University of Southern Queensland

FACULTY OF HEALTH ENGINEERING AND SCIENCES

Finite Element Analysis of Fillet

Welded Joint

A dissertation submitted by

Rafiqul Islam Mohammed

In fulfilment of the requirements of

ENG4111/ENG4112 Research Project

Towards the degree of

Bachelor of Engineering (Mechanical)

Submitted: October 2015

Abstract
T-joint fillet welding is the most common welding in engineering applications.
Transport vehicles, marine ships, mobile plant equipment are few examples where
fillet welding are used extensively. Analysis of welded structures are still remains a
challenge for the designer to produce desired output results. In welding process rapid
heating and cooling introduced residual stress and geometrical deformations. Heat
effected zone play pivotal role in determining the strength of a welded joint which
changes the properties of parent material and reduce the strength after welding
operation. There are many case which structures are continuously under cyclic loading
when the fatigue life of the welded joints are a major design consideration

The aim of this project is to analyse the normal stress and fatigue life of fillet welded
joints using computer modelling and experiments. Finite element based tool ANSYS
Workbench 15.0 was been used to analyse the normal stress and the fatigue life under
cyclic loading. Computer model of the joint developed using three different types of
material which was parent metal, heat affected zone metal and weld metal.
Experimental tests were carried out at USQ laboratory on double side welded T-joints.
Grade 250 Structural steel was used to prepare specimen and gas metal arc welding
(GMAW) process applied to welding the joints.

The ultimate purpose of the project has been achieved with developing techniques of
the finite element analysis of fillet welded joint. The experimental investigation
validate the performance of the FEA analysis results were found 1.2% error on tensile
test. The experiment yield stress was found 263.4 MPa and simulation yield stress at
the same location appears 266.7 MPa. In order to calculate fatigue life of welded joint
used iterative process to define stress at one million cycle. The analysis found 274
MPa stress and 7740 cycle fatigue life applying yield load. After reduced load at 12kN
and found the fatigue life one million cycle where shows 88 MPa stress which is 35%
of yield stress. So that designer can consider 35% of yield strength when design
structure for fluctuating and repeated loading conditions.

i
 University of Southern Queensland

FACULTY OF HEALTH ENGINEERING AND SCIENCE

ENG4111/ENG4112 Research Project

Limitation of Use

The council of the University of Southern Queensland, its Faculty of Health,

Engineering & Sciences, and the staff of the University of Southern Queensland, do
not accept any responsibility for the truth, accuracy or completeness of material
contained within or associated with this dissertation.

Persons using all or any part of this material do so at their own risk, and not at the risk
of the Council of the University of Southern Queensland, its Faculty of Health,
Engineering & Sciences or the staff of the University of Southern Queensland.

This dissertation reports an educational exercise and has no purpose or validity beyond
this exercise. The sole purpose of the course pair entitled “Research Project” is to
contribute to the overall education within the students chosen degree program. This
document, the associated hardware, software, drawings, and other material set out in
the associated appendices should not be used for any other purpose: if they are so used,
it is entirely at the risk of the user.

Dean

Faculty of Health, Engineering & Sciences

ii
 Certification

I certify that the ideas, designs and experimental work, results, analysis and
conclusions set out in this dissertation are entirely my own effort, except where
otherwise indicated and acknowledged.

I further certify that the work is original and has not been previously submitted for
assessment in any other course or institution, except where specifically stated.

Rafiqul Mohammed

-----------------------
Signature
16/09/2105
-----------------------
Date

iii
 Acknowledgements

This project was carried out under the principal supervisor Mr. Chris Snook. I would
like to offer my sincere appreciation to Mr. Chris for his support and guidance
throughout the completion of this dissertation. I am very grateful for his supports and
innovative ideas which make this project in reality.

I would like to thank my friends and family for their support and assistance throughout
this project, whether deliberate or not. I like would also to thank Dr. Ray Malpress for
his generous time for useful tips to using ANSYS workbench.

Rafiqul Mohammed

University of Southern Queensland

October 2015

iv
Table of Contents

Chapter 1 – Introduction .......................................................................................... 1

1.1 Overview ....................................................................................................... 1

1.2 Background ................................................................................................... 1

1.3 Project topic ................................................................................................... 5

1.4 Scope of the research ..................................................................................... 6

1.5 Project objectives .......................................................................................... 7

1.6 Expected outcomes and significance of the study ......................................... 7

1.7 Dissertation outline........................................................................................ 8

Chapter 2 – Literature Review .................................................................................. 9

2.1 General .......................................................................................................... 9

2.2 Welding process ............................................................................................ 9

2.3 Brief History of FEA ................................................................................... 10

2.4 Effect of residual stress ............................................................................... 10

2.5 Effect of welding sequence ......................................................................... 18

2.6 Fatigue life calculation ................................................................................ 21

2.7 Fatigue life improvement ............................................................................ 24

2.8 Computational challenges in welding simulation ....................................... 26

2.9 Thermal and mechanical finite element analysis......................................... 26

2.10 Heat affected zone ....................................................................................... 27

2.11 Welding deformation ................................................................................... 28

2.12 Weld metal property .................................................................................... 30

2.13 Visualise stress and deformation ................................................................. 31

2.14 Adaptive mesh technique ............................................................................ 32

2.15 Further Research .......................................................................................... 34

Chapter 3 – Methodology and Project planning ..................................................... 35

v
 3.1 Introduction ................................................................................................. 35

3.2 Methodology ............................................................................................... 35

3.2.1 Experimental Investigation .................................................................. 35

3.2.2 Finite Element analysis ........................................................................ 38

Chapter 4 – Finite Element Simulation ................................................................... 45

4.1 Introduction ................................................................................................. 45

4.2 Background ................................................................................................. 45

4.3 ANSYS simulation ...................................................................................... 47

4.3.1 Stress analysis ...................................................................................... 49

4.3.2 Fatigue life analysis.............................................................................. 57

Chapter 5 – Results and Discussion ........................................................................ 65

5.1 Introduction ................................................................................................. 65

5.2 FEA Simulation results ............................................................................... 65

5.3 Experiment results ....................................................................................... 66

5.4 Discussions .................................................................................................. 69

Chapter 6 – Conclusions ......................................................................................... 71

6.1 Chapter overview ........................................................................................ 71

6.2 Achievement of the project ......................................................................... 71

6.3 Further work and recommendation ............................................................. 72

6.4 Conclusion ................................................................................................... 72

List of References ...................................................................................................... 73

List of Appendices ..................................................................................................... 77

A - Project specification ......................................................................................... 77

B - Resource analysis ............................................................................................. 78

Cost involves ...................................................................................................... 78

C - Risk management ............................................................................................. 79

D - Project timeline ................................................................................................ 86

vi
Table of Figures
Figure 1-1: fillet welding diagram (Welded Connections)........................................... 1
Figure 1-2: Structure failed due to fatigue (Leap Australia 2014) .............................. 3
Figure 1-3: Fatigue cracking in a bridge girder from weld defects (Haghani 2012) .. 3
Figure 1-4: Example of radial and tangential cracks (Haghani 2012) ......................... 4
Figure 1-5: Analysis of fillet weld using FEA (Aygul 2012) ...................................... 5
Figure 2-1: H-type beam(Seok, Suh & Park 1999) .................................................... 11
Figure 2-2 Longitudinal residual stress vs location from the centreline of weld bead
(Seok, Suh & Park 1999) ........................................................................................... 11
Figure 2-3:Longitudinal residual stress from the bottom on weld metal and parent
metal.(Seok, Suh & Park 1999).................................................................................. 12
Figure 2-4: Finite element meshes of the T-joint fillet weld(Teng et al. 2001) ......... 13
Figure 2-5: Transverse residual stress distribution along the X direction (Teng et al.
2001) .......................................................................................................................... 13
Figure 2-6: Longitudinal residual stress distribution along the X-direction .............. 14
Figure 2-7: Transverse residual stress distribution for different penetration depths . 14
Figure 2-8: Geometry of T joint (Perić et al. 2014) ................................................... 15
Figure 2-9: : Deflection and residual stress of fillet welding at direction along the line
C-C, from figure 10 (Perić et al. 2014) ...................................................................... 16
Figure 2-10: Transverse stress distribution at the top surface (a) and the bottom surface
(b) (PadmaKumari 2013) ........................................................................................... 17
Figure 2-11: Residual stress profile at cross section of flange (Chang & Lee 2009). 17
Figure 2-12: Welding sequences (Gannon et al. 2010) .............................................. 18
Figure 2-13: Deformation shapes with von misses stress contours (Gannon et al. 2010)
.................................................................................................................................... 19
Figure 2-14: Angular change in T-fillet joint (A) free restrained (B) Restricted....... 19
Figure 2-15: Variation of welding sequence, (a) WS-1, (b) WS-2, (c) WS-3, (d) WS-4
(Syahroni & Hidayat 2012). ....................................................................................... 20
Figure 2-16: Peak temperature for each welding sequence (Syahroni & Hidayat 2012).
.................................................................................................................................... 21
Figure 2-17: Distribution of transverse residual stress along the fillet weld for each
welding sequence (Syahroni & Hidayat 2012) .......................................................... 21
Figure 2-18: Linear (Left) Quadratic (Right) extrapolation based on stress derived
from strain gage measurement (Rother & Rudolph 2011) ......................................... 22

