289 81 20
Cumulative Defective
3,488 6,400 8,394 10,131 11,815 13,383 14,622 15,838 16,832 17,439 17,908 18,275 18,564 18,645 18,665
Garments
A total of 1,57,573 garments were inspected and a total of 18,665 garments were found defective.
Table 1 shows the defects frequently occurring in the sewing section of every garment factory
along with the frequency of defective garments as well as cumulative defective garments. Before
finding the major defects in the sewing section of garment factories, first, it is needed to find out
if all the defects shown in table 1 are independent of each other or not.
Table 2: Defective rate and no-defective rate of occurred defects in the garment factories.
Skip Broken Uneven Raw-edge Open Joint Missing Off Over Needle Fabric Poor
Others Tension Damages
Stitch Stitch Stich Out stitch Stitch Stich Placement Stich Cut Fault Shape
% Defective 2.2 1.8 1.3 1.1 1.1 1 0.8 0.8 0.6 0.4 0.3 0.2 0.2 0.1 0
% Non-defective 97.8 98.2 98.7 98.9 98.9 99 99.2 99.2 99.4 99.6 99.7 99.8 99.8 99.9 100
Pearson Chi-Square χ2 = 11,764.60351 P=0.0000
P = 0.0000 < 0.05 Significant differences exists somewhere among these defects.
Page | 1354
In table 2, there exists a significant difference among the defects i.e., these defects are independent
of each other. Now, the major defects that occurred in the sewing section will be found using the
Pareto analysis in figure 1. The major defects frequently happened in any garment factory are
‘Skip Stitch’ and
‘Broken Stitch’. In
the next section, this
study will find out
the root causes
behind these defects
and remedial actions
of them to improve
the processes.
Figure 1: Pareto Analysis of Defects frequently occurring in garment factories.
4 Analysis and Finding
In every garment factory, there are production lines and each has two types of quality checkpoints.
These are called end line QC table (100% inspection) and sampling inspection table for finished
garments. Now, the DMAIC methodology of Six Sigma is going to be applied to minimize the
defect rate.
4.1 Define phase
The garment factories in Bangladesh selected for this study are suffering from a high rate of
rejections of their products due to defects. Therefore, the goal is to reduce the defect percentage
to minimum level and thereby improve the quality, reduce waste and increase productivity.
SIPOC (Suppliers, Inputs, Processes, Outputs, and Customers), an effective communication tool,
is constructed for ensuring that all members of the improvement team are viewing the process in
the same way in the selected garment factories, shown in table 3.
Table 3: SIPOC Flow of selected garment factories.
Suppliers Inputs Process Outputs Customers
 Zaber & Zubair Fabrics  Unstitched Cloth  Fabric Laying  Cargo Short  VF Asia
LTD.  Machinery  Cutting  Cargo Long  Lee & Wrangler Asia
 Kohinoor Mills LTD.  Thread  Sewing  Girls Denim  Walmart
 YKK Group  Needles  Washing  Denim Long Pant  GAP
 A&E Bangladesh Ltd.  Zipper  Ironing and Finishing  H&M
 R-Pack  Packaging
4.2 Measurement phase

The defect counts and defect rates are shown in the calculation section. In this phase, this study
measures the existing DMPO (Defect per Million Opportunity) and Sigma Level of the selected
garment factories based on the collected data, shown in table 4. As end line QC table (100%
inspection) is conducted after each production line, thus the defect per unit opportunity is one.
Table 4: DMPO and Sigma Level of Existing Process in the selected garment factories.
Garment A Garment B Garment C Garment D Garment E
Total Checked Items 26,071 28,193 16,617 18,465 68,227
No. of Conforming items 23,180 24,495 14,822 16,128 60,283
No. of Non-conforming items 2,891 3,698 1,795 2,337 7,944
% Defects 11.09 13.12 10.80 12.66 11.64
DPO 0.110889 0.131167 0.108022 0.126564 0.116435
DPMO 1,10,889 1,31,167 1,08,022 1,26,564 11,6435
Sigma Level 2.72 2.62 2.74 2.64 2.69
Page | 1355
4.3 Analysis phase
In the calculation section, it is already noticed that ‘Skip Stitch’ and ‘Broken Stitch’ are the major
defects among all the sewing defects that occurred in any garment factory. Five why analysis, one
of the many brainstorming methods of asking “why” five times repeatedly, is used to help in
identifying the root causes of the major defects. Then, a cause-and-effect diagram is used to
organize the identified causes of major defects (Skip Stitch and Broken Stitch) into cause types
of man, machine, method, and material, shown in figure 2.
Figure 2: Cause and effect diagram of major defects (Skip Stitch and Broken Stitch).
Now, a Pareto chart is constructed using the frequency of overall causes (man, machine, method,
and material related, shown in figure 3.
Figure 3: Pareto chart for identifying the top occurring root causes with the frequency of occurrences (mean).
The Pareto chart shows that the first 13 underlying root causes have an 80.77% effect on the
occurrence of the major defects.
Page | 1356
4.4 Improvement phase
After having the Root Causes of major defects from the Cause-Effect Diagram in several garment
factories of Bangladesh, solutions of Root Causes of major defects have been suggested in this
phase. Remedial action procedures for the Root Causes of major defects are given in table 5.
Table 5: Suggested solutions of root causes of major defects (Skip Stitch and Broken Stitch).
Cause Root causes Remedial Action Procedures

types
 The needle thread needs to be tightened if the stitching thread shows loops on the underside.
Thread is too tight or loose.
 The needle thread needs to be loosed if the stitching thread shows loops on the topside.
Incorrect needle size or type.  Change needle size or style.
MAN
Inappropriate thread tension  Tension of the thread properly adjusted.
Needle plate, pressure foot, needle  Inspect the needle point at regular intervals and check for sharp or burred points (T. Islam
holes may have sharp edges. et al., 2017).
 Follow the instruction of the guide book to setup the m/c properly.
Faulty machine setup.
 Skilled operator should be assigned to setup the machine.
 Clean and oil the machine per manufacturer directions and make sure throat plate is properly
Broken needle head. aligned.
MACHINE  Use pins to hold fabric, pull the pins before the fabric goes under the foot.
Wrong needle size and thread size.  Needle size and thread size should be synchronized.
Excessive abrasion or chemical
 Special care should be taken during washing.
degradation.
Wrong needle selection for
 Select appropriate needle for sewing.
MATERIAL sewing.
Wrong thread.  Select good quality thread which is free from flaws.
 Control the timing between loading steps and cycles (Misra et al., 2010).
Needle bend.
METHOD  Use a reinforced needle, reset the needle guard (Coats, 2019).
Incorrect sewing tension.  Re-adjust the tensions.
4.5 Control phase

The control stage is the last and final stage and its sole purpose is to preserve the optimized
response obtained from the study. After having permission from the authority of one garment
factory (Garment A), remedial actions had been taken to remove root causes of major defects
mentioned in Analysis phase. Then defect data was collected again to calculate the defect rate and
sigma level, shown in table 6.
Table 6: Calculation of defect rate for ‘Garment A’ after implementation.
Total checked items 18,000 DPO 0.0901
No. of conforming items 16,378 DPMO 90,111
No. of non-conforming items 1,622 Sigma Level 2.84
% Defects 9.01
4.6 Findings
The defects rate and sigma level results after implementing six sigma in Garment A is respectively
9.01% and 2.84. The defect rate decreased from 11.09% to 9.01%. The Sigma level is increased
from 2.72 to 2.84.
Table 7: Comparison between before and after implementation in Garment A.
Defect Rate Sigma Level
Before implementation 11.09% 2.72
After implementation 9.01% 2.84
Page | 1357
5 Conclusion
The selected garment manufacturers taken for this study are focused on exports where quality is
major concern. This study uses the DMAIC Six Sigma technique to identify the main flaws and
their underlying causes, and then it recommends rational ways to reduce those flaws. Quality is
being improved in one garment factory by raising the sigma level from 2.72 to 2.84 and also
bringing the defect rate down from 11.09% to 9.01%. Good quality raises a product's worth,
creates a brand name, and produces a good reputation for the garment exporter, all of which lead
to satisfied customers, large sales, and foreign currency for the nation. This study suggests that in
order to more effectively reduce the defects in garment factories, additional industrial areas
including as knitting, dyeing, cutting, and finishing should be addressed.
References
Bala, S., & Gupta, T. (2009). Factors influencing costing of woven fabrics. The Indian Textile
Journal, vol. 1. no. 2, pp. 57–68.
Bañuelas, R., & Antony, J. (2003). Going from six sigma to design for six sigma: An exploratory
study using analytic hierarchy process. The TQM Magazine, 15(5), 334–344.
https://doi.org/10.1108/09544780310487730
BGMEA, 2019. Export Performance, Available in:
https://www.bgmea.com.bd/page/Export_Performance, (accessed 14 September 2022).
Coats, 2019. Solving Your Sewing Problems, Available in: https://coats.com/en/information-
hub/Solving-Your-Sewing-Problems, (accessed 8 October 2022)
Dean, J. W., & Bowen, D. E. (1994). Management Theory and Total Quality: Improving Research
and Practice through Theory Development. The Academy of Management Review, 19(3),
392–418. JSTOR. https://doi.org/10.2307/258933
Garments Merchandising, 2016. List of Defects Found in Sewing Section | Sewing Defects List,
Available in: https://garmentsmerchandising.com/list-of-defects-found-in-sewing-section/,
(accessed 21 March 2016)
Islam, M. M., Khan, A. M., & Khan, M. M. R. (2013). Minimization of reworks in quality and
productivity improvement in the apparel industry. International Journal of Engineering, 1(4),
2305–8269.
Islam, T., Khan, S., Sakib, H., Sakib, A., & Siddiquee, M. (2017). Analysis of Major Defects
Position and Percentage in Sewing Lines of a Garments Factory with the Help of Pareto
Chart, Cause Effect Diagram and Sigma Level. International Journal of Scientific and
Engineering Research, 8.
Kayaalp, İ. D., & Erdoğan, M. Ç. (2008). DECREASING SEWING DEFECTS BY USING
STATISTICAL PROCESS CONTROL METHODS IN THE APPAREL FACTORY.
TEKSTİL ve KONFEKSİYON, 6.
Kolarik, & William, J. (1999). Creating quality: Process design for results. Boston:
WCB/McGraw-Hill, c1999.
Kumar, C. s, & Naidu, N. (2011). Performance improvement of manufacturing industry by
reducing the Defectives using Six Sigma Methodologies. IOSR Journal of Engineering, 1,
1–9. https://doi.org/10.9790/iosrjen11A01100010009
Misra, S., Reed, K., Schafer, B., Ramesh, K. T., & Okamura, A. (2010). Mechanics of Flexible
Needles Robotically Steered through Soft Tissue. I. J. Robotic Res., 29, 1640–1660.
https://doi.org/10.1177/0278364910369714
Page | 1358
PAPER ID: 249
Assessment of Fire Risk Conditions of High-Rise Commercial

Buildings Located in Dhaka City
Jahid Hasan1*, Md. Mehedi Hasan Kibria2 and Md. Omar Faruq3
1,2
Assistant Professor, Department of Industrial and Production Engineering, Shahjalal
University of Science & Technology, Sylhet-3114, Bangladesh. Email:
1
j.hasan.ipe@gmail.com, 2kibria-ipe@sust.edu
3
Department of Industrial and Production Engineering, Shahjalal University of Science
& Technology, Sylhet-3114, Bangladesh. Email: 3faruq334032@gmail.com
*Corresponding author [Jahid Hasan]
Keywords: Abstract: Dhaka is a densely populated and centralized city of trade

and commerce for the entire country, making it difficult to
● Dhaka City; accommodate so many people in such a small space. In response to
● Fire-risk; the housing shortage, the number of high-rise buildings in Dhaka city
is growing day by day. As a result, high-rise commercial structures
● High-rise building;
are a common sight in this densely populated city. Such busy
● BNBC; commercial buildings must follow the safety provisions like the
● FRI. Bangladesh National Building Code (BNBC) to ensure the safety of
the building and its occupants. Among the many hazards endangering
the lives of people living and working in Dhaka's commercial
buildings, fire hazard is becoming an increasingly common
phenomenon. As a result, fire risk analysis for high-rise commercial
buildings is critical. According to BNBC, there are four major
parameters on which a building's fire safety system is entirely
dependent: By assigning grade points to these parameters, this study
attempts to quantify overall building safety. The study has been
attempted on a questionnaire-based survey, including 30 commercial
high-rise buildings in Dhaka city. Based on the analysis, the study has
to identify the most important parameters that are most affecting the
fire safety condition, as well as determine the current scenario of fire
protection of commercial high-rise buildings.
Furthermore, according to the FRI value and the correlation test with
other parameters, the escape facility and hazard management have
been found to have the highest priority of the four major parameters.
Most buildings failed to score a minimum of 3.0(on the basis of a
5.0(very good) scale) in all FRI importance parameters, which
indicates that most of the high-rise buildings in the study are at high
risk of fire hazard. Poor fire safety record and lack of information on
fire safety is common scenario in most of high-rise commercial
building in Dhaka city can be concluded from the study
Page | 1359
1. Introduction
Overpopulation is a major cause of most of the world’s problems. Population in the world is
currently (2020) growing at a rate of around 1.05% per year [1]. Population growth has a negative
and significant effect on land. Bangladesh population is equivalent to 2.11% of the total world
population while Dhaka city remains the most densely populated city in the whole country as well
as in the world, with a density of 47,400 people per square kilometer [2]..So, high-rise structures
are a common sight in Dhaka streets. A high-rise building is one that has a small footprint, a small
roof area, and very tall facades. To ensure the safety of habitant of high-rise building BNBC
introduced various parameters on fire hazard. Fire hazard is one of the key issues in high-rise
buildings. If safety can be determined of structures, precautionary measures can be taken to
prevent the hazard. So, measuring fire safety of Commercial High-rise buildings of Dhaka city
will be useful for government and concerning personnel. Results. Wadud, et al (2014) developed
an influential method which can be used to determine the specific fire safety condition of any
commercial building.
2. LITERATURE REVIEW
Safety culture is the collection of the beliefs, perceptions, and values that employees share in
relation to risks within an organization, such as a workplace or community [3]. Furthermore, fire
safety measures include those that are planned during the construction of a building or
implemented in structures that are already standing, and those that are taught to occupants of the
building [4]. Threats to fire safety are commonly referred to as fire hazards. A fire hazard may
include a situation that increases the likelihood of a fire or may impede escape in the event of a
fire [5]. On the other hand, the term "fire code" refers to the fire prevention code or safety code.
Bangladesh National Building Code (BNBC) is a government organization concerned with
building safety in general.The BNBC has established some defined procedures for construction,
and it consists of several aspects of safety, including fire safety, earthquake safety, etc. [6]. BNBC
has stated that there are four major parameters on which the fire safety system of a building is
soulfully dependent: a. Structural safety, b. Escape facility, c. Built-in firefighting capability, and
d. Hazard management [7].
All the parameters have their own sub-parameters where the sub-parameters were given weight
points by studying the previous fire incidents in Dhaka city, according to Wadud and Noor
(2014)[8]. Besides, the fire risk index (FRI), also known as risk ranking, point system, or
numerical grading, is the oldest quantitative method for fire risk assessment, and is still the most
widely used. In their research paper, Wadud et al. (2014) developed a fire risk index (FRI) for
soft parameters in the garment industry through surprise inspection. The study found a U-shaped
relationship between FRI and factory size [9]. Ming Lo (1999) proposed a fuzzy tire safety
assessment approach based on fire risk ranking techniques [10]. Sarraz and Islam (2012)
determined a fire risk estimate based on the results of data analysis and previous studies using an
Event Tree Analysis (ETA) [11]. Islam and Hossain (2018) explored the fire hazard risk and
effectiveness of the mitigation measures in Dhaka City [12].
3. RESEARCH METHODOLOGY
A survey research approach was performed for this study. So, on the basis of the literature
review and physical observation, a structured questionnaire was formed and analyzed based on
BNBC 2015. In this study, 42 variables have been introduced, which belong to 4 main
components according to BNBC 2015. There are two major approaches followed to collect data
from commercial buildings, which are: a) Primary data: Building characteristics. b) Secondary
data: Safety personnel or security officers' data.
Page | 1360
4. CALCULATION & RESULT:
4.1 Analysis of Data:
4.1.1 Reliability Test: Cronbach's alpha is the measurement of internal consistency of data, that
is, how
Table-1 Item -Total Statistics
closely a (Overall)
set of Scale Mean If Scale Variance If Corrected Item- Cronbach’s
data are Parameters Item Deleted Item Deleted Total Alpha if
related Correlation ItemDeleted
within a Structural Safety 12.3910 2.830 .699 .886
group. Escape Facility 11.5750 2.967 .801 .848
The Inbuilt Firefighting 11.8132 2.727 .777 .854
general Ability
thumb Hazard Management 11.3271 2.916 .782 .853
rule of Cronbach's alpha of 0.70 and above is considered good, 0.80 and above is considered as
better, and 0.90 and above is considered as best. [13]
4.1.2 Correlation Test: Pearson correlation (2 tailed) is performed to identify the relationships
among the four Table-2: Pearson Correlation among all four categories
major
parameters [14]. Parameters Structural Escape Inbuilt Hazard
The results are Safety Facility Firefighting Management
shown in the Ability
**
following tables. Structural Safety 1 .685 .639** .573**
Escape Facility 1 .671** .756**
Inbuilt Firefighting 1 .750**
Ability
Hazard Management 1
**
Correlation is significant at the .001 level (2 tailed)
4.1.3 Criteria Importance

Rating: To ensure that the rating scale (1-5) for measuring the criteria gives the same result over
time, a reliability analysis using the internal consistency method was firstly tested. Before ranking
the factors, the relative importance of Fire Risk Index components was quantified using the
relative importance index (RII) method and a total of X factors was ranked by category.
RII = ∑W / A*N Here, W-Weight given to each factor, A-Highest weight
N=Total number of samples
To identify the relative importance of causes based on the surveyed data relative importance index
is performed using the above-mentioned formula.[15]
4.2 Fire Risk Index (FRI) Value: For the current work, a linear additive model according to
Wadud, Z., et al (2014), of the following form to determine the FRI for the soft parameters:
FRI= ∑Wi* Xi /∑Wi
Here, Xi is the dimensionless score or grade points for parameters i. The grade points or scores
Xi indicate the 'measurement' of the risk, the outcome of the accident or protection in fire safety
given by a parameter i (BNBC 2015) and Wi is the weight of parameter.
Page | 1361
For this study, data was collected on 30 high-rise commercial buildings where the weight point is
Table 3: Calculation of FRI value given according
Building ID 8 to BNBC (2015) .
Parameters Weight(w) Grade(X) W×X For calculating
Structural Safety
Non loading Walls: 1 5 5
FRI value fewer
Structural Frame: 1 5 5 than four major
Corridor Enclosures: .75 5 3.75 parameters
False ceilings: 1.5 4 6 which are
Vertical Fire resistive Material .75 3 2.25 defined various
Service Duct Enclosure 2 5 10 sub-parameters.
Exposed Electrical or gas line 2 5 10
Smoke Vents 2 5 10
A sample
Smoke Arrest System 2 4 8 calculation on
Total 13 60 Structural safety
FRI 4.62 has been shown.
Table 4: FRI value for thirty commercial Buildings
Inbuilt Inbuilt
Building Structural Escape Hazard Building Structural Escape
Firefighting Fire
No. Safety Facility Management No. Safety Facility
Ability fighting
Ability
1 3.21 1.95 2.52 2 16 1.96 1.69 1.84
2 3.02 1.76 2.71 1.56 17 2.23 1.75 2.17
3 3.17 1.86 2.71 1.56 18 1.69 1.64 1.36
4 3.58 1.44 1.95 1.56 19 1.77 1.74 1.86
5 3.05 1.35 1.85 1.58 20 2.43 1.5 1.64
6 3 1.39 1.47 1 21 2.5 1.55 2.12
7 3.09 1.97 3.1 1.29 22 1.62 1.52 1.93
8 4.62 4.47 3.57 3.55 23 1.88 1.74 1.76
9 2.65 1.87 2.78 1.91 24 2.5 1.55 2.02
10 2.36 1.2 1.9 2.17 25 1.25 1.15 1.19
11 2.69 1.85 2.85 1.55 26 3 2.15 3.76
12 1.38 1.29 1.28 1 27 2.42 1.7 2.24
13 2.56 1.87 3.85 2.38 28 1.87 1.8 2.36
14 1.77 1.34 1.47 1 29 3.08 1.98 1.79
15 2.94 2.39 1.85 2.14 30 1.98 1.79 2.17
Table 2 shows the FRI values of 30 buildings in four major categories. To maintain confidentiality
name of buildings are not mentioned.
4.3 Frequency Distribution: Frequency distribution determiners show many buildings
(frequency) fall under each category. Frequency bands are in the range of Likert scale. Therefore
5 frequency bands are found where, 1= very poor, 2=poor, 3= average, 4=good, 5= very good.
Page | 1362
FRI Range Structural Escape Inbuilt Fire Hazard
Safety Facility Fighting Management Table 3: Frequency
Ability
(1.00≤FRI<2.00) 10 27 15 24 distribution of FRI
(2.00≤ FRI <3.00 12 2 11 4 value ranges.
(3.00≤ FRI <4.00) 7 0 4 2
(4.00≤FRI < 5.00) 1 1 0 0
5 0 0 0 0
Figure 1: Combined Frequency Distribution
FRI values of Individual Buildings

