You are on page 1of 84

ANALYSIS OF BLIND-BOLT CONNECTIONS IN

HOLLOW SECTIONS: FAILURES IN STATIC


LOADINGS

by
Ahmad Qarirul U’yun bin Mohd Yusoff

session 2021/2022

Final year research project report submitted to the


Faculty of Engineering
University of Malaya
In partial fulfilment of the requirement to the
Bachelor’s degree of Engineering

DEPARTMENT OF CIVIL ENGINEERING


FACULTY OF ENGINEERING
UNIVERSITY OF MALAYA

June 2022
PENGAKUAN OLEH PELAJAR

Saya, Ahmad Qarirul U’yun Bin Mohd Yusoff, dengan ini mengakui bahawa kecuali

di mana perakuan telah dilakukan, karya yang dipersembahkan di dalam tesis ini

adalah karya asal saya, dan tidak pernah dikemukakan secara sepenuhnya atau

sebahagiannya, sebelum ini untuk melayakkan mana-mana anugerah akademik lain.

Kandungan latihan ilmiah ini adalah hasil daripada kerja yang telah saya jalankan sejak

daripada tarikh permulaan rasmi untuk projek tesis yang disahkan tersebut.

Tarikh : 24/6/2022 Tandatangan :

Nama Penuh : Ahmad Qarirul U’yun Bin Mohd Yusoff

No. K.P. : 950802-03-5125

No. Matrik : 17151222/1

i
DECLARATION BY THE CANDIDATE

I, Ahmad Qarirul U’yun Bin Mohd Yusoff, hereby declare that except where due

acknowledgement has been made, the work presented in this thesis is my own, and has

not been submitted previously in whole or in part, to qualify for any other academic

award.

The content of this graduation exercise is the result of the work I have been carrying

out since the official commencement date of the approved thesis project.

Date : 24/6/2022 Signature :

Full Name : Ahmad Qarirul U’yun Bin Mohd Yusoff

NRIC No. : 950802-03-5125

Matric No. : 17151222/1

ii
ABSTRAK

Sebagai bahan struktur, keluli mempunyai banyak kelebihan. Salah satunya ialah

kawalan kualiti pembuatan untuk bahan ini boleh dipastikan dengan ketat. Di samping

itu, keluli struktur adalah bahan yang sempurna untuk digunakan dalam kaedah

pembinaan pasang siap. Dalam struktur, kebanyakan anggota struktur keluli

disambungkan dengan bolt dan nat. Walau bagaimanapun, susunan bolt dan nat secara

konvensional menyebabkan cabaran yang ketara terutamanya semasa kerja

penyambungan anggota struktur ke bahagian berongga keluli. Untuk mengatasinya,

jenis bolt baharu dicipta – bolt buta (blind bolt). Bolt buta membolehkan pemasangan

dari satu sisi dengan menyediakan mekanisme pengapit sendiri. Kertas ini menyiasat

bagaimana bolt jenis ini berfungsi di bawah beban statik. Ini dilakukan dengan

melakukan sejenis teknik berangka yang dipanggil Kaedah Unsur Terhingga (Finite

Element Method). Proses ini dilakukan dengan menggunakan pakej perisian FEM yang

dipanggil Ansys Mechanical. Sebelum analisis, beberapa model 3D sambungan bolt

buta dibina dengan menggunakan alat pelukis CAD terbina dalam yang dipanggil

SpaceClaim. Kemudian, proses pencerakinan (discretisation) dilakukan pada model.

Selepas itu, pelbagai syarat sempadan penting ditetapkan. Proses ini diikuti dengan

kajian parametrik yang bertujuan untuk menyiasat pengaruh saiz bolt, kekuatan bolt,

dan ketebalan plat hujung ke atas kekakuan sambungan. Daripada parameter ini,

kekuatan bolt dan ketebalan plat hujung menunjukkan beberapa tahap pengaruh dalam

kekakuan sambungan bolt buta. Lebih banyak kajian boleh dicadangkan untuk

mengesahkan keputusan daripada analisis berangka ini.

iii
ABSTRACT

As a structural material, steel poses numerous advantages. One of them is the

manufacturing quality control for this material can be stringently ensured. In addition,

structural steel is a perfect material to be used in the prefabrication construction

method. In structures, most of the steel structural members are connected with bolts

and nuts. However, the conventional bolt and nut arrangement causes a significant

challenge when it comes to connecting a structural member to a steel hollow section.

To overcome this, a new type of bolt is invented – a blind bolt. A blind bolt enables

installation from one side by providing a mechanism of self-clamping. This paper

investigates how this type of bolt performs under a static load. This is done by

performing a type of numerical technique called the Finite Element Method (FEM).

This process is done by using a FEM software package called Ansys Mechanical.

Before the analysis, some 3D models of the blind-bolted connection are built by using

the built-in CAD draughting tool called SpaceClaim. Then, the process of

discretisation is performed on the models. After that, various important boundary

conditions are set. This process is followed by a parametric study which intends to

investigate the influence of the bolt size, bolt strength, and the endplate thickness on

the stiffness of the connection. Out of these parameters, the bolt strength and the

endplate thickness show some degree of influence in the stiffness of the blind-bolted

connection. More studies can be proposed to verify the results from this numerical

analysis.

iv
ACKNOWLEDGEMENTS

Praise Allah the Almighty for all the blessings and guidance to me so that I am able to

achieve this project. I would like to express my thankfulness to my parents for all the

guidance throughout this project. Special thanks also to my brothers for all the moral

support. Lastly, my deepest gratitude to my supervisor, Dr Fazaulnizam for giving me

the opportunity to conduct this research and for lending the useful knowledge and the

experience to guide me through this project.

v
TABLE OF CONTENTS

Title Page

Pengakuan oleh Pelajar i


Declaration by the Candidate ii
Abstrak iii
Abstract iv
Acknowledgements v
Table of Contents vi
List of Figures vii
List of Tables x
List of Symbols and Abbreviations xii

CHAPTER 1: INTRODUCTION 1
1.1 Background 1
1.2 Objectives of this Research 5
1.3 Scope of this Research 5
1.4 Thesis structure 6
CHAPTER 2: LITERATURE REVIEW 8
CHAPTER 3: METHODOLOGY 17
3.1 Overview 17
3.2 Setting Up the Simulation for FEM 19
3.3 Performing Static FEM Analysis in Ansys 21
3.4 Conducting a Parametric Analysis 30
CHAPTER 4: STATIC FEM RESULTS 31
4.1 Development of Stresses in the Model 31
4.2 Propagation of Deformation in the Model 34
4.3 Relative Stiffnesses of the Model Components 35
CHAPTER 5: PARAMETRIC ANALYSIS 39
5.1 Part 1 – Effect of Bolt Size on the Stiffness 39
5.2 Part 2 – Effect of Bolt Steel Strength on the Stiffness 48
5.3 Part 3 – Effect of Endplate Thickness on the Stiffness 56
CHAPTER 6: CONCLUSIONS 65
6.1 Summary and Recommendations 65
6.2 Contribution to entrepreneurship, economic, environmental, and
socio-cultural progress 66
CHAPTER 7: REFERENCES 67
APPENDICES A1

vi
LIST OF FIGURES

Figure 1.1: Comparison between a blind bolt and a standard bolt ............................... 2

Figure 1.2: Typical stress-strain curve of a material of a ductile material ................... 4

Figure 3.1: The process flow of the methodology of this study ................................ 17

Figure 3.2: Geometric details of the blind bolts ......................................................... 18

Figure 3.3: Screenshot of the main interface of Ansys Mechanical .......................... 19

Figure 3.4: Screenshot of SpaceClaim interface with the simulation model ............. 20

Figure 3.5: The illustration of the bilinear model of the stress-strain relationship in the

Ansys material library ................................................................................................ 22

Figure 3.6: The discretisation of the model ............................................................... 23

Figure 3.7: Cross-section of the bolt connection and the contact conditions involved

.................................................................................................................................... 25

Figure 3.8: The loading step of bolt pretension in Ansys .......................................... 27

Figure 3.9: Front and rear view of the connection model showing the supports and the

external load ............................................................................................................... 28

Figure 3.10: Force convergence plot for M16 blind bolted connection model .......... 30

Figure 4.1: The results from the analysis of the M16 blind bolt, showing the location

of the yielding and the true deformation geometry .................................................... 31

Figure 4.2: F-Dt relationship of different components in M10 bolted connection .... 36

Figure 4.3: F-Dt relationship of different components in M12 bolted connection .... 36

Figure 4.4: F-Dt relationship of different components in M16 bolted connection .... 37

Figure 4.5: F-Dt relationship of different components in M20 bolted connection .... 37

Figure 4.6: F-Dt relationship of different components in M24 bolted connection .... 38

vii
Figure 5.1: F-Dt relationship of all the blind bolts in the models .............................. 39

Figure 5.2: F-Dt relationships with polynomial regression lines ............................... 41

Figure 5.3: The plot of the average stiffness against bolt diameter showing the region

of common stiffnesses ................................................................................................ 43

Figure 5.4: F-Dt relationship of the SHS with different bolt diameters ..................... 44

Figure 5.5: The plot of the average stiffness of SHS against bolt diameter............... 45

Figure 5.6: F-Dt relationship of the global reaction with different bolt diameters .... 46

Figure 5.7: The plot of the average global stiffness against bolt diameter ................ 47

Figure 5.8: F-Dt relationship of the M16 bolt shank with different bolt strengths .... 48

Figure 5.9: The plot of the average stiffness of the M16 bolt against the bolt strength

.................................................................................................................................... 49

Figure 5.10: F-Dt relationship of the endplate with different bolt strengths .............. 50

Figure 5.11: The plot of the average stiffness of the endplate against the bolt strength

.................................................................................................................................... 51

Figure 5.12: F-Dt relationship of the SHS with different bolt strengths .................... 52

Figure 5.13: The plot of the average stiffness of the SHS against the bolt strength .. 53

Figure 5.14: F-Dt relationship of the global reaction with different bolt strengths ... 54

Figure 5.15: The plot of the average global stiffness against the bolt strength ......... 55

Figure 5.8: F-Dt relationship of the M16 bolt shank with different endplate thicknesses

.................................................................................................................................... 57

Figure 5.9: The plot of the average stiffness of the M16 bolt against the endplate

thicknesses ................................................................................................................. 58

Figure 5.8: F-Dt relationship of the endplates ............................................................ 59

viii
Figure 5.9: The plot of the average stiffness of the M16 bolt against the endplate

thicknesses ................................................................................................................. 60