vii
Figure 2-19: Classification of some weld improvement methods(Kirkhope et al. 1999)
.................................................................................................................................... 24
Figure 2-20: Typical improvement in fatigue strength of mild steel fillet weld resulting
from selected weld toe improvement technic (Maddox 2002)................................... 25
Figure 2-21: Basic design curve for welded tubular plates in the air (Joshi & Price
2009) .......................................................................................................................... 25
Figure 2-22: Heat distribution of melted zone(Rabih Kamal Kassab1 2012) ............ 27
Figure 2-23: High temperature HAZ showing grain coarsened region (Woolling &
Carrouge 2002)........................................................................................................... 28
Figure 2-24: Heat affected zone after welding (Welding Inspection Cswip 2010) .. 28
Figure 2-25: deformation after welding (Deng, Liang & Murakawa 2007) .............. 29
Figure 2-26: Combined heat source in the fillet welding joint (Deng, Liang &
Murakawa 2007) ........................................................................................................ 29
Figure 2-27: Distribution of angular distortion along welding line (Deng, Liang &
Murakawa 2007) ........................................................................................................ 30
Figure 2-28: Experimental stress strain carves of parent pipe and GMAW, FCAW weld
metal (Pisarski, Tkach & Quintana 2004) .................................................................. 31
Figure 2-29: Bureau Veritas use FE map to help designs conform to regulations
(SIEMENS 2015) ....................................................................................................... 32
Figure 2-30: Stress sensitivity to mesh density (Hale 2014)...................................... 33
Figure 2-31: Relative difference in stresses at shared nodes for coarse mesh (L) and
finer mesh (R) (Hale 2014) ........................................................................................ 33
Figure 3-1: Geometry of test sample .......................................................................... 36
Figure 3-2: Direction of load applied ......................................................................... 37
Figure 3-3: basic model of welded joint .................................................................... 38
Figure 3-4: Exploded view of the model ................................................................... 39
Figure 3-5 Proof of fillet welding .............................................................................. 40
Figure 3-6: Mesh development .................................................................................. 41
Figure 3-7: Mesh development .................................................................................. 42
Figure 3-8: Mesh element size Vs stress .................................................................... 43
Figure 3-9: Model set-up tree..................................................................................... 44
Figure 4-1: General procedure of finite element analysis (Hobbacher 2009) ............ 46
Figure 4-2 : S/N curves for welded and un-welded carbon steel (TWI Fatigue testinng
2015) .......................................................................................................................... 46

viii
Figure 4-3 Model with force direction and fixed support .......................................... 48
Figure 4-4: Stress distribution on the fillet welded specimen .................................... 49
Figure 4-5: Stress distribution in the welded region .................................................. 50
Figure 4-6: Stress distribution with 60kN load .......................................................... 50
Figure 4-7: Stress distribution with 50kN load .......................................................... 51
Figure 4-8: Stress distribution with 40kN load applied ............................................. 52
Figure 4-9: Stress distribution with 35kN load applied ............................................. 52
Figure 4-10: Stress distribution with 30kN load ........................................................ 53
Figure 4-11: Stress distribution with 25kN load ........................................................ 54
Figure 4-12: Stress distribution with 20kN load ........................................................ 54
Figure 4-13: Stress distribution with 18kN load ........................................................ 55
Figure 4-14: Stress distribution with 16kN load ........................................................ 55
Figure 4-15: Stress distribution with 14kN load ........................................................ 56
Figure 4-16: Stress distribution with 12kN load ........................................................ 56
Figure 4-17 Fatigue life with 60kN load .................................................................... 57
Figure 4-18 Fatigue life with 50kN load .................................................................... 58
Figure 4-19 Fatigue life with 40kN load .................................................................... 58
Figure 4-20 Fatigue life with 35kN load .................................................................... 59
Figure 4-21 Fatigue life with 30kN load .................................................................... 60
Figure 4-22 Fatigue life with 25kN load .................................................................... 61
Figure 4-23 Fatigue life with 20kN load .................................................................... 62
Figure 4-24 Fatigue life with 18kN load .................................................................... 62
Figure 4-25 Fatigue life with 16kN load .................................................................... 63
Figure 4-26 Fatigue life with 14kN load .................................................................... 64
Figure 4-27 Fatigue life with 12kN load .................................................................... 64
Figure 5-1: S-N carve of welded joint........................................................................ 66
Figure 5-2: Motion of tensile test ............................................................................... 66
Figure 5-3: Brocken test pieces after tensile test ....................................................... 67
Figure 5-4: Load elongation graph of item 2 & 3 ...................................................... 67
Figure 5-5: Load elongation graph of item 4,5 and 6 ................................................ 68
Figure 5-6: Comparison of experiment test with model simulation .......................... 69
Figure 5-7: Zoomed view of stress concentrated area ............................................... 70

ix
List of Table
Table 2-1:Computed structural notch stress including derived fatigue life (Rother &
Rudolph 2011)............................................................................................................ 22
Table 2-2: Fatigue test result of fillet welded A515 Steel specimen (Shen & Clayton
1996) .......................................................................................................................... 23
Table 2-3: Tensile properties of grade X100 pipe (Pisarski, Tkach & Quintana 2004)
.................................................................................................................................... 30
Table 4-1: Calculation of stress with 60kN load ........................................................ 51
Table 4-2: Calculation of stress with 50kN load ........................................................ 51
Table 4-3: Average stress of 40kN load ..................................................................... 52
Table 4-4: Average stress of 35kN load ..................................................................... 53
Table 4-5: Average stress of 30kN load ..................................................................... 53
Table 4-6: Average stress of 25kN load ..................................................................... 54
Table 4-7 Average stress of 20kN load ...................................................................... 54
Table 4-8: Average stress of 18kN load ..................................................................... 55
Table 4-9: average stress of 16kN load ...................................................................... 55
Table 4-10: Average stress of 14kN load ................................................................... 56
Table 4-11: Average stress of 12kN load ................................................................... 56
Table 4-12: Calculation of fatigue life with 60kN load ............................................. 57
Table 4-13: Calculation of fatigue life with 50kN ..................................................... 58
Table 4-14: Calculation of fatigue life with 40kN ..................................................... 59
Table 4-15: Calculation of fatigue life with 35kN ..................................................... 59
Table 4-16: Calculation of fatigue life with 30kN ..................................................... 60
Table 4-17: Calculation of fatigue life with 25kN ..................................................... 61
Table 4-18: Calculation of fatigue life with 20kN ..................................................... 62
Table 4-19: Calculation of fatigue life with 18kN ..................................................... 63
Table 5-1: Results of stress and fatigue life of different load combination ............... 65
Table 5-2: Test results of experiment......................................................................... 68
Table 6-1: project timeline ......................................................................................... 86

x
 Chapter 1 – Introduction

1. Chapter 1 – Introduction

1.1 Overview
The aim of this chapter is to give an insight on the background of fillet welding,
purpose of the study, and research objectives of the project. An understanding on the
welding behaviour under real world situation is presented. This information will be
relevant throughout the dissertation and underlining the comparison of experiment test
results with simulation results.

1.2 Background
Welding is the most commonly used process for permanent joining of machine parts
and structures. Welding is a fabrication process which joins materials (metals) or
thermoplastics, by causing union (A. Thirugnanam 2014). In the joining process of
welding application uses heat and/or pressure, with or without the addition of filler
material. Various auxiliary materials, e.g. shielding gases, flux or pastes, may be used
to make the process possible or to make it easier. The energy required for welding is
supplied from outside sources.

Fillet welding is the process of joining two pieces of metal together whether
perpendicular or in angle between 80-100 degrees (AWS 2010). This welds are
commonly referred to as Tee joints which is perpendicular to each other’s or lap joints
which are overlap one another and welded at the edges. Figure 1-1 is the example of
fillet weld.

Figure 1-1: fillet welding diagram (Welded Connections)

1
Chapter 1 – Introduction

Due to the influence of the welding residual stress, residual plastic deformation, heat
affected zone and stress concentration effect the fatigue life of welded components is
far lower than the parent metal (Yang, Zou & Deng 2015).

Fatigue life is most common word in engineering design, it simply means life of
structure under repeated or fluctuating loading. Fatigue failure of the welded structures
remains the most common type of failure (Richards 1969). Repeated or fluctuating
load disturbs the material strengths and it causes the initiation of cracks. Neither metals
nor the welds joining them are as smooth as they look, they have pits, grooves and
cracks (Hicks 1987). The pits, grooves and cracks in a metal under load causes high
strains over very small areas. Fluctuating loads will create small tears or fractures
which increases in length with each application of the load.

In general, welded joints are more susceptible to fatigue cracking in comparison with
bolted joints (A. Thirugnanam 2014). Residual stress and stress concentrations due to
weld geometry induce micro-cracks that’s are often accelerate fatigue damage. Failure
caused by fatigue can be minor or catastrophic. Fatigue is a failure under repeated or
otherwise varying load which never reaches a level sufficient to cause failure in single
application. The consequences of failure are often very costly and it is estimated that
80-90% of all structural failures are caused by fatigue (Leap Australia 2014). Another
statics shows 70-90% of the welded structures invalidation accidents in the past several
years were caused by fatigue failure (Yang, Zou & Deng 2015). Leap Australia shows
an example in Figure 1-2 the damage of fatigue failure. This is a structural failure due
to fatigue.