5
4
FRI Value
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Building ID
Structural Safety Escape Facility Inbuilt Firefighting Ability Hazard Management
Figure 2: FRI value for individual buildings

5. DISCUSSION
The present fire safety conditions in selected thirty commercial high-rise buildings in Dhaka
city are almost similar, with a very few exceptions. Most of the buildings lacked the average
requirements suggested by the BNBC to be acknowledged as fireproof buildings. From the FRI
calculation, it indicates that no building scored a perfect 5 in any category.
From Pearson correlation (2-tailed) analysis, results show that escape facility and hazard
management should be the prime concerns as they both have the highest correlation with other
parameters. However, based on the correlation value with respect to component-1, the ranking
of major parameters can be determined as-
Page | 1363
Escape facility > Hazard management > Inbuilt firefighting ability > Structural safety
So, it can be concluded that the condition of the escape facilities of a selected 30 commercial
high-rise buildings is in a poorer and more dangerous condition.
6. CONCLUSION
A fire risk index provides a quantitative safety condition of a building. The present fire safety
conditions in most of the commercial buildings in Dhaka city are quite similar. Most of the
areas of safety culture are alarming in those commercial buildings. Only a few of the buildings
have better fire safety conditions. Some limitations have been found in this study, and at the
same time, some scopes for future work have been found. As the escape facility is the most
concerning parameter, the escape facility of each building should be improved first. The number
of fire exits, exit location, width, etc. must be redesigned with the BNBC 2015 standard.
7. REFERENCES:
1. Bangladesh Population- World Population review 2020 Retrieved from:
https://worldpopulationreview.com/
2. Dhaka, densely populated city-Dhaka Tribune 2019. Retrieved from:
http://www.dhakatribune.com/bangladesh/dhaka/2018/10/14/dhaka-remains
3. Safety culture. (2019). Retrieved from https://en.wikipedia.org/wiki/Safety
4. Fire Safety (2019). Retrieved from https://en.wikipedia.org/wiki/Fire
5. Common Fire Hazards Retrieved from https://en.wikipedia.org/wiki/Fire
6. Fire Code (2019) Retrieved from https://ico.wikipedia.org/wiki/Fire safety code
7. . BNBC. (2015). Bangladesh National Building Code Housing and Building Research
Institute, 3, 301-312. Retrieved from huwww.apsel.com/download/Banglades National
Building Code: 2015 Vol 3 320(Draft).pdf
8. Wadud. Z. & Noor, S. E., (2014). Development of Fire Safety Rating for Commercial
High-rise Buildings of Dhaka City. April 2014.
9. Wadud. Z. Huda. F. Y., & Ahmed, N. U. (2014). Assessment of Fire Risk in the
Readymade Garment Industry in Dhaka, Bangladesh: Fire Technology, 50(5), 1127
1145. https://doi.org/10.1007/s10694-013-0349-2
10. Ming Lo, S. (1999). System for Fire Technology, 35(2), 602-607
11. Sarraz, A., Mohammod, E., & Islam, A. (2012) Proceedings of the 1st International
Conference on Civil Engineering for Sustainable Development PERFORMANCE
BASED FIRE SAFETY MANAGEMENT IN COMMERCIAL MIXED-USE
BUILDINGS OF BANGLADESH Professor, Department of Civil and Environmental
Engineering , Shahj
12. Islam, Z., & Hossain, K. (2018). "Fire Hazards in Dhaka City: An Exploratory Study
on Mitigation Measures." IOSR Journal of Environmental Science, Toxicology und
Food Technology (IOSR-JESTFT), 12(5), 46-56. https://doi.org/10.9790/2402-
1205014656
13. Cortina, J. M. (1993). What is coefficient alpha? An examination of theory and
applications. Journal of applied psychology, 78(1), 98.
14. Pearson Correlation 2019. Available at: https://www.spss-tutorials.com/pearson
correlation-coefficient/
Page | 1364
PAPER ID: 260
Experimental Investigation of PV Cooling Methods using a

Finned Heat Sink and Cotton Wicks
Rajesh S. Rejon1*, Muhammad M. Hasan2, Ahsanul Kabir3, Md. T. Firdaus4,
and A.B.M. A. Malek5
1,2,3,4,5
Department of Industrial and Production Engineering, Shahjalal University of Science
and Technology, Sylhet-3114, Bangladesh.
Email: 1rajeshsarker1498@gmail.com; 2muhammad.hasan-ipe@sust.edu;
3
nabilkabir82@gmail.com; 4tasneemfd.sust.ipe@gmail.com; 5bashar@sust.edu
* Corresponding author: 1Rajesh S. Rejon (rajeshsarker1498@gmail.com)
Abstract: In today's world, it is crucial to replace fossil-fuel-based energy

Keywords: with renewable energy, particularly solar energy. Photovoltaics (PV) is the
easiest and simplest way to harness solar energy, though it suffers
● Solar energy; deterioration and power output reduction due to panel heating effects.
Therefore, an effective cooling method is necessary to be integrated with PV
● Photovoltaics (PV) modules. In the present work, four cooling methods have been investigated
modules; using a finned heat sink for a polycrystalline silicon (Poly-Si) module of 50W.
The heat sink has been fabricated using 3 mm aluminum sheet and attached
● Polycrystalline-Si; with the module using a number of clamps. The cooling techniques tested are:
(a) natural cooling of the PV module without a heat sink, (b) passive cooling
of the PV module with a heat sink, (c) active cooling of the PV module with
● Monocrystalline-Si; heat sink and forced air cooling (fan), and (d) water cooling with cotton wicks
on the finned heat sink. An experimental setup has been made and installed on
● Cotton wicks. the roof-top of an academic building at Shahjalal University of Science and
Technology, Sylhet. For measuring different parameters to evaluate power
output and temperatures, a data logging device was designed and fabricated
using a ESP 32 microprocessor and different sensors (current, voltage and
temperature). Several outdoor tests have been conducted in August and
September 2022, with the majority of the days being mostly sunny. Around
noon, when solar irradiance is at its highest, the Poly-Si module's (50W) power
output for the water spray cooling method reaches its maximum.
1. Introduction
Renewable energy sources including solar, wind, biomass, and hydropower can potentially be
used as alternatives to enable Bangladesh and the rest of the globe deal with their current energy
crisis. In Bangladesh, renewable energy now accounts for only 1.45% of total energy production,
with solar energy accounting for the majority of that amount (62.9%) (SREDA, 2022). Solar
energy is one of the most prospective renewable energy sources because it is the most abundant
source and has the most potential for global energy production. One hour of sunlight hitting the
planet produces more energy than an entire year's worth of human use (Lewis, 2007). Direct solar
energy harvesting with photovoltaic panels is one of the most promising and expanding markets
in the renewable energy sector due to its falling cost (Lewis, 2016; Sampaio & González, 2017).
Page | 1365
The International Energy Agency (IEA) projects that by 2050, photovoltaics (PV) will generate
around 11% of the world's electricity, preventing 2.3 Gt of carbon dioxide (CO 2) emissions
annually (Touati et al., 2016).
In reality, only 15–25% of incident sun irradiation is converted to electricity, with the remainder
becoming heat (Teo et al., 2012). It has been claimed that when module temperature rises, PV
module efficiency falls (Rajvikram et al., 2019; Radziemska, 2003; Rahman et al., 2015). For
every 1 °C increase in module temperature, the PV module's output power is reported to fall by
roughly 0.65% (Radziemska, 2003). Depending on the kind of PV materials, the reduction in
module efficiency ranges from 0.25 to 0.5% per degree Celsius (Rahman et al., 2015). Several
studies (Bayrak et al., 2020; Chandrasekar et al., 2013; Lim et al., 2017; Rajvikram et al., 2019;
Rahman et al., 2015) were found to investigate various cooling strategies where PV modules were
tried to cool by active and passive cooling methods. To increase cooling effects, a variety of
methods were used in these studies, including aluminum fins, spraying water with ice blocks,
water cooling with cotton wicks, phase-change materials, etc.
The output power of PV module can be increased by finding an optimum and cheap cooling
method, which is still worth looking into. Therefore, the purpose of this study is to apply a few
straightforward cooling solutions to investigate the effects of temperature on the output power of
a commonly used Poly-Si PV module.
2. Experimental details
Throughout the whole experimental run, the solar irradiance (E) was measured by a Pyranometer
(SP Lite2, Kipp & Zonen, Netherlands) that was interfaced with a Mooshimeter (Mooshim
Engineering, USA). Every 10 seconds, the solar irradiance was recorded using a Mooshimeter
that was Bluetooth-connected wirelessly to a cellphone. The following equation is used to
calculate the irradiance (E) in W/m2:
Uemf
E (1)
S
where, Uemf (µV) is the output voltage and S is the sensitivity of the Pyranometer (73.4 μV/W/m²).
On the roof of academic building "C" at Shahjalal University of Science & Technology (SUST),
Sylhet, an experimental setup has been created for testing solar photovoltaic panels at a fixed 20°
angle to the horizontal. The experimental procedure has been detailed in our prior work (Hasan
et al., 2022).
The effects of cooling on PV power output were examined in the current work using one
polycrystalline Si module with a peak power of 50W. The total cell area of the selected solar
panel and its fill factor are 3.088×10-1 m2 and 0.747, respectively. Fill Factor (FF) of a PV module
is the ratio its maximum peak power (PMPP) with the product of short-circuit current (Isc) and open
circuit voltage (Voc). Four different cooling techniques had been studied in this research: (a)
natural cooling, (b) natural cooling with the finned heat sink, (c) forced cooling with the finned
heat sink, and (b) water cooling using cotton wicks covering the finned heat sink. Figure 1 depicts
the developed heat sink that is covered with cotton wicks to allow for water cooling. The forced
air-cooling method, which employed a conventional electric fan to move ambient air at a speed
of 4-6 m/s to cool the module, was intendd to be contrasted with the natural air-cooling method.
Page | 1366
The peak power and the efficiency of a PV module are determined using the procedure detailed
in our previous work (Hasan et al., 2022).
(a) (b)
Figure 1. The developed finned heat sink covered by cotton wicks for water cooling: (a) top view and (b)
backside view of the PV module.

3.1 Natural air cooling of PV module alone
When using the natural air cooling method, Figure 2a depicts the variations in solar irradiation,
panel temperature, and power output for the poly-crystalline Si module. The temperature of the
panel is observed to be between 40 and 64 °C. The amount of solar irradiation is found to have a
significant impact on peak power output. The peak power output of the studied PV module is
calculated to be 44W. Figure 2b shows a distinct fluctuation in efficiency for the investigated
solar panel in relation to panel temperature. As the temperature of the module increases, the
efficiency of the panel drops. For the temperature range of 45–62 °C, the efficiency of the Poly-
Si panel is determined to be between 11–14%.
(a) (b)
Figure 2 (a) Solar irradiance, panel temperature and peak power of the PV module during natural air
cooling and (b) the effect of panel temperature on its efficiency.
3.2 Natural air cooling with the finned heat sink

To take advantage of passive cooling, a finned heat sink attached to the back side of the PV
module was developed. Figure 3a demonstrates the variations in solar radiation, panel
temperature, and power output during natural air cooling with the finned heat sink. Here, the
temperature of the panel is roughly found to be between 36 and 52 °C. The peak power output of
the PV module is calculated to be 42W. The temperature of the module increases, the efficiency
Page | 1367
of the panel drops as shown in Figure 3b. For the temperature range of 43.5-50.5°C, the efficiency
of the Poly-Si panel is estimated to be between 11.5-17.0%.
(a) (b)
Figure 3 (a) Solar irradiance, panel temperature and peak power of the PV module during natural air
cooling with the finned heat sink and (b) the effect of panel temperature on its efficiency.
3.3 Forced air cooling with the finned heat sink

In order to observe changes in solar radiation, panel temperature, and power output, forced air
cooling with a finned heat sink was examined using an electric fan, as illustrated in Figure 4a.
Throughout the duration of the experiment, the lowest panel temperature was recorded in the
range of 36.4 to 42°C. The power output of the PV module is determined to be a maximum of
48W due to the lowest panel temperature. Figure 4b depicts the efficiency of the panel in relation
to temperature and shows a similar downward trend. The maximum efficiency of the Poly-Si
panel is estimated to be about 16.4%.
(a) (b)
Figure 4 (a) Solar irradiance, panel temperature and peak power of the PV module during forced air
cooling with the finned heat sink and (b) the effect of panel temperature on its efficiency.
3.4 Water cooling using cotton wicks with the finned heat sink
Water spray cooling method was reported to necessitate a large volume of water as well as electric
energy to run a water pump. To eliminate these requirements, water cooling using cotton wicks
is a viable option. In this research work, we have designed a cheap system of wicks that was made
of martin clothes. As an easy and simple cooling system, this method is expected perform better
than forced air cooling in which large electric energy is required. Figure 5a illustrates the changes
in solar radiation, panel temperature, and power output during this water cooling with cotton
wicks. In this test, the temperature of the panel is roughly found to be between 35 and 47 °C.
From the figure, it is found that the panel temperature remains at around 40 °C for the most of
Page | 1368
test duration. Figure 5b shows the efficiency of the panel in relation to temperature and shows a
similar downward trend. The efficiency of the Poly-Si panel ranges between 10.5-15.4%.
(a) (b)
Figure 5 (a) Solar irradiance, panel temperature and peak power of the PV module during water cooling
using cotton wicks with the finned heat sink and (b) the effect of panel temperature on its efficiency.
3.5 Comparison of the cooling methods

There is a general fluctuation among all indicators due to various operating conditions. Table 1
displays the operating temperatures, the efficiency range, and peak power attained during our
studied tests. The forced air-cooling approach produced the lowest temperature range (36.4-42.0)
with an efficiency range (13.0-16.4), although there was a problem with the electric fan's power
consumption. There is evidence of a various rate of deterioration that was observed with an
average of about 0.5 percent per degree centigrade.
Table 1: Summary of the effects of the cooling techniques studied in the present work.
Natural air Natural air Forced air Water

cooling cooling cooling cooling
with fins with fins with wicks
Operating temperature (°C) 43.0-61.5 43.5-50.5 36.4-42.0 37.7-46.5
Maximum peak power (W) (at solar 35.5-44.0 28.0-42.0 39-48 31.0-44.0
radiation>950 W/m2)
Efficiency range (%) 10.5-15.0 11.5-17.0 13.0-16.4 10.5-15.4
Efficiency deterioration rate (%/°C) 0.14 0.89 0.51 0.37
4. Conclusion
In this present work, a variety of cooling techniques using an aluminum finned heat sink, as well
as a simple and easier method of water cooling with cotton wicks on the finned heat sink, have
been investigated. Although there was an issue with the electric fan's power consumption, the
forced air-cooling approach produced the lowest temperature range (36.4-42.0 °C) with an
efficiency range (13.0-16.4%). There is evidence of a variable rate of deterioration, with an
average of about 0.5 percent per degree Celsius. In this research work, we have designed a cheap
system of wicks that was made of martin clothing. As an easy and simple cooling system, this
method is found to perform close to forced air cooling in which large electric energy is required.
The temperature of the panel is roughly observed to be between 35 and 47 °C during the water
cooling with cotton wicks, and the efficiency is found to be a maximum of 15.4%. Therefore,
Page | 1369
water cooling with cotton wicks can be considered the most effective, environmentally
acceptable, simple to use, and cost-efficient.
Acknowledgment
The authors would like to acknowledge the financial support provided by the SUST Research
Centre, Shahjalal University of Science and Technology, Sylhet under the Project Code:
AS/2022/1/42.
References
Bayrak, F., Oztop, H. F., & Selimefendigil, F. Experimental study for the application of different
cooling techniques in photovoltaic (PV) panels. Energy Conversion and Management. 2020,
212; 112789.
Chandrasekar, M.; Suresh, S.; Senthilkumar, T. et al. Passive cooling of standalone flat PV
module with cotton wick structures. Energy Conversion and Management. 2013, 71; 43–50.
Hasan, M. M., Rifat, A. A., Safwan, O. F. et al. Effects of simple cooling methods on power
output and efficiency of solar panels in outdoor conditions. ARPN Journal of Engineering
and Applied Sciences. 2022, 17(3); 288-294.
Lewis, N. S. Toward Cost-Effective Solar Energy Use. Science. 2007, 315(5813); 798–801.
Lewis, N. S. Research opportunities to advance solar energy utilization. Science. 2016,
351(6271); aad1920.
Lim, J.-H., Lee, Y.-S., & Seong, Y.-B. Diurnal thermal behavior of photovoltaic panel with phase
change materials under different weather conditions. Energies. 2017, 10(12); 1-14.
Rajvikram, M., Leoponraj, S., Ramkumar, S. et al. Experimental investigation on the abasement
of operating temperature in solar photovoltaic panel using PCM and aluminium. Solar
Energy. 2019, 188; 327-338.
Radziemska, E. (2003). The effect of temperature on the power drop in crystalline silicon solar
cells. Renewable Energy. 2003, 28(1); 1–12.
Rahman, M. M., Hasanuzzaman, M., & Rahim, N. A. (2015). Effects of various parameters on
PV-module power and efficiency. Energy Conversion and Management, 103, 348–358.
Sampaio, P. G. V., & González, M. O. A. (2017). Photovoltaic solar energy: Conceptual
framework. Renewable and Sustainable Energy Reviews, 74, 590–601.
SREDA. Present status of Renewable Energy in Sustainable and Renewable Energy Development
Authority, available in: http://www.sreda.gov.bd/index.php/site/re_present_status (Accessed
29 August 2022)
Teo, H. G., Lee, P. S., & Hawlader, M. N. A. (2012). An active cooling system for photovoltaic
modules. Energy Solutions for a Sustainable World, Special Issue of International Conference
of Applied Energy, ICA2010, April 21-23, 2010, Singapore, 90(1), 309–315.
Touati, F., Al-Hitmi, M. A., Chowdhury, N. A., Hamad, J. A., & San Pedro Gonzales, A. J. R.
(2016). Investigation of solar PV performance under Doha weather using a customized
measurement and monitoring system. Renewable Energy, 89, 564–577.
Page | 1370
PAPER ID: 288
Effects of Investment on Industry 4.0 implementation
Mahfuzul Haque 1*, Choudhury Abul Anam Rashed 1 and Mst. Nasima Bagum 1*
1
Department of Industrial and Production Engineering, Shahjalal University of Science and
Technology, Sylhet
Email: mahfuzhaque1995@gmail.com, rashed-ipe@sust.edu,
* mahfuzhaque1995@gmail.com, * nasima-ipe@sust.edu
*Corresponding author [Author name]
Abstract: The governing parameters of any business for sustainable growth