Figure 5.8: F-Dt relationship of the SHS with different endplate thicknesses ........... 61

Figure 5.9: The plot of the average stiffness of the SHS against the endplate

thicknesses ................................................................................................................. 62

Figure 5.8: F-Dt relationship of the global reaction with different endplate thicknesses

.................................................................................................................................... 63

Figure 5.9: The plot of the average global stiffness against the endplate thicknesses64

ix
LIST OF TABLES

Table 3.1: Dimensions of the blind bolts ................................................................... 18

Table 3.2: Material properties of the nonlinear structural steel (bolts) ...................... 22

Table 3.3: Tightening factors for different bolt types ................................................ 26

Table 3.4: Bolt pretension magnitudes....................................................................... 27

Table 3.5: External forces applied to different bolt shank diameters ........................ 28

Table 4.1: The stress development of the blind-bolted connection model ................ 32

Table 4.2: Propagation of strain in the model ............................................................ 34

Table 5.1: The R2 values and the polynomial regression equations of each bolt

diameters .................................................................................................................... 40

Table 5.2: The bolt diameters and the calculated average stiffness ........................... 42

Table 5.3: The R2 values and the polynomial regression equations of SHS with

different bolt diameters .............................................................................................. 44

Table 5.4: The bolt diameters and the calculated average stiffness of SHS .............. 45

Table 5.5: The R2 values and the polynomial regression equations of the global with

different bolt diameters .............................................................................................. 46

Table 5.6: The bolt diameters and the calculated average global stiffness ................ 47

Table 5.7: The polynomial regression equations and the R2 values of the M16 bolt

shank with different strengths .................................................................................... 49

Table 5.8: The bolt strengths and the calculated average bolt stiffness ..................... 49

Table 5.9: The polynomial regression equations and the R2 values of the endplate with

different strengths....................................................................................................... 50

Table 5.10: The bolt strengths and the calculated average endplate stiffness ........... 51

x
Table 5.11: The polynomial regression equations and the R2 values of the SHS with

different strengths....................................................................................................... 52

Table 5.12: The bolt strengths and the calculated average SHS stiffness .................. 53

Table 5.13: The polynomial regression equations and the R2 values of the global

reaction with different strengths................................................................................. 54

Table 5.14: The bolt strengths and the calculated average global stiffness ............... 55

Table 5.15: The polynomial regression equations and the R2 values of the M16 bolt

shank with different endplate thicknesses .................................................................. 57

Table 5.16: The endplate thicknesses and the calculated average bolt stiffness ........ 58

Table 5.17: The polynomial regression equations and the R2 values of the endplates

.................................................................................................................................... 59

Table 5.18: The endplate thicknesses and its calculated average stiffness ................ 60

Table 5.19: The polynomial regression equations and the R2 values of the SHS with

different endplate thicknesses .................................................................................... 61

Table 5.20: The endplate thicknesses and the average stiffness of the SHS.............. 62

Table 5.21: The polynomial regression equations and the R2 values of the SHS with

different endplate thicknesses .................................................................................... 63

Table 5.22: The endplate thicknesses and the average global stiffness ..................... 64

xi
LIST OF SYMBOLS AND ABBREVIATIONS

CAD Computer-aided design

CFT Concrete-filled tubular section

D Diameter of blind bolt

Dt Total deformation

EC0 Eurocode – Basis of structural design

EC3 Eurocode 3 – Design of steel structures

EHB Extended Hollo-Bolt

F Force

FEM Finite Element Modelling

K Tightening factor of the blind bolt

k Stiffness of a material

P Pretension of the bolt

T Required torque to tighten the blind bolt

ε Strain, i.e., the ratio of the elongation of a material per its initial

dimension

σ Stress, i.e., force per unit area experienced by a material

xii
CHAPTER 1: INTRODUCTION

1.1 Background

Recently, steel construction is gaining popularity across the globe. There is an

increasing number of applications of structural steel, especially on bridges and high-

rise buildings. This situation is enhanced by the era of prefabrication, where different

structural elements are fabricated off-site. Steel structures present significant

advantages in this method of construction as the fabrication of steel can be tightly

regulated, and the mechanical properties of structural steel can be strictly ensured. In

terms of member support, especially for building designs, steel structures typically use

either a combination of bolt systems or welding. The conventional type of bolts which

are extensively used in building design come in pairs – one bolt and one nut. This

requires every bolt to be installed from both sides. Despite being guaranteed in terms

of its strength, the method of installing the conventional bolt poses one major

limitation. And this limitation leads to the advent of a blind bolt. A blind bolt is a

special type of bolt which has a unique mechanism of installation that enables

installation from one side. This feature is important as it enables the installation where

access from both sides of the section is limited, especially in tubular sections and

concrete-filled sections.

1
(a) Blind bolt (b) Standard bolt and nut
Figure 1.1: Comparison between a blind bolt and a standard bolt

The understanding of the failure mechanism of a blind bolt as well as the

interaction with the main structure is still in progress. Therefore, this study intends to

contribute to the advancement of the understanding of this new type of bolt. This can

be achieved by investigating how this new type of bolt responds to different types of

loadings up until failure. Whilst there are numerous ways to observe the failure

mechanisms of the bolt, this study focuses on the numerical simulation through a

method called Finite Element Modelling (FEM).

The choice of numerical simulation over physical experimentation has some

considerable merits. Simulations are easy to be repeated. This is because the result

from this method is virtually free from instrumental error and human error if the

simulated model is built correctly. The financial investment for this method is

significantly less than the experimentation method. This is because there are no

laboratory work and material consumption required. However, most importantly, since

the purpose of this study is to investigate the performance of relatively newly

developed structural members, numerical simulation can be the precursor of physical

2
experimentations. Fundamentally, in this study, numerical simulation refers to the

application of a numerical method in solving a series of differential equations related

to physical quantities, such as stress and strain, which describe the condition of the

bolt after it is subjected to loading. The Finite Element Method (FEM) is one of the

most extensively used numerical simulations both in research and industry. It involves

the process of discretisation, where the model geometry is divided or meshed into

smaller parts with the same shapes such as triangles and quadrilaterals. Then, all the

differential equations about the stiffness of the model element and all the forces

applied are employed in these discrete geometries. Then, all the individual equations

for each mesh element are assembled into a larger system of equations that simulate

the entire properties of the system. In addition, similar to every typical differential

equation, boundary conditions must be set so that there will be meaningful solutions

to the equations. In structural analysis, the boundary condition can be in the form of

the forces applied and the fixity of the structural member, such as whether the

structural member is simply supported or fixed to other members.

In several parts of the world, including Malaysia, the structures are designed

with the reference to the specifications in Eurocode. There are 10 volumes of

Eurocode, numbered from Eurocode 0 (EC0) to Eurocode 9 (EC9). However, the three

volumes that are going to be referred to in this study, are EC0 and EC3. EC0 provides

an overview of the entire Eurocode publications. It outlines the fundamental concepts

of the structural analysis and the design methods applied in the later volumes.

Meanwhile, EC3 stipulates the guide for designing steel structures. In terms of the

3
actions on the structures, there are three main types, namely permanent action, variable

action, and accidental actions (European Committee, 2005). Permanent actions in a

structure are the summation of the structure self-weights and fixed equipment and

finishing. Meanwhile, variable action is the loads which vary throughout the lifetime

of the structure. This includes imposed loads, wind loads, and snow loads. Lastly, the

accidental load is the load which occurs in a significantly small chance and usually

happens in a brief time such as explosions or vehicle collisions. This study, however,

focuses on the loading conditions which represent the ultimate state design. Such a

limit state primarily focuses on the magnitude of loading until the structural material

passes its elastic limit, therefore, experiences yielding. In structural analysis

especially, it is useful to express the reaction of material upon experiencing external

loading as stress (σ), which equals the force per unit area, and strain (ϵ), which

represents the ratio of the deformation of the material per its original dimensions.

Figure 1.2 shows a stress-strain relationship of a ductile material.

Ul ima e s reng h

rac ure
Yield s reng h

Elas ic las ic
region region

Figure 1.2: Typical stress-strain curve of a material of a ductile material

4
The next three subchapters of this report briefly list the main objectives, the scope of

this study, and the structure of this thesis.

1.2 Objectives of this Research

• To investigate how does a selected type of blind bolt perform under

increasing static stresses

• To conduct a numerical analysis of the selected blind bolts by using the

Finite Element Method (FEM)

• To establish some important correlations between different parameters

during the numerical simulation

1.3 Scope of this Research

• This study will only involve some literature reviews to identify the gap in

knowledge from the past and recent research

• Only numerical analysis and some brief parametric analysis will be involved

in this study

• Physical experimentation is not covered in this study

• The parametric study focuses on the effect on different parameters

considered:

o The bolt sizes

o The structural steel strength

o The thickness of endplates

5
1.4 Thesis structure

This study will have 7 main chapters which can be outlined as follows.

CHAPTER 1 – INTRODUCTION

This is where the theoretical foundations pertinent to this study are laid. All the

important concepts will be explained briefly in this chapter.

CHAPTER 2 – LITERATURE REVIEW

Several past research will be outlined to determine the possible research gaps and to

learn about the methodologies used.

CHAPTER 3 – METHODOLOGY

This chapter explains the process of the FEM employed in this study. The explanation

includes all the necessary steps and the conditions of the simulations.

CHAPTER 4 – STATIC FEM RESULTS

This chapter presents the basic output of the FEM process with Ansys Mechanical.

Some important discussions are also made in terms of the failure mechanism of the

chosen blind bolts.

CHAPTER 5 – PARAMETRIC ANALYSIS

This chapter investigates and discusses the potential parameters which affect the

stiffness of the blind-bolted connection setup.

6
CHAPTER 6 – CONCLUSIONS

This chapter concludes this research by outlining the key takeaways as well as

recommendations for future research. In addition, there are also some outlines of the

entrepreneurship opportunity related to this research as well as the economic,

environmental, and socio-cultural impacts in professional practice related to this study.

CHAPTER 7 – REFERENCES

This chapter lists all the materials referred to in this study.

7
CHAPTER 2: LITERATURE REVIEW

Various research has been conducted to test the mechanical properties and the failure

mechanisms of blind bolts. In general, when comparing the previous studies, the

variations can be explained in the form of the research methodologies, the structural

members involved, and the types of failure of the bolts. There were a few research

methodologies that were identified. Some of which include reviews, experimentations,

analyses, and combinations of two or more methods mentioned. In terms of the types

of blind bolts, it was found that several studies utilised certain modified versions of

blind bolts available in the industry, whereas the others used the standard ones. For the

failure types, a large number of studies focused on both static and dynamic loadings,

whereas some of them investigated the performance of the bolts under fire conditions,

fatigue, and seismic loadings.