2
Chapter 1 – Introduction

Figure 1-2: Structure failed due to fatigue (Leap Australia 2014)

Fatigue cracking on bridge structures are quite often occurred where always induced
dynamic loads. The crack started from weld defects at the intersection between the
filleted welds connecting to the longitudinal stiffeners to the girder web and the butt
welds made for transverse splices in the longitudinal stiffener. Fatigue damage
involving web crack on a girder, as well as in flange to web weld caused by the radial
stress in Figure 1-3.

Figure 1-3: Fatigue cracking in a bridge girder from weld defects (Haghani 2012)

The damage was caused by the fatigue cracking on the structure was found and
reported that the connection between flanges to the web was made by fillet welding
(Haghani 2012). The cracks on the girder web of the bridge can grow with radial stress
component as seen in Figure 1-4

3
Chapter 1 – Introduction

Figure 1-4: Example of radial and tangential cracks (Haghani 2012)

As the stories of building getting higher and the spans of bridge are getting longer,
there are greater demands for high performance joints capability. T-joint fillet welds
are extensively employed in various engineering applications. The welded joints has
residual stress which interrupt the life of structures.

Processes for the fatigue analysis of metallic structures are now well defined, both in
the stress-life or strain-life regimes (Aygul 2012). It has been recognized for some
considerable time that the design of fabricated welded structures can be heavily
dependent on the fatigue life of welded regions. Finite Element tools are often used to
analyse complex welded joints. Figure 1-5 shows an analysis of a multi-sided weld
joint using FEA. The optimization of such designs is therefore dependant on having
good predictions of the fatigue life at the weld.

The fatigue life of welded metallic structures is dependent on the increased levels of
stress found at the weld toe, root or throat due to weld geometry and the reduction in
material properties with the heat affected zone. In addition to this, the amount of fusion
of the weld into the parent metal and whether there is hot or cold fusion at a weld end
all add to the inherent variability of welded structures. In short, the prediction of a
robust fatigue life at a weld is extremely difficult.

This project was carried out FEA analysis with an overview of the work in the area of
weld fatigue. Its includes industrial real world applications of current weld life
prediction and modelling methods which gave an insight into the decisions to be taken
to achieve robust results.

4
Chapter 1 – Introduction

Figure 1-5: Analysis of fillet weld using FEA (Aygul 2012)

The way finite element analysis obtains the temperatures, residual stresses, and the
load application, or other desired unknown parameters in the finite element model are
by minimizing an energy functional. An energy functional consists of all the energies
associated with the particular finite element model. Based on the law of conservation
of energy, the finite element energy functional must equal zero. The finite element
method obtains the correct solution for any finite element model by minimizing the
energy functional. The minimum of the functional is found by setting the derivative of
the functional with respect to the unknown grid point potential for zero. Thus, the basic
𝜕𝐹
equation for finite element analysis is 𝜕𝑃 = 0

Where F is the energy functional and p is the unknown grid point potential in
mechanics, the potential displacement to be calculated. This is based on the principle
of virtual work, which states that if a particle is under equilibrium (Chang & Lee 2009),
under a set of a system of forces, then for any displacement, the virtual work is zero.
Each finite element will have its own unique energy functional.

1.3 Project topic

The topic of the project is to develop and evaluate techniques to analyse the strength
and stiffness of welded joints. The analysis of welding with the software to validate
simulation method compare with experiment testing of fillet welded T-joints. The aim
of the project is to develop techniques for simplifying finite element analysis such as
stress analysis and dynamic properties of structures under vibration excitation on non-
complete joint penetrated (Fillet) welding. The analysis allow us to understand and
compare experiment test result and software analysis result.

5
Chapter 1 – Introduction

1.4 Scope of the research

In welding process used electric current and produce arc to melt the consumable
electrode and the work piece to be welded. Electric current flow through high
resistance air gap generates an intense arc with temperature running from 3000o C –
6000o C

Residual stresses caused by welding can have various kinds of influence on the welded
structure, e.g. increasing the susceptibility of a weld to fatigue damage, stress crossing
cracking and fracture. Moreover, residual stresses developed in T-joint fillet welds
made of steels are probably different from those of full penetrated welds in magnitude.
Residual stresses are unavoidable, and the effects on welded structures are cannot be
disregarded. Therefore, it is very important to clarify the characteristics of residual
stresses in T-joint fillet welds in the structures. Welded steel joints always susceptible
to fatigue damage when subjected to repetitive loading. Fatigue failure may occur even
under modest in-service stresses. Furthermore, fatigue lives exhibit considerable
scatter even under constant amplitude loading in controlled laboratory conditions. This
phenomenon makes statistical methods indispensable and fatigue life has to be
predicted at given probability levels of failure for a given welded detail under defined
environment and loading conditions.

The driving factor for conducting this research project is need for a more confidently
using of finite element analysis tool to predict life of the welded structures. This is the
time for engineers to design more sustainable, safe and desirable component for the
society and finite element analysis tool is more convenient and very less time required
to solve the most complex structures.

The comparison of physical test with finite element analysis are not much available
which can refer to predict the life of structures. Finite element analysis (FEA) is a
computer based method of simulating/analysing the behaviour of engineering
structures and components under variety of conditions.

Fillet weld is most common welding in the engineering field and this weld used almost
everywhere in building structures. Finite element analysis of simple fillet welding

6
Chapter 1 – Introduction

compare with physical testing will allow us to analyse more complex structures with
more accurately with minimum error.

1.5 Project objectives

 The primary objective of this project is to develop techniques of finite

element analysis to define normal stress and fatigue strength of fillet welded
joint.
 Compare experimental test results with finite element analysis results of the
investigated joint
 Characterise the strength of fillet welded joints in comparison with parent
metals.

1.6 Expected outcomes and significance of the study

The research outcome will help designer to estimate strength of welded structural
component under dynamic load. Welding is the common engineering join method
which used for various structures. The major construction projects in all around the
worlds are using cranes, earth moving equipment, those also always under dynamic
load, the marine ships another heavy welded structure which are also subjected to
dynamic loads at all the time. The body of transportation equipment are highly
impacted by dynamic load which is focused in this analysis. Fatigue test and analysis
of a T joint with fillet welding in a typical connection is presented in this study.
Relevant finite element analysis of the joint is also established. The project will help
to characterise the behaviour of fillet welded joints.

This study is expected to provide with an in-depth understanding of the behaviour of

fillet welded joint under cyclic loading. The comparison of the test results with the
finite element analysis will also give an indication about the assessment capability of
the analysis tools for this type of joint. A simple and useful graphic representation in
the form of chart, which will aid the readers to actualise the effect of cyclic loading on
the joint is also expected.

7
Chapter 1 – Introduction

1.7 Dissertation outline

The dissertation is arranged in the following sequence of the chapters:

 Chapter 1 introduces the study and states the objectives of the study followed
by the expected outcomes from the study.
 Chapter 2: Literature Review contains the recap of the available literature.
Potential gap in the studies is identified and reason for this study is highlighted.
 Chapter 3: Methodology and Project planning composed of the details of the
methodology of the tests and simulations to be carried out in this investigation.
 Chapter 4: Physical Investigation provides with the test properties of the fillet
weld joints in the laboratory.
 Chapter 5: Finite Element Analysis and simulation with ANSYS program on
the fillet welded joints for comparison with the physical testing.
 Chapter 6: Discussion of Outcomes narrates analysis of the results and findings
in terms of the parameters involved.
 Chapter 7: Conclusion gives a summary of the results and scope of further
study and outlined.

8
 Chapter 2 – Literature Review

2. Chapter 2 – Literature Review

2.1 General
The following chapter reviews literature relevant to the welding process and welding
analysis using finite element methods. To calculate strength of welded structure and
to find behaviour under cyclic load, it is very important to do analysis according to
real life problems, so that analysis results can be used to predict the life of structures.
After welding of structures there are many factor involve defining parameter in the
finite element analysis. In this this report will go through and underline the associating
maters with the welding. Material behaviour will influence by the heat effected zone,
residual stress, types of welding, types of joint and different types of stress applications
also included in the report. Fatigue analysis will take place to determine life of
structure under cyclic load.

2.2 Welding process

Welding is the process of permanently joining two or more materials together, usually
metals, by heat or pressure or both. When heated, the material reaches molten state
and maybe joined together with or without additional filler materials being added
(Welding fundamentals and processes 2011).

Welding is an expedient by which metals may be joined by increasing the temperature

of the work pieces to their fusion point and allowing the molten metal formed to flow
together and solidify (Morris 1955). There are many different types of welding process
available now a days. The common welding processes are manual metal arc (MMA)
welding, gas metal arc (GMA) welding, tungsten inert gas (TIG) welding, and
submerged arc (SA) welding.

All of those processes used electric arc to melt electrode and parent metals. Due to
high productivity GMA welding is widely used in the industry for general fabrications.
In GMA welding process electrode is melted and molten metal is transferred to the
work piece. The transfer of molten metal from the electrode to the work piece can be
divided in to three modes (Welding : theory and practice 1990). Olson D L. et al.
found increasing current increases weld penetration and weld size thereby distortion
and residual stress is more which is affect to the welded structures.

9
Chapter 2 – Literature Review

Heat induced by welding has metallurgical effect on the metals it caused grain growth
found observing microstructural feature (Dodge et al. 2014). Based on observations in
the number of study found, a coarse-grained HAZ, martensite grain in the welded
region but post weld heat treatment can be developed the HAZ materials. Normally
very high percentage of welding structures used without heat treatment however, most
of the structures induce stress in the joint which called residual stress.