Keywords: and development are technology and innovation. The goal of Industry 4.0 is to
achieve a standardized level of productivity and operational efficiency. The
● Industry 4.0; fourth industrial revolution is interchangeably known as Industry 4.0, which
● Investment; is a control of the industrial value chain. The objective of the study is to
investigate the impacts of investment on the implementation of Industry 4.0
● Manufacturing;
in the manufacturing sector. A semi-structured questionnaire was used to
● Technology; perform the research work. The respondents were chosen based on their
relevancy to the management of technology in their respective sectors. Both
● Innovation;
face-to-face interviews and web-based questionnaires were used to collect
● Productivity. the data. The whole analysis was performed using SPSS 25 and Microsoft
Office Excel 2016. The developed questionnaire was validated through
consultation with academics and industrial arena experts. A hypothesis was
developed to testify to the relevance of the investment with the
implementation of Industry 4.0. It was observed that in the past 2 years
investment of the organization for technology and innovation was about
medium scale. Over the next 5 years investment of the organization for
technology and innovation go through medium scale as well as large scale. In
this case, the investment criteria (small, medium, large) were identified based
on the percentage of the investment of total revenue.
1. Introduction
Industry 4.0 must reach the complicated level through productivity and effectiveness in order to
move through the next level of automatization. The fourth industrial revolution, or "Industry
4.0," is a new phase of the industrial value chain. Due to a lack of resources, adopting new
technologies is a frightening task for Small and Medium-sized Enterprises (SMEs).
Industry 4.0 is crucial for helping manufacturers with current problems because it increases
their ability to adapt to market changes. Innovation that is customer-focused and has a quicker
design cycle is accelerated by industry 4.0.
Bangladesh began its development as a small non-traditional export sector of RMG and
manufacturing in the late 1970s. The first manufacturing facility opened its doors in 1976. For
Page | 1371
three reasons, the sector deserves special treatment: The industry employs roughly 4.2 million
people and generates (a) about 77 percent of the nation's annual foreign exchange earnings (b)
the fastest growing industry in recent years. In spite of the difficulties, it has faced in the past, it
has consistently displayed reliable performance, a commitment to the community, and
competitive strength [1].
Innovation and technology are crucial to any organization's success. A company's technological
assets are any technology that it possesses. Therefore, this asset may also include technological
know-how, brainpower, hardware and software. They represent the organization's collective
technical know-how, which includes its personnel, machinery, and systems [2]. For the
development of a competitive strategy for a country depends on the proper linkage
among investment criteria, technology and the system of trade. Therefore, it is needed
to investigate the relationship between implementation of technology and innovation
with investment on that particular technology.
2. Objectives of the Research

i. To identify the effects of investment on the implementation of Industry 4.0 in
manufacturing sector.
3. Literature Review
Industry 4.0 is important for the development of a country because it automates and facilitates
data exchange in manufacturing-related technologies [3]. Countries with a high level of
innovation and technological integration can increase their productivity and production
capacity.
Uddin and Tangem's [4] mentioned that, the RMG sector accounts for more than 80% of export
profits in Bangladesh as the sector is the primary source of foreign income. The purpose of the
study was to identify the link between labor unrest in Bangladesh and workplace unrest in the
RMG sector.
Bangladesh is one of the major nations producing apparel, competing with China. As the world
adjusts to Industry 4.0, it is important to evaluate the prospects for developing nations,
particularly in the ready-made garments (RMG) sector. Islam et al. [5] conducted a study on
I4.0 and its potentiality in the developing countries. They identified some of the issues faced by
the indutry, including inadequate infrastructure, readily accessible, inexpensive labor, expensive
technology setup, insufficient knowledge, and improper government support.
Ullah et al. [6] provided a summary of the current consistency management in the RMG
manufacturing plants in their study. In their research, they found evidence of the value of
current consistency in management practices on RMG among employers and employees.
Rantala et al [7]'s work was centered on the interaction between sustainability and adoption-
related innovations. By describing the manageability factors that underlie operators' eagerness
to adopt new technologies, services, and business models in their study, they close a research
gap.
Page | 1372
According to Saiful & Chandra [8], the Bangladeshi garment industry is an emerging tiger in
the global economy. Their research was focused on the economy of the sector. They tried to
evaluate the influence of macroenvironmental, competitive, internal, and external variables on
the Bangladeshi apparel industry.
Saha & Mazumder's [9] proposed on their study, the readymade garments industry (RMG)
sector serves as the economic backbone of Bangladesh and serves as a catalyst (impetus) for the
development of the nation. Through the use of some significant statistical tools Analysis of
variance (ANOVA), and hypothesis testing, various parameters of workplace prompting
(environment leading) less efficiency of RMG industries was distinguished and analyzed in
their research work.
4. Methodology
The current research is survey-based research. At first research problem was formulated. In
preliminary step, appropriate research area was framed. At the same time all available literature
was examined to get acquainted with the selected area. Then the research questionnaires were
developed. The developed questionnaire is validated through consultation with the experts from
both academics and industrial arena. Based on the collected data scientific data analysis
technique is performed. To identify the relationship between investments with implementation
of industry 4.0 a hypothesis for both RMG and Plastic processing industry was developed and
tested.
5. Analysis and Result

5.1 Companies Investment in the Implementation of Industry 4.0
The studied organizations investment in technology and innovation management over last 2
years and next 5 years are depicted respectively in figure 1 and 2. Here, investment of the
organizations for technology and management in the past 2 years was about medium scale. Over
the next 5 years organizations investment in technology and management sector is moving
towards medium scale to some extent in large scale. Investments have been categorized as
small, medium and large based on the percentage being invested of the total revenue.
Investment that is <5% as small scale, <10% as medium scale and >10% as large scale.
Figure 1: Investment status in the past 2 years
Page | 1373
Figure 2: Investment status in the next 5 years
5.2 Analysisof the Hypothesis
Figure 3 illustrates the link between Implementation of industry 4.0 with the Investment of the
organizations.
Investment H1 Implementation of
Industry 4.0
Figure 3: The Conceptual Model.

Hypothesis:
Null Hypothesis (H0): Investment has an impact on implementation of I4.0.
Alternative Hypothesis (HA): Investment has no impact on implementation of I4.0.
Investment in the required departments of the studied factories like Research & Development,
Production, Purchasing, Logistics, Sales, Services, IT etc. with proper synchronization for the
implementation of I4.0.
Hypothesis analysis for RMG
Table 1: Model Fitting Information
Model Fitting Information
Model Fitting Likelihood Ratio Tests
Criteria
Model -2 log likelihood Chi-square df Sig. (α)
Intercept only 179.356
Final 4.159 175.197 112 0.000
Using the chi-square statistic, the model fitting was assessed. Probability of a statistical
hypothesis test is p-value. The p-value is greater than 0.05 so, the null hypothesis is accepted.
Page | 1374
Table 2: Goodness of Fit test
Goodness of Fit
Ch-square df Sig.
Pearson 0.000 136 1.000
Deviance 0.000 136 1.000
From this table Pearson & Deviance chi-square indicates the model is fit [p = 1.00]. So, this test
is statistically significant.
Table 3: Pseudo R-square
Pseudo R-square
Cox and Snell 0.915
Nagelkerke 0.990
McFadden 0.955
The model accounts for 95.5% to 99% of the variance and represents relatively decent-sized
effects.
Likelihood Ratio Test
The independent variables of hypothesis were significant, which proves that these predictors
contribute significantly to the final model.
Hypothesis analysis for Plastic Product Manufacturing
Table 4: Model Fitting Information
Model Fitting Information
Model Fitting Likelihood Ratio Tests
Criteria
Model -2 log likelihood Chi-square df Sig. (α)
Intercept only 65.884
Final 3.008 62.876 57 0.276
Using the chi-square statistic, the model fitting was assessed. Probability of a statistical
hypothesis test is p-value. The p-value is greater than 0.05 so, the null hypothesis is accepted.
Table 5: Goodness of Fit test
Goodness of Fit
Ch-square df Sig. (α)
Pearson 0.000 12 1.000
Deviance 0.000 12 1.000
From this table Pearson & Deviance chi-square indicates the model is fit [p = 1.00]. So, this test
is statistically significant.
Table 6: Pseudo R-square
Pseudo R-square
Cox and Snell 0.877
Page | 1375
Nagelkerke 0.973
McFadden 0.905
The model accounts for 90.5% to 97.3% of the variance and represents relatively decent-sized
effects.
6. Discussion
The whole analysis is based on the Manufacturing Industries of Bangladesh. At the investment
phase, it seems that, investment of the organizations for technology and management in the past
2 years was about medium scale which is seen in the next 5 years moving towards large scale.
This is the reflection that the manufacturing sector is not yet ready to adopt industry 4.0 from
the perspective of investment.
7. Conclusion
From the obtained results it can be concluded that,
i. Investment indicates that the selected industrial sector is not yet ready.
ii. There is a strong relationship between investment and implementation of industry 4.0
References
[1] Lasi, H., Fettke, P., Kemper, H., Feld, T., & Hoffmann, M. (2014). Industry 4.0. Business &
Information Systems Engineering, 6(1), 239-242.
[2] Khalil, T., (2000). Management of Technology-The key to Competitiveness and Wealth
Creation. Chapter-1.
[3] Buhr, D. (2015). Social innovation policy for Industry 4.0. Friedrich-Ebert-Stiftung,
Division for Social and Economic Policies.
[4] Uddin, M. I. & Tangem, S. A Study on the Relationship between Working Environment and
Labor Unrest in Ready-Made Garment (RMG) Industry of Bangladesh. International Journal
of Management Sciences and Business Research, Sep-2015, ISSN-(2226-8235), 4(9).
[5] Islam, M. A., Jantan, A. H., Hashim, H., Chong, C. W., & Abdullah, M. M. (2018). Fourth
Industrial Revolution in Developing Countries: A Case on Bangladesh. Journal of
Management Information and Decision Sciences (JMIDS), 21(1).
[6] Ullah, N., Sunny, M. M. A. & Rahman, M. H. Compliance Management Practices on RMG
Industry in Bangladesh: An exclusive Study.
[7] Rantala, T., Ukko, J., Saunila, M. & Havukainen, J. (2017). The Effect of Sustainability in
The Adoption of Technological, Service, and Business Model Innovations. Journal of
Cleaner Production, Vol. 172, 20 January 2018, Pages 46-55.
[8] Saiful, I. M. & Chandra, P. D. (2017). Impact of Macro Environmental Factors on Garments
Industry That Drives Export in Bangladesh. Studies in Business and Economics no. 12(2),
2017.
[9] Saha, P. & Mazumder, S. (2015). Impact of Working Environment on Less Productivity in
RMG Industries: A Study on Bangladesh RMG Sector. Global Journal of Management and
Business Research: G Interdisciplinary, 15(2), Version 1.0, 2015. Type: Double Blind Peer
Reviewed International Research Journal, Publisher: Global Journals Inc. (USA), Online
ISSN:2249-4588 & Print ISSN: 0975-5853.
Page | 1376
PAPER ID: 328
Study on Musculoskeletal Disorders in Farmers Involved in

Rice Cultivation in Selected Villages of Sylhet, Bangladesh
Mohammad Iqbal 1*, Rafid Ahmed Chowdhury 2* and Md. Isfar Khan 3*
1
Professor, Department of Industrial and Production Engineering, Shahjalal University of
Science & Technology, Sylhet 3114, Bangladesh; Email: iqbalm_ipe@yahoo.com
2
Technology, Sylhet 3114, Bangladesh; Email: rafidahmed92@gmail.com
3
Technology, Sylhet 3114, Bangladesh: Email: isfarkhan37@gmail.com
*Corresponding author [Mohammad Iqbal, Rafid Ahmed Chowdhury and Md. Isfar Khan]
Keywords: Abstract: In Bangladesh, farming is a significant and important industry but

the farmers frequently experience musculoskeletal issues. The objective of the
● Musculoskeletal
research is to study and analyze the most prevalent musculoskeletal disorders in
Disorder;
farmers who are involved with the cultivation of rice in three villages situated in
● Prevalence; Sylhet, Bangladesh. Slovin’s formula was used to estimate sample size from a
given population size and a margin of error. A total of 200 rice farmers were
● Factor;
selected randomly from the three selected villages. A questionnaire-based
● Farmer; interview was conducted to collect data. Percentages of the most prevalent
musculoskeletal effects for different operations (ploughing, weeding, fertilizing
● Bangladesh.
and cutting) of rice cultivation were plotted by using Microsoft Excel. Chi
Square test was used to find the significant relationships between the
demographic factors (age, weight and height) and the most prevalent
musculoskeletal pains during the aforementioned operations of rice cultivation.
Findings show that lower back pain is the most prevalent during ploughing
(41%), weeding (47%) and cutting (53%). Hand pain is the highest reported pain
during fertilizing (54%). Chi Square test concluded that lower back pain
experienced during cutting and weeding is not influenced by any significant
demographic factors while lower back pain during ploughing is influenced by
the height (p=0.00) and weight (p=0.00) of farmers. The occurrence of hand pain
during fertilizing is significantly influenced by the weight (p=0.004) of the
farmers. Height (p=0.038) and weight (p=0.00) are responsible for the
prevalence of hand pain during cutting while age (p=0.015) and weight
(p=0.003) are responsible for hand pain during weeding.
1 Introduction
Species of grains and cereals, which are the primary source of food for billions of people globally,
include wheat and rice (Welch & Graham, 2004). Increasing demand from domestic and
Page | 1377
international consumers is key to this industry's expansion. Globally, agriculture production
employs an estimated 1.3 billion people. Asia is home to half of the world's workforce and a
significant portion of the agricultural workforce (International Labour Office., 2011). In both
developing and developed nations, agricultural work has long been acknowledged as one of the
most dangerous professions for human health (Naeini et al., 2014). Farmers may be at risk of
getting musculoskeletal disorders (MSDs) because farming is typically seen as a physically
demanding occupation. Numerous agricultural chores are frequently performed in hazardous and
unmanaged environments, which increases the risk of accidents, illnesses, and fatalities,
especially those connected to musculoskeletal disorders (Bhattarai et al., 2016) (Palmer &
Goodson, 2015). According to Fathallah (2010), most farmers face a number of occupational
dangers when cultivating grains and cereals like wheat and rice, especially when they must work
in a stooping and crouching position. Various researches highlight the correlation between the
occupation and the prevalence of MSDs, while other investigations have found a link between
farming and the development of MSDs (Hartman et al., 2006; Rosecrance et al., 2006). Farmers
who work long hours in various tasks with awkward postures are more likely to develop
musculoskeletal disorders (Jain et al., 2018).
A set of diseases known as musculoskeletal disorders include those that affect the musculoskeletal
system, which includes the nerves, muscles, tendons, joints, and supporting structures including
intervertebral discs, etc. (Osborne et al., 2010). Musculoskeletal disorders can lead to discomfort,
musculoskeletal structure damage, poor health, a poor quality of life, and decreased productivity
(Bihari et al., 2011). They are the most frequent cause of severe, ongoing pain and impairment,
and it is currently estimated that hundreds of millions of people worldwide are affected by them
(Ghasemkhani et al., 2008). Work-related musculoskeletal disorders (WMSDs) are a significant
contributor to occupational health problems as well as economic costs, according to the
International Labor Organization (ILO), which estimates that there are over 160 million work-
related illnesses worldwide each year. Musculoskeletal Disorders (MSDs) are a widespread issue
that significantly affect a nation's health and economy (Woolf & Pfleger, 2003). Work-related
issues like MSDs may not be resolved until farmers are subjected to a risk assessment for MSDs.
These conditions may further result in disabilities (Stankevitz et al., 2016).
Agriculture is the main industry in Bangladesh. Agricultural activities are primarily practised by
the majority of the population, who live in rural areas. Since they provide the biggest contribution
to our economy, keeping them fit is a major concern in Bangladesh. This study focuses on rice
farmers to analyze the prevalence of musculoskeletal problems among them while conducting
various farming operations, including ploughing, weeding, fertilizing, and cutting.
2 Research Methodology
2.1 Sample preparation
Using the survey method, data were collected from three selected villages namely Ashighar
dakkin, Ashighar uttar and Gilachara of Fenchuganj Upazila, Sylhet, Bangladesh. Farmers were
selected randomly to prepare the sample.
2.2 Materials
Statistical Package for Social Sciences and Microsoft Excel were used for statistical analysis.
Page | 1378
2.3 Method
An action plan was constructed so that logical and sequential progress can be made throughout
the study. The action plan shown in Fig.1 describes the stages involved in this study.
Fig 1: Action plan of this research

An idea was generated to work on determining the prevalence of musculoskeletal disorders
among farmers. Some research papers about MSD and farmers were reviewed. The search terms
were ‘musculoskeletal disorder’ or ‘musculoskeletal symptom’ and ‘farmer’ or ‘agriculture’.
2.4 Questionnaire formulation
A two-part questionnaire was made by following the objectives of this research. The first part of
the questionnaire featured questions about the demographic data. The second part included
questions about musculoskeletal issues experienced by the selected farmers.
2.5 Sample size
Slovin’s formula was used to estimate sample size from a given population size and a margin of
error. The farmers in these three selected villages namely Ashighar dakkin, Ashighar uttar and
Gilachara of Fenchuganj Upazila were approximately 400. So, by calculating with a confidence
level of 95 percent and rounding to a whole number, the sample size was 200.
1. Calculation
The most prevalent musculoskeletal effects for different operations (ploughing, weeding,
fertilizing and cutting) of rice cultivation are identified from the following bar graph.
60% 54% 53%

47%
41% 37.50%
40%
21% 25%
20% 15%
0%
Ploughing Weeding Fertilizing Cutting
Lower Back Pain Hand Pain Wrist Pain
Fig 2: Most occurred musculoskeletal disorders during different operations of rice cultivation
Page | 1379
Fig. 2 shows that the most prevalent pains experienced by farmers during ploughing were lower
back pain (41%) and hand pain (21%). The most frequent pain in weeding was lower back pain
(47%) which was followed by hand pain (37.5%) according to Fig. 2. The most occurred pains
during fertilizing were hand pain (54%) and wrist pain (15%) as indicated by Fig. 2. The most
prevalent pain during cutting was lower back pain (53%) which was followed by hand pain (25%)
according to Fig. 2.
According to the hypothesis testing with a 95% confidence of interval, when p value is smaller
than 0.05, it indicates that the null hypothesis is rejected, that is, there is a significant relationship
between the variables (factors and pains). When p value is greater than 0.05, it indicates that the
null hypothesis is accepted, that is, there is no significant relationship between the variables
(factors and pains). Chi Square test was used in this study to calculate the p value.
Table 1: Chi Square test between the demographic factors and the most prevalent pains during ploughing
and weeding.
Ploughing Weeding
Factors
p-value for p-value for p-value for p-value for
lower back pain hand pain lower back pain hand pain
Age 0.552 0.023 0.118 0.015
Height 0 0.316 0.462 0.734
Weight 0 0.198 0.687 0.003
According to table 1, age is a significant factor (p=0.023) for occurring hand pain while height
(p=0.00) and weight (p=0.00) are responsible for associating lower back pain during ploughing.
Age (p=0.015) and weight (p=0.003) are significant factors for occurring hand pain during
weeding as indicated by table 1.
Table 2: Chi Square test between the demographic factors and the most prevalent pains during fertilizing.
Fertilizing
Factors
p-value for wrist pain p-value for hand pain
Age 0.49 0.087
Height 0.109 0.101
Weight 0.133 0.004
From table 2, it can be seen that weight is a significant factor (p=0.004) for occurring hand pain
during fertilizing. According to table 3, height (p=0.038) and weight (p=0.00) are significant
factors for associating hand pain during cutting.
Page | 1380
Table 3: Chi Square test between the demographic factors and the most prevalent pains during cutting.
Cutting
Factors
p-value for lower back pain p-value for hand pain
Age 0.848 0.403
Height 0.87 0.038
Weight 0.848 0