One of the examples of a review study is conducted by M. Cabrera, W. Tizani

and J. Ninic (2021), where they investigate the trends of previous research on several

types of blind bolt systems that were currently available in the industry. In the study,

there were four types of blind bolts critically described, which are Ajax ONESIDE

fastener, thread-fixed one-side bolt, T-shaped one-side bolt, and Lindapter® Hollo-

bolt. the author of this study compares the performance characteristics of these bolts

based on previous studies. The Ajax ONESIDE fastener is a type of blind bolt where

it has a split step-washer, a solid washer, and a structural nut. From the study, it was

explained that the strength characteristics of the bolt comply with the American

8
Standard AS4291.1 which pertained to the mechanical properties of steel bolts. The

thread-fixed one-side bolt was also found to be compliant with the resistance to seismic

loadings. For T-shaped one-side bolt, the author explained that the initial stiffness of

the bolt increases when it is slotted vertically, but decreases when slotted horizontally,

as compared to the standard high-strength bolts. Whereas the Lindapter® Hollo-bolt

shows some potential in the application of the moment-resisting connection (Cabrera

et al., 2021). In another study, T. Pokharel (2016) conducted a review on developing

structural systems for low to mid-rise buildings using moment resisting frames with

blind bolted connections with sufficient strength and stiffness to create a lateral force-

resisting system suitable for low to moderate seismic regions. In this study, Pokharel

provided some description of the design methods for structures that implements blind

bolts into the design. His research included some experiments and analyses to support

the design practice (Pokharel, 2016).

Besides, there are also several research which involve some experimentations

with some selected blind bolts available on the market. Li and Zhao (2021) conducted

an experimental study on blind bolted T-stub to square hollow section (SHS)

connections made by stainless steel. The authors identified several parameters which

affect the performance of the blind bolt T-stubs namely the blind bolt system, the

layout of the bolts, the thickness of the SHS, and the specimen configuration. The

experimental program featured in the study includes tests for a single bolt, T-stub, the

hollow section, and T-stub to hollow section. From the study, they concluded that the

mechanical properties of bind bolts are greatly affected by their interlocking

9
mechanisms (Li & Zhao, 2021). Meanwhile, Y. Liu, J. Chen, X. Zhang, and D. Tan

(2018) conducted an experimental study on the fatigue performance of a blind bolt

under axial-tension cyclic loads. From the results, it was found that different modes of

failure were observed at the standard bolt and the blind bolt. For the standard bolt, the

shank fracture was observed, whilst for the blind bolt, a sleeve fracture was observed

when it was subjected to static tension (Liu et al., 2018).

In addition, W. Tizani and T. Pitrakkos (2015) conducted experimental research

to investigate the performance of a newly developed blind bolt, intended for use in

constructing bolted moment-resisting connections to concrete-filled tubular steel

profiles. From their research, it was found that the connections to the steel hollow

unfilled section could not develop the full tensile capacity of the bolt. However,

different conditions were observed for the concrete-filled tubes. When the concrete is

filled into the tubes after the blind bolts were installed, the overall connection stiffness

increases, and the bending and the deformation of the tube decreases (Tizani &

Pitrakkos, 2015).

Apart from experiments, there are also some analysis studies. Under this type of

research, there are two sub-categories, which are parametric study and, Finite Element

Modelling (FEM). The two software applications mostly used are Ansys and Abacus.

Meanwhile, a parametric study is the study of the influence of different geometric or

physical parameters or both on the solution of the problem. This typically involves the

derivation of equations. One of the parametric studies surveyed is by Javora and Skejić

10
(2017), who studied the interactions between different bolt systems in beam-to-column

joints. In the study, 1,155 combinations of parameters define the behaviour of 5 chosen

blind bolt systems, namely generic blind bolt, Ajax ONESIDE fastener, Flowdrill,

Huck Ultratwist, and Hollo-bolt. These parameters include the strength of the columns,

the bolt size, the bolt grade, and the distance between the bolts. In the study, three main

failure modes were identified, which are the tensile force on the bolt, the bending of

the column face, and the bending of the endplate (Javora & Skejić, 2017).

Meanwhile. Zheng and Xia (2009) provided one of the earliest studies on the

FEM simulation of the blind bolt. The FEM simulation was conducted in a 3D FEM

software called Ansys. In the study, the author explained the method of idealisation of

the blind bolt thread. Instead of modelling the thread in 3 dimensions, the author used

node coupling to simulate the interaction between the bolt surface with the connected

block. The simulation procedures are as follows: the author defined the pretension

force. The pretension force has one degree of freedom of translation and has a

magnitude of 55 kN. After the simulation was completed, the author then conducted a

theoretical analysis of the bolt. The theoretical analysis involves the derivation of

principal equations, especially about the stress and strain of bolts loaded with axial

loads. In this process, the interaction of the bolt thread is explicitly modelled by the

existing equations. Then, the author proceeded with a comparison between the

theoretical analysis and the simulation and eventually concluded that the simulation is

consistent with the existing theories of bolt loading (Zheng & Xia, 2009).

11
In another study, A. M. Pascual, M. L. Romero and W. Tizani, (2015) provided

an advanced numerical model (FEM) to predict the fire behaviour of blind bolts in the

tension area of endplate connections between I-beams and concrete-filled tubular

(CFT) columns. A few analyses leading to the FEM were conducted by the author,

including nonlinear heat transfer analysis to test the bolt when it is subjected to thermal

load through conduction, convection, and radiation, as well as nonlinear stress analysis

to test the bolt under tensile stress. The results showed that the addition of concrete to

the inside of the column improves the connection's responsiveness in terms of FRR

and stiffness at high temperatures. When anchored blind bolts were used instead of

regular blind bolts, stiffer connections were achieved, although the FRR improvement

was dependent on plate thickness and steel bolt characteristics (Pascual et al., 2015).

Another FEM study was conducted by Cabrera, Tizani, Mahmood and

Shamsudin, (2020), which investigates the combined failure mode of the Extended

Hollo-Bolt (EHB) and the effect of the column thickness on the tensile behaviour of

the blind fastener by using 3D FEM. The authors constructed three full-scale 3D

models by using drafting software and conducted the simulation on the combined

loadings in finite element software called Abaqus. The study proves the reliability of

FEM in understanding the behaviour of blind bolts when subjected to loadings. This

can be shown from the comparison between the deformation in experimental data and

the FE analysis, the marginal error in initial stiffness is between 4 to 16%. In addition,

from the global displacement model, the FE model can predict the force-displacement

curve with different slenderness ratios of the column with 90% accuracy. However,

12
the author clarified that when the column is slender, the concrete tends to reach failure

first before the steel column itself. Therefore, it was suggested a study be conducted

to investigate the effect of differing infill concrete strengths and the column

slenderness ratio on the failure of the bolt (Cabrera et al., 2020).

Besides sole experimentations and simulations, there were also some studies

which incorporated both methodologies in one study. The purpose of this is often to

compare the results from these two methodologies. This way, the authors can verify

the validity of the simulations with the real data from the experiments. For example,

Mesquita et al. (2021) conducted a comparison study between 3D simulation with

experimental results on the performance of steel joints with Hollo-bolt. In the study,

the author compared numerical results with experimental data by using full-scale

prototypes under monotonic load. For this purpose, various 3D models were created

by using a mechanical CAD drafting software called Autodesk Inventor, which were

then imported to Ansys to perform numerical simulations. Then, the author used the

true geometrical and mechanical parameters of the individual profiles to perform a

numerical calibration. From the numerical simulation, it was found that the sleeves of

the hollo-bolts were the first component to fail in all tests, implying that they were

responsible for the joint's integrity. The author finally suggested that a parametric

study can be conducted to investigate the effect of varying the bolt's tightening torque

as well as the width and thickness of the tube profiles on he bol ’s performance

(Mesquita et al., 2021).

13
Meanwhile, Lee et al. (2011) researched to test the performance of T-stub

connection with blind bolt subjected to static loading. The test specimen of the research

was Ajax ONESIDE blind bolt. In the experiment set-up, a series of Ajax ONESIDE

blind bolts are attached to a T-stub and a square hollow section (SHS). The assembly

of the blind bolt included the bolt itself, a collapsible washer, a shear sleeve, a solid

washer, a nut, and an installation tool. The whole setup was also equipped with sensors

that can detect and record the strain of the specimen relative to the ground. From the

results, it was clear that the incorporation of bolt sleeves contributes to the strength

and the stiffness of the bolt system as it can reduce the bending of the bolt. Due to this

reason, the author recommended the use of sleeves, especially for the blind bolt of this

variant. The tolerance between the outside diameter of the sleeves and the bolt hole

determines their effectiveness. The sleeves are more effective in reducing bolt bending

and slide between the endplate and the tube face when the tolerance is tighter.

Additionally, the author also conducted a non-linear FEM simulation and found that

the FEM can accurately replicate the experimental results (Lee et al., 2011b).

Furthermore, significant variations in terms of the structural members under

study were found in previous research. The most apparent one is the use of steel hollow

sections, both unfilled and filled with concrete. Typically, the researchers who

conducted this study aim to investigate the deformation pattern of the bolt, known as

a local displacement, and the deformation of the entire connection system including

the members, known as global displacement. The influence of the structural member

on the performance of the bolt is also one of the key interests in this type of study. One

study conducted by Tahir et al. (2018) investigated the interactions of the moment

14
connection between an I-beam and a square hollow column with a blind bolt by the

method of numerical analysis and experiments. In both of these procedures, eight full-

scale specimens with four extended end plates and four-flush end plate external

connections were examined by subjecting them to monotonic stress. The joint's

moment-rotation curves, failure mechanism, starting stiffness, and connection

classification were measured and discussed. The results showed that the thickness and

kinds of end plates, as well as the size of the beam, can increase the moment capacity

and rotational stiffness of connections. However, the author also noted that there were

a few exceptions where for certain SHS dimensions, the ability for the connection to

obtain a partial strength connection status according to the EC3 classifications is

reduced when the beam size and end plate thickness are increased beyond a certain

amount (Tahir et al., 2018).

For hollow sections, it is apparent that the moment connection is one of the most

commonly tested connections in the previous studies. Moment connection is a type of

connection which allows the transfer of the bending moment of a beam to a column.