2.3 Brief History of FEA

The finite element method is a numerical procedure that can be applied to obtain
solution to many different type of engineering problem. The modern finite element
method can be traced back to early 1900s. However, Courant has been credited with
being first person to develop the finite element method in paper published in early
1940s (Moaveni 2007). The next significant step in the utilisation of finite element
methods was taken by Boeing in 1950s after that 1960 Clough made the term finite
element popular.

ANSYS is a comprehensive general purpose finite element computer program which

can use for structural mechanics, fluid dynamics, electromagnetics, system and
multiphasic analysis and simulations this program was released in 1971 for first time.
ANSYS has more than 100,000 lines of code and it capable of performing static,
dynamic, heat transfer, fluid flow etc. ANSYS has been leading FEA program over 35
years.

2.4 Effect of residual stress

Due to varying temperature distribution of welding process, thermal stress is
generated. It is known that thermal stress leads to residual stress. At higher temperature
when metal melted as liquid stage and solidified very sort of time, it has change grain
structure of surrounding of weld metal which causes shrinkage and introduce residual
stress. The welding residual stress has an effect on deformation of structures,
concentration of structure, fatigue fracture etc.

The investigation done by Seok et al. (Seok, Suh & Park 1999) for the residual stress
of H-type beam in Figure 2-1: H-type beam on butt welding made of high tensile steel
by finite aliment method, they found the case of longitudinal component of the residual
stress.

10
Chapter 2 – Literature Review

Figure 2-1: H-type beam(Seok, Suh & Park 1999)

The tensile residual stress increases from the centreline of weld bead to 5mm in the
transverse direction and decreases to converge to zero after the location of 5mm off.
That is the maximum tensile residual stress occurs in HAZ (heat effected zone), which
is the location of 5mm apart from the centre line of weld bead, and magnitude is 62.5%
of yield stress shown in Figure 2-2

Figure 2-2 Longitudinal residual stress vs location from the centreline of weld bead (Seok, Suh & Park 1999)

There are many different stress levels in the welded joint, in the Figure 2-3 as
following;

11
Chapter 2 – Literature Review

Figure 2-3:Longitudinal residual stress from the bottom on weld metal and parent metal.(Seok, Suh & Park 1999)

The location A and B are corresponding to the weldment and HAZ respectively in
addition C, D and E are for base metal. In figure 5 it is observed that residual stress
distribution of B on HAZ is higher than the A weldment.

T-joint fillet welding is extensively using in manufacturing industry such as

shipbuilding and bridge constructions. During welding localized heating and rapid
cooling causes tensile residual stress which develops close to the toe of the welding.
Finite element analysis done by (Teng et al. 2001) on T-joint fillet welded joint. In
order to analyse the Tee section Mr. Teng consider symmetric and used mesh
refinement in the Figure 2-4.

12
Chapter 2 – Literature Review

Figure 2-4: Finite element meshes of the T-joint fillet weld(Teng et al. 2001)

A stress acting on the direction of the weld bead is known as transverse residual stress,
Figure 2-5 represents distribution of residual stress along X direction. A very large
tensile residual stress is produced at the toe of the welding where stress shows 25 MPa
(Teng et al. 2001).

Figure 2-5: Transverse residual stress distribution along the X direction (Teng et al. 2001)

A very large tensile residual stress is produced at the surface of the base plate near the
fillet weld toes. The value of the residual stress near the weld toe is 25 MPa and
decreases to zero as the distance from the weld toes increases. A stress acting parallel
to the direction of the weld bead is called longitudinal stress denoted at Figure 2-6

13
Chapter 2 – Literature Review

depicts the distributions of the residual stress along the X-directions. The residual
stress value found is 110 MPa approaching the yield stress of the material.

Figure 2-6: Longitudinal residual stress distribution along the X-direction

In the welding process, different weldment thicknesses required different weld

penetration depths to controls the heat input to investigate the effect of different weld
penetration depths on residual stress and distortions. In Figure 2-7 it can be seen that
residual stress of fillet weld is higher than the full penetration joint.

Figure 2-7: Transverse residual stress distribution for different penetration depths

14
Chapter 2 – Literature Review

The following research investigated both an experimental and finite element

simulations for welding distortion in a T-joint fillet welding. There are many research
and investigation done on the above topics. In principle, a finite element (FE)
simulation of the welding process consist of two main parts: thermal analysis and
mechanical stress analysis (Perić et al. 2014). In thermal analysis, the temperature field
is determined as a function of time for each integration point. This temperature time-
history is used as an input into the thermal stress analysis. Herein, the thermal solution
can be sequentially or fully coupled with the mechanical solution of the structure. As
presented in the literature, the use of three dimensional (3D) models is required for
accurate prediction of post-weld deformation and residual stress distribution. Peric et
al. 2014 done experiment according to Figure 2-8

Figure 2-8: Geometry of T joint (Perić et al. 2014)

The welding experiments are conducted and the measured temperatures and
displacements are compared with the results obtained by the numerical analysis by
(Perić et al. 2014). It is shown that the numerical results agree well with the
experimental results in Figure 2-9.

15
Chapter 2 – Literature Review

Figure 2-9: : Deflection and residual stress of fillet welding at direction along the line C-C, from figure 10 (Perić et al.
2014)

To control residual stress in the welded joint clamping method is given significant
different result from unclamped joint. Three different clamping condition has been
analysed by Padma Kumari.T and Venkata Sairam.S the condition (i) is unclamped (ii)
is clamped but released immediate after welding and (iii) clamped and released after
cold down at ambient temperature. Among the three cases considered, the third case
yields the minimum transverse residual stresses at the upper surface of the plate while
it is reversed with regard to the bottom surface. Since the plate is unclamped after
cooling down to ambient temperature, it increases the longitudinal stresses at the
bottom surface due to the self-weight of the plate as it acts as a fixture to the bottom
surface found (PadmaKumari 2013). Hence the third case is the most preferable
condition in regard to the residual stresses in Figure 2-10.

16
Chapter 2 – Literature Review

Figure 2-10: Transverse stress distribution at the top surface (a) and the bottom surface (b) (PadmaKumari 2013)

To predict residual stress of T-joint fillet welding (Kyong-Ho Chang & Chin-Hyung
Lee) described in Figure 2-11 thermal analysis and mechanical stress analysis where
they found stress are higher on the weld toe area and its get lower while distance from
centre is increased.

Figure 2-11: Residual stress profile at cross section of flange (Chang & Lee 2009).

17
Chapter 2 – Literature Review

2.5 Effect of welding sequence

Influence of welding sequences on the distribution of residual stress and distortion
generated when welding flat bar as a stiffener on to the steel plate. Gannon et al. 2010
found In the case of longitudinal residual stresses, welding sequence did not have a
significant influence on the distribution pattern of the stress; however it did affect the
peak values. In the Figure 2-12 welding Sequences (a) and (d) caused the notably
higher compressive residual stresses in the plate than sequences (b) and (c). In the
stiffener, the maximum compressive residual stress due to sequence (b) was
approximately 3.5 times greater than the next lowest value. The distribution and peak
values of residual stress were similar to measured values and predictions used in
literature. Maximum tensile residual stresses equal to the material yield strength were
predicted in the vicinity of the weld and maximum compressive residual stresses from
_97 MPa to _58 MPa in the plate and stiffener respectively (Gannon et al. 2010).

Figure 2-12: Welding sequences (Gannon et al. 2010)

In the case of welding-induced distortion, welding sequence B resulted in the largest

out-of-plane deflection of the plate which may result in a reduced plate effectiveness.
Sequence B resulted in the largest out-of-plane deformation of the plate along the axis
of the stiffener. These predicted distortions were of lower magnitude than typical
values suggested in literature. Welding sequences A and B resulted in the largest lateral
deflections of the stiffener. Sequences C and D resulted in the least distortion of both
plate and stiffener for the geometry considered. Considering both residual stress and
distortion as a result of welding, welding sequence D is identified as the preferred
welding sequence with the lowest welding-induced residual stress and distortion in
Figure 2-13.

18
Chapter 2 – Literature Review

Figure 2-13: Deformation shapes with von misses stress contours (Gannon et al. 2010)

Overall, the finite element model including both thermal and mechanical procedures
has provided simulation results in reasonably good agreement with the experimental
measurements. The variations observed between the numerical and experimental
results were considered to be within the acceptable limits (Gannon et al. 2010).

Fillet welded joint has very common angular distortion, restricted stiffeners welding
and free restrained stiffeners welding are given different angular distortion and
residual stress see Figure 2-14. Type of weld, joint preparation, thickness of plate, size
of weld, joint restraint, heat input as well as welding sequence are effect of residual
stress on welded joint.

Figure 2-14: Angular change in T-fillet joint (A) free restrained (B) Restricted

19
Chapter 2 – Literature Review

(Syahroni & Hidayat 2012) consider in his study and the numerical investigation
according to the following sequence in Figure 2-15 welding sequences considered
were the one direction welding (WS-1), the contrary direction welding (WS-2), the
welding from centre of one side (WS-3), and the welding from centres of two sides
(WS-4).

Figure 2-15: Variation of welding sequence, (a) WS-1, (b) WS-2, (c) WS-3, (d) WS-4 (Syahroni & Hidayat 2012).