The findings of the study are as follows: During ploughing, the most occurred MSDs were lower
back pain (41%) and hand pain (21%). Height (p=0.00) and weight (p=0.00) of farmers are the
significant factors that are responsible for occurring lower back pain while age (p=0.023) is
responsible for occurring hand pain during ploughing. In weeding, lower back pain (47%) and
hand pain (37.5%) were the most prevalent pains. No demographic factor is responsible for
occurring lower back pain while age (p=0.015) and weight (p=0.003) are responsible for
occurring hand pain during weeding. During fertilizing, the most reported MSDs were hand pain
(54%) and wrist pain (15%). The weight (p=0.004) of the farmer is responsible for occurring hand
pain during fertilizing. During cutting, the most occurred MSDs were lower back pain (53%) and
hand pain (25%). No demographic factor is responsible for occurring lower back pain while
height (p=0.038) and weight (p=0.00) are responsible for occurring hand pain during cutting.
From this research, it can be concluded that lower back pain and hand pain are the most common
pains during the cultivation of rice. Most of the time, weight is responsible for occurring hand
pain during different operations. Demographic factors (height and weight) are found responsible
for associating lower back pain, only during ploughing.
3. Conclusion
The high prevalence of musculoskeletal effects (lower back pain, hand pain and wrist pain) among
the farmers involved in rice cultivation highlights an urgent need for ergonomic interventions to
reduce exposure to musculoskeletal risks. Age, height and weight of farmers influenced the risks
for musculoskeletal effects in different body regions, suggesting that both individual and work-
related factors should be considered in the assessment and treatment of such complaints.
References
Bhattarai, D., Singh, S. B., Baral, D., Sah, R. B., Budhathoki, S. S., & Pokharel, P. K. (2016).
Work-related injuries among farmers: a cross-sectional study from rural Nepal. Journal of
Page | 1381
Occupational Medicine and Toxicology, 11(1), 48. https://doi.org/10.1186/s12995-016-
0137-2
Bihari, V., Kesavachandran, C., Mathur, N., Srivastava, A., & Pangtey, B. (2011).
Musculoskeletal pain and its associated risk factors in residents of national capital region.
Indian Journal of Occupational and Environmental Medicine, 15(2), 59.
https://doi.org/10.4103/0019-5278.90375
Fathallah, F. A. (2010). Musculoskeletal disorders in labor-intensive agriculture. Applied
Ergonomics, 41(6), 738–743. https://doi.org/10.1016/j.apergo.2010.03.003
Ghasemkhani, M., Mahmudi, E., & Jabbari, H. (2008). Musculoskeletal Symptoms In Workers.
International Journal of Occupational Safety and Ergonomics, 14(4), 455–462.
https://doi.org/10.1080/10803548.2008.11076784
Hartman, E., Oude Vrielink, H. H. E., Huirne, R. B. M., & Metz, J. H. M. (2006). Risk factors
for sick leave due to musculoskeletal disorders among self-employed Dutch farmers: A
case-control study. American Journal of Industrial Medicine, 49(3), 204–214.
https://doi.org/10.1002/ajim.20276
International Labour Office. (2011). Safety and health in agriculture : ILO code of practice.
International Labour Office.
Jain, R., Meena, M. L., Dangayach, G. S., & Bhardwaj, A. K. (2018). Association of risk factors
with musculoskeletal disorders in manual-working farmers. Archives of Environmental &
Occupational Health, 73(1), 19–28. https://doi.org/10.1080/19338244.2017.1289890
Osborne, A., Blake, C., McNamara, J., Meredith, D., Phelan, J., & Cunningham, C. (2010).
Musculoskeletal disorders among Irish farmers. Occupational Medicine, 60(8), 598–603.
https://doi.org/10.1093/occmed/kqq146
Palmer, K. T., & Goodson, N. (2015). Ageing, musculoskeletal health and work. Best Practice &
Research Clinical Rheumatology, 29(3), 391–404.
https://doi.org/10.1016/j.berh.2015.03.004
Rosecrance, J., Rodgers, G., & Merlino, L. (2006). Low back pain and musculoskeletal symptoms
among Kansas farmers. American Journal of Industrial Medicine, 49(7), 547–556.
https://doi.org/10.1002/ajim.20324
Naeini, H. S., Karuppiah, K., Tamrin, S. B., & Dalal, K. (2014). E rgonomics in agriculture: An
Approach in Prevention of Work-related Musculoskeletal Disorders (WMSDs). Em Journal
of Agriculture and E nvironmental Sciences (Vol. 3, Issue 2). Online.
Stankevitz, K., Schoenfisch, A., de Silva, V., Tharindra, H., Stroo, M., & Ostbye, T. (2016).
Prevalence and risk factors of musculoskeletal disorders among Sri Lankan rubber tappers.
International Journal of Occupational and Environmental Health, 22(2), 91–98.
https://doi.org/10.1080/10773525.2016.1168073
Welch, R. M., & Graham, R. D. (2004). Breeding for micronutrients in staple food crops from a
human nutrition perspective. Journal of Experimental Botany, 55(396), 353–364.
https://doi.org/10.1093/jxb/erh064
Woolf, A. D., & Pfleger, B. (2003). Burden of major musculoskeletal conditions. Bulletin
of the World Health Organization, 81(9), 646–656.
Page | 1382
PAPER ID: 330
CHARACTERIZATION OF NANOSTRUCTURE SEDIMENTARY

CLAY MATERIAL OF NETROKONA DISTRICT
T. M. S. A. Hossain1*, M. Hasanuzzam2, M. A. Hakim2 and M. F. Islam2
1*
Department of Mechanical Engineering (ME), University of Global Village, Barishal-8200
2
Department of Nanomaterials and Ceramic Engineering (NCE), Bangladesh
University of Engineering and Technology (BUET), Dhaka-1000, Bangladesh
1*
E-mail: Tarazul.gce@gmail.com
Keywords: Abstract: Clay is an essential industrial mineral which often needs

improvement for particular applications. Due to huge demand in ceramic
 Nanostructure
industry, nanostructure clay materials can be modified to use as advanced
materials,
materials such as porous filtration, heat resistance insulators and other
 kaolinite, applications. The excellent alternating layered crystal sheet structure of clay
advanced minerals allows to comprehensively modifying its properties. In this
applications research, local clay was chosen as potential material and its properties are
thoroughly investigated. Chemical composition analysis by X-ray
 minerals, fluorescence revealed that the clay contains Al2O3 (26.67 %) and SiO2 (66.32
 microstructure %) as major oxides and impurity oxides include mainly Fe2O3 (2.01%) and
and TiO2 (2.47 %) with acceptable amount. Mineralogical study employing X-ray
sedimentary diffraction (XRD) exhibited mainly kaolinite and quartz phases with minor
clay contents of illite and rutile phases. Field Emission Scanning Electron
Microscopy (FE-SEM) displayed the morphological features of clay mineral
structural. Particle size of kaolinite platelets varied from < 2 μm. In addition,
particle shape of clay structure was also discerned with size spans from
50−500 nm. EDX analysis also gives the elemental analysis of nanostructure
sheet. The local sedimentary clay resembled white clay type characteristics
though XRD contradicts with well-crystalline structure of kaolinite.
Differential thermal analysis (DSC- TG) confirmed phase transition occurs
with increasing temperature. The obtained results from various
characterization techniques provide an in- depth understanding of its
fundamental characteristics to suit it for applications.
1. INTRODUCTION
Clay nanostructures materials are a group of hydrous aluminium silicates with a repeating
layered structure which ensue considerable demand as advanced materials in diversified
applications. It is the expression of very small, crystalline particles dominated by planar arrays
of [SiO4]4- structural units and many structural hydroxyls and water of the clay minerals,
kaolinite has shown profound industrial applications such as ceramic raw materials, filler
pigment and coating for paper, filler for paint, formulation of medicine/cosmetics and catalyst
manufacture. Kaolinite is composed mainly of silica sheets [Si2O5]2- linked to modified gibbsite
sheets [Al2 (OH) 4]2- [1]. The silica sheet is called tetrahedral layer due to tetrahedral shape of
the [SiO4]4- group (each side is 0.26 nm). The oxygens/hydroxyls are arranged to form the
corners of octahedra (each side is 0.29 nm) having six corners and eight faces in the gibbsite
sheet. In the atomic lattices of kaolinite, one unit length (basal spacing) is 0.72 nm. In kaolinite,
a gibbsite sheet is constructed directly over a silica sheet in such a manner that one in three of
the OH radicals are replaced by the unsaturated apical oxygens of the silica sheet. The apical
Page | 1383
oxygens thus form a bridge between the two sheets resulting a composite unit layer of kaolinite,
Al2Si2O5(OH)4. An outstanding feature of kaolinite is that in gibbsite sheet only two out three
plausible sites are occupied by Al3+ keeping one vacant and the resulting structure is called
dioctahedral. A kaolin crystal is consisted of a large number of composite layers (typically 70 to
100 layers). The entire crystal is united together by hydrogen bonds between OH- radicals of
dioctahedral layer and oxygen atoms of adjacent tetrahedral layer. In kaolinite, the layers are
stacked so that the atoms in one silicon layer are displaced regularly along the a-axis. The unit
cell is triclinic with   91.80,   104.50,   90 and a  5.16 Å, b  8.94 Å and c  7.37 Å
[2].
Clays are broadly classified in two types: a) residual clay and 2) sedimentary clay. Residual
clays (commonly known as China clays) are located near to source altered igneous rocks and
obtained in a purest state of the mineral kaolinite. In general, mineralogical composition shows
80-95 % kaolinite and 5-15 % mica (muscovite). The remainder is quartz, paragonite, and trace
amount of montmorillonite, different oxides of magnesium, calcium and iron. The content of
silica and alumina is near to 46.5 % and 39.5 % respectively of pure mineral. The iron content
(as Fe2O3) is between 0.5-1.2 % and the alkalis are less than 2 %. In general, the total impurities
of China clay amount to about 5-15 %. A typical composition of China clay is: SiO2 (46 %),
Al2O3 (38 %), Fe2O3 (0.3 %), TiO2 (0.03 %), CaO (0.6 %), MgO (0.6 %), Na2O (0.15 %), K2O
(0.5 %). In contrast, sedimentary clays are transported by natural agencies and thus impurities
are picked up and retained during their transportation. Fine impurities may render difficulty to
remove. The main sedimentary clays are ball clay, fire clay and brick clay. The chief mineral of
sedimentary ball clay is disordered kaolinite with main impurities as quartz and mica (illite) and
minor impurities as oxides of iron and titanium and organic matter. Various deposits of clays
were identified by Geological Survey of Bangladesh (GSB) namely Bijoypur in Maymenshing,
Maddhyapara and Barapukuria in Dinajpur, Patia in Chittagong, Sylhet [3]. However, no
systematic complete study has been carried out to characterize these deposits for its suitability
in applications as advanced materials such as porous fiber, structural heat insulation etc. In this
context, for the first time a detail cutting-edge-technology based characterization and analysis
was conducted to probe its composition, mineralogical test, crystallinity, phase changes,
morphology (size and shape) kaolinite nanoparticles.
2. EXPERIMENTAL
Geological Survey of Bangladesh (GSB) indicated that Bangladesh has deposits of sedimentary
types of clays [4]. Based on this report a number of deposits across the country have been
identified as the sources of various types of clay. Important deposits are Bijoypur in
Mymensingh, Barapukuria and Maddhyapara in Dinajpur. In this research work, Durgapur
(Bijoy-07) clay collected from Netrokona district which is commonly known as Durgapur
(Bijoy-07) clay in that area of deposits is used for characterization and beneficiation. Durgapur
clay is of high quality and used predominantly by the ceramic, paper and rubber industry of the
country. These regions are totally sediment area. The clay deposits are found in different river
side areas. Generally, the clay is light greyish white to bluish white with light yellow and
slightly soapy to feel, massive and soft to medium hard. The colour of Durgapur (Bijoy-07) clay
was found to be whitish yellow. The clay was collected in necessary amount and prepared for
characterization technique. The aim of the present research is the investigation the properties of
Page | 1384
clay and their thermal behaviour. The identification and quantification of natural clay were
conducted by Field Emission Scanning Electron Microscopy (FESEM), X-ray Fluorescence
Spectroscopy (XRF), and X-ray Diffractometry (XRD). Clay minerals often adopt the form of
mixed-layer structure with various ratios of individual components. Thus, IR spectra and
Differential Scanning Calorometry (DSC)/Thermogravimetric (TG) analysis were also
conducted.
3. RESULT AND DISCUSSIONS

In this research to expose Durgapur clay as prospective advanced materials, detail analysis was
conducted employing several experimental tools. Morphology of clay kaolinite in particular was
determined employing Field Emission Scanning Electron Microscopy. X-ray fluorescence was
used for compositional analysis of clay materials. X-ray diffraction will lead to an
understanding of the structural characteristics of the clay minerals. Thermal behaviour and
weight loss of clay sample were studied by Differential Scanning Calorometry to obtain
information on phase changes. Obtained results are presented below under each technique.
A. Optical Microscopy
The theoretical resolution of optical microscopy is about 0.2 m. The length scale of clay
particles are below this value thereby rendering limited information from optical microscopy of
clays. However, overall size (and shape) of clay aggregates and its colour can valuably be
discerned from it. Fig. 1 shows two clay chunks of which one is broken into two parts. The
color of clay was found to be whitish. A tint of iron-rich mineral has shown to exist along with
this clay.
Fig 1: Optical image of Bijoy-07 natural clay

B. X-ray Fluorescence Analysis
To determine the percentage of oxides present, X-ray fluorescence (XRF) was employed as
widely used versatile tool for elemental and chemical analysis of clays. Basically, XRF emits
characteristic secondary x-rays from a material that has been excited by bombarding with high-
energy x-rays.
Page | 1385
Table 1 shows the composition (major) of (Bijoy-07) clay. The clay has found to be Al2O3 rich
than other natural clays obtained from different areas.
TABLE 1: COMPOSITIONAL ANALYSIS OF CLAY
Oxide SiO2 Al2O3 Fe2O3 TiO2
Weight (%) 66.32 26.67 2.01 2.47
The clay also contains a small amount of CaO (0.15), MgO (0.38), ZrO2 (0.23), P2O5 (0.06),
Cr2O5 (0.05) and K2O (1.56). The presence of iron and titanium oxide bearing materials may
impair the usefulness of clay [5].
C. X-ray Diffraction Analysis
To quantitatively analyze minerals present in natural clays, X-ray diffraction (XRD) is an
indispensible tool. The high intensity X-ray beam (PAN ANALYTICAL EMPYREAN) was
diffracted on clays within a scanning range (2) from 10 to 700 with CuK radiation. Fig. 2
presents a XRD pattern of natural clay. Kaolinite and quartz are found to be major mineral
constituents of clay. Rutile was detected as main titanium bearing impurities along with small
amount of Illite.
Fig. 2: XRD pattern of Bijoy-07 natural clay

Obtained two strong reflections are well matched with the standard prominent basal reflections
at 12.390 and 24.930. The peaks were found to be sharp representing a well crystalline kaolinite
[10-11]. To determine quantitative percentage of these minerals Rietvelt refinement was
conducted using HIGH SCORE PLUS Software. Table 2 shows the quantitative mineralogical
analysis of major minerals. In addition to this list, calcium sodium aluminium silicate and
trimagnesium dioxide dihydroxide were also detected as 4.6 % and 0.7 % respectively.
TABLE 2: RIETVELT ANALYSIS
Minerals Kaolinite Quartz Illite Muscovite
Wt (%) 46.0 23.4 21.6 3.7
Page | 1386
D. Secondary Electron Microscopy Analysis
To reveal the approximate size and shape of clay particles, Scanning Electron Microscopy
(SEM) can be employed with its ability to obtain very high resolution and magnified image of
individual grains of clay minerals. In this context, Field Emission Scanning Electron
Microscopy (FESEM) (Model: JEOL JSM-7600F) was used.
(a) (b)
Fig. 3: SEM micrographs of Bijoy-07 natural clay showing: (a) kaolinite hexagonal sheet plates
and (b) kaolinite particles stacking.
Samples were scanned for secondary electrons (SE) mode for morphological contrast and back
scattered electrons (BSE) mode for phase contrast [9]. Fig. 3 shows the FESEM micrographs of
nanoparticles of natural clay. High magnified images revealed both plate (Fig. 3(a)) and particle
shaped (Fig. 3(b)) clay particles.
E. Energy Dispersive X-ray (EDX) Analysis

Energy Dispersive X-ray (EDX) can provide valuable information on major and minor elements
allowing highly reliable compositional identification [6].
Fig. 4: EDX spectra of Bijoy-07 natural clay
Page | 1387
Basically, EDX is used for qualitative as well as quantitative (the percentage of the
concentration of each element of the sample) analysis. Fig. 4 shows captured EDX spectra from
clay sample. Evidently peaks of major elements namely Si and Al with allied oxygen were
identified. Table 3 shows summary of elemental analysis of clay in atomic (and mass)
percentage.
Al and Si contents were found to be higher in conformity with compositional analysis of XRF
(Table 1). Except detected minor elements namely Mg (K), K (K), Ti (K), Cr (K), and Mn (K)
concentration of major elements are only presented. In analyzed clay, the ratio of Al to Si is
lower around half. In addition, some minor elements namely Ti, Ca, Mg, Mn, Fe, K, Fe and Cr
are also identified in clay sample.
TABLE 3: EDX ELEMENTAL ANALYSIS OF NATURAL CLAY
Element O(K) Al(K) Si(K) Fe(K)
Mass (%) 39.31 19.98 38.34 6.40
Atom (%) 53.26 16.05 29.59 0.40
F. FTIR Analysis
The structural differences of clay can be ascertained by vibration spectroscopic investigations
yielding useful information about hydration characteristics, interlayer cations and moisture
content in clays. Fig. 5 shows infrared spectra of the typical associated minerals present in
clays. The band at 533 cm−1 shows deformation vibration of Si–O and Si–O–Al whereas the
band at 469 cm−1 indicates the presence of amorphous silica.
Fig. 5: FTIR spectrum of Bijoy-07 natural clay

Bands observed at 780-798 cm-1 is due to Si-O-Si inter tetrahedral bridging bonds in SiO2.
Bands at 1620 and 2642 cm–1 could be assigned to H-O- H bending of water, which is
observed in almost all the natural hydrous silicates like illite minerals. The absorption bands
observed at 1813 cm-1 is related to carbonate and at around 2358 cm-1 related to calcite. The
bands between 3450 and 3670 cm–1 are attributed to the OH stretching mode [7-8]. The band at
3620 cm-1 may be ascribed to the inner hydroxyls and the other three characteristic bands are
generally ascribed to vibrations of the external hydroxyls.
Page | 1388
G. Differential Scanning Calorimetric (DSC) Analysis
Differential thermal analysis is an important method to detect reaction due to dehydroxylation
of clay minerals, decomposition of carbonates, loss of combined water, loss of sulphur and
decomposition of organic matter [12-13].
Fig. 6: DSC and TG curves of Bijoy-07 natural clay

Fig. 6 shows DSC and TG curves of clay analyzed at room temperature to 1400oC depicting the
effect of energy changes (endothermic or exothermic reactions) and weight changes in the
sample. A first endothermic reaction was observed at temperature below 100oC indicating
desorption of surface H2O (e.g. H2O on exterior surfaces) and dehydration (e.g. interlayer H2O)
at low temperatures. Kaolinite shows a weight loss starting at just above 400oC due to
dehydroxylation, extending to about 650oC (TG-curve of Fig. 6). A significantly high
endothermic peak observed at 456oC is related to the formation of meta-kaolinite (Al2Si2O7).
Quartz transformation has shown to take place at 573oC. Next, an exothermic peak appeared at
1000oC is related to the formation of mullite (Al6Si2O13) phase. An allotropic phase transition
from quartz to cristobalite was detected by a small exothermic peak at 1200oC.
4. CONCLUSION
We successfully characterized the sedimentary deposit of local clay. The investigated clay
contains silica and alumina as predominant oxides and iron, titanium, zirconium, calcium,
potassium and magnesium oxides are present as minor constituents. The major minerals include
kaolinite, quartz and illite with a trace of rutile mineral. Obviously, clay sample contains some
key impurities namely iron, titanium oxides that may be removed by magnetic separation
technique. The characteristic four bands (al bait small though) of kaolinite were confirmed from
IR spectroscopy. The nanosized kaolinite sheet plate and particles would allow this material to
suit its properties for applications as an advanced material for porous ceramic filter and
structural heat insulators.
Page | 1389
ACKNOWLEDGMENT
The authors greatly acknowledge the use of lab facilities in GCE, BUET. This research was
fully financially supported by Higher Education Quality Enhancement Project (HEQEP), UGC,
and Government of the Peoples Republic of Bangladesh under University-Industry
Collaboration Project CP-3823.
REFERENCES
[1] W. E. Worrall , "Clay and Ceramic Raw Materials", Springer, 2012
[2] J. G. Thompson and C. Cuff, "Crystal structure of kaolinite: dimethylsufoxide intercalate",
Clays and Clay Minerals, Vol. 33, No. 6, 490-500, 1985.
[3] A. H. Dewan, S. Mustafi, M. Ahsan and M. S. Ullah, "Investigation on physical properties
of patia clay (Chittagong), Bangladesh", Bangladesh J. Sci. Ind. Res. 49(4), 255-262, 2014.
[4] Lee V-G., Yeh T-H., Materials Sci. and Eng.: A 485: 5-13, 2008.
[5] Mahmoudi S., Srasra E., Zargoune F., Applied Clay Science, 42: 125-129, 2008.
[6] Misra M.L and Hummel F.A., Bull. Amer. Ceram. Soc., 28:235, 1949.
[7] Huber J. M., “Kalolin clays and their industrial uses”, Corporation, 100 Park Ave., N.Y. 17,
U.S.A., 1956.
[8] Ralph E. Grim, “Clay Mineralogy” (New York; McGraw Hill Book Co., 1968.
[9] P. Sengupta, P. C. Saikia and P. C. Borthakur, "SEM-EDX characterization of an iron-rich
kaolinite clay", Journal of Scientific and Industrial Research, Vol. 67, pp. 812-818, 2008.
[10]Moore DM and Reynolds RC Jr (1997) “X-ray Diffraction and Identification andAnalysis
of Clay Minerals”, 2nd edn. New York: Oxford University Press.
[11]Klug, H. P. and Alexander, L. G. (1972) “X-ray Diffraction Procedures for
Polycrystalline and Amorphous Materials”: John Wiley & Sons, London.
[12] Rolad and Joseph Clark, J. Am. Ceram. Soc., 10: 98, 1927.
[13] Stull R. T. and Bole, G. A., J. Am. Ceram. Soc., 6: 730, 1923.
Page | 1390
PAPER ID: 340
Pseudo-random number generation-a comparative study