For example, Lee et al. (2011) conducted another study which comprises an

experiment and a numerical analysis on the beam-to-column moment connection with

blind bolts. The research was conducted in cooperation with two of the major

Australian industrial steel manufacturers which are Ajax Engineered Fasteners and

Australian Tube Mills to develop a range of blind bolts which can be used in the

construction industry. In the experiment setup, a T-stub blind bolted connection was

used, which consisted of a top and a bottom flanges and channels which are installed

15
to a backplate inside the SHS column connecting the flanges. The purpose of the

incorporation of channels inside the SHS is to shift the tension stress from the beam

flange to the rear face of the hollow section column, minimising strain on the flexible

column face, which can affect the connection's rigidity. As a result, with the same

thickness of the endplate, the extended T-stub connection with rear face support is

about five times stiffer than the T-stub connection. This demonstrated that the T-stub

blind bolt configuration is stiffer than a typical end plate connection, therefore is

suitable to be used in rigid connections in compliance with EC3 (Lee et al., 2011a).

16
CHAPTER 3: METHODOLOGY

3.1 Overview

Figure 3.1 shows the process flowchart of the methodology of this study.

Start

Retrieving the
blind bolt, the
endplate, and the
SHS geometries

Produce 3D model
of the connections

Perform dynamic
analysis with FEM

Conduct parametric
analysis

Interpret the
results

End

Figure 3.1: The process flow of the methodology of this study

The methodology of this study starts with selecting the specific blind bolt already

available in the industry. in this case, the blind bolt selected is from BlindBolt Asia.

There are three main products by this company which are “The Blind Bol ”, “Thin

17
Wall Bol ”, and “Heavy Du y Bol ”. The first one is chosen for modelling in this study

because it is the standard version of the blind bolt as shown in Figure 1.1 in Chapter

1. Then, the catalogue of the chosen blind bolt series is examined to obtain the required

geometries as well as the recommended bolt pretension. In addition, five blind bolt

diameters were chosen, namely M10, M12, M16, M20, and M24, to be analysed. The

geometric details of the blind bolt can be presented in Figure 3.2 and Table 3.1 (Blind

Bolt Company, 2020).

Figure 3.2: Geometric details of the blind bolts

Table 3.1: Dimensions of the blind bolts

Bolt Anchor Fixing thicknesses Depth


Bolt length
diameter Clearance Min Max Clearance
M10 95 23 25 65 30
M12 120 26 30 85 35
M16 90 36 13 43 43
M20 110 44 21 56 56
M24 130 53 21 66 64
Note: All dimensions in this table are in mm.

18
3.2 Setting Up the Simulation for FEM

There are many software packages available which are capable of performing a FEM

simulation. However, in this study, Ansys Mechanical was chosen for this purpose as

demonstrated in Figure 3.3. Before running the simulation in Ansys, the simulation

model had to be constructed. This can be done in multiple ways with different CAD

draughting methods and then can be imported into Ansys Workbench, the main

interface of Ansys. Ansys also provides a built-in CAD tool called SpaceClaim, which

is used in this study. Figure 3.4 shows a screenshot of the SpaceClaim package. In

addition, the simulation model can be constructed in several ways depending on the

desired accuracy aimed for the simulation. For example, a 3-dimensional object can

be modelled as it is with varying levels of geometrical precision or be modelled as a

2D surface or 2D line. This study aimed to observe the failure mechanisms in great

detail. Therefore, all the steel connection setups were modelled in 3D.

Figure 3.3: Screenshot of the main interface of Ansys Mechanical

19
Figure 3.4: Screenshot of SpaceClaim interface with the simulation model

In the 3D model, the bolt was fixed to an unfilled SHS column with an extended

endplate to idealise a semi-rigid beam-to-column connection. For the column, an SHS

200/12 (square hollow section with the section dimensions of 200 × 200 mm and the

thickness of 12 mm) with an extended plate of the dimensions of 300 × 150 × 12 mm

is used. Also, both the endplate and the SHS would have holes which follow the blind

bolt diameter with an additional 1 mm tolerance. For example, in the setup for the M16

blind bolt, the holes on the endplate and the SHS would have a diameter of 17 mm.

This was done according to specifications from the Blind Bolt Company itself.

20
3.3 Performing Static FEM Analysis in Ansys

Generally, the methodology of the static FEM analysis in this study can be listed as

follows:

1. Identifying the material properties

2. 3D model discretisation

3. Identifying all boundary conditions

4. Setting up the analysis

5. Retrieve the results

3.3.1 Material properties

The chosen material was structural steel with a yield strength of 250 MPa. This was

the default material in the Ansys programme. However, it was intended to apply the

external load past the yield strength of the model. Hence, a non-linear structural steel

material was chosen from the material library in Ansys. The non-linear material differs

from the linear counterpart due to the stress-strain model that was attached to the

material properties. For linear structural steel, only a linear stress-strain relationship is

available, whereas, for the non-linear structural steel, the stress-strain relationship was

approximated by a bilinear model in the software, which can be demonstrated in

approximating the non-linear relationship by splitting the stress-strain curve into two

parts, separated at the yield point on the curve. On the left side of the yield point, the

gradient of the curve is Young’s modulus of he ma erial, whereas, on he righ side of

the yield point, the gradient is called the tangent modulus of the material. Young’s

modulus of the material is 200 GPa and the tangent modulus is just 1.45 GPa. In the

21
parametric analysis, which will be discussed in Chapter 5, two other structural steel

grades will also be used for bolts, which are S275 and S355.

Yield
poin

Tangen modulus

E E D
True s ress s rain curve
Young s Bilinear model
modulus S ress
S rain

Figure 3.5: The illustration of the bilinear model of the stress-strain relationship
in the Ansys material library

The following table describes the properties of the nonlinear structural steel used.

Table 3.2: Material properties of the nonlinear structural steel (bolts)

Parameters Default S275 S355


Density (kg/m3) 7,850
Young’s modulus (GPa) 200
Tangent modulus – in the 1.45
plastic region (GPa)
Poisson’s ratio 0.3
Bulk modulus (GPa) 166.67
Shear modulus 76.92
Yield strength (MPa) 250 275 355
Ultimate Strength (Not available) 430 510

22
3.3.2 Discretisation

Similar to any typical FEM analysis, it started with the process of discretisation. This

involved creating subdivisions or meshing to the steel connection 3D model. It was

determined that the mesh size varied with the different elements in the steel

connection. This condition can be created in the Ansys Workbench by using the feature

called “Body Sizing”. These different mesh sizes can be listed as follows:

1. For the bolt shanks and the nuts – use 2.5 mm mesh, except for the M24 bolt

shank.

2. For the M24 bolt shank – use 5 mm mesh

3. For the endplate – use 10 mm mesh

4. For the SHS – use 25 mm mesh

The decision to vary the mesh sizing was pursued due to the limitation of the

computation resource and the lesser requirement for the accuracy of larger structural

members. Figure 3.6 shows the applied mesh sizing of the steel connection model.

Figure 3.6: The discretisation of the model

23
After the mesh was generated, various boundary conditions were specified before the

analysis was started.

3.3.3 Boundary Conditions

3.3.3.1 Boundary condition 1 – Contact conditions around the connection

After the meshing was generated, some of the individual meshes between different

bodies would be close together or touch each other. This would be detected by Ansys

algorithms and then would be listed in the interface. These are called contact

conditions. There are five types of contact conditions and each of them represents

different physical conditions of the interactions between a pair of bodies. The five

contact conditions can be listed as follows:

1. Bonded contact – the two bodies cannot separate and slide between each other

2. Rough contact – the friction coefficient between the surfaces of the two bodies

is assumed to be infinite. The two bodies are glued to each other

3. Frictional contact – the two surfaces of the bodies have a finite static friction

coefficient

4. Frictionless contact – the two bodies can separate and slide between each other

5. No separation contact – the two bodies can slide but not separate between each

other

However, in this study, only two contact conditions were utilised, which are the rough

and the bonded contact. The bonded contact was applied to the surfaces between the

bolt shank and the SHS as well as between the shank and the endplate. Meanwhile, the

24
rough contact condition was applied to all other contact surfaces in the model. This

can be illustrated in Figure 3.7.

Bol shank
u

Endpla e
layer E E D
SHS layer Rough con ac
Bonded con ac
Bol anchor

Figure 3.7: Cross-section of the bolt connection and the contact conditions
involved

3.3.3.2 Boundary Condition 2 – Bolt Pretension

Bolt pretension, also known as bolt preload, represents the force applied during the

tightening of the bolt which eventually creates compression inside the connection from

the clamping effect. From the technical sheet of the blind bolt, the bolt preload is not

specified directly. Instead, the manufacturer recommends the torque required to

tighten the bolts. The bolt pretension can be derived from the following equation:

𝑇 =𝐾 ×𝐷×𝑃 [3.1]

Where,

𝑇 Required orque o igh en he bol

𝐾 Tigh ening fac or (see Table 3.3)

𝐷 Diame er of he bol shank

𝑃 re ension of he blind bol

25
The following table shows the tightening factors commonly used:

Table 3.3: Tightening factors for different bolt types

K Type of bolt
0.2 Steel bolt without plating
0.15 Steel bolt with cadmium plating
0.28 Steel bolt with zinc plating
0.18 Steel bolt with lubrication

For the simulation setup, 𝐾 = 0.2 (steel bolt without plating) is used.

By applying the tightening factor, the bolt pretension force can be derived from the

equation 3.1, which results in the following equation:

𝑇
𝑃= [3.2]
0.2𝐷

For the first test subject, an M10 bolt is used. From the BlindBolt catalogue, the

recommended torque for the bolt of this diameter is 24 Nm. Therefore, the bolt

pretension is calculated as follows:

24
𝑃= = 12,000 N
0.2(0.01)

All the bolt preloads can be tabulated as follows:

26
Table 3.4: Bolt pretension magnitudes

Bolt size Shank Diameter (mm) Torque (Nm) Bolt Pretension (N)
M10 10 24 12,000
M12 12 30 12,500
M16 16 50 15,625
M20 20 65 16,250
M24 24 75 15,625

Then, under the Static Structural analysis settings, the loading step is set as 3. This is

to ensure that the external load is applied once the bolt pretension was fully applied.

The bolt preload was applied in step 1. Then, in steps 2 and three, the load is ‘locked’.

This can be demonstrated by the following figures.