According to above preparation finite element procedure was employed to simulate

the thermos mechanical response of welding problem Syahroni & Hidayat 2012 found
results on the problem considered are presented in this section. The finite element
simulation for all the variation of welding was completed in 45 load-steps (LS). During
the number of load-steps, the welding process took for 40 load-steps, while the cooling
one took for the rest of the LS. For the presentation of welding simulation, the results
of the LS which respectively represent the conditions of the peak temperature and the
beginning of cooling processes were taken and plotted. Note that the temperature went
down towards the room temperature after the LS of 41. Accordingly, the longitudinal
and transverse residual stresses and the distortions occurred due to the welding
sequences. Peak temperature is always varying during welding in different sequencing
found in the Figure 2-16

20
Chapter 2 – Literature Review

Figure 2-16: Peak temperature for each welding sequence (Syahroni & Hidayat 2012).

Figure 2-17 describes the transverse residual stress distribution along the fillet weld
for each WS. The maximum values of longitudinal and transverse stresses as well as
von Mises stress for each welding sequence were summarized in Table 2. The ratio
between the longitudinal and the transverse residual stress values for the problem
considered varies from 1. 06 to 1. 22 (Syahroni & Hidayat 2012).

Figure 2-17: Distribution of transverse residual stress along the fillet weld for each welding sequence (Syahroni &
Hidayat 2012)

It can be clearly observed that the distributions of transverse residual stresses produced
by WS-3 and WS-4 and WS-1 and WS-2, respectively, are in consistent nature with
respect to the welding sequences.

2.6 Fatigue life calculation

Welded steel joints are vulnerable to fatigue damage when subjected to repetitive
loading. Fatigue failure may occur even under modest in-service stresses. Furthermore,
fatigue lives exhibit considerable scatter even under constant amplitude loading in
controlled laboratory conditions (Lassen, Darcis & Recho 2006). Fatigue strength of

21
Chapter 2 – Literature Review

weld seams can be evaluate according to nominal stress or the notch stress approach.
Determination of the stress of welding seams have to be generated a concept of
compatible way (Rother & Rudolph 2011) that means correct stress need to be
assigned on finite element analysis.

Characteristic of structures after welding with the method of hot spot stress Rother et
el. found in the Figure 2-18 where distance can be calculated based on plate thickness.

Figure 2-18: Linear (Left) Quadratic (Right) extrapolation based on stress derived from strain gage measurement
(Rother & Rudolph 2011)

Rother & Rudolph done seven test specimens according notch stress method to predict
fatigue life they found the following results in Table 2-1

Table 2-1:Computed structural notch stress including derived fatigue life (Rother & Rudolph 2011)

(Shen & Clayton 1996) have done testing before stress relief and after stress relief,
Table 2-2 shows the fatigue test result under pulsed-tension there was no consistence
effect of residual effect on fatigue strength. With tension-compression cyclic loads,
stress relieved specimens exhibited a superior fatigue life.

22
Chapter 2 – Literature Review

Table 2-2: Fatigue test result of fillet welded A515 Steel specimen (Shen & Clayton 1996)

The most recent design code in this regard is the ASME Boiler and Pressure Vessel
Code, Section VIII-2 (2007) (Joshi & Price 2009). Annex 3.F in Part III gives
equations for the number of allowable design cycles for non-welded and welded joint
fatigue curves based on empirical constants. In the following eq. code states the
formula for N, the number of design cycles for welded joint fatigue curve, as

f = Fatigue improvement factor, when the structure has been burr grinded, TIG dressed,
or hammer peened,

fE = Environmental modification factor, which is typically a function of the fluid

environment, loading frequency, temperature, and material variables such as grain size
and chemical composition, for structures operating in environments other than ambient
air,

fMT = Temperature adjustment factor, which is required for materials other than carbon
steel and/or for temperatures above 21 oC (70 oF),

∆𝑆𝑟𝑎𝑛𝑔𝑒 = Structural stress range.

Residual stresses are unavoidably generated in the component after welding (Zhang et
al. 2013). To investigate the effect of residual stress Zhang et al. used 16 inch pipe, the
result from the centre-hole measurements suggested that there are no clear effect of

23
Chapter 2 – Literature Review

residual stress, however, he found high tensile residual stress on the surface of the weld
root where welding started and stopped.

2.7 Fatigue life improvement

In many cases, the fatigue life can be improved employing good detail design of
welding profile. However, there are two main improvement method addresses by
(Kirkhope et al. 1999) which is weld geometry modification methods and residual
stress methods. Following Figure 2-19 is the layout of the fatigue improvement
methods.

Figure 2-19: Classification of some weld improvement methods(Kirkhope et al. 1999)

The Welding Institute recommended that after welding fatigue life can be improved
with some technic, such as Hammer peening, Machining, Shot peening, Plasma
dressing and Disc grinding locally burr machining seen in the Figure 2-20

24
Chapter 2 – Literature Review

Figure 2-20: Typical improvement in fatigue strength of mild steel fillet weld resulting from selected weld toe
improvement technic (Maddox 2002)

The offshore technology report provides information of selection design process for
specially oil platforms. In the Figure 2-21 from OT report produce basic design curve
for welded tubular and plate in the air.

Figure 2-21: Basic design curve for welded tubular plates in the air (Joshi & Price 2009)

25
Chapter 2 – Literature Review

2.8 Computational challenges in welding simulation

Welding analysis is an important topic in engineering research and it is widely
employed in the fabrication due to their advantage of improved structures
performance, cost savings and easy implementation. However, welding application
cause undesirable permanent distortion and residual stress in the material (Deng, Liang
& Murakawa 2007). Welding has about 26 imperfections e.g. cracks, porosity, worm
hole, inclusions, lack of penetration, lack of fusion, lack of fit, undercut, excessive
weld overfill, insufficient weld throat, root overfill, misalignment, weld sag,
incomplete root, cold lap, arc strike, sputter etc. (Hobbacher 2009). With all this
imperfection simulation parameter will be in challenge. Material modelling is, together
with the uncertain net heat input, one of the major problems in welding simulation
(Lindgren 2001)

2.9 Thermal and mechanical finite element analysis

Finite element analysis (FEA) has been used widely by many researchers Bibby et al.
(1992), Goldak et al. (1991), Zhang et al. (2006). Wikander et al. (1996), Breiguine et
al.(1992), Gundersen et al. (1997), Lindgren L-E et al.(1988)], to perform welding
simulations and to predict residual stresses in different types of welded joints and
materials. Prediction is very difficult due to the complex variations of temperature,
thermal contraction and expansion, and variation of material properties with time and
space (PadmaKumari 2013).

Over the last few years, the technology of laser direct metal deposition (LDMD) has
gained increasing attention in the industry for the rapid manufacture, repair, and
modification of metallic components especially those involving high cost material
such as superalloys. The technology is used in various industrial disciplines for its
demonstrated possibility in the production of parts with complex internal structures
that could not be achieved by machining. Furthermore, its associated lower material
wastage ratios in low volume manufacturing applications is among those attributes
leading to its high level of utilization in the aerospace industry.

The prediction of residual stress in laser deposited Waspaloy parts was found that
experimental result and simulation result has some relation. While the results of the
study indicated that modelling the deposition process with slight overestimation about
20% in the width of the deposited wall has no significant effect on predicted residual

26
Chapter 2 – Literature Review

stress, the effect (A. M. Kamara 2011), which an underestimation in the width of the
deposited wall could have on predicted residual stress, is also an interesting issue to
be investigated.

(Brickstad & Josefson 1998) simulate the residual stresses due to welding using
ABAQUS to perform finite element analysis, their analysis consist of two main parts
the thermal and the structural. The analysis is two-dimensional and axisymmetric. The
thermal analysis models the heat input from the welding torch into the weld elements
causing the weld to melt. Heat losses allow the weld region to solidify. The
temperature reliefs obtained from this part of the analysis are used in the sequential,
structural analysis to derive the stresses generated as the material heats up and cools
down again. The behaviour of the material involves non linearity and therefor residual
stresses remain in the welded joint after cooling.

The mechanical effect of the material due to heat source, during analysis need to
consider the following effect in Figure 2-22 two dimensional model of heat source.
This model gave a more realistic model with better heat distribution in the melted zone.

Figure 2-22: Heat distribution of melted zone(Rabih Kamal Kassab1 2012)

2.10 Heat affected zone

Heat affected zone is the most critical location in welded joint where the most residual
stress develop due to metal grain coarsened. Heat affected zone is very hard to predict
heat effected zone property (Wang et al. 2007) however, the region of HAZ partially

27
Chapter 2 – Literature Review

ferrite and austenite at peak temperature (Woolling & Carrouge 2002). Figure 2-23
shows the grain growth in the ferrite phase.

Figure 2-23: High temperature HAZ showing grain coarsened region (Woolling & Carrouge 2002)

The next to the weld is assigned strength properties slightly lower than the values given
in the design codes in order to trigger strain localization. The width of this zone should
be approximately equal to the plate thickness and discretised with one or several
elements. The remaining part of the HAZ is assigned the strength properties given in
the design code, and discretised with several elements. The accuracy of this procedure
remains to be validated. After welding heat affected zone are completely change
geometry of grain structure as Figure 2-24, due to the affect material strength has been
change.