Polash Kumar Guptho, Abul Mukid Md. Mukaddes, and Nahidul Haque Abir
Department of Industrial and Production Engineering
Shahjalal University of Science and Technology
Email: mukaddes-ipe@sust.edu, polashguptho@gmail.com and nahidulabir22@gmail.com.
Keywords: Abstract: Random numbers are essential part of analyzing a system. It helps to include
uncertainty in the system. There are various methods available to generate pseudo-
 Random number random numbers. In this paper, we have proposed a new version of LCG (Linear
 Simulation Congruential) method and compared our method over different random number
generation techniques, i.e., LCG method, Mid square method, and WH (Wichmann Hill)
 RNG method method. We tested the performance, uniformity and processing time required to run the
 Analyzer code by each method which are major prerequisites of the RNG (Random Number
 Performance Generation) method. Our proposed modified LCG method performs better than LCG,
Mid-square and WH methods in terms of repetitions cycle period. Moreover, we have
found some valuable insights from the comparative study.
1. Introduction
Random events are uncertain and unpredictable, and random numbers are the element that defines
randomness. By using specific computational algorithms to generate random numbers, the system's
randomness can be incorporated. This algorithm or approach is known as a pseudo-random number generator
[1]. The important pre-requisites or properties that any RNG (Random Number generation) methods should
be followed are [2] –
 They should generate uniformly distributed numbers between the continuous range zero and one.
 The number they generate should be as independent as possible of one another.
 The generator must be fast and should not require excessive computer memory.
 The generator should have a long cycle before their sequence repeats.
Khoshnevis and Behrokh (1994) explained two different methods such as LCG and Mid –square method [3].
E G Sesari et al (2019) used LCG method for Randomization of Test Item in Computer-Based Psychological
Edwards Personal Preference Schedule (EPPS) [4]. Although these two methods are simple and fulfill the
criteria of random number formation, they cannot be able to generate a large number of random numbers
bounded by some factors (e.g., c, m) moreover have a zero-repetition problem. K. Malik, J. Pulikkotil, and
A. Sharma (2021) compared different methods, including the above two methods (LCG & mid square
method). Based on some tests, they evaluate the performance of those methods. That paper introduces a new
method called “WH (Wichmann Hill) method”. After performing all the tests, they claim that the WH method
performs better than all the other methods [5]. B. Wichmann and D. Hill (1982) also claimed that WH method
is reasonably short, fast, machine-independent, and statistically sound [6]. Fransiskus Panca Juniawan et al
(2019) compared Two data randomization algorithms (Linear Congruent Method and Fisher-Yates Shuffle).
Comparison was done to determine the performance between the two algorithms.[7]
However, in this paper, we have proposed a new modified version of the LCG method and compare its
performance, uniformity, and processing time required to run the code over the other methods. The code was
developed using the C language and run in the C platform. The modified LCG method shows better
performance in terms of repetition.
Page | 1391
2. Materials and Method
Materials:
In this paper, we have used C-programming, Arena’s Input Analyzer tool, Notepad, and Excel for
developing, comparing, and validating the RNG methods.
1. C Programming: C program is helpful to iterate the model and gives the repetition cycle period quickly.
2. Arena’s Input Analyzer: This tool helps us to check the uniformity of random numbers.
3. Notepad: We made the text file of random numbers to import them into Arena's input analyzer.
4. Excel: We have validated our C-code using Excel software.
Method:
The considered random number generation method in this paper are-
I. Mid-square method:
The mid-square method is the simplest method for generating pseudorandom number. Although this method
generally has very poor performance, the simplicity of its algorithm makes it suitable for demonstration of
the pseudorandom number generation concept. In this method, only one parameter is needed to be called the
seed [3]. The working procedure of the Mid-square method is shown in Fig-1.
II. Linear Congruential Method (LCG):
A more widely used method for pseudorandom number generation is the linear congruential method. This
method uses the following recursively evaluated equation with a random number generated at each iteration:
Zi = (a*Zi-1+c) mod m; ri = Zi / m
where Z0 is the seed. Parameters a, c, and m, as well as the seed, are nonnegative integers and must satisfy 0
< m, a < m, c < m, and Z0 < m [3]. The working procedure of the LCG method is shown in Fig-2.
III. WH method:
The WH generator is an extension of LC generator, as it uses three linear congruential equations with different
values of the moduli of the form of LCG equation i.e.,
Xin+1 = (ai * Xin) * mod(mi); for i = 1, 2, 3
Here Xi0; (i = 1, 2 and 3) are the three seed values. The values generated by each of these equations are
summed, and their modulo with 1 is taken to get the final value between 0 and 1 [5]. The working procedure
of the WH method is shown in Fig-3.
IV. Modified Linear Congruential Method:
This is our purpose method of random number generation. It is a modified version of the LCG method.
We have used the equation of the LCG method for generating random numbers i.e.,
Zi = (a*Zi-1+c) mod m; ri = Zi / m
When Zi = Zi-1 , then ri = 1- Zi / m and iteration goes on. The working procedure of our modified LCG method
is shown in Fig-4.
Page | 1392
Fig. 1: Flow chart of Mid-square method Fig. 2: Flow chart of Linear Congruential method
Fig. 3: Flow chart of WH method Fig. 4: Flow chart of modified LCG method.
In this paper, we have checked performance, uniformity and time required to complete the iterations by
each method.
Page | 1393
3. Results and Discussions
3.1 Repetition cycle test
We have compared the performance of all these methods using C programs. The result is given in below-
Table 1: Random number repetition cycle period of each method
Method Parameter RN repetition cycle period Ranking
1. LCG method Seed=19, a=3, c=1, m=30000 1000 3
2. Mid-square Only one parameter, Seed=19 6 4
3. WH method Seed=19,21,18, a=3,2,4, 40000 2
m=30000,31000,29000
4. Modified LCG Seed=19, a=3, c=1, m=30000 As user needed (No-repetition) 1
method
3.2 Uniformity test

3.2.1 LCG method
If we take Seed=19, a=3, c=1, m=30000 and analyze the RN data generated by LCG method, we can see that
the data follow Uniform distribution. Note that, here, RN values are between 0 and 1.
Number of iterations
Fig. 5: Uniformity test of LCG method.
3.2.2 Mid-square method
If we take a 3-digit seed = 465 and analyze the random number generated by Mid-square method, we will
observe that the data follow Uniform distribution. Note that, here, RN values are between 0 and 1.
Fig. 6: Uniformity test of Mid-square method.
Page | 1394
3.2.3 WH method
If we take- a1 17 a2 4 a3 3
m1 11 m2 44 m1 60
and analyze the RN data generated by WH, we can see that the data follow Uniform distribution. Note that,
here, RN values are between 0 and 1.
Fig. 7: Uniformity test of WH method.
3.2.4 Modified LCG method
If we take Seed=19, a=3, c=1, m=30000 and analyze the RN data generated by LCG method, we can see that
the data follow Uniform distribution. Note that, here, RN values are between 0 and 1.
Fig. 8: Uniformity test of Modified LCG method.
Therefore, we can say that all these methods pass the Uniformity test.
3.3 Processing time test
Table 2: Processing time required by each method
Iteration=1000 times
Method 1st run 2nd 3rd run 4th run 5th run Average Rank
(s) run (s) (s) (s) time
(s) (s)
Mid-square 3.056 3.041 3.058 3.102 3.037 3.0588 1
LCG method 3.058 3.053 3.06 4.399 3.65 3.444 2
Modified LCG method 4.37 4.38 3.06 3.525 3.56 3.779 3
WH method 3.88 3.55 4.447 4.41 4.437 4.145 4
Although our modified LCG method takes longer time as compared to Mid-square & LCG method, but it
has performed better than WH method.
Page | 1395
3.4 Discussions
After conducting performance, uniformity & processing time requirement test, we have got the following
results:
Table 3: Results obtained by all tests
Method Performance test Rank Uniformity Processing Rank
(Repetition cycle period) test time taken
(s)
LCG method 1000 3 Pass 3.444 2
Mid-square 6 4 Pass 3.0588 1
method
WH method 40000 2 Pass 4.145 4
Modified LCG As user needed (No-repetition) 1 Pass 3.779 3
method
From the above table, we can see that, Although the mid-square method takes the least processing time, Our
Modified method performs better in performance test and also follow the uniform distribution.
4. Conclusion
In this paper, we have compared different RNG methods in terms of performance, the processing time is
taken, etc., and tried to develop a new random number generation method that would allow us to analyze a
system in a better way. After conducting all the tests, we have got some key insights such as-
1. Modified LCG method performs better than the WH method.
2. WH method takes more run time than the other methods.
3. The modified LCG method obeys uniformity criteria like the other RNG methods.
4. It can generate an infinite number of random numbers.
References
[1] Budiman, A., Bulolo, E., & Saputra, I. (2020). Middle Square Method Analysis of Number Pseudorandom
Process. The IJICS (International Journal of Informatics and Computer Science), 4(2), 35-38.
[2] Averill M.Law, W.Dabid Kelton (2003). Simulation modeling and analysis. McGraw-Hill Higher
Education
[3] Khoshnevis, B. (1994). Discrete systems simulation. McGraw-Hill College.
[4] Sesari, E. G., Dirgantoro, B., & Setianingsih, C. (2019, May). Linear Congruential Method for
Randomization of Test Item in Computer-Based Psychological Edwards Personal Preference Schedule
(EPPS) Test. In Journal of Physics: Conference Series (Vol. 1201, No. 1, p. 012054). IOP Publishing.
[5] Malik, K., Pulikkotil, J., & Sharma, A. (2021). Comparison of pseudorandom number generators and their
application for uncertainty estimation using Monte Carlo simulation. Mapan, 36(3), 481-496.
[6] Wichmann, B. A., & Hill, I. D. (1982). An efficient and portable pseudo‐random number generator.
Journal of the Royal Statistical Society Series C, 31(2), 188-190.
[7] Juniawan, F. P., Pradana, H. A., & Sylfania, D. Y. (2019, March). Performance comparison of Linear
Congruent method and Fisher-Yates Shuffle for data randomization. In Journal of Physics: Conference Series
(Vol. 1196, No. 1, p. 012035). IOP Publishing.
Page | 1396
PAPER ID: 341
Mechanical characterization of novel sandwich structures

made of perlite/epoxy core with JFRP facings
Hossain, G. M. I., Karua, P. and Arifuzzaman*, M.

● Department of Mechanical Engineering, Khulna University of Engineering & Technology,
Khulna-9203, BANGLADESH
● *Corresponding author [Md Arifuzzaman]
Abstract: A novel sandwich panel is fabricated using expanded

Keywords: perlite/epoxy-based core and jute fiber reinforced polymer (JFRP) as
skins. The core is prepared following the compression molding technique
● Sandwich composites;
and JFRP skin is made separately following the hand layup method. The
core density of the sandwich panels is varied during manufacturing to
● Expanded perlite;
investigate the effect of density on the mechanical properties of the
sandwich structure. The prepared panels were subjected to flexural and
● Flexural properties;
compressive tests to investigate the strength, modulus, and failure
behavior under flexural and compressive loading. The results show that
● Compressive properties; the flexural and compressive properties are highly dependent on the core
density. The failure sequence during flexural loading is core cracking
● Failure behavior.
followed by skin delamination and finally skin tearing. The occurrence of
skin delamination is highly dependent on the core density. The failure
behavior during compression provides a clear indication of the significant
load-bearing and energy absorption capacity of the sample even after
crack initiation.
1. Introduction
Sandwich composites are vastly used in engineering applications nowadays. In aerospace,
automobiles, and offshore constructions, the use of sandwich panels is growing. This is because
the sandwich panels are inherently lightweight, can withstand high loads, and has relatively high
flexural strength and rigidity (Hassouna et. al, 2022). Sandwich composites are a special class of
composites formed by attaching two thin sheets of higher strength on both sides of a lightweight
core. The core material is usually a low-strength material, but the increased thickness gives the
sandwich composite an overall mechanical stiffness (Fazzolary, 2017). Such materials include
gypsum, perlite, particle straw, different metal matrices, etc. (Doleželová et. al, 2017), whereas
skin materials can be of different types, including steel, aluminum, carbon fiber reinforced
polymers, fiberglass reinforced polymers, and fiber reinforced polymers (Masuelli, 2013).
Synthetic fibers like carbon, glass, Kevlar, etc. are used frequently as the skins of sandwich panels
for better strength and rigidity of the overall structure (Gupta et. al, 2021). But considering the
biodegradability and cost-effectiveness, natural fibers are preferred by many researchers
Page | 1397
(Jeyapragash et. al, 2020). Jute fiber, itself holds a great strength and jute fiber reinforced polymer
can be an alternative to those synthetic fibers. It is bio-friendly as well as a cheap material. While
dealing with the core, researchers introduce metals, metallic fibers, gypsum, expanded perlite
composites, etc. (Karua and Arifuzzaman, 2022; Arifuzzaman and Kim, 2017a; Arifuzzaman and
Kim, 2017b; and Soloveiko et. al, 2018). Different matrices are used to prepare the core for
example sodium silicate solution (Karua and Arifuzzaman, 2022; Arifuzzaman and Kim, 2017a;
and Arifuzzaman and Kim, 2017b), epoxy (Allameh-Harey et. al, 2017), starch (Shastri and Kim,
2014), recycled polyester (Al Abir et. al, 2020), etc. Since the expanded perlite is a glassy volcanic
rock and holds excellent thermal insulation and fire-resistant properties (Sing and Garg, 1991), it
can be an excellent candidate to be used as a core for designing the multifunctional sandwich
structure. The epoxy resin may be used as the matrix for manufacturing expanded perlite-based
core for better strength. Jute fiber reinforced polymer may be used as skins of the sandwich
structure considering easy availability, low cost, biodegradability, etc.
Therefore, in this work, the expanded perlite/epoxy-based core and JFRP skins are used to prepare
the sandwich panels. The prepared sandwich panels are characterized for their compressive and
flexural properties for various core densities along with the failure mechanism.
2. Experimental details
2.1 Materials
Perlite is an amorphous volcanic glass with a relatively high water content, usually formed by the
hydration of obsidian rock (Sing and Garg, 1991). It has the unusual property of occurring
naturally and expanding significantly when heated sufficiently. It is an industrial mineral and is a
useful product after processing due to its low density. It has a bulk density of about 65–75 kg/m3
(Pavlik and Bisaha, 2018) and extremely low thermal conductivity with high insulation property
up to a temperature of 650 °C (Pavlik and Bisaha, 2018). In this study, the expanded perlite
particles were bought from King Caster, China with a particle size range of 4-5.6 mm. The bulk
density of the EP particles was measured to be 66–71 kg/m3.
The two most common jute fiber mats available in Bangladesh are white jute and Tossa jute.
Tossa jute is softer, silky, and stronger than white jute (Islam, 2019). That is why fiber mats
prepared with Tossa jute were used in this study for skin preparation. The mats were collected
from Ajax Jute Mill, Khulna. The mass of the mat was measured to be 253.33 g/m2.
Epoxy was used as the matrix for the preparation core and skins. The mixture of resin
and hardener cause a chemical reaction, which transforms the binding agents into a solid
state within 24 hours. In this study, Araldite AW-106 and HV-953 were used as resin and hardener
respectively to prepare the skin, and a crystal-clear AB glue resin hardener was chosen for the
core preparation.
2.3.1 Specimen preparation
The compression molding technique was adopted to manufacture the core material. Firstly, the
epoxy resin (AB glue) was mixed with hardener maintaining a weight ratio of 3:1 as
recommended by the manufacturer. The prepared resin mix was then poured into expanded perlite
in a container and mixed properly using a plastic spoon for 5 minutes. Before pouring the wet
Page | 1398
perlite into the mold, an anti-sticking agent was applied to the bottom and walls of the mold so
that the prepared core could be easily removed. The mold was designed in such a way that the
final thickness of the prepared core becomes 15 mm. A 10 kg mass was placed on the top of the
top plunger to ensure no spring back occurs. The density of the perlite/epoxy core was varied by
increasing the quantity of perlite. The densities of the prepared perlite/epoxy panels were found
to be 0.61, 0.68, and 0.75 g/cm3.
(a) (b) (c)
(d) (e) (f)

Figure 1: Sandwich composite fabrication process, (a) addition of adhesive to the skin, (b) placement of
lower skin into the mold, (c) placement of core into the lower skin, (d) placement of upper skin, (e) applying
force on top of the top skin to ensure proper contact between the skin and the core, and (f) fabricated
sandwich composites after 24 hours of curing
The JFRP skins were prepared using the hand layup method (Karua and Arifuzzaman, 2021).
After the completion of skin manufacturing, a layer of adhesive (epoxy resin and hardener) was
added to the skins. Then, the skins were attached on both sides of the expanded perlite/epoxy core
to prepare the sandwich panels. The preparation technique is shown in Figure 1.
2.3.2 Mechanical testing
Flexural and compression tests were carried out using a Universal Testing Machine (Shimadzu
AGX-300kN) following the ASTM C393 and ASTM C365 standards respectively. The crosshead
speed for both tests was 5 mm/min. For the flexural test, the samples were cut to the size of 150
mm in length and 20 mm in width for flexural tests, and 20 mm × 20 mm for compression tests
from the fabricated sandwich panels. Three samples from each category (core densities 0.61, 0.68,
and 0.75 g/cm3) were tested for both flexural and compression tests. At least three samples were
tested for each test from each category.
Page | 1399
3. Results and Discussions
3.1 Flexural properties
The values of the flexural strength (FS) and modulus (FM) of the sandwich composites are shown
as a function of core density in Figure 2. As expected, the FS and FM increased with the increasing
core density of the sandwich panels. The least-square line for flexural strength and core density
relationship is found to be y = 118.51x - 62.433, with a correlation coefficient, R² = 0.9941 [see
Figure 2 (a)] and for flexural modulus the least-square line is y = 3464.5x - 1297.1 and R² =1 [see
Figure 2(b)]. The high correlation coefficients are indicating the high linearity of flexural
properties with the core density.
30 1500
Flexural strength, MPa
Flexural modulus, MPa