Figure 3.8: The loading step of bolt pretension in Ansys

3.3.3.3 Boundary Condition 3 – Supports and External Load in Static Analysis

Due to the scope of this study, the simulation must not allow rigid body motion in the

simulation models. This condition can be achieved by limiting the external load to the

point where the test models only experienced elastic and plastic deformation, but not

total failure or rupture. In Ansys, the sub-program used for this purpose was Static

Structural. Whereas, if the material failure is intended, then Explicit Dynamics was

27
used instead. The external load was applied as a line load across the width of the

endplate to represent the force exerted from the transfer of load from a beam to a

column between the second and the third step of load application as demonstrated in

Figure 3.9: Front and rear view of the connection model showing the supports and the

external load. This virtual setup was essentially mimicking the standard pull-out test

of the standard bolts but without the total failure of the bolts. Different external load

is applied to each bol ’s diameter. These different forces can be tabulated in Table 3.5.

In addition, to ensure that the steel connection setup has adequate stability, fixed

supports were set at both ends of the SHS.

oad oad

i ed suppor

Figure 3.9: Front and rear view of the connection model showing the supports
and the external load

Table 3.5: External forces applied to different bolt shank diameters

Bolt diameter (mm) M10 M12 M16 M20 M24


Force applied (kN/m) 1,250 1,400 1,500 1,650 1,800

28
3.3.4 Setting Up the Static Failure Analysis in Ansys

After all the necessary boundary conditions were established, the solution can be set

up. To ensure the accuracy of the final results of the FEM, a few preliminary

adjustments were made in he “Analysis Se ings” in the left pane of the Ansys

Mechanical interface. The adjustments made are listed as follows:

1. Under “S ep Controls”

a. Number of steps :3

b. Auto time-stepping : On

c. Initial sub-steps : 30

d. Minimum sub-steps : 20

e. Maximum sub-steps : 150

2. Under “Solver Controls”

a. Large deflection : On

3. Under “Advanced”

a. Nonlinear data : On

Note that the “ arge deflec ion” se ing was urned on because i was in ended o

subject the bolted connection model under plastic strain conditions. Then, the analysis

was run. To ensure the reliability of the results, the Force Convergence plot was

checked. This can be done by going o he “Solu ion Informa ion” in he lef pane of

the interface and setting the “Solu ion Ou pu ” o “ orce Convergence”. The resul s

from the FE analysis are considered reliable when the force convergence line (the blue

line in Figure 3.10) falls below the force criterion line (the cyan line). An example of

a force convergence plot is displayed in the Force convergence plot for M16 blind

29
bolted connection model. All the force convergence plots are included in Appendix A

of this report. Then, under the solutions, all the intended results such as the Von-Mises

equivalent stress and strain were set up.

Figure 3.10: Force convergence plot for M16 blind bolted connection model

3.4 Conducting a Parametric Analysis

In this study, the parametric analysis refers to the investigation of different

independent parameters to the overall stiffness of the blind-bolted connection models

and the local stiffness of their components.

30
CHAPTER 4: STATIC FEM RESULTS

4.1 Development of Stresses in the Model

It is intended to locate the development of stresses in the various components of the

bolted connection model. Since the material fracture is not covered in this study, all

the results from the FE analysis will only indicate the locations of the yielding of the

material. An example of this is shown in the following diagram, where the red region

represents the area where the equivalent stress exceeded the 250 MPa yield strength of

the structural steel.

Figure 4.1: The results from the analysis of the M16 blind bolt, showing the
location of the yielding and the true deformation geometry

In addition, it is also important to look at how the stresses develop inside the model

with the gradual increment of the external load imposed on it. Based on the simulation

conducted, it was observed that the stress development for all blind bolt diameters was

31
almost completely identical. Table 4.1 shows an example of the development of stress

in the M16 bolted connection setup. The left column of the table shows the longitudinal

section view of the simulated model. Similar to Figure 4.1, the red region represents

the area where the steel yielding had occurred.

Table 4.1: The stress development of the blind-bolted connection model

FEM results Explanation

Stage 1 Stress developed inside the shank due to the

compression effect created by the bolt

pretension from the tightening process. At this

point, the bolt preloading is locked, and the

external force started to be imposed on the

model. The endplate and the SHS were

minimally affected by the bolt pretension

process.

Stage 2 After the external load was applied to the

model, the stresses started to develop further

in the shank and the nut as well as all other

elements. The load tended to be concentrated

at the location of the external load and spread

horizontally at the contact surfaces. All the

elements are still under elastic deformation at

this point.

32
FEM results Explanation

Stage 3 The yielding started to occur in the bolt shank

and the top surface of the endplate. The

yielding formed a shear pattern starting at the

surface facing the centre of the endplate. This

was due to the eccentricity of the external load

applied relative to the position along the bolt

shank.

Stage 4 The endplate started to deform excessively

due to the external load. The yielding of the

steel covered almost the entire top surface of

the endplate. The excessive deformation

(Section view) created a pair of forces which acted

diagonally with the respect to the global

coordinate of the model. These diagonal

upward forces pushed towards the shank and

the nut, where at this point, the nut started to

(Plan view) experience elastic stress.

33
4.2 Propagation of Deformation in the Model

Because structural steel tends to have a significant magnitude of stiffness, the strain is

not necessarily visible before it yields. This is true for the connection model used in

this study. However, similar to stress development, it is essential to observe how the

strain is developing inside the bolted connection. This can be demonstrated in Table

4.2.

Table 4.2: Propagation of strain in the model

FEM results Explanation

Stage 1 This was the early stage of the simulation,

where the bolt pretension was fully applied

and locked. Since the stress caused by the

pretension was relatively small, the strain

caused by it was almost zero. This condition

continued even after the external force was

applied.

Stage 2 At this frame, there was a significant

deflection observed at the endplate. The

maximum deflection at this location exceeded

2 mm. However, it is interesting to note that

the stress inside the endplate, as well as the

bolt shank, corresponds to Stage 3 in the stress

development; it already exceeded the yield

limit of the steel. This was because the bolt

34
FEM results Explanation

shank is constrained by the bolt hole whilst

the endplate is free to move and separate.

Stage 3 This was at the end of the simulation, similar

to Stage 4 in stress development. The total

deformation at the endplate reached a

maximum of almost 10 mm. The deformation

was absolutely visible at this stage. However,

inside the bolt shank, significant total

deformation was only observed at the top part

of the shank, and the distribution of the

deformation was perpendicular to the

direction of the deformation of the endplate.

4.3 Relative Stiffnesses of the Model Components

The observed propagation of failure can also be further verified by investigating the

relationship between the applied force (F) and the total displacement (Dt) for each

component in the bolted connection. From this relationship, we can obtain the relative

stiffness, which is represented by its gradient. After some analyses were made, it was

found that all the bolted connection models show similar variation in terms of

stiffnesses in the connection components. In addition, when the total deflection for the

entire model and the endplate were obtained, it was found that the two data matched

completely in all models. This suggests that the endplate has the least stiffness among

35
all other components in all models. Therefore, this condition confirms the observation

made in subchapters 4.1 and 4.2. Figure 4.2 until Figure 4.6 show the force-

displacement relationship for all blind-bolted connection models.

200
180
160
140
120
Force (kN)

Bolt Shank
100
Series2
80
Endplate
60
SHS
40
20
0
0 1 2 3 4 5 6 7 8
Total Deformation (mm)

Figure 4.2: F-Dt relationship of different components in M10 bolted connection

250

200

150
Force (kN)

Bolt Shank
Global
100
Endplate
SHS
50

0
0 2 4 6 8 10
Total Deformation (mm)

Figure 4.3: F-Dt relationship of different components in M12 bolted connection

36
250

200

150
Force (kN)

Bolt Shank
Global
100 Endplate
SHS

50

0
0 2 4 6 8 10 12
Total Deformation (mm)

Figure 4.4: F-Dt relationship of different components in M16 bolted connection

300

250

200
Force (kN)

Bolt shank
150
Global
Endplate
100 SHS

50

0
0 2 4 6 8 10 12
Total Deformation (mm)

Figure 4.5: F-Dt relationship of different components in M20 bolted connection

37
300

250

200
Force (kN)

Bolt shank
150
Global
Endplate
100
SHS

50

0
0 2 4 6 8 10 12 14
Total Deformation (mm)

Figure 4.6: F-Dt relationship of different components in M24 bolted connection

The effect of the bolt diameter on the global stiffness and the stiffnesses of the

components are going to be discussed in the next part.

38
CHAPTER 5: PARAMETRIC ANALYSIS

5.1 Part 1 – Effect of Bolt Size on the Stiffness

Figure 5.1 shows the stiffnesses of the bolts in all the bolted connections. From the

figure, it was observed that there were two pairs of bolts which have slightly similar

stiffness, which is the M10-M12 and M16-M20 pairs. The M24 has the highest

stiffness among all other fours. However, this observation needs to be verified further.

This can be done by obtaining the average stiffnesses. Due to the nonlinearity of the

F-Dt relationship, polynomial regression equations were obtained for all the bolt

diameters.

300

250

200
Force (kN)

150

100

50

0
0 0.5 1 1.5 2 2.5 3 3.5
Total Deformation (mm)

M10 M12 M16 M20 M24

Figure 5.1: F-Dt relationship of all the blind bolts in the models

39
To generate the polynomial model of all the F-Dt relationships, it must be ensured that

the equation fits the original data well. The general form of a polynomial equation of

the nth order is shown as follows

𝑛
𝑦 = 𝑓(𝑥) = ∑ 𝑎𝑟 𝑥𝑟 [5.1]
𝑟=0

Where n is the order (the highest power) of the polynomial.

To ensure the reliability of the polynomial equation, the R2 value was also obtained.

This value represents how well the polynomial model fits with the original data. In this

study, a minimum of 0.8 is considered acceptable for this value. In addition, since the

y-intercept should be 0, then the term in equation 4.1 was set to be 0. Table 5.1

shows the generated polynomial equation and the R2 values for all bolt diameters.

Table 5.1: The R2 values and the polynomial regression equations of each bolt
diameters

Bolt

M10 0.9997

M12 0.9999

M16 0.9998

M20 0.9997

M24 0.9993

The following figure shows the F-Dt relationships for all bolt diameters and their

respective polynomial regression equation.