Figure 2-24: Heat affected zone after welding (Welding Inspection Cswip 2010)

2.11 Welding deformation

The deformation of welded structure in Figure 2-25 result from non-uniform
expansion and contraction of weld and the surrounding base material due to heating

28
Chapter 2 – Literature Review

and cooling cycle during welding process. Welding deformation play an important role
in sealing capabilities and service life of welded joints says (Abid & Siddique 2005).
There are lots of numerical and analytical model has been available for butt welding
process. However, only very limited literature describing welding deformation of fillet
welds are available (Deng, Liang & Murakawa 2007). Deng et al. found that the flange
thickness of T-joint has influence on welding deformation and the simulated result
demonstrates that the temperature gradient through thickness is a main factor that
strongly governs the generation of angular distortion in fillet-welded joint.

Figure 2-25: deformation after welding (Deng, Liang & Murakawa 2007)

The heat source used in fillet welding joint in Figure 2-26

Figure 2-26: Combined heat source in the fillet welding joint (Deng, Liang & Murakawa 2007)

Distortion after welding in Figure 2-27 shows 500mm of welding can angle difference
.022 rad.

29
Chapter 2 – Literature Review

Figure 2-27: Distribution of angular distortion along welding line (Deng, Liang & Murakawa 2007)

The distortion of welded structure also be subject to welding sequence as well as

number of pass of welding. Analysis of a welding thermo-mechanical response due to
welding is the important factor. Welding distortions depend on the geometry, welding
condition as well as the material properties (WANG Rui 2008). Sulaiman et al. 2011
has done investigation of simulation and experimental distortion of welded T-joint,
where they found 20.9 % error of the result (Sulaiman et al. 2011) which is reasonable
agreement.

2.12 Weld metal property

Weld metal properties are not exactly same as a parent metals, it has different strength
and different grain structures. Pisarski et al. investigated on a grade X100 pipe, they
found the tensile property of the weld metal after different type of weld process used
which shown in Table 2-3 and fi can see that weld metal strength are lower than the
parent metal.

Table 2-3: Tensile properties of grade X100 pipe (Pisarski, Tkach & Quintana 2004)

30
Chapter 2 – Literature Review

Figure 2-28: Experimental stress strain carves of parent pipe and GMAW, FCAW weld metal (Pisarski, Tkach &
Quintana 2004)

2.13 Visualise stress and deformation

Founded in 1828, Bureau Veritas is a global leader in testing, inspection and
certification. In the marine industry, Bureau Veritas as an official certification body
(SIEMENS 2015). Bureau Veritas provides numerical stress analysis services based
on the Finite Element Method. FEA calculates and visualises stresses and deformations
resulting from applied loads in Figure 2-29. With stresses accurately calculated, actual
safety margins over material strength can be determined. This approach to design is
recognised within the design standards as being an appropriate alternative to manual
calculation. Complex geometries may be accurately assessed and time dependant load
histories may be applied. Loads may include force, restraint, pressure, temperature,
gravity and dynamic loads.

31
Chapter 2 – Literature Review

Figure 2-29: Bureau Veritas use FE map to help designs conform to regulations (SIEMENS 2015)

2.14 Adaptive mesh technique

(Qingyu et al. 2002) have developed an adaptive mesh technique applied in the three-
dimensional numerical simulation of the welding process on the basis of the
commercial software MARC. The adaptive mesh technique generates a dense mesh
and makes it move simultaneously with the heat source. Any part of the mesh away
from the heat source is much coarser, significantly saving CPU time. The calculation
time comparison shows that the adaptive mesh technique can reduce the CPU time by
almost one-third.

In traditional finite element analysis, as the number of element increases the accuracy
of solution will be improves. Mesh density will be chosen when the result difference
will be very less from one elements size to another elements size. Example of the

32
Chapter 2 – Literature Review

Figure 2-30 shown, where, a 2D bracket model is constrained at its top end and
subjected to a shear load at the edge on the lower right. This generates a peak stress
in the fillet, as shown. The curve shows that as the mesh density increases, the peak
stress in the fillet increases. Ultimately, increasing the mesh density further produces
only minor increases in peak stress. In this case, an increase from 1134 elements per
unit area to 4483 elements per unit area yields only a 1.5% increase in stress (Hale
2014).

Figure 2-30: Stress sensitivity to mesh density (Hale 2014)

The difference results between coarse mesh and finer mesh found from above
investigation in Figure 2-31

Figure 2-31: Relative difference in stresses at shared nodes for coarse mesh (L) and finer mesh (R) (Hale 2014)

33
Chapter 2 – Literature Review

2.15 Further Research

Fillet weld joints are one of the mostly used welding joints in the industry. Current
body of literatures suggests considerable effect of thermal and cyclic loading on fillet
weld joints. Numerous studies are dedicated to characterise the behaviour of these
types of joints. Complexity of dimensions and physical parameters possess a challenge
for the researchers. However, there are lack of study on fillet weld joints comparing
experiment test and finite element analysis in light of nominal stress method and hot
spot method. This study exerts an effort to actualise this behaviour to further enrich
this domain of knowledge. This research will allow us to justify uses of finite element
tools for this kind of applications.

34
 Chapter 3 – Methodology and Project planning

3. Chapter 3 – Methodology and Project planning

3.1 Introduction
This chapter details the specimens used as well as the testing methods and processes
used in both simulation and physical investigation of fatigue life of T-joint fillet
welded structure. In the simulation define parameter of thermal and mechanical
behaviour of the joints. In the physical test will be used maximum load to disturb the
object and compare with simulation result. Also covered in this chapter is a guideline
into the finite element analysis methods and procedures used to computationally
investigate/correlate the grating properties and data. The design process is to over view
of the project research that carried out to complete the finite element analysis and
physical investigation.

3.2 Methodology
This study comprised of experimental investigations and computer modelling of fillet
welding. The double sided T-section of structural steel welded specimen were prepared
for tensile test. Computer modelling used the finite element based commercial package
of ANSYS Static Structural.

3.2.1 Experimental Investigation

The preparation of experimental specimens and the procedure of experiment is
discussed in this section.

[Link] Experiment preparation

The experiment made from standard structural steel grade 250, it has 250MPa yield
stress. Three pieces of plate cut from full plate with automatic gas cutting and the size
chosen for the base plate one is 400mm X 150mm and other two 400mm X 150mm.
All the edges of plate has been grinded to remove any kind of slug or rust. Three plate
now join together using tack weld and before welding fitted three back-up plate to
control deformations. Welding has been carried out using gas metal arc welding
(GMAW). In the following Figure 3-1 can see the shape of specimen which is created
using AutoCAD.

35
Chapter 3 – Methodology and Project planning

Figure 3-1: Geometry of test sample

After welding completed, the object cut-off small sections (15mm) with bend saw and
grind all sharp edges.

[Link] Procedure
Tensile test is the fundamental test in the material science where we can find yield
stress and ultimate stress. The test has been carried out to find yield stress of the welded
joint where mostly heat passes through during welding. The load will be applied on
two axis, one is on the base metal which is continue plate, this will allow us to find
stress on the heat affected zone Another loading axis will be normal to the welded
metal this will us to find welding strength or bonding strength. Details of applying load
on the two different directions can see Figure 3-2

36
Chapter 3 – Methodology and Project planning

Figure 3-2: Direction of load applied

Following procedure taken to consider for the tensile test:

 Six specimens where chosen and marked 1-6 for the records and identifications
 Record down cross section area of individual pieces, where 6 pieces are not
same sizes
 Visual inspection conducted to ensure that, the specimens did not have any
notching or cracks
 Clean up machine parts and clamping area then started the machine
 Wait for warm up and set graph paper in to the barrel for load deflection graph
and make sure pen is working condition
 Before loading the specimen make sure safety glasses is worn
 Load specimen and set dial indicator to zero position then apply load, slowly
increase the load and wait until the necking started and increase the load to
break the specimen
 Switch down to idle position and take off graph paper and rest the machine for
next test.

37
Chapter 3 – Methodology and Project planning

3.2.2 Finite Element analysis

The section presents finite element model development, creating fillet weld,
methodology of mesh element size development and model set-up. Meshing element
size define with various number of elements and graphical presentation also made.

[Link] Model development

Model development is the crucial part of the analysis, if the model parameter is not
match with the experiment then desire results will not match with the simulation
results. The dimension of the model is taken 15mm width and 10mm thick of double
sided T-joint. The model separated in three main parts which is body, heat effected
zone and weld material. Reason to do that its can apply different type of material
property. Figure 3-3 drawn in PTC Creo parametric and for analysis Creo model will
be exported to ANSYS for the simulation.

Base metal

Weld metal

Heat effected zone

Material not connected

Figure 3-3: basic model of welded joint

The following Figure 3-4 is the exploded view of the model where can find the
different parts of the model.

38
Chapter 3 – Methodology and Project planning

Figure 3-4: Exploded view of the model

[Link] Creating fillet weld

In order to create fillet welding in the model, three pieces of plate will not connected
together with one another. The connected area will be the welded material only. In this
model base plate and welded metal will be different parts. In the connection tab from
model tree can find contact region and suppressed plate connection to disconnect the
surface. After applying load Figure 3-5 can see that there is no connection between
plates except welding.

39
Chapter 3 – Methodology and Project planning

Figure 3-5 Proof of fillet welding

[Link] Meshing
In traditional finite element analysis we know that, the number of element increases
the accuracy of solution will be improved also it is not necessary to put element size
very small. Smaller element size will takes longer time to analysis and some time it is
almost impossible to run the model.