25 y = 118.51x - 62.433 y = 3464.5x - 1297.1
R² = 0.9941 1300 R² = 1
20
1100
15
900
10
5 700
0.60 0.70 0.80 0.60 0.70 0.80
Core density, g/cm3 Core density, g/cm3
(a) (b)
Figure 2: (a) Flexural strength and (b) modulus of the sandwich composites for different core densities
(a) (b)
Figure 3: (a) Typical load vs displacement curves and (b) photographs of different failures during the
flexural test (A, B, and C showing core shear, delamination, and skin tearing respectively).
Typical load-displacement curves under flexural loading are shown in Figure 3 along with the
photographs taken during the test showing the specimen failure. The curves appear to have similar
characteristics, showing a gradually increasing linear portion before the first peak where the first
failure has taken place and the load starts to drop with displacement. The first failure may be
Page | 1400
denoted as core shear failure (Arifuzzaman and Kim, 2017a) as shown in Figure 3(b)-A. Once the
core shear cracking is finished the load starts to increase again up to a second peak where the
delamination of the skin takes place as shown in Figure 3(b)-B. After that, the load is carried by
the skins until the bottom skin tears as shown in Figure 3(b)-C. Therefore, the failure sequence
during the flexural test is identified to be core shear cracking-skin delamination-bottom skin
tearing. It is interesting to see that the beginning of skin delamination after core shear cracking is
significantly affected by the core density i.e. the skin delamination of the sandwich composites
for high-density core takes place at a shorter displacement after the shear cracking than the low-
density cores.
3.2 Compressive properties
The compressive strength (CS) and modulus (CM) of the sandwich composites are plotted as a
function of different core densities in Figure 4. Similar to the flexural properties, the CS and CM
also increased with the increasing core density of the sandwich panels.
10
500
y = 1699.2x - 867.68
Compressive strength, MPa
Compressive modulus, MPa

y = 43.214x - 24.819 400 R² = 0.9671
R² = 0.9946
6 300
4 200
2 100
0 0
0.60 0.65 0.70 0.75 0.80 0.60 0.65 0.70 0.75 0.80
Core density, g/cm3 Core density, g/cm3
(a) (b)
Figure 4: (a) Compressive strength and (b) modulus of the sandwich composites for different core densities.
Before
After failure
(a) (b)
Figure 5: (a) Typical load vs displacement curves and (b) photographs of the specimen before and after the
failure during the compression test.
Page | 1401
The least-square line for compressive strength and core density relationship is found to be y =
43.214x - 24.819 with a correlation coefficient, R² = 0.9946 [see Figure 4(a)] and the least-square
line for the compressive modulus and density relationship is found to be y = 1699.2x - 867.68
with a correlation coefficient, R² = 0.9671 [see Figure 4(b)]. Again the high correlation
coefficients are indicating high linearity between compressive properties and core density.
Typical load-displacement curves under compressive loading are shown in Figure 5 along with
the photographs taken during the test showing the specimen before and after failure [see Figure
5(b)]. The curves appear to have similar characteristics, showing a gradually increasing linear
portion, then a slight drop indicating the failure and after that a long plateau region before
densification. The sandwich composites with high-density core showed high load-bearing
capacity as expected. The specimen under compressive load showed considerable bulging in the
middle of the core before the initiation of several cracks with random orientations [see Figure 5
(b)]. The long plateau region in the load-displacement curves indicates a significant load-bearing
and energy absorption capacity of the sandwich composites even after the crack initiation.
4. Conclusion
In this research, sandwich composites were fabricated using expanded perlite/epoxy core and
JFRP as skins. The prepared composites were characterized for flexural and compressive
properties along with the failure mechanism. The major findings of the work can be summarized
as –
 The flexural and compressive properties of the sandwich panels are significantly dependent
on the core density. The relationship between the properties and the core density is highly
linear.
 The flexural failure sequence is identified to be core shear cracking – skin delamination –
bottom skin tearing. The occurrence of skin delamination is dependent on the core density
in such a way that the skin delamination for the sandwich composite with a high-density
core takes place at a smaller displacement after the shear cracking than the sandwich
composite with a low-density core.
 The failure nature during compression provides an indication of the significant load-
bearing and energy absorption capacity of the sample even after crack initiation.
Acknowledgments
The Authors would like to acknowledge Khulna University of Engineering & Technology,
Bangladesh for the partial financial support and lab facilities for this research work.
References
Al Abir, A., Faruk, M. O., Arifuzzaman, M., Novel expanded perlite based composite using
recycled expanded polystyrene for building material applications, International Conference on
Mechanical, Industrial and Energy Engineering. 2020, 19-21 December, 2020, Khulna,
Bangladesh.
Allameh-Haery H, Wensrich CM, Fiedler T, et. al., Novel cellular perlite-epoxy foams: Effects
of particle size, Journal of Cellular Plastics. 2017, 53(6); 597-622.
Page | 1402
Arifuzzaman, M., Kim, H. S., Novel flexural behaviour of sandwich structures made of perlite
foam/sodium silicate core and paper skin, Construction and Building Materials. 2017a, 148;
321-333.
Arifuzzaman, M., Kim, H. S., Flatwise compression and flexural behaviour of perlite/sodium
silicate composite foam, Applied Mechanics and Materials. 2017b, 860; pp. 19-24.
Doleželová, M., Krejsová, J., Vimmrová, A., Lightweight gypsum based materials: Methods
of preparation and utilization, International Journal of Sustainable Development and
Planning. 2017, 12(2); 326-335.
Fazzolari, F. A., 2017. Sandwich structures. In Stability and Vibrations of Thin Walled
Composite Structures (pp. 49-90). Woodhead Publishing, Sawston. UK.
Gupta, M. K., Ramesh, M., Thomas, S., Effect of hybridization on properties of natural and
synthetic fiber‐reinforced polymer composites (2001–2020): A review, Polymer Composites.
2021, 42(10); 4981-5010.
Hassouna, S., Allah, M. J., Timesli, A., 2022. Crashworthiness applications of the composite
sandwich structures. In Sandwich Composites (pp. 321-348). CRC Press. London. UK.
Islam, M. M., Varietal advances of jute, kenaf and mesta crops in Bangladesh: A review. Int J
Bioorganic Chem. 2019, 4(1); 24-41.
Jeyapragash, R., Srinivasan, V., Sathiyamurthy, S. J. M. T. P., Mechanical properties of natural
fiber/particulate reinforced epoxy composites–A review of the literature, Materials Today:
Proceedings. 2020, 22; 1223-1227.
Karua, P., Arifuzzaman, M., Compressive behavior of perlite/sodium silicate composite foam
modified by boric acid, Metallurgical and Materials Engineering. 2022, 28(1); 103-124.
Karua, P., Arifuzzaman, M., Influence of fiber orientation on the tensile and flexural properties
of jute fiber reinforced polymer composites, International Conference on Industrial &
Mechanical Engineering and Operations Management. 2021, Dhaka, Bangladesh.
Masuelli, M. A., 2013. Introduction of fibre-reinforced polymers− polymers and composites:
concepts, properties and processes. In Fiber reinforced polymers-the technology applied for
concrete repair. IntechOpen. London, UK.
Pavlík, V., Bisaha, J., Lightweight mortars based on expanded perlite. Key Engineering
Materials. 2018, 776; 104-117).
Singh, M., Garg, M., Perlite-based building materials—a review of current
applications. Construction and Building Materials. 1991, 5(2); 75-81.
Shastri, D., Kim, H.S., A new consolidation process for expanded perlite particles,
Construction and Building Materials. 2014, 60; 1-7.
Soloveiko, S., Pulkis, K., Skujans, J., et. al., Composite sandwich-type panel made of foam
gypsum, In Proceedings of the International Scientific Conference. 2018, Volgograd, Russia.
18-20.
Page | 1403
PAPER ID: 384
Analysis and Optimization of Fabric Defects Through the

Application of Lean Tools
Tanvir Ahmed1*, and Md Mostakinur Rahman Chowdhury

1
Shahjalal University of Science and Technology, 2 Square Textile Division
Email: tanviripe30@gmail.com, m.r.chowdhury.sust.ipe@gmail.com
*Corresponding author [Tanvir Ahmed]
Abstract: Readymade garment (RMG) industry is making great contribution

Keywords: to rebuild the economy of Bangladesh where 83% of total export earnings
come from it. One of the major wastages in this industry is “Cut Panel
● Readymade Rejection” due to fabric defects. The purpose of this study is optimizing and
Garment; reducing fabric defects by applying lean tools. Data of the fabric defects were
distributed with U chart to show the defect variation. After analyzing total six
● Fabric defects;
month cut panel rejection data due to fabric defects, six major defects are
● Lean tool; identified by pareto analysis. “Five Whys”, a lean tool, is applied to find out
the causes and effects of the fabric defects. Further, the application of the
● QC chart;
remedies made the cut panel rejection within 5% by the application of 5S,
● 5S; Logical solution and Control chart which was almost 6.5% previously. This
study was conducted by applying specific lean tools combined with QC tool
● 5Why.
which can be vary with other lean tools.
1. Introduction:
The motto “Made in Bangladesh”, pride of the nation, is commonly seen on the tag of readymade
garments around the world. Readymade garment is the top foreign currency earning source of
Bangladesh. The journey of the RMG industry was started in 1978 when Reaz Garments exported
its first consignment in USA and earned 69 thousand USD. But the actual milestone was led by
the Desh Garments Ltd established in 1979, the first 100 percent export-oriented company [1]. At
present, Bangladesh is the second largest exporting country of readymade garment [2].
Despite having on a leading position, Bangladesh is facing challenges to keep it because of
different challenges. To compete with the close competitor, industry should have more concern
about productivity and quality. The most important aspect of increasing productivity is reducing
wastages. The major percentage of wastages occur at cutting process because of fabric defects. A
huge amount of fabric throws away as wastage. Fabric defect is the flaws of fabric standard
quality requirements.
Page | 1404
To reduce wastage and improve quality control in industry an effective method is using lean tools.
Lean manufacturing has minimized waste and in turn reduced costs by applying reforms
introduced by “Toyota” in the 1990s, and from which the technical term “lean” was derived and
became popular [3]. The main goal of a lean manufacturing system is to produce products of
higher quality at the lowest possible cost and in the least time by eliminating wastes [4].
2. Methodology:
To minimize cut panel rejection quantity due to fabric defect particular lean tools are used in this
study. This study was occurred based on the seventh month fabric rejection data of a renowned
RMG industry of Bangladesh. Quality control charts are used to analyze and identify fabric
defects which help to control and reduce wastes. Control chart is the display of numerical and
graphical details to represent the sample variation from the standard with by using sample
distribution. There are different types of control chart available for measuring sample data such
as R chart, X bar, S chart, U chart, Moving range chart, Part chart etc. In our study we used U
chart for analyzing defects percentage range and Pareto chart to identify major defects. Also, lean
tools such as Root Cause Analysis, 5S, 5 Why also applied to get an effective result.
3. Data analysis
3.1 U Chart Analysis
U chart is used to identify defects of different sizes of sample based on unit. In this study w
considered daily fabric rejection quantity of six months. By using U chart, the sample data of the
rejection were taken for understanding is it in control or out of control. The quantity of rejection
during cutting are demonstrated below U charts:
Fig 1. U chart of rejection quantity (November-April)
Page | 1405
From the figure 1 it can be easily understood that the rejection quantity is out of control where
Upper Control Limit of the defects was 7.11% and Lower Control Limit of the defects was 5.73%
where, mean of the defects was 6.42%. It is Concluded that, if per day defects between 7.11% to
5.73% are acceptable & Defects are in control.
3.2 Pareto Chart Analysis
A basic quality tools Pareto Chart used to identify most frequent defects. It is a vertical bar chart.
To find out the major fabric defect types pareto chart is used in this study.
Fig 2. Pareto chart of defects (November-April)

From the pareto analysis six major defects were identified: Hole – 32.7%, D. Need-12.8%, Cut –
10.8%, Dirty spot – 7.6%, Slub – 6.7%, Join 5.6%.
3.3 Root Cause Analysis
After analyzing the root cause of the defects and applying six sigma tools following remedies
were found for the abnormality:
3.3.1 Hole
Major reasons for creating holes are High Yarn Tension, Yarn Overfeed or Underfeed, High
Fabric Take Down Tension, Obstructions in the yarn passage, clogging of eyelets, yarn guides &
tension discs, with wax & fluff, Incorrect gap between the Dial & Cylinder rings.
Remedies for reducing defects are ensure uniform yarn tension on all the feeders, with a Tension
Meter, Rate of yarn feed should be strictly regulated, as per the required Stitch Length, the fabric
tube should be just like a fully inflated balloon, not too tight or too slack, the yarn being used,
Page | 1406
should have no imperfections, like slabs, naps & big knots etc., The gap between the cylinder &
the dial should, be correctly adjusted, as per the knitted loop size.
3.3.2 Drop Needle
Reasons behind this defect are High Yarn Tension, count variation, mixing of the yarn lots,
package hardness variation, bend needle.
Remedies for decreasing drop needle are ensuring uniform yarn tension on all the feeders, the
average count variation in the lot, should not be more than (+-) 0.3, ensuring that the yarn being
used for knitting is of the same lot / merge no., ensuring that the hardness of all the yarn packages
is uniform, using a hardness tester, all needles should be straight.
3.3.3 Cut
Causes of cut defects are excess amount of fabric cuts during dyeing operations, excess amount
of babric cuts during GSM, Test & Shrinkage test, carelessness of workers.
Need more counseling, avoid unnecessary fabric cuts, management level should have needed to
Gemba practice to avoid this defect.
3.3.4 Dirty Spot
Unclean knitting machine bed, unclean stenter machine, poor chemical use in dyeing and
finishing, uncover fabric batch after each operations finished are the cause of dirty spot
Maintaining 5S on the production floor is most effective solution for this defect.
3.3.5 Knot/Slab
Causes of knot/slab are tying spools of yarn together, high yarn tension, count variation, mixing
of the yarn lots, package hardness variation.
Maintaining uniform Yarn Tension on all the feeders, the average count variation in the lot, should
not be more than (+-) 0.3, ensuring the yarn being used for knitting is of the same Lot / Merge
no. are useful for solving this issue.
3.3.6 Join
High rate of yarn lot, batch mixing by overlock sewing for fabric slitting make join.
Decreasing the change of yarn lot as more as can, cuts join during fabric final inspection
can help against join problem.
3.4 Five Why Analysis
Five why is lean tool of identifying causes of a problem by asking questions “Why?” for five
times. The answers to the questions help to identify the causes of defects. From Five Why analysis
the following causes were found:
various types of defects found in fabric cutting section, not properly inspecting fabric after fabric
finishing, shade problem in dyeing section due to poor fabric knitting, scheduling & chemicals,
Page | 1407
poor fabric knitting occurred due to poor machine operating & destitute yarn quality, poor supply
chain & logistic support in procurement division.
4. Result and Discussion

After analyzing the data and identifying the causes of defects by applying lean tools related
remedies were implemented to lessen the cut panel quantity. A fruitful output was found within
four months and target was achieved after the implementation of lean tools. Quality control chart
of the sample defects after the implementation was given below:
0.056
0.054
0.052
Sample Count Per Unit
0.050
U CL=0.04829
_
0.048
U=0.04757
LCL=0.04685
0.046
0.044
0.042
0.040
1 9 17 25 33 41 49 57 65 73
Sample
Tests performed with unequal sample sizes
Fig 3. U chart of defects (June-August)

From the fig 3. it is clear that cut panel rejection quantity is under the limit means defects are in
control. The average rejection quantity is 4.75%.
This study showed a significant change in cut panel quantity due to fabric defects by applying
lean tools. Further this study can be improved by using other lean tools.
5. Conclusion
Cut panel rejection because of fabric defects increase the cost of the garments, which affects the
benefit of the industry. Day by day price of the raw materials and other costs are increasing. So,
to keep the products profitable it is important to eliminate unnecessary costs. Lean manufacturing
is the keyway to reduce the costs like decreasing wastage. Without the application of lean tools,
it’s tough to keep the leading position for Bangladesh competing with other countries.
6. References
Page | 1408
[1] S. Zaman, "Researching the garment sector in Bangladesh: fieldwork challenges and
responses," in Field Guide for Research in Community Settings, Edward Elgar Publishing,
2021, pp. 75-84.
[2] M. A. R. A. A. M Saiful Islam, "Ready-made garments sector of Bangladesh: Its

contribution and challenges towards development," Stud, vol. 5, no. 2, 2016.
[3] L. C. N. H. M. M. A. Majo George, "Lean Concept in Fashion and Textile Manufacturing,"

in Lean Supply Chain Management in Fashion and Textile Industry, Springer, 2022, pp.
67-94.
[4] S. K. J. Shaman Gupta, "A literature review of lean manufacturing," International Journal
of Management Science and Engineering Management, pp. 241-249, 2013.
Page | 1409
PAPER ID: 386
Effect of Adhesive Layer Thickness and Adherent Geometry