40
300
y = 18.338x3 - 108.34x2 + 264.54x
R² = 0.9993
250
y = 18.645x3 - 110.69x2 + 262.55x
R² = 0.9997
200 y = 36.83x4 - 105.82x3 + 8.1342x2 + 234.81x
R² = 0.9998
Force (kN)

150 y = 13.885x3 - 90.043x2 + 216.16x


R² = 0.9999
100
y = 11.095x3 - 86.726x2 + 221.71x
R² = 0.9997
50

0
0 0.5 1 1.5 2 2.5 3 3.5
Total Deformation (mm)

M10 M12 M16 M20


M24 Poly. (M10) Poly. (M12) Poly. (M16)
Poly. (M20) Poly. (M24)

Figure 5.2: F-Dt relationships with polynomial regression lines

rom Hooke’s law, the stiffness is defined by the change in the force concerning the

elongation. Therefore, the average stiffness can be obtained by summing all the

individual stiffnesses across the deformation and then dividing with the change in the

deformation. The formula to find the average stiffness can be derived as follows:

rom Hooke’s law,

(ignoring he sign conven ion) [5.2]

Then, the tangent stiffness of an F-Dt relationship is,

𝑑𝐹
𝑘=
𝑑𝐷𝑡

41
Therefore, the average stiffness between an interval of deformation, 𝑘avg is,

𝑑𝐹
∫𝑎𝑏 𝑑𝐷
𝑑𝐷𝑡 𝑡 𝐹(𝑏) − 𝐹(𝑎) [5.3]
𝑘avg = =
𝑏−𝑎 𝑏−𝑎

An example of the average stiffness calculation can be demonstrated as follows:

For the M10 bolt:

𝑎 = 0, 𝑏 = 2

𝐹(𝑏) = 𝐹(2) = 11.095(2)3 − 86.726(2)2 + 221.71(2) = 185.276 kN

𝐹(𝑎) = 𝐹(0) = 0

𝐹(𝑏) − 𝐹(𝑎) 185.276 − 0


𝑘avg = = = 92.638 kN/mm
𝑏−𝑎 2−0

Then, the same calculations were repeated for the rest of the bolts. This can be

presented in the following table and figure.

Table 5.2: The bolt diameters and the calculated average stiffness

Bolt diameter (mm) 10 12 16 20 24

Interval 0 - 2.2 0 - 2.8 0 - 1.7 0 - 2.2 0 - 2.4

(kN/mm) 84.612 72.898 123.764 109.274 113.031

42
140

120
Average Stiffness (kkN/mm)

100

80

60

40

20

0
10 12 14 16 18 20 22 24
Bolt diameter (mm)

Figure 5.3: The plot of the average stiffness against bolt diameter showing the
region of common stiffnesses

From this analysis, it was observed that the relationship between the average stiffness

of the bolts obtained from the polynomial equations was not significantly clear. The

only pattern that can be seen from Figure 5.3 was that the stiffnesses of the bolts form

two distinct regions, where the smaller two bolts were in proximity to each other, and

the rest have almost the same stiffness. However, this confirms the nature of the F-Dt

relationships of the bolts as demonstrated in Figure 5.2. The same analytical procedure

can be applied to the F-Dt relationships for the global stiffness and the SHS.

43
For SHS element:

300
y = 154.11x4 - 344.66x3 + 164.69x2 + 249.73x
R² = 1
250
y = 304.26x4 - 631.23x3 + 316.76x2 + 226.59x
R² = 0.9999
200
Force (kN)

150

100 y = 373.92x4 - 717.82x3 + 332.98x2 + 225.24x


R² = 1

50 y = 393.61x4 - 641.11x3 + 239.03x2 + 241.87x


R² = 1
y = 173.16x4 - 334x3 + 102.47x2 + 254.01x
0 R² = 1
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Total Deformation (mm)

M10 M12 M16 M20


M24 Poly. (M10) Poly. (M12) Poly. (M16)
Poly. (M20) Poly. (M24)

Figure 5.4: F-Dt relationship of the SHS with different bolt diameters

Table 5.3: The R2 values and the polynomial regression equations of SHS with
different bolt diameters

Bolt 𝑭 𝑹𝟐

M10 𝐹 = 173.16𝐷𝑡 4 − 334𝐷𝑡 3 + 102.47𝐷𝑡 2 + 254.01𝐷𝑡 1

M12 𝐹 = 393.61𝐷𝑡 4 − 641.11𝐷𝑡 3 + 239.03𝐷𝑡 2 + 241.87𝐷𝑡 1

M16 𝐹 = 373.92𝐷𝑡 4 − 717.82𝐷𝑡 3 + 332.98𝐷𝑡 2 + 225.24𝐷𝑡 1

M20 𝐹 = 304.26𝐷𝑡 4 − 631.23𝐷𝑡 3 + 316.76𝐷𝑡 2 + 226.59𝐷𝑡 0.9999

M24 𝐹 = 154.11𝐷𝑡 4 − 344.66𝐷𝑡 3 + 164.69𝐷𝑡 2 + 249.73𝐷𝑡 1

44
Table 5.4: The bolt diameters and the calculated average stiffness of SHS

Bolt diameter (mm) 10 12 16 20 24

Interval 0 - 0.93 0 - 1.03 0 - 0.93 0 - 1.12 0 - 1.22

(kN/mm) 199.713 238.026 214.834 217.010 217.500

240

235

230
Average Stiffness (kN/mm)

225

220

215

210

205

200

195
10 12 14 16 18 20 22 24
Bolt diameter (mm)

Figure 5.5: The plot of the average stiffness of SHS against bolt diameter

45
For the global reaction:

300 y = -0.1068x4 + 3.0346x3 - 30.142x2 + 130.74x


R² = 0.9996
250
y = 0.7279x3 - 15.419x2 + 104.21x
R² = 0.997
200
Force (kN)

y = -0.1526x4 + 3.7689x3 - 32.763x2 + 125.39x


150
R² = 0.9998

100 y = -0.1307x4 + 3.2402x3 - 28.737x2 + 114.31x


R² = 0.9998
50 y = 1.9807x3 - 25.864x2 + 110.2x
R² = 0.9998
0
0 2 4 6 8 10 12 14
Total Deformation (mm)

M10 M12 M16 M20


M24 Poly. (M10) Poly. (M12) Poly. (M16)
Log. (M20) Poly. (M20) Poly. (M24)

Figure 5.6: F-Dt relationship of the global reaction with different bolt diameters

Table 5.5: The R2 values and the polynomial regression equations of the global
with different bolt diameters

Bolt 𝑭 𝑹𝟐

M10 𝐹 = 1.9807𝐷𝑡 3 − 25.864𝐷𝑡 2 + 110.2𝐷𝑡 0.9998

M12 𝐹 = −0.1307𝐷𝑡 4 + 3.2402𝐷𝑡 3 − 28.737𝐷𝑡 2 + 114.31𝐷𝑡 0.9998

M16 𝐹 = −0.1526𝐷𝑡 4 + 3.7689𝐷𝑡 3 − 32.763𝐷𝑡 2 + 125.39𝐷𝑡 0.9998

M20 𝐹 = 0.7279𝐷𝑡 3 − 15.419𝐷𝑡 2 + 104.21𝐷𝑡 0.997

M24 𝐹 = −0.1068𝐷𝑡 4 + 3.0346𝐷𝑡 3 − 30.142𝐷𝑡 2 + 130.74𝐷𝑡 0.9996

46
Table 5.6: The bolt diameters and the calculated average global stiffness

Bolt diameter (mm) 10 12 16 20 24

Interval 0-7 0 - 9.3 0 - 9.8 0 - 10.9 0 - 11.5

(kN/mm) 26.206 22.171 22.652 22.625 22.733

26.5

26

25.5
Average Stiffness (kkN/mm)

25

24.5

24

23.5

23

22.5

22

21.5
10 12 14 16 18 20 22 24
Bolt diameter (mm)

Figure 5.7: The plot of the average global stiffness against bolt diameter

Similar to the individual average stiffness of the bolts, the average global stiffness, and

the stiffness of the SHS in all setups show no correlation with the bolt diameter. This

is especially true for the SHS force versus displacement relationship, where the data

for all bolt diameters were relatively close to each other. In addition, there were some

anomalies found in the stiffness versus bolt diameter relationship. For example, in

Figure 5.7, the M10 bolt setup has the highest stiffness compared to the rest.

47
Meanwhile, the average stiffness of the SHS in the M12 bolt setup is the highest. Even

if the anomalies were omitted in the analysis, it was still impossible to establish any

meaningful relationship between the bolt diameter and the stiffness of the connection.

Therefore, more parameters need to be studied to look for any discernible pattern.

5.2 Part 2 – Effect of Bolt Steel Strength on the Stiffness

This part explores the use of two more bolt steel strengths, namely S275 and S355.

Here, the same methodology with the previous part was employed. However, the

independent parameter was the steel strength, and the controlled parameter was the

bolt diameter and the dimensions of the endplate and the SHS. Only an M16 bolt shank

is used in this section.

For the bolt shank:

250

200 y = 106.5x4 - 275.56x3 + 115.75x2 + 225.57x


R² = 0.9998
Force (kN)

150 y = 28.06x3 - 147.74x2 + 285.57x


R² = 0.9995

100 y = 36.83x4 - 105.82x3 + 8.1342x2 + 234.81x


R² = 0.9998

50

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Total deformation (mm)

S275 S355 Default


Poly. (S275) Poly. (S355) Poly. (Default)

Figure 5.8: F-Dt relationship of the M16 bolt shank with different bolt strengths

48
Table 5.7: The polynomial regression equations and the R2 values of the M16
bolt shank with different strengths

Bolt
strength 𝑭 𝑹𝟐
(MPa)
250 𝐹 = 36.83𝐷𝑡 4 − 105.82𝐷𝑡 3 + 8.1342𝐷𝑡 2 + 234.81𝐷𝑡 0.9998

275 𝐹 = 28.06𝐷𝑡 3 − 147.74𝐷𝑡 2 + 285.57𝐷𝑡 0.9995

355 𝐹 = 106.5𝐷𝑡 4 − 275.56𝐷𝑡 3 + 115.75𝐷𝑡 2 + 225.57𝐷𝑡 0.9998

Table 5.8: The bolt strengths and the calculated average bolt stiffness

Bolt strength (MPa) 250 275 355

Interval 0 - 1.84 0 - 1.55 0 - 1.38

(kN/mm) 120.945 123.987 139.63

145

140
Average Stiffness (kkN/mm)

135

130
y = 0.1822x + 74.731
R² = 0.994
125

120

115
250 270 290 310 330 350 370
Bolt strength (MPa)

Figure 5.9: The plot of the average stiffness of the M16 bolt against the bolt
strength

49
For the endplate:

250

y = -0.0167x4 + 2.31x3 - 28.562x2 + 124.19x


R² = 0.9999
200

150
Force (kN)

y = -0.0773x4 + 3.0551x3 - 31.221x2 + 125.5x


R² = 0.9999

100 y = -0.1564x4 + 3.8434x3 - 33.184x2 + 125.99x


R² = 0.9998

50

0
0 2 4 6 8 10 12
Total deformation (mm)

S275 S355 Default


Poly. (S275) Poly. (S355) Poly. (Default)