In order to find optimum element size I have chosen different element size and analyse
the model, where the stress are not changing dramatically assume that is the optimum
size. Following Figure 3-6 and Figure 3-7 can see how different element size vary the
stress with same applying load.

40
Chapter 3 – Methodology and Project planning

Default mesh size Mesh size 10mm

Mesh size 5mm Mesh size 4mm

Figure 3-6: Mesh development

41
Chapter 3 – Methodology and Project planning

Mesh size 3mm Mesh size 2.5mm

Mesh size 2mm Mesh size 1.5mm

Figure 3-7: Mesh development

The stress concentration on red colour location is very high, its dose not display actual
results because the area is very small compare with cross section and its only 0.5 mm.
The grinding operation can be removed sharp concentration areas. In the Figure 3-8
shows the stresses are changing due to different size of element.

42
Chapter 3 – Methodology and Project planning

Mesh element size vs stress

480
470
460
Max Stress (MPa)

450
440
430
420
410
400
390
Default 10mm 5mm 4mm 3mm 2.5mm 2mmm 1.5mm
Mesh element size

Figure 3-8: Mesh element size Vs stress

Above figure can identify that element size between 2.5mm to 1.5mm where the stress
haven’t been changes. There for the adequate mesh element size can be considered
2.5mm to 1.5mm.

[Link] Model set-up

There are many different parameter has been set-up for the model in the Figure 3-9 is
the model tree of the model set-up.

43
Chapter 3 – Methodology and Project planning

Figure 3-9: Model set-up tree

There are 11 different parts in the model and its sets different material property. Base
metal used structural steel which tensile yield strength 250 MPa, for heat affected zone
used 350 MPa due to rapid heating and cooling process, and for the weld metal strength
used 450 MPa (Austmig ES6) which provided Austmig ES6 material datasheet.
Contact region sets two suppress to make disconnect surface between the plates. In the
mesh set-up tab, all mesh is patch conforming method and different body has sets
different element size.

44
 Chapter 4 – Finite Element Simulation

4. Chapter 4 – Finite Element Simulation

4.1 Introduction
In this chapter included basic simulation on a double sided T-joint fillet welding using
ANSYS workbench. The property of base metal chosen structural steel material of 250
MPa, for heat affected zone used 350 MPa yield strength and for weld material used
450 MPa yield strength. Result set-up for stress chosen equivalent stress and fatigue
life. Adequate mesh applied to the model.

4.2 Background
Finite element analysis (FEA) is the modelling of products and systems in a virtual
environment, for the purpose of finding and solving potential (or existing) structural
or performance issues (SIEMENS 2015). FEA is the practical application of the finite
element method (FEM), which is used by engineers and scientist to mathematically
model and numerically solve very complex structural, fluid, and multi physics
problems.

A finite element model comprises a system of points, called “nodes”, which form the
shape of the design. Connected to these nodes are the finite elements themselves which
form the finite element mesh and contain the material and structural properties of the
model, defining how it will react to certain conditions. The density of the finite element
mesh may vary throughout the material, depending on the anticipated change in stress
levels of a particular area. Regions that experience high changes in stress usually
require a higher mesh density than those that experience little or no stress variation.
Points of interest may include fracture points of previously tested material, fillets,
corners, complex detail, and high-stress areas.

FE models can be created using one-dimensional (1D beam), two-dimensional (2D

shell) or three-dimensional (3D solid) elements. By using beams and shells instead of
solid elements, a representative model can be created using fewer nodes without
compromising accuracy. FEA is originally developed for solving solid mechanics
problem (Qi 2006). The Finite Element Analysis is offers a mean to find the
approximation, this is not an exact solution. The general procedure of FEA in Figure
4-1 shows how does it works:

45
Chapter 4 – Finite Element Simulation

Figure 4-1: General procedure of finite element analysis (Hobbacher 2009)

In the past, there are many different way welding joints considered and analysed with
different types of simulation tools such as ANSYS, ABAQUS, and SOLIDWORKS
etc. Using those tools and analysed normal stress, residual stress, fatigue life
calculations and deformations etc. where mostly found after welding of steel strength
of metal specially fatigue life are less than original metal. The welding institute tested
welding joints applying axial force and they found the design of welded joint has a
dominant effect on fatigue life. Figure 4-2 can see the S/N curves for welded and un-
welded specimens where welded carbon steel has less fatigue strength than un-welded
steel.

Figure 4-2 : S/N curves for welded and un-welded carbon steel (TWI Fatigue testinng 2015)

46
Chapter 4 – Finite Element Simulation

4.3 ANSYS simulation

In order to analyse the model with ANSYS the nominal stress and fatigue life
calculations are rely on the ASME Boiler and Pressure Vessel Code, Section III (and
Section VIII, Division 2) for guidelines on range counting, simplified elastic-plastic
adaptations, and cumulative fatigue summation by Miner's rule. There are few number
of method available in fatigue assessment of welded joint;

 Nominal stress method

 Hot spot method
 Fracture mechanic analysis
 Effected notch stress method

Simulation details as following

Project name Finite element analysis of fillet welding

FEA tools & ANSYS 15.0.7 Release
version
Analysis type Static structural
Material data  Structural steel
 HAZ material
 Weld material
Unit system Metric (mm, kg, N, mV, mA)
Length on X 200mm
direction
Length on Y 210mm
direction
Length on Z 15mm
direction
Applied force 40000N
Connections Total 14 connections, region 1 and 3 suppressed
Mesh Method applied Tetrahedrons, element size 3mm for base metal,
2mm for heat affected zone and 1.5mm for welded material.
Nodes 50230
Elements 29694

47
Chapter 4 – Finite Element Simulation

Results Maximum principal stress and Fatigue life

Loaded model

Figure 4-3 Model with force direction and fixed support

48
Chapter 4 – Finite Element Simulation

4.3.1 Stress analysis

Figure 4-4: Stress distribution on the fillet welded specimen

Static structural analysis has been carried out applying maximum principal stress as
shown above in Figure 4-4. Maximum stress was found at 606.8 MPa with a appled
yield load of 40 kN. However, the stress in the vertical axis was 266.65 MPa which
demonstrate that those regions were not affected by welding.

The maximum stress was concentrated at the toe of the welding which can identify
very small area (red coloured in Figure 4-5). Approximately 1.5 mm away from the
maximum stress point, the stress are significantly less than the maximum stress which
was below of 260 MPa. The gap in the figure demonstrate that some part of the material
are not bonded in the fillet welding.

49
Chapter 4 – Finite Element Simulation

Figure 4-5: Stress distribution in the welded region

Above Figure 4-5 shows stress distribution in zoomed view. Stress appears 250-260
MPa which is very close to the yield stress. However, maximum stress 606 MPa
distributed only 0.5 mm areas.

Figure 4-6: Stress distribution with 60kN load

50
Chapter 4 – Finite Element Simulation

Figure 4-6 (above) applied 60kN load, where found maximum 910 MPa stress at the
concentrated area. However, the average stress found in the randomly selected location
close to the concentrated area (no. 1-6 is from lower to upper value) as follows:

Table 4-1: Calculation of stress with 60kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 393 444 438 468 398 438 430

Figure 4-7: Stress distribution with 50kN load

Figure 4-7 (above) applied 50kN load and maximum stress found 758 MPa at the
concentrated area. However, the average stress found as follows:

Table 4-2: Calculation of stress with 50kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 351 332 297 364 352 374 345

51
Chapter 4 – Finite Element Simulation

Figure 4-8: Stress distribution with 40kN load applied

Figure 4-8 (above) applied 40kN load where found 606 MPa stress at the concentrated
area on red coloured marked. 40kN load was found in the experiment as its yield load.
However, the average stress in surrounding found as follows:

Table 4-3: Average stress of 40kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 276 275 273 282 262 277 274

Figure 4-9: Stress distribution with 35kN load applied

52
Chapter 4 – Finite Element Simulation

Now load reduction interval reduced by 5kN for better result. Figure 4-9: (above)
applied load is 35kN, where maximum 530 MPa stress found at the concentrated area.
However, the average stress in surrounding found as follows:

Table 4-4: Average stress of 35kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 245 242 273 246 225 284 252

Figure 4-10: Stress distribution with 30kN load

Figure 4-10 (above) applied load is 30kN, maximum 430 MPa stress found at the
concentrated area. However, the average stress in surrounding found as follows:

Table 4-5: Average stress of 30kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 225 220 234 234 232 226 235

53
Chapter 4 – Finite Element Simulation

Figure 4-11: Stress distribution with 25kN load

Figure 4-11 found maximum stress 358 MPa with applied load 25kN. The maximum
stress shows at the inner part of the welding where weld metal started fusion. Actual
stress occurs surrounding to the red coloured marks. The average stress as follows:

Table 4-6: Average stress of 25kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 166 164 163 219 197 174 180

Figure 4-12: Stress distribution with 20kN load

Figure 4-12: (above) applied 20kN load, where found 303 MPa stress at the
concentrated area. However, the average stress in surrounding found as follows:

Table 4-7 Average stress of 20kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 141 140 189 144 138 168 153

54
Chapter 4 – Finite Element Simulation

Figure 4-13: Stress distribution with 18kN load

Figure 4-13: (above) applied load is 18kN, maximum 273 MPa stress found at the
concentrated area. However, the average stress in surrounding found as follows:

Table 4-8: Average stress of 18kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 129 126 138 132 152 139 136