on the Tensile Properties of Adhesively Bonded T- Joint
Raju Ahammad*1, Md. Omar Faruk2, Arup Kumar Debnath3, Md. Shariful
Islam4
1,2,3,4
Department of Mechanical Engineering, Khulna University of Engineering & Technology,
Khulna- 9203, Bangladesh (rajuahammad@me.kuet.ac.bd, omarfaruk1605078@gmail.com,
akdebnath@me.kuet.ac.bd, msislam@me.kuet.ac.bd)
*Corresponding author [Raju Ahammad]
Abstract: Adhesive bonding is the process of uniting materials with the aid
Keywords: of an adhesive; its development accelerated acutely over the last century,
especially in the aeronautic industry. Similar or dissimilar metals can be
● Adhesive; bonded by adhesive. T-joint is one of the oldest joining techniques. This work
● Adherent; aims to investigate the impact of adherent shape and adhesive layer thickness
on the tensile properties of adhesively bonded T-joints. By using the two-part
● FEM;
epoxy adhesive Araldite 2011, two stainless plates of steel were joined. Four
● Tensile Load; specimens (0.5, 1.0, 1.5, and 2.0) mm were tested by the Finite Element
Method (FEM). Numerical Analysis was conducted in ABAQUS/CAE 6.14-
● Failure Load;
5 software. The numerical result shows, a T-joint specimen with an adhesive
● Stress. thickness of 1 mm yields maximum tensile load. Failure load found 1.085 KN
for adhesive thickness 1 mm. The minimum failure load was 0.847 KN for an
adhesive thickness of 2.0 mm. Different adherent geometry like rectangular,
square, cylindrical, elliptical, tapered, and fillet geometry was analyzed to
investigate the effect of tensile property on T-joint. The stress-volume ratio
was determined to compare different geometry. Square geometry withstands
maximum tensile load and elliptical geometry minimum. Stress to volume
ratio for elliptical geometry was 0.47 MPa/mm3 and for square geometry is
0.08 MPa/mm3.
1. Introduction
Adhesive bonding is one of the oldest assembly processes. There are many types of adhesive
joints which are butt joint, lap joint, single layer joint, double layer joint, and T-joint, etc. T-joint
means right-angle joint to each other by forming the shape of the letter T. The adhesive joint can
provide uniform distribution of stress and large bearing-stress areas. The adhesive joint provides
a smooth surface finish compared to a welded joint. Adhesive joints are mainly used where
welding operation is very difficult.
Many researchers investigated how adhesive thickness affected T-joints that were adherently
bonded. The influence of adhesive thickness on an adhesively bonded T joint is studied by A. R.
Page | 1410
Abdullah, Mohd Afendi, and M. S. Abdul Majid (Abdullah et al., 2013). The failure load of the
T joint was determined experimentally. Four specimens were taken where two steel plates are
bonded by two-part epoxy with different adhesive thicknesses. T-joints that have triangular blocks
on either side, were investigated by Khalili and Ghaznavi (Khalili & Ghaznavi, 2011). T-joint
specimens with adhesive cleats were built by Guo and Morishima (Guo & Morishima, 2011).
They looked into various composite sandwich T-joint designs and failure mechanisms. Silva and
Adams (da Silva et al., 2009) tested T-joints with thin and thick sides glued together with different
stiffener combinations. Major analytical models for T-joint were experimented with in this
research. Elastic adhered and adhesive, elastic adhered with nonlinear adhesive, and nonlinear
analyses for both adhered and adhesive were the three main scenarios that were taken into
account. The impact of a T-joint on bond-line length was investigated by Hu et al. (Hu et al.,
2012). The goal of this work was to examine the adhesive-bonded T-joints' load capacity under
tension loads. The paper includes an experimental investigation into the impact that the bond-line
length has on these T-joints built of steel and aluminum. The effect of the geometry of the
specimens on the strength of flexible adhesive joints was studied by WR Broughton, E Attanz,
and RD Mera (Broughton et al., 2001). When measuring the strength of single-lap and scarf joints
joined with elastomeric adhesives, specimen geometry is taken into account. The influences of
the material and geometric properties on the effectiveness of adhesive T-joints were studied by
Le Crocker and WR Broughton (Crocker & Broughton, 2002). In this experiment, the strength of
single-lap and scarf joints joined with rubber-toughened adhesives was examined with specimen
shape and the adherend's elastic characteristics. This represented the results of the effects of
adhesive layer thickness, and adhered properties. Epoxy adhesive lap joints with adhesive
thicknesses of 0.5 mm, 1.3 mm, and 2.4 mm were explored by Kant and Narayanan (Kant &
Narayanan, 2012). Here single lap joint was studied which was bonded by adhesive. Four
different thickness specimens were used to investigate the effect of thickness. Yi Wang and When
Yi Wang and Hongguang LIU (WANG et al.) examined the breakdown process analysis of T-
joints under tensile load using DIC experiment and numerical simulation, they looked at the
distribution of the strain of the stress-concentration zone of a Carbon Fiber Reinforced Polymer
(CFRP) T-shaped joint.and observed that the triangle region always undertakes the most serious
strain concentration. Numerical and experimental research was conducted by Madjid Mokhtari,
Morteza Shahravi, and Mahmood Zabihpoor (Mokhtari et al., 2018) on the development of
dynamic behavior of the innovative composite T-joints where they investigated dynamic behavior
of novel sandwich T-joint. For joining purposes and to create the step-by-step behavior of the
adhesive region, an epoxy adhesive was chosen.
This present study examined adhesively bonded T-joints that were 0.5, 1.0, 1.5, and 2.0 mm thick.
Two-part epoxy Araldite 2011 was used to unite the adherent SUS 304 stainless steel plates. The
tensile test was conducted numerically by the Finite Element Method (FEM) using simulation
software ABAQUS/CAE 6.14. The effect of adhesive layer thickness on the tensile property of
the T-joint was investigated. The effect of different adherent geometry on the tensile properties
of the T-joint was also analyzed.
2. Computational Modelling
2.1 Materials property
SUS 304 stainless steel as adherent and Araldite 2011 as an adhesive is used for this numerical
analysis. The modulus of elasticity and poison ratio of SUS 304 stainless steel and Araldite 2011
are 210000 MPa, 1600 MPa, and 0.30, 041 respectively.
Page | 1411
2.2 Computational details
SUS 304 stainless steel plate, 10 mm thick in the upper adherent and 1.25 mm in the lower
adherent, is used to simulate the model. Araldite 2011 is used for forming a T-joint between them
which is acted as an epoxy adhesive. The height of the upper adherent is 100 mm and the width
is 10 mm as shown in Fig.2(a). Adhesive and adherent parts are assembled, the upper and lower
parts are stainless steel and the middle part is adhesive Araldite 2011, after that the three parts are
combined into one part. The assembled part is shown in Fig.1(b). For the initial step, the boundary
condition applied at the lower surface of the lower part is fixed and at the upper surface of the
upper part, the load is applied as shown in Fig.1(c).
Fig.1 (a) T-joint specimen; (b) Assemble of arts; (c) Loading and boundary condition applied
on the model
A8-node C3D8R and an 8-node linear brick are used for meshing the part. The part was meshed
by 155640 elements. Double bias is used by the number of elements 15 as shown in Fig.2(a).
Mesh sensitivity is evaluated for a more perfect result. A graph of stress vs the number of elements
was plotted as shown in Fig.2(b). From the graph, it is found that from element 17080 to 47800
stress increase very rapidly, then from number 63228 to 145130 stress increase almost linearly.
It shows almost constant after element number 155640 as shown in Fig.2(b). So, element number
155640 was taken to simulate the model.
(a) (b)
Fig.2 (a) Meshing of the model; (b) Von-Mises stress vs the number of elements
Page | 1412
3.1 Model validation
The model was validated by comparing the present work and reference paper works as shown in
Fig.3. There was little deviation between the present numerical analysis and reference paper.
Fig.3 Failure load vs thickness of the adhesive layer graph

3.2 Effect of adhesive layer thickness
In this study, the effect of adhesive layer thickness on the tensile property of the T-joint was
investigated. Two steel plates were joined by Araldite 2011 adhesive as shown in Fig.1(b). Four
specimens of adhesive thickness (0.5, 1.0, 1.5, and 2.0) mm were analyzed by fixing the lower
surface, and the load was applied at the top surface as shown in Fig.1(c). From numerical analysis,
it was observed that with an increase in adhesive layer thickness, failure load also increased as
shown in Fig.3. After reaching a maximum value, failure load decreased slightly with an increase
in adhesive layer thickness. So, the failure load increased to the optimum thickness and then
reduced slightly with adhesive thickness according to Kinloch’s theory (Kinloch & Kinloch,
1987), a thicker adhesive layer will result in increased failure load and optimum thickness was
found 1 mm.
(a) (b) (c) (d)

Fig.4 Stress contour of thickness (a) 0.5, (b) 1.0, (c) 1.5, (d) 2.0 mm
Page | 1413
3.3 Effect of adherent geometry
Different adherent geometry was used to investigate the effect of tensile property on the T-Joint
as shown in Fig.5.
Fig.5 Stress contour for different adherent geometry
Table 1 Comparison between different adherent geometry
Applied force Stress/Volume

Geometry Volume(mm3) Stress(MPa)
(kN) (MPa/mm3)
Rectangular 80 10.61 0.13
Cylindrical 78.50 11.08 0.14
Elliptical 47.10 22.41 0.47
Square 100 8.22 0.08

1
Fillet 88 10.61 0.12
Tapered 90 11.39 0.13
Rectangular, cylindrical, elliptical, square, tapered, and fillet geometry of different adhesive
thicknesses were analyzed. A fixed load of 1 kN was applied in all geometry. For comparison,
stress to volume ratio was determined as shown in Table 1. From this work, it is found that the
elliptical geometry has a higher value of 0.47 MPa/mm3 for stress to volume ratio and square
geometry has a lower value of 0.08 MPa/mm3. So, square geometry can withstand the maximum
load. As a result, the square geometry is the optimum geometry.
4. Conclusion
In this study, the effect of adhesive layer thickness and the effect of adherent geometry were
investigated. Four specimens were tested in the Finite Element Method. Among four specimens,
an adhesive thickness of 1 mm could withstand the maximum load. It was found that an adhesive
thickness of 1 mm withstands 1.085 KN load. The lowest load was carried by a thickness of 2.0
mm. Failure load increased with thickness and after reaching an optimum value, it decreased
Page | 1414
slightly. For different geometry, stress to volume ratio was compared. The elliptical geometry has
a higher value of 0.47 of stress to volume ratio. So, it can be concluded that-
1. As adhesive thickness grows, the failure load rises until it reaches the optimum thickness.
Then the failure load decreases slightly.
2. Where the failure load was greatest, 1.0 mm of optimal thickness was found.
3. The most efficient adherent geometry is square geometry.
References
Abdullah, A., Afendi, M., & Majid, M. A. (2013). Effect of adhesive thickness on adhesively
bonded T-joint. IOP Conference Series: Materials Science and Engineering,
Broughton, W., Attanz, E., Mera, R., & Duncan, B. (2001). Effects of specimen geometry on the
strength of flexible adhesive joints.
Crocker, L. E., & Broughton, W. (2002). Geometric and material property effects on the
performance of adhesive T-joints.
da Silva, L. F., das Neves, P. J., Adams, R., Wang, A., & Spelt, J. (2009). Analytical models of
adhesively bonded joints—Part II: Comparative study. International Journal of Adhesion
and Adhesives, 29(3), 331-341.
Guo, S., & Morishima, R. (2011). Numerical analysis and experiment of composite sandwich T-
joints subjected to pulling load. Composite Structures, 94(1), 229-238.
Hu, P., Shao, Q., Li, W., & Han, X. (2012). Experimental and numerical analysis on load capacity
and failure process of T-joint: Effect produced by the bond-line length. International
Journal of Adhesion and Adhesives, 38, 17-24.
Kant, R., & Narayanan, R. G. (2012). In-plane testing behavior of adhesive-bonded steel sheets:
influence of sheet surface roughness, adhesive thickness, and bonding width. Journal of
Testing and Evaluation, 40(2), 292-301.
Khalili, S., & Ghaznavi, A. (2011). Numerical analysis of adhesively bonded T-joints with
structural sandwiches and study of design parameters. International Journal of Adhesion
and Adhesives, 31(5), 347-356.
Kinloch, A. J., & Kinloch, A. J. (1987). Adhesion and adhesives: science and technology.
Springer Science & Business Media.
Mokhtari, M., Shahravi, M., & Zabihpoor, M. (2018). Development of dynamic behavior of the
novel composite T-joints: Numerical and experimental. Advances in aircraft and
spacecraft science, 5(3), 385.
WANG, Y., LIU, H., MA, X., & XIONG, K. Failure process analysis of T-joints under tensile
load by DIC experiment and numerical simulation.
Page | 1415
PAPER ID: 393
Finite Element Analysis of Carbon Epoxy Composites

Pressure Vessels
Md Mafidul Islama, Md Foisal Hossainb*, Muhammed Sohel Ranac, and

Md Shafiul Ferdousd
a, b .
Dept. of Materials Science and Engineering., Khulna University of Engineering and
Technology, Khulna-9203, Bangladesh
b
*Dept. of Electronics and Communication Engineering., Khulna University of Engineering
and Technology, Khulna-9203, Bangladesh
c
Institute of Nuclear Science. and Technology, Bangladesh Atomic Energy Commission,
Savar, Dhaka-1349, Bangladeh
d
Frederick, Maryland-21702, USA
a
islam1727042@stud.kuet.ac.bd, b*foisal@ece.kuet.ac.bd, csohelcef@gmail.com,
d
munazeer_218@yahoo.com
Abstract: The main target of the present work is to preserve the structural
Keywords: integrity of the composite pressure vessels (PVs) used in storing high
pressurized liquid and gas under internal pressure loading. Composite
● Burst pressure; Overwrapped Pressure Vessels (COPVs) are widely used in many fields
● Optimization; including aeronautics and by companies such as Tesla, SpaceX to contain high
pressure fluids. Their lighter weight makes them applicable than all-metal
● Lamina; vessels, although they demand special design, manufacturing, and testing
● Mesh; requirements. A detailed design optimization of a composite over wrapped
pressure vessel (COPV) is performed through finite element analysis FEA
● Integrity; (ABAQUS). The PVs are made of carbon/epoxy with (0,45,90,0) lay-up of 4
● Overwrapped. ply. Here lamina property and mechanical properties are taken from prepreg
Carbon fiber cloth of T700. The work consists of three main cases. First is
changes in different kinds of stress like circumferential, axial and radial
stresses with respect to meshing. Second, the maximum burst pressure with a
certain internal pressure and third is to use the failure criteria to find the burst
strength of Composite pressure vessel.
1. Introduction
Due to its significance in numerous sectors and applications, pressure vessels (PVs) have been the
subject of extensive research. PVs frequently use composite materials to enhance their design and
produce a lighter, more controllable material behavior. The minimal thickness of the pressure
vessel under a certain internal pressure was determined by design and failure considerations of
four standard sizes of pressure vessels (SWRO PVs). The PVs were constructed from C/E
AS4/3501-6 and C/E IM7/977-3 composite material and stainless steel SST 316L [1].
Page | 1416
An internal pressure of 8 MPa loaded the PVs. For SST 316L PV, the von Mise criteria was used
and for composite materials, Tsai-Wu criterion was used [1]. There are two design case studies
were used in a computation of 14 models. The expected conditions that COPV models are to
experience in a performance test involve a composite overwrapped pressure vessel with an internal
pressure of 30 MPa. There are many parameters to assess the performance and here fiber
orientation and ply sequence were used as performance metrics for the produced COPVs. Also,
different fiber orientation at same thickness is used to work the impact of the ply stacking order
using FEA. The goal of the current work is to improve the type IV polymer lined composite
overwrapped pressure vessel's ASTM D2585 COPV design [2].
The key results are as follows:
i. In the majority of instances, cylindrical (hoop) portions show signs of tensile failure.
ii. The burst pressure determined by the maximum stress and strain criteria is more liberal
(less conservative) than that determined by the Hashin damage criterion.
iii. The composite ply thickness and winding angle have a significant impact on the burst
pressure.
iv. As hoop and polar angles increase, the burst pressure also rises.
In this work, the effects of damage evolution on burst pressure and auto-frettage pressure were
examined. The matrix fracture damage was found on the head of the vessel and the fiber tensile
damage was found on the center part of the cylinder when the cylinder was put under service
pressure. The findings indicate that the carbon fiber composite component weighs the least and is
equivalent to 17% of the corresponding steel component. The strong non-linearity between the
deterministic coefficient and the level of dependability is highlighted in terms of reliability, and
the component built of composite materials performs similarly to those constructed of isotropic
materials [4]. An analytical and numerical analysis in FEA were used to optimize the pressure
vessel design parameters. The in-plane damages increase steadily until a sudden catastrophic
failure that indicates the vessel's rupture appears [5]. The fiber orientation, lay-up sequence and
material type determine the thickness of the pressure vessel. The composite pressure vessel offers
lower weight compared to stainless steel pressure vessels. A 3D parametric finite element model
was used to analyze the damage profile and failure strength of composite storage tanks under a
certain internal pressure. The structural damage was determined based on four failure criteria
where the Tsai-Wu failure criteria was found to be the most accurate. Along with these, the
mechanical behavior of composite pressure vessels under internal pressure to failure was
investigated using the ABAQUS software. Some examples of the aerospace industry uses are
rocket propellant tanks and solid rocket motor casings and in automotive industry uses are high-
pressure fuel storage tanks for hydrogen-powered cars. Then the design is optimized of the
composite pressure vessel was studied by FEA. The main objective is to examining the design of
composite overwrapped pressure vessels to provide optimum variables with a focus on two key
challenges is the goal of the current work. (1) Performing FEA on a composite overwrap pressure
vessel that has been designed with a variety of layers and orientations to withstand a maximum
burst pressure, and (2) Examining the stress and damage characteristics of COPV designs with a
uniform fiber angle orientation and ply sequences.
2. Composite Pressure Vessel
Engineers have the option to create structures that are lighter and stronger than those made of bulk
materials thanks to composite materials. Engineers can create pressure vessels that are extremely
adjustable to specific requirements by changing the orientation of the fibers' angles. The longer
Page | 1417
lifetime of composites because they do not corrode is another benefit [3]. In this work, the
cylindrical shell revolution model of a pressure vessel was considered. The component has an
internal diameter of 200 mm, and length of 600 mm. The vessel is subjected to 10 MPa.
2.1 Finite element model of the composite vessel
The vessel consists of T700 carbon fiber reinforced epoxy composite layer. Specific geometry
parameter of the vessel is listed in Table 1. The finite element model is loaded with internal
pressure at the inner surface. Each composite layer is modeled using the three-dimensional
continuum shell element.
60
Axial stress, S11
Hoop stress, S22
Shear stress, S12
Stress (MPa)
40
20
0
0 2 4 6 8 10
Global Size
Fig. 1. Meshed vessel with global size 1 and 0.5 Fig. 2. Max. S11, S22 and S12 for different mesh
In the current finite element model, the carbon fiber reinforced epoxy composite is assumed to be
orthotropic and linear elastic. The mechanical properties of materials are listed in Table 1. we
performed a number of simulations to validate these parameters. Actually, these values were
determined by repeated calculations and comparing the simulation results with experiments, until
the most significant fit parameters were obtained. The strength parameters of the composite ply
are listed in Table 2.
Table 1. Mechanical Properties
Material Property Value Material Property Value
E1 138 GPa ν 23 0.4
E2 15 GPa G12 5.5 MPa
E3 15 GPa G13 5.5 MPa
ν 12 0.3 G23 5.5 MPa
ν 13 0.2
Table 2. Strength parameter
Properties Value (MPa)
Longitudinal Tensile strength 1950

Longitudinal Compressive strength 1500
Transverse Tensile strength 50
Transeverse Compressive strength 200
Shear strength 80
To improve the accuracy of the FEA simulation, MESH research was conducted. The aim of this
section, aside from demonstrating accuracy, is to choose a MESH that will provide the best
accuracies while at the same time having an acceptable computation time. However, as the model
continues to grow, the time spent running the job dramatically increases. The first meshes used
Page | 1418
had relatively few components, which meant that the work took significantly less time to complete.
However, the computational analysis produced by these meshes showed significant inaccuracies,
thus finer meshes were used to find the best compromise. At first the initial seed values of around
10 were chosen randomly. Then the following data involved global mesh sizes of 5, 3, 1, and 0.5.
The data from this work was deemed inconclusive and hence excluded from the final mesh analysis
because, when the global size approached 1, the element sizes were fairly large for the student
version of the software. In order to resolve the meshing problem, the stress value (S 11, S12, S22)
were taken for simulation. All these values are taken from visualization as the maximum value.
The time also taken to run the job to be affected by the computer activity to achieve the accuracy
when same computer used in the iteration. From the data obtained in Fig. 2, mesh size versus
various stress like axial, hoop and shear stresses are plotted to find best mesh sizes. It is important
to not to increase the mesh around a particular region. If the stress values were taken from a middle
element, steady results would have been expected unlike the drastic jumps as shown in the figure
for global sizes between 2 and 4. Again from the above graphs an approximate meshing of 3 will
be used throughout the work. Here it is important to say that analysis error for meshing is zero for
all the global sizes examined. So relatively lower time requiring global size is taken for which
satisfying result is also be found.
3. Numerical model and failure criteria
(a) (b)
(c) (d)
Page | 1419
(e) (f)
Fig. 3. (a) Maximum stress failure index (MSTRS) less than 1 at load step 0.9125, (b) Maximum stress
failure index (MSTRS) greater than 1 at load step 1.0, (c) Tsai Wu failure index (TSAIW) smaller than 1 at
load step 0.9125, (d) Tsai Wu failure index (TSAIW) greater than 1 at load step 1.0, (e) Tsai Hill failure
index (TSAIH) smaller than 1 at load step 0.9125, (f) Tsai Hill failure index (TSAIH) greater than 1 at load
step 1.0.
The failure criteria for this analysis are maximum stress, maximum strain, Tsai Hill, and Tsai Wu
and Hashin. In the following work, maximum stress, TsaiH and TsaiW are studied. Three of these
failure criteria are used to compare the resulting burst pressure. The internal pressure is applied
gradually (in boundary condition option) with a little load increments and the failure index is
observed. If the failure index is more than 1, then the applied load is considered the pressure at
which the COPV fails called burst pressure. The boundary conditions are assigned using the local
cylindrical coordinate system. The radial and circumferential displacements were fixed. The inside
face of the model was subjected to a homogeneous 18 MPa pressure.
The maximum stress failure indices (MSTRS) are shown in Fig. 3(a) for the step time of 0.9125.
The maximum stress failure index values are lower than 1 in all the areas of the model. So, this
step time of load is not taken as the burst pressure. In Fig. 3(b), the maximum stress failure indices
(MSTRS) at step time after 0.9125 is 1.0 are shown. The failure indices are more than 1 in the
large hoop areas (red). Therefore, this step time of load is taken as the burst pressure. As
maximum stress are calculated by multiplying the time step of failure and the applied stress,
therefore, The burst pressure at this step time as per maximum stress failure criterion is (18×1.0)
18 MPa. Next, at step time 0.9125 in Fig. 3(c), the Tsai Wu failure index (TSAIW) distribution
is displayed. Here, all model areas have TSAIW failure indices that are lower than 1. Therefore,
the load's step time is not used to determine the burst pressure. Fig. 3(d) illustrates the Tsai Wu
failure index (TSAIW) distribution at step time 1.0, and it reveals that the large hoop areas have
values greater than 1. (red). The burst pressure is determined by the load's step time. At this step
time, the burst pressure is (18×1) 18 MPa. Fig. 3(e) depicts the Tsai Hill failure index (TSAIH)
contour for a step time of 0.9125. All regions of the model have TSAIH failure index values that
are less than 1. As a result, the burst pressure is not determined by the load's step time. The values
in the large hoop portions of Fig. 3(f)'s Tsai Hill failure indices (TSAIH) at step time 1.0 are more
than 1. (red). The burst pressure is determined by this load's step time. At this step time, the burst
Page | 1420
pressure is (18×1) 18 MPa. Here it is found that both the failure indices MSTRS and TsaiH are
equally dominating for the failure of the pressure vessel and the index TsaiW is less effective
failure Criteria as it gives lower values. There also possible to see the Von Mise stress distribution
according with burst pressure. Which can explain that in which location the maximum stress
distributed or applied. On the other hand, it can be said that in which portion of the pressure vessel
will bear the maximum load as well as other portions load bearing profile is seen. Fig 4 shows
the Von Mises stress distribution at burst pressure.
5. Conclusion
Fig. 4. Von Mises stress distribution at burst pressure