Figure 5.10: F-Dt relationship of the endplate with different bolt strengths

Table 5.9: The polynomial regression equations and the R2 values of the
endplate with different strengths

Bolt
strength 𝑭 𝑹𝟐
(MPa)
250 𝐹 = −0.1564𝐷𝑡 4 + 3.843𝐷𝑡 3 − 33.184𝐷𝑡 2 + 125.99𝐷𝑡 0.9998

275 𝐹 = −0.0773𝐷𝑡4 + 3.0551𝐷𝑡 3 − 31.221𝐷𝑡 2 + 125.5𝐷𝑡 0.9999

355 𝐹 = −0.0167𝐷𝑡 4 + 2.31𝐷𝑡 3 − 28.562𝐷𝑡 2 + 124.19𝐷𝑡 0.9999

50
Table 5.10: The bolt strengths and the calculated average endplate stiffness

Bolt strength (MPa) 250 275 355

Interval 0 - 9.83 0 - 6.29 0 - 5.96

(kN/mm) 22.577 30.755 32.48

35

30
Average Stiffness (kkN/mm)

25
y = 0.0766x + 6.1406
R² = 0.6304
20

15

10

0
250 270 290 310 330 350 370
Bolt strength (MPa)

Figure 5.11: The plot of the average stiffness of the endplate against the bolt
strength

51
For the SHS:

250

200 y = 402.89x4 - 773.12x3 + 337.19x2 + 233.91x


R² = 1

150
Force (kN)

y = 302.59x4 - 605.62x3 + 246.13x2 + 246.41x


R² = 1

100
y = 373.92x4 - 717.82x3 + 332.98x2 + 225.24x
R² = 1

50

0
0 0.2 0.4 0.6 0.8 1 1.2
Total deformation (mm)

S275 S355 Default


Poly. (S275) Poly. (S355) Poly. (Default)

Figure 5.12: F-Dt relationship of the SHS with different bolt strengths

Table 5.11: The polynomial regression equations and the R2 values of the SHS
with different strengths

Bolt
strength 𝑭 𝑹𝟐
(MPa)
250 𝐹 = 373.92𝐷𝑡4 − 717.82𝐷𝑡 3 + 332.98𝐷𝑡 2 + 225.24𝐷𝑡 1

275 𝐹 = 302.59𝐷𝑡4 − 605.62𝐷𝑡 3 + 246.15𝐷𝑡 2 + 246.41𝐷𝑡 1

355 𝐹 = 402.89𝐷𝑡 4 − 773.12𝐷𝑡 3 + 337.19𝐷𝑡 2 + 233.91𝐷𝑡 1

52
Table 5.12: The bolt strengths and the calculated average SHS stiffness

Bolt strength (MPa) 250 275 355

Interval 0 - 1.03 0 - 1.03 0 - 0.96

(kN/mm) 215.267 188.07 201.556

220

215
Average Stiffness (kkN/mm)

210

205

200

195

190

185
250 270 290 310 330 350 370
Bolt strength (MPa)

Figure 5.13: The plot of the average stiffness of the SHS against the bolt
strength

53
For the global reaction:

250

200 y = 0.0044x4 + 2.0544x3 - 27.643x2 + 123.34x


R² = 0.9998

150 y = -0.0558x4 + 2.7962x3 - 30.307x2 + 124.64x


Force (kN)

R² = 0.9998

y = -0.1526x4 + 3.7689x3 - 32.763x2 + 125.39x


100 R² = 0.9998

50

0
0 2 4 6 8 10 12
Total deformation (mm)

S275 S355 Default


Poly. (S275) Poly. (S355) Poly. (Default)

Figure 5.14: F-Dt relationship of the global reaction with different bolt strengths

Table 5.13: The polynomial regression equations and the R2 values of the global
reaction with different strengths

Bolt
strength 𝑭 𝑹𝟐
(MPa)
250 𝐹 = −0.1526𝐷𝑡4 + 3.7689𝐷𝑡 3 − 32.763𝐷𝑡 2 + 125.39𝐷𝑡 0.9998

275 𝐹 = −0.0558𝐷𝑡4 + 2.7962𝐷𝑡 3 − 30.307𝐷𝑡 2 + 124.64𝐷𝑡 0.9998

355 𝐹 = 0.0044𝐷𝑡 4 + 2.0544𝐷𝑡 3 − 27.643𝐷𝑡 2 + 123.34𝐷𝑡 0.9998

54
Table 5.14: The bolt strengths and the calculated average global stiffness

Bolt strength (MPa) 250 275 355

Interval 0 - 9.83 0 - 6.29 0 - 5.96

(kN/mm) 22.565 30.752 32.495

40

35

30
Average Stiffness (kkN/mm)

25
y = 0.0768x + 6.0675
20 R² = 0.6317

15

10

0
250 270 290 310 330 350 370
Bolt strength (MPa)

Figure 5.15: The plot of the average global stiffness against the bolt strength

For all the force-displacement relationships with these different bolt strengths, the

polynomial equations were successfully generated. The polynomial equations in this

part almost completely fit with the raw data generated from Ansys. This can be shown

by the R2 values which range from 0.9995 to 1. Unlike the individual bolt sizes, the

bolt strength as a parameter shows some degree of correlation between the strength

55
and the global as well as the local stiffnesses of most components, except the SHS.

This is especially true for the blind bolt itself, where the R2 value generated from the

linear correlation shows the value close to 1. For the global stiffness and the endplate,

both the force-displacement and the stiffness-bolt strength relationships show

similarities between each other. The correlation between the average stiffness and the

bolt strength was fair but significantly less than that of the bolt strength and the bolt

stiffness. This suggests that even by increasing the bolt strength, without the proper

support from the endplate, the stiffness for the entire blind-bolted connection would

not be improved. One of the ways to increase the endplate stiffness is by increasing its

thickness. This will be discussed in the next part of this parametric study.

5.3 Part 3 – Effect of Endplate Thickness on the Stiffness

This part discusses how the stiffness of the blind-bolted connection is affected by the

endplate thickness. Similarly, the force-displacement relationships were obtained for

the entire model and its individual components. There were four different endplate

thicknesses were used, including the one used in the previous parts, which are 10 mm,

16 mm, 20 mm, and 24 mm.

56
For the bolt shank:

250

y = 433.25x4 - 908.35x3 + 528.11x2 + 168.03x


R² = 0.9997
200
y = 269.66x4 - 666.59x3 + 414.97x2 + 172.98x
R² = 0.9997 y = 36.83x4 - 105.82x3 + 8.1342x2 + 234.81x
R² = 0.9998
150
Force (kN)

y = 199.36x4 - 446.33x3 + 204.52x2 + 191.58x


R² = 0.9997

100

50

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Total deformation (mm)

10 mm 16 mm 20 mm 24 mm
Poly. (10 mm) Poly. (16 mm) Poly. (20 mm) Poly. (24 mm)

Figure 5.16: F-Dt relationship of the M16 bolt shank with different endplate
thicknesses

Table 5.15: The polynomial regression equations and the R2 values of the M16
bolt shank with different endplate thicknesses

Endplate
thickness 𝑭 𝑹𝟐
(mm)
10 𝐹 = 199.36𝐷𝑡4 − 446.33𝐷𝑡 3 + 204.52𝐷𝑡 2 + 191.58𝐷𝑡 0.9997

16 𝐹 = 36.83𝐷𝑡4 − 105.82𝐷𝑡 3 + 8.1342𝐷𝑡 2 + 234.81𝐷𝑡 0.9998

20 𝐹 = 269.66𝐷𝑡4 − 666.59𝐷𝑡 3 + 414.97𝐷𝑡 2 + 172.98𝐷𝑡 0.9997

24 𝐹 = 433.25𝐷𝑡4 − 908.35𝐷𝑡 3 + 528.11𝐷𝑡 2 + 168.03𝐷𝑡 0.9997

57
Table 5.16: The endplate thicknesses and the calculated average bolt stiffness

Endplate thickness (mm) 10 16 20 24

Interval 0-1 0-1 0-1 0-1

(kN/mm) 139.800 173.954 191.020 223.040

250.000

200.000
Average Stiffness (kkN/mm)

150.000

y = 5.7746x + 80.899
100.000 R² = 0.988

50.000

0.000
10 12 14 16 18 20 22 24 26
Endplate thickness (mm)

Figure 5.17: The plot of the average stiffness of the M16 bolt against the
endplate thicknesses

58
For the endplate:

250
y = -6.0421x4 - 6.5351x3 + 11.692x2 + 161.55x
R² = 1

200 y = 7.8758x4 - 40.665x3 + 35.339x2 + 130.26x


R² = 1

y = -0.1564x4 + 3.8434x3 - 33.184x2 + 125.99x


150 R² = 0.9998
Force (kN)

y = -0.1282x4 + 3.2001x3 - 27.625x2 + 102.17x


R² = 0.9998
100

50

0
0 2 4 6 8 10 12
Total deformation (mm)

10 mm 16 mm 20 mm 24 mm
Poly. (10 mm) Poly. (16 mm) Poly. (20 mm) Poly. (24 mm)

Figure 5.18: F-Dt relationship of the endplates

Table 5.17: The polynomial regression equations and the R2 values of the
endplates

Endplate
thickness 𝑭 𝑹𝟐
(mm)
10 𝐹 = −0.1282𝐷𝑡4 + 3.2001𝐷𝑡 3 − 27.625𝐷𝑡 2 + 102.17𝐷𝑡 0.9998

16 𝐹 = −0.1564𝐷𝑡4 + 3.8434𝐷𝑡 3 − 33.184𝐷𝑡 2 + 125.99𝐷𝑡 0.9998

20 𝐹 = 7.8758𝐷𝑡4 − 40.665𝐷𝑡 3 + 35.339𝐷𝑡 2 + 130.26𝐷𝑡 1

24 𝐹 = −6.0421𝐷𝑡4 − 6.5351𝐷𝑡 3 + 11.692𝐷𝑡 2 + 161.55𝐷𝑡 1

59
Table 5.18: The endplate thicknesses and its calculated average stiffness

Endplate thickness (mm) 10 16 20 24

Interval 0 - 9.8 0 - 9.8 0 - 2.6 0 - 1.7

(kN/mm) 18.122 22.705 85.671 132.855

140

120
Average Stiffness (kkN/mm)

100

80

60 y = 8.4838x - 83.628
R² = 0.854

40

20

0
10 12 14 16 18 20 22 24 26
Endplate thickness (mm)

Figure 5.19: The plot of the average stiffness of the M16 bolt against the
endplate thicknesses