Figure 4-14: Stress distribution with 16kN load

Figure 4-14: found maximum stress 229 MPa with applied load 16kN. Now reduce
load by 2kN only to get better results. The maximum stress shows at the inner part of
the welding where weld metal started fusion. Actual stress occurs surrounding to the
red colour marks. The average stress as follows:

Table 4-9: average stress of 16kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 112 111 114 115 123 119 115

55
Chapter 4 – Finite Element Simulation

Figure 4-15: Stress distribution with 14kN load

Figure 4-15 (above) applied load is 14kN, maximum stress found 212 MPa at the
concentrated area. However, the average stress in surrounding found as follows:

Table 4-10: Average stress of 14kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 103 101 112 100 109 112 106

Figure 4-16: Stress distribution with 12kN load

Figure 4-16 (above) applied load is 14kN, maximum stress found 182 MPa at the
concentrated area. However, the average stress in surrounding found as follows:

Table 4-11: Average stress of 12kN load

Location 1 2 3 4 5 6 Average

Stress (MPa) 84 85 89 85 94 91 88

56
Chapter 4 – Finite Element Simulation

4.3.2 Fatigue life analysis

In order to predict fatigue life of the structure, in this analysis chosen fatigue tool from
solution tab and selected life. The loading type used fully reversed constant amplitude
and stress life with Goodman mean stress theory. Finite element analysis starts with a
structural simulation to calculate the how many repeated load can be applied without
disruption of structures. Following figure shows the fatigue life in terms of cycle and
result taken using probe selection on the model.

Figure 4-17 Fatigue life with 60kN load

Figure 4-17 (above) shows the calculated fatigue life, applying 60kN forces which is
ultimate load. The red spotted line shows minimum 351 cycles of alternating load can
be applied. Average number of cycles in Table 4-12 found in the surrounded area of
red coloured line which is more realistic and area is more than the concentrated area.

Table 4-12: Calculation of fatigue life with 60kN load

Location 1 2 3 4 5 6 Average

No. of cycle 2713 2433 2578 2016 2344 2689 2462

57
Chapter 4 – Finite Element Simulation

Figure 4-18 Fatigue life with 50kN load

Figure 4-18 (above) shows the calculated fatigue life from simulation at 50kN load
which is ultimate load. Minimum 564 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-13: Calculation of fatigue life with 50kN

Location 1 2 3 4 5 6 Average

No. of cycles 3353 3452 3046 3231 2932 2746 3127

Figure 4-19 Fatigue life with 40kN load

Figure 4-19 (above) shows the calculated fatigue life from simulation at 40kN load
which is ultimate load. Minimum 1008 cycles of alternating load can apply and its

58
Chapter 4 – Finite Element Simulation

shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-14: Calculation of fatigue life with 40kN

Location 1 2 3 4 5 6 Average

No. of cycles 5181 6018 5407 5417 4860 6882 5627

Figure 4-20 Fatigue life with 35kN load

Figure 4-20 (above) shows the calculated fatigue life from simulation at 35kN load
which is ultimate load. Minimum 1427 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-15: Calculation of fatigue life with 35kN

Location 1 2 3 4 5 6 Average

No. of cycles 13407 13026 12751 13998 13737 11233 13025

59
Chapter 4 – Finite Element Simulation

Figure 4-21 Fatigue life with 30kN load

Figure 4-21 (above) shows the calculated fatigue life from simulation at 30kN load
which is ultimate load. Minimum 2157 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-16: Calculation of fatigue life with 30kN

Location 1 2 3 4 5 6 Average

No. of cycles 20124 18612 25528 16008 20228 20132 20105

60
Chapter 4 – Finite Element Simulation

Figure 4-22 Fatigue life with 25kN load

Figure 4-22 (above) shows the calculated fatigue life from simulation at 25kN load
which is ultimate load. Minimum 3789 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-17: Calculation of fatigue life with 25kN

Location 1 2 3 4 5 6 Average

No. of cycles 44917 47749 44113 57148 47517 40465 46985

61
Chapter 4 – Finite Element Simulation

Figure 4-23 Fatigue life with 20kN load

Figure 4-23 (above) shows the calculated fatigue life from simulation at 20kN load
which is ultimate load. Minimum 7553 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows;

Table 4-18: Calculation of fatigue life with 20kN

Location 1 2 3 4 5 6 Average

No. of cycles 84164 81529 71657 78125 81235 74912 78603

Figure 4-24 Fatigue life with 18kN load

62
Chapter 4 – Finite Element Simulation

Figure 4-24 (above) shows the calculated fatigue life from simulation at 18kN load
which is ultimate load. Minimum 10512 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot as
follows:

Table 4-19: Calculation of fatigue life with 18kN

Location 1 2 3 4 5 6 Average

No. of cycles 116640 152980 114060 119500 103230 118270 120780

Figure 4-25 Fatigue life with 16kN load

Figure 4-25 (above) shows the calculated fatigue life from simulation at 16kN load
which is ultimate load. Minimum 15736 cycles of alternating load can apply and its
shows in red spot area only. Average cycles found in the surrounding of red spot is
174975 cycles.

Location 1 2 3 4 5 6 Average

No. of cycles 218260 188570 151320 152350 152560 186790 174975

63
Chapter 4 – Finite Element Simulation

Figure 4-26 Fatigue life with 14kN load

Figure 4-26 (above) shows the calculated fatigue life from simulation at 14kN load
which is ultimate load. Minimum 25248 cycles of alternating load can apply and its
shows in red spot area only. Average cycle found in the surrounding of red spot is
425023 cycles.

Location 1 2 3 4 5 6 Average

No. of cycles 419560 411060 384180 407700 477200 450440 425023

Figure 4-27 Fatigue life with 12kN load

Figure 4-27 (above) shows the calculated fatigue life from simulation at 12kN load
which is ultimate load. Minimum 44445 cycles of alternating load can apply and its
shows in red spot area only. Average cycle found in the surrounding of red spot is
1000000 cycles. The million cycle consider to be infinite life according to ASME.

64
 Chapter 5 – Results and Discussion

5. Chapter 5 – Results and Discussion

5.1 Introduction
This chapter includes computer simulations results, experiment test results and
discussions. Discussions carried out for how the results are interpreted and justify the
methods that set up in the methodology. Limitations of the project was the critical part
of the discussion.

5.2 FEA Simulation results

In order to find stress of the welded joint, FEA simulations set up for maximum
principal stress. There are many different load combination has been used in the
simulations. However, this project compare yield stress of the experiment yielding
location. During simulation two different results has been deployed one is stress and
other is fatigue life. Table 5-1 shows the stresses and fatigue life of welded joint using
FEA tools (ANSYS).

Table 5-1: Results of stress and fatigue life of different load combination

Maximum principal Fatigue life No.

Force (kN)
Stress (MPa) of Cycle
60 430 2462
50 345 4299
40 274 7740
35 252 14258
30 235 21082
25 180 45426
20 153 90508
18 136 136166
16 115 274938
14 106 545923
12 88 1000000

The following Figure 5-1, S-N carve obtain from simulation data. In this graph shows
88 MPa stress can be consider for the fillet welded joints.

65
Chapter 5 – Results and Discussion

S-N Graph
1000

430
274
Stress (MPa) 235
153
115
100 88 88

10
1.00E+03 1.00E+04 1.00E+05 1.00E+06 1.00E+07
No. of cycle

Figure 5-1: S-N carve of welded joint

5.3 Experiment results

Tensile test has been carried out in the experiment where disruption occurs in the base
metal. The breaking point is far from the welding joint following Figure 5-2 and Figure
5-3 can see the break happen 60mm apart from welding joint.

Figure 5-2: Motion of tensile test

66
Chapter 5 – Results and Discussion

Figure 5-3: Brocken test pieces after tensile test

Above figure shows broken location and the reading from the tensile test machine.
Five different specimen has been tested and the specimen was numbered as 2, 3, 4, 5
and 6.

Figure 5-4: Load elongation graph of item 2 & 3

67
Chapter 5 – Results and Discussion

Figure 5-5: Load elongation graph of item 4,5 and 6

Total six number of item carried out tensile test. Item 1 has failed after loaded.
Everyone has different size of width due to manual cutting process. Table 5-2 given
all the test result and average yield stress found 263.4 MPa

Table 5-2: Test results of experiment

Item Cross section area of Yield Ultimate Yield stress Ultimate

no experiment load (N) load (N) (MPa) stress (MPa)

1 10x15.66 = 156.6mm2 Fail Fail Fail Fail

2 10x15 = 150mm2 39000 59900 260 400
3 10x16 = 160mm2 42000 63000 263 394
4 10x15.3 = 153mm2 40500 61800 265 404
5 10x14.5= 145mm2 37800 57900 261 399
6 10x14.2 = 142mm2 38000 56400 268 397
Average stress 40000 60000 263.4 399

68
Chapter 5 – Results and Discussion

5.4 Discussions
Fatigue damage occurs when stress at a point changes over time. Fatigue life set up in
the simulation as constant amplitude loading with fully reversed (+1 to -1). Fatigue
life can be over the time (no. of day hour) or number of cycle, to develop SN carve we
chose number of cycle.

The main aim of the project was to compare physical data of fatigue life with finite
element analysis data. However, fatigue life test of the specimen could not be
completed because of time and resource constraints from the technical support area
due to their commitments to post graduate