The ABAQUS FEA model was able to simulate several conditions, including the effects of
gravity, temperature, and discovering the appropriate mesh values to be employed in the
construction of a specific pressure vessel. The accuracy was affected for particularly small mesh
size values because to the shifting mesh variables, but a compromise was made in order to obtain
the optimum values at a manageable running time. The goal was the current work is to observe
the burst pressure of composite overwrapped pressure vessels of a certain design. For this work,
prepreg carbon fibers and resin are considered. The maximum pressure containment capability is
found to be achieved by the model with a combination of ply winding (0°, 45°, 90°, 0°). Main
findings of this work are the distribution of burst pressure with a certain fixed parameter, the burst
pressure from MSTRS is greater than the TsaiW failure criteria, burst pressure distribution
changes a little with mesh refinement and failure is observed in the circumferential areas in most
cases.
References
[1] M. Halawa and N. Al-Huniti, “Optimum design of carbon/epoxy composite pressure vessels including
moisture effects,” Journal of Composites Science, vol. 3, no. 3, Sep. 2019, doi: 10.3390/jcs3030065.
[2] Y. Regassa, J. Gari, and H. G. Lemu, “Composite Overwrapped Pressure Vessel Design Optimization
Using Numerical Method,” Journal of Composites Science, vol. 6, no. 8, Aug. 2022.
[3] O. Mouhat, A. Bybi, A. el Bouhmidi, and M. Rougui, “Effects of ply orientation on nonlinear buckling
of Aircraft composite stiffened panel,” Frattura ed Integrita Strutturale, vol. 13, no. 50, pp. 126–140, 2019,
doi: 10.3221/IGF-ESIS.50.12.
[4] A. Mukuvare, “Pressure Vessel Analysis Computational Fluid Dynamics View project Solid Mechanics
View project FEA COURSEWORK REPORT UFMFWD-30-M Modelling and simulation”, doi:
10.13140/RG.2.1.1580.7763.
[5] S. Alam, G. R. Yandek, R. C. Lee, and J. M. Mabry, “Design and development of a filament wound
composite overwrapped pressure vessel,” Composites Part C: Open Access, vol. 2, Oct. 2020, doi:
10.1016/j.jcomc.2020.100045.
Page | 1421
PAPER ID: 394
Numerical Analysis and Design Optimization of Kevlar Epoxy

Pressure Vessels
Pranti Sahaa, Md Foisal Hossainb*, Muhammed Sohel Ranac, and
Md Shafiul Ferdousd
a,b .
Dept. of Materials Science and Engineering., Khulna University of Engineering and
Technology, Khulna-9203, Bangladesh
b
*Dept. of Electronics and Communication Engineering., Khulna University of Engineering
and Technology, Khulna-9203, Bangladesh
c
Institute of Nuclear Science. and Technology, Bangladesh Atomic Energy Commission,
Savar, Dhaka-1349, Bangladeh
d
Frederick, Maryland-21702, USA
a
saha.pranti2000@gmail.com, b*foisal@ece.kuet.ac.bd, csohelcef@gmail.com,
d
munazeer_218@yahoo.com
Abstract: In this study, Kevlar reinforced Epoxy pressure vessels were

Keywords: analyze by finite element method (FEM) with Abaqus. Several scenarios with
variable ply layers of Kevlar have been simulated to minimize the longitudinal
● Pressure vessels; and hoop stress of the pressure vessel. Simulation shows that the maximum
● Kevlar; circumferential stress of the composite pressure vessel can be reduced by
fifteen times when six layers of Kevlar is used. Furthermore, simulation
● Composite; suggests that using minimum of four layers of Kevlar is sufficient to improve
● FEM. the stress performance of the pressure vessel.
1. Introduction
The Composite Overwrapped Pressure Vessels (COPVs) are widely used in fields including
aeronautics, automobiles and by companies such as SpaceX to hold high pressure fluids [1]. They
are favored for these applications because they are far lighter than all-metal vessels, although they
demand special design, manufacturing, and testing requirements. Currently, the pressure vessels'
outer shells are constructed of common metals like steel and aluminum alloys. The weight affects
the payload's performance, speed, and operational range. Performance improves as weight
decreases; one way to do this is by lessening the weight of the shell structure. The performance of
the vessel is enhanced by the use of composite materials, which also result in significant material
savings [2]. Composite pressure vessels are typically designed with a low bulk while meeting
strength requirements.
Polymer composites have been employed in building, home appliances, and furniture during the
past few years. Natural waste or fibers, also known as natural fiber polymer composites, are used
Page | 1422
to create the composite. These composites are lightweight, inexpensive, and resistant to corrosion
[3]. Based on how they are built, pressure vessels are divided into four kinds. Type I: Steel- or all-
metal construction. The most affordable vessels are of type I. The knowledge and tools required to
manufacture them are widely dispersed around the world. Type I vessels are the heaviest, which is
their disadvantage. Type II: mostly made of steel or aluminum with an overwrap made of glass-
fiber composite in the hoop direction. The structural loads on the metal vessel and the composite
components are roughly equivalent. Although they weigh 30 to 40% less than Type I vessels, Type
II vessels cost roughly 50% more to produce. Type III: Metal liner with a complete composite
overwrap, often made of aluminum and carbon fiber. The structural loads are carried by the
composite materials. Type IV: An all-composite design that combines carbon fiber or a hybrid
carbon/glass fiber composite with a polymer liner (usually made of high-density polyethylene, or
HDPE). All structural loads are supported by composite materials. Weight savings of 0.75 lb/L to
1.0 lb/L (0.3 kg/L to 0.45 kg/L) are achieved by Type III and Type IV vessels. The price of Type
III and Type IV vessels, however, is nearly twice as much as Type II vessels and 3.5 times as high
as the all-metal Type I tanks.
Here, Kevlar fiber was used for reinforcement. An example of an aramid fiber is Kevlar. One of
the most significant organic fibers having an aromatic polyamide structure is known as aramid,
and it is frequently utilized as reinforcement in composite materials. Typically, condensation of
aromatic diacid derivatives and aromatic diamines results in the formation of this fiber. Kevlar is
exceptionally strong and lightweight, resistant to heat and corrosion, and is woven into textile
materials. Due to their low weight, great strength, and extraordinary performance at high
temperatures, polymer-based composites offer a unique potential to replace traditional structural
materials like steel and aluminum [4]. Using ABAQUS software and the finite element method, a
type IV polymer-lined composite overwrapped pressure vessel's design is optimized. A finite
element model is used to study sustainable and renewable alternative fibers for use in high-pressure
vessels as a replacement for the conventional carbon fiber reinforced epoxy composite. Hybrid
vessels incorporating carbon and alternative fibers are suggested to increase physical,
environmental, and economic performances in accordance with the storage pressure and with
respect to mechanical requirements [5]. Hybrid composites have been used for a few years now to
obtain superior qualities under various working conditions. In light of macro mechanics, it
combines more than one composite layer. It has a variety of fibers, taking into account micro
mechanics. In comparison to traditional fiber reinforced polymer composites, hybrid composites
are more advanced composites. Compared to other fiber reinforced composites, they are more
flexible. Typically, it includes both high- and low-modulus fibers. While the low-modulus fiber
makes the composite more damage-tolerant and helps keep the material cost low, the high-modulus
fiber gives rigidity and load bearing capabilities [6]. In this paper, finite element modeling was
used to conducted stress assessments on a Kevlar reinforced pressure vessel with different number
of composite layers.
2. Numerical Methodology
Abaqus software is used in this study for numerical analysis. For modeling the composite
overwrapped pressure vessel, first the part module is created. Following the earlier work done by
Sulaiman et al. [7], we set the length of the pressure vessel to 1200 mm, diameter to 300 mm, and
shell thickness to 0.3 mm. The liner of the material is Epoxy resin, upon which multiple layers of
Kevlar fibers are applied. We vary the composite layer number from 0 to 6. Each layer thickness
is 0.762mm. For the epoxy liner material, the Young’s modulus of resin is 5000 MPa, whereas the
Poisson ratio is set to 0.4 [4]. Table I shows the corresponding lamina properties of Kevlar fiber
Page | 1423
layers used in the model [6]. Only 0o ply orientation angle of the Kevlar fibers is considered in this
study.
Table 1. Lamina properties of Kevlar fiber
Lamina Properties Kevlar
E1 76000 MPa
E2 5500 MPa
E3 5500 MPa
Nu12 0.34
Nu23 0.31
Nu13 0.3
G12 2100 MPa
G13 2100 MPa
G13 1500 MPa
Fig. 1. Shell model of the pressure vessel.
Fig. 1 shows the axisymmetric shell model of the pressure vessel. Only one-eighth potion of the
structure is modeled, whereas the rest is handled through applying boundary condition. Following
periodic boundary conditions are applied: After applying the boundary condition, 11.5 MPa inner
pressure [7] is applied to analyze the stress deformation profile of the vessel.
After generating the model structure and applying all the boundary condition and material
properties, we first analyzed the sensitivity of the stress profile with respect to mesh density. We
used epoxy-only pressure vessel to check mesh convergence. Figure 2 shows how the hoop stress
evolves with different mesh dimensions. Mesh dimensions of 0.80mm, 0.95mm, 1mm, 2mm,
5mm, 7mm, 10mm, 12mm, 15mm, 20mm, 30mm, 50mm, 70mm, 100mm are considered for
testing mesh sensitivity. The maximum hoop stresses from the simulation are 6.088E+03 MPa
for 50mm mesh dimensions, whereas for 10mm, it is 5.907E+03 MPa. So, it is 1.0293 times
smaller to 50mm mesh’s hoop stress. For 1mm mesh dimension, the hoop stress we observed
5.847E+03 MPa. It is 1.0103 times smaller to the hoop stress that was observed for 10 mm mesh
dimensions. It is seen that from 1 mm mesh dimension, the hoop stress remains constant. That
means for 1mm or less than 1mm mesh dimension, the value of the hoop stresses is well
converged. The total number of mesh points depends on the mesh dimension. For example, for
Page | 1424
25mm, 10mm, and 1mm mesh dimension results in 272, 1872 and 183372 elements, respectively.
The high mesh density requires higher computational resources. For this work, we select mesh
dimension of 1mm to run all the Kevlar composite models and for extracting the stress results.
6200
Hoop Stress, S22 (MPa)

6100
6000
5900
Converged
5800 0 1 2
10 10 10
Mesh Size (mm)
Fig. 2. The convergence of the circumferential stress S22 for different mesh size.
Fig. 3. Mesh convergence for three cases: (a) 25mm, (b) 10mm, (c) 1mm.
1500
Hoop Stress, S22 (MPa)
1000
500
1 2 3 4 5 6
Layer No.
Fig. 4. Maximum Hoop stress profile from multiple layers of Kevlar reinforced epoxy pressure vessel.
(After 4 layers of Kevlar layers, the Hoop stress does not change much drastically).
Page | 1425
Fig. 5. Hoop stress of epoxy pressure vessel Fig. 6. Hoop stress of 3 layers kevlar epoxy pressure vessel
Fig. 7. Hoop stress of 6 layers Kevlar reinforced epoxy pressure vessel
Next, we optimize the number of ply layers for better stress distribution. This layer optimization
helps us to decide how much Kevlar layer is needed to reduce stress for pressure vessel and
increase its load bearing capacity. Figure 5 shows the hoop stress distribution over the whole
simulation geometry for only epoxy lined pressure vessel. As expected, the stress is higher at the
equatorial region of the pressure vessel. On the other hand, the cap region sees minimum hoop
stress, which is almost half of the value of the maximum stress. The seam between the cap region
and the circumferential shell section sees a sharp stress variation, indicating a potential region of
failure concern.
Figure 6-7 show the analyzing result of the effect of Kevlar reinforcement on the epoxy pressure
vessel. Without any Kevlar winding, the maximum hoop stress on the epoxy layer was 5.847E+03
MPa, which gets reduced to 7.239E+02 MPa, when 3 Kevlar layers are applied as reinforcement
with the epoxy matrix. This is an 8 times reduction on maximum stress. The stress around the cap
region also goes down by almost 10 times. Moreover, the stress at the seam region has become
more gradual, as seen in Fig. 5. When 6 layers of Kevlar are applied, the hoop stress is reduced to
3.85E+02 MPa, as seen in Fig. 7. This stress is almost 15 times smaller to the pressure vessel stress
which had no Kevlar fiber layer. The stress around the cap region goes down by almost 15 times.
These results suggest that large number of Kevlar layers can significantly increase the load bearing
capacity of composite overwrapped pressure vessel and hence improve the overall design. To
Page | 1426
understand how many of the Kevlar layers are needed to fine tune the design of the pressure vessel,
we simulate a layer variation study, starting from 1 layer to 6 layers. Figure 4 shows how the
maximum hoop stress varies with layer numbers. From the figure, it is evident that when there is
only one Kevlar layer around the epoxy matrix, the maximum hoop stress is 1.739E+03 MPa. The
value of the hoop stress reduces by almost 2 times when just one additional Kevlar layer is
introduced. Moreover, when the number of the Kevlar layer is increased to 6, the hoop stress
reduces by 4.5 times. The value of the hoop stress seems to be reaching convergence at that point,
as seen in figure 4, where the hoop stresses for the layer 4, 5, 6, are quite close to each other
(between 400-600MPa). This suggests that 4 layer of Kevlar is a reasonable number of composite
layers that can provide better stress performance without increasing the mass of the pressure vessel
by unnecessarily introducing any additional layers of Kevlar fibers.
3. Conclusion
Here, Kevlar reinforced composite pressure vessel was modeled for improving stress profile. It is
seen that only epoxy layer provides a larger amount of stress. For reducing this large amount of
stress, Kevlar layers were optimized here. Six piece of Kevlar layers or plies provided minimum
amount of stress. The addition of six layers of Kevlar, according to simulation, can reduce the
maximum circumferential stress of the epoxy pressure vessel by fifteen times. Therefore, from
our numerical analysis it was found that with increasing of layer number, the load bearing capacity
of the pressure vessel is increased.
4. References
[1] Bouvier, Mathilde., Vincent, Guiheneuf., and Alan, Jean-marie., "Modeling and simulation
of a composite high-pressure vessel made of sustainable and renewable alternative fibers."
2019,International Journal of Hydrogen Energy no. 44 (23):11970-11978.
[2] Khalid, Muhammad Yasir., Ans, Al Rashid., Zia, Ullah Arif., Muhammad, Fahad Sheikh.,
Hassan, Arshad., and Muhammad ,Ali Nasir., "Tensile strength evaluation of glass/jute
fibers reinforced composites: An experimental and numerical approach."2021, Results in
Engineering no. 10:100232.
[3] Krishna., GR, Sanjay., "FINITE ELEMENT ANALYSIS OF LAMINATED HYBRID
COMPOSITE PRESSURE VESSELS." 2017, no. 8 (4):916-934.
[4] Kss, RY., R Krishna, Mohan., and B Vijay, Kiran.,"Composite pressure vessels."2002,
International Journal of Research in Engineering and Technology no. 1 (4).
[5] Maheshwari, Peeyush., A Misra., TS, Samant., PK, Vishnoi., and PS, Gautam., Simulation
of centrally loaded GrFRP and analysis of fixed (both ends) supported stress concentrated
JFRP beam.2021, Paper read at IOP Conference Series: Materials Science and Engineering.
[6] Oliveira., Luiz, G., Rafael, B Leite., Pablo, A Munõz-Rojas., and Ricardo, De Medeiros.,
"Design Optimization of a Composite Pressure Vessel Using FEA and Derivative Based
Method."2018.
[7] Regassa, Yohannes., Jema, Gari., and Hirpa, G. Lemu., "Composite Overwrapped Pressure
Vessel Design Optimization Using Numerical Method."2022, Journal of Composites
Science no.6 (8):229.
[8] Sulaiman, Shamsuddin., Soroosh, Borazjani., and Amir, Roshan., Finite element analysis of
aluminum-kevlar/epoxy pressure vessel. Paper read at Advanced Materials Research.2014.
Page | 1427

Part E - Proceedings Articles Theme 4 - Mechanical, Manufactuirng and Industrial Engineering - Final

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

Part E - Proceedings Articles Theme 4 - Mechanical, Manufactuirng and Industrial Engineering - Final

Uploaded by

Copyright:

Available Formats

7th International Conference on Engineering Research, Innovation and Education

January 12-14, 2023, Sylhet, Bangladesh

PAPER ID: 032

Study the impact of SINGLE MINUTE EXCHANGE OF DIE (SMED)

Mrinmoy Kanti Dey*, Syed Misbah Uddin

Industrial and Production Engineering Department

Abstract: The basic idea behind lean manufacturing is to eliminate

Figure 2.1: SMED process steps [www.whatissixsigma.net]

2.2 Identify Elements

Table 2.1 External and Internal activities

2.3 Separate External Elements

2.4 Streamline remaining elements

3.2 Analysis of internal data

Internal Activity line5 line 6 line 7 line 8

3.3 Overall time in four lines

4. Results and discussion

Before After Time Save Change

Total 66.27 50.84 15.42 93% 0

PAPER ID: 041

DESIGN AND CONSTRUCTION OF HYBRID COOKSTOVE AND

Riffat Jahan Juthi, Muhammad Rashed Al Mamun*, Gulam Kibria Robbani,

Abstract: One of the prime sources of indoor air pollution is household

2.2 Design Consideration

2.3 Design Parameters

It was measured by the following equation:

The efficiency was measured by Eq. (6):

Efficiency, η = 𝑂𝑢𝑡𝑝𝑢𝑡 𝐸𝑛𝑒𝑟𝑔𝑦

4. Result and Discussions

Figure 2: Hybrid Stove when tray is closed and opened

Figure 4: Comparison of temperatures of flame while cooking rice

Time to cook (min)

Figure 5: Comparison on the time to cook rice

Causes and Consequences of Bullwhip Effect on Boutique

KEYWORDS: Supply chain Management; Bullwhip Effect; Boutique Industries; Causes;

Figure 1: Order variability during times

1.1. CAUSES OF BULLWHIP EFFECT

3. DATA ANALYSIS & FINDINGS

Cronbach’s Alpha No of Item

Table 2: Reliability statistics of wholesales’ opinion

Cronbach’s Alpha No of Item

Highly Responsible Responsible Neutral Less Responsible Not Responsible

Figure 2: The responsibilities of the identified causes behind bullwhip

3.1. CROSS TABULATION

3.2. CORRELATON AMONG THE VARIABLES:

Table 3: KMO and Bartlett's Test

Approx. Chi-Square 118.679

Bartlett's Test of Sphericity Df 78

3.4. FACTOR ANALYSIS DATA REDUCTION FOR CAUSES

Table 4: Communalities for factor analysis

3.4.2. ROTATED COMPONENT MATRIX

Table 5: Rotated Component Matrix

3.4.3. SCREE PLOT

Figure 3: Scree Plot

Regression analysis is the process of estimating the dependency of on variable (dependent

3.5.1. MODEL SUMMARY

Table 6: Model Summary

b. Dependent Variable: Bullwhip Effect

4. CONSEQUENCES OF BULLWHIP EFFECT

Figure 4: Consequences those boutique industries face due to bullwhip

Table 9: Significant Causes after analysis

Ben-Daya M, R. A. (1994). Inventory models involving lead time as a decision variable.

Blinder, A. S. (1982, June). The American Economic Review, 72(3), 334-348.

EH, B. (1963). Consistency and optimality in managerial decision making. Management

MD Shafikul Islam1, Azmine Tousik Wasi2

Mst. Nasima Bagum1, Mustofa Tanbir Kuhel2, Md. Ahsan Habib3