60
For the SHS:

250

y = -0.6164x4 - 104.28x3 + 66.859x2 + 274.35x


R² = 1
200

150
Force (kN)

y = -98.612x2 + 317.35x
R² = 0.9994
y = 373.92x4 - 717.82x3 + 332.98x2 + 225.24x
100 R² = 1

y = 1333.3x4 - 1630.2x3 + 562.67x2 + 206.89x


50 R² = 0.9997

0
0 0.2 0.4 0.6 0.8 1 1.2
Total deformation (mm)

10 mm 16 mm 20 mm 24 mm
Poly. (10 mm) Poly. (16 mm) Poly. (20 mm) Poly. (24 mm)

Figure 5.20: F-Dt relationship of the SHS with different endplate thicknesses

Table 5.19: The polynomial regression equations and the R2 values of the SHS
with different endplate thicknesses

Endplate
thickness 𝑭 𝑹𝟐
(mm)
10 𝐹 = 1333.3𝐷𝑡4 − 1630.2𝐷𝑡3 + 562.67𝐷𝑡2 + 206.89𝐷𝑡 0.9997

16 𝐹 = 373.92𝐷𝑡4 − 717.82𝐷𝑡 3 + 332.98𝐷𝑡 2 + 225.24𝐷𝑡 1

20 𝐹 = −98.612𝐷𝑡 2 + 317.35𝐷𝑡 0.9994

24 𝐹 = −0.6164𝐷𝑡4 − 104.28𝐷𝑡 3 + 66.859𝐷𝑡 2 + 274.35𝐷𝑡 1

61
Table 5.20: The endplate thicknesses and the average stiffness of the SHS

Endplate thickness (mm) 10 16 20 24

Interval 0 - 0.6 0 - 0.6 0 - 0.6 0 - 0.6

(kN/mm) 245.613 247.386 258.183 276.791

280.000

275.000

270.000
Average Stiffness (kkN/mm)

265.000

y = 2.1628x + 219.14
260.000 R² = 0.8136

255.000

250.000

245.000
y = 0.2418x2 - 5.9956x + 281.39
240.000 R² = 1

235.000
10 12 14 16 18 20 22 24 26
Endplate thickness (mm)

Figure 5.21: The plot of the average stiffness of the SHS against the endplate
thicknesses

62
For the global reaction:

250 y = -27.681x3 + 34.812x2 + 153.77x


R² = 0.9999
y = 16.071x4 - 76.967x3 + 79.735x2 + 115.37x
R² = 0.9999
200
y = -0.1526x4 + 3.7689x3 - 32.763x2 + 125.39x
R² = 0.9998

150
Force (kN)

y = -0.1254x4 + 3.1433x3 - 27.286x2 + 101.67x


R² = 0.9998
100

50

0
0 2 4 6 8 10 12
Total deformation (mm)

10 mm 16 mm 20 mm 24 mm
Poly. (10 mm) Poly. (16 mm) Poly. (20 mm) Poly. (24 mm)

Figure 5.22: F-Dt relationship of the global reaction with different endplate
thicknesses

Table 5.21: The polynomial regression equations and the R2 values of the SHS
with different endplate thicknesses

Endplate
thickness 𝑭 𝑹𝟐
(mm)
10 𝐹 = −0.1254𝐷𝑡4 + 3.1433𝐷𝑡3 − 27.286𝐷𝑡2 + 101.67𝐷𝑡 0.9998

16 𝐹 = −0.1526𝐷𝑡4 + 3.7689𝐷𝑡 3 − 32.763𝐷𝑡 2 + 125.39𝐷𝑡 0.9998

20 𝐹 = 16.071𝐷𝑡4 − 76.967𝐷𝑡 3 + 79.735𝐷𝑡 2 + 115.37𝐷𝑡 0.9999

24 𝐹 = −27.681𝐷𝑡 3 + 34.812𝐷𝑡 2 + 153.77𝐷𝑡 0.9999

63
Table 5.22: The endplate thicknesses and the average global stiffness

Endplate thickness (mm) 10 16 20 24

Interval 0 - 9.8 0 - 9.8 0 - 2.6 0 - 1.7

(kN/mm) 18.124 22.652 84.848 132.952

140

120
Average Stiffness (kkN/mm)

100

80

60 y = 8.471x - 83.599
R² = 0.853

40

20

0
10 12 14 16 18 20 22 24 26
Endplate thickness (mm)

Figure 5.23: The plot of the average global stiffness against the endplate
thicknesses

In this part, all the polynomial regression equations almost perfectly fit with the raw

data from the Ansys FEM. After the stiffnesses were obtained from the regression

equations, it was apparent that the endplate thickness did affect the stiffness of the

other components in the blind-bolted connection. This can be supported by the R2

values ranging between 0.8 until 0.988.

64
CHAPTER 6: CONCLUSIONS

6.1 Summary and Recommendations

This research has successfully performed the numerical analysis of a series of blind-

bolted connection systems. The process of discretisation of the 3D model was

performed effectively by using appropriate variable mesh sizes, ranging from 2.5 mm

for the bolt to 25 mm for the SHS. The analyses in the Ansys Mechanical were

performed under the convergence criteria, where the force convergence falls below the

force criterion line. These are the main takeaways from this study:

• The development of stresses inside the bolted connection was observed, and

there were some patterns on how the stresses propagate. The bolt pretension

created internal compressive stress which was induced by the tightening

process of the bolt. The endplates tended to yield first as expected as the

resistance or the support at the endplate were the least. The yielding of the

endplate pushed the top part of the bolt sideways, therefore inducing shear

stresses to the bolt.

• From the parametric analysis, all the parameters show some variation in the

correlation with the stiffness of the blind-bolted connections. There was no

noticeable effect from the changing of the bolt size to the overall stiffness.

Meanwhile, when the bolt strength increases, the overall stiffness also

increases, except for the SHS. However, it was found that the endplate

thickness had the strongest effect on the stiffness of the blind-bolted

connection.

65
• More studies can be proposed to establish relationship between the bolt size

and other parameters.

6.2 Contribution to entrepreneurship, economic, environmental, and socio-

cultural progress

In terns of the entrepreneurship opportunity, the advent of blind bolt can benefit in a

number of ways. One of them is it can path some new ways to the development of

proprietary bolt technology. This will accelerate the development of more

improvements to the existing bolted connection method and technology.

In terms of the economy, the blind bolts are capable to simplify some steel

construction methods, particularly the ones that involve steel hollow sections. This can

be achieved by eliminating the need to install the bolted connection from both sides of

the hollow sections. This will in turn save the construction time, and therefore, the

construction cost as well as these two quantities are highly dependent on each other.

From the simplification of the construction method, more and more structures

can be constructed by using steel. This poses some advantages to the environment.

This is because structural steel is highly recyclable. The recyclability of this material

is becoming more crucial in this modern era as many structures are aging and showing

a lot of serviceability problems. The blind bolts can aid this process from easier process

of disassembly.

Lastly, blind bolt can ease the assembly of building materials by prefabrication.

This is perfect for the socio-cultural progress as this enables custom modular buildings

with beautiful architecture to be designed and constructed.

66
CHAPTER 7: REFERENCES

Blind Bolt Company. (2020). Blind Bolt Technical Data.

https://www.blindboltasia.com/blind-bolt-technical-data/

Cabrera, M., Tizani, W., Mahmood, M., & Shamsudin, M. F. (2020). Analysis of

Extended Hollo-Bolt connections: Combined failure in tension [Article].

Journal of Constructional Steel Research, 165.

https://doi.org/10.1016/j.jcsr.2019.105766

Cabrera, M., Tizani, W., & Ninic, J. (2021). A review and analysis of testing and

modeling practice of extended Hollo-Bolt blind bolt connections. Journal of

Constructional Steel Research, 183, 106763.

https://doi.org/https://doi.org/10.1016/j.jcsr.2021.106763

European Committee. (2005). EN 1990:2002 Eurocode: Basis of structural design

Javora, A., & Skejić, D. (2017). Resis ance assessmen of beam-to-column joints with

different blind bolt systems. Tehnicki vjesnik/Technical Gazette, 24(4).

Lee, J., Goldsworthy, H. M., & Gad, E. F. (2011a). Blind bolted moment connection

to sides of hollow section columns. Journal of Constructional Steel Research,

67(12), 1900-1911. https://doi.org/https://doi.org/10.1016/j.jcsr.2011.06.007

Lee, J., Goldsworthy, H. M., & Gad, E. F. (2011b). Blind bolted moment connection

to unfilled hollow section columns using extended T-stub with back face

support. Engineering Structures, 33(5), 1710-1722.

https://doi.org/https://doi.org/10.1016/j.engstruct.2011.02.008

Li, Y.-L., & Zhao, X.-L. (2021). Experimental study on stainless steel blind bolted T-

stub to square hollow section connections [Article]. Thin-Walled Structures,

167. https://doi.org/10.1016/j.tws.2021.108259

67
Liu, Y., Chen, J., Zhang, X., & Tan, D. (2018). Fatigue behaviour of blind bolts under

tensile cyclic loads [Article]. Journal of Constructional Steel Research, 148,

16-27. https://doi.org/10.1016/j.jcsr.2018.05.019

Mesquita, A., da Silva, L. S., & Jordão, S. (2021). 3D numerical models of steel joints

with hollo-bolts–a comparison with experimental results. In Modern Trends in

Research on Steel, Aluminium and Composite Structures (pp. 328-334).

Routledge.

Pascual, A. M., Romero, M. L., & Tizani, W. (2015). Fire performance of blind-bolted

connections to concrete filled tubular columns in tension. Engineering

Structures, 96, 111-125.

Pokharel, T. (2016). Development and evaluation of building frames with blind-bolted

connections http://hdl.handle.net/11343/128062

Tahir, M. M., Mohammadhosseini, H., Ngian, S. P., & Effendi, M. K. (2018). I-beam

to square hollow column blind bolted moment connection: Experimental and

numerical study. Journal of Constructional Steel Research, 148, 383-398.

https://doi.org/https://doi.org/10.1016/j.jcsr.2018.06.012

Tizani, W., & Pitrakkos, T. (2015). Performance of T-stub to CFT joints using blind

bolts with headed anchors. Journal of Structural Engineering, 141(10),

04015001.

Zheng, X., & Xia, W. (2009). Numerical Simulation of Blind Hole Bolt Connection

with 3-D Finite Element Approach. 2009 Second International Conference on

Information and Computing Science,

68
APPENDICES

Force convergence for all models

M10 Bolt:

M12 Bolt

M16 bolt:

A1
M20 bolt:

M24 bolt:

A2
A3

You might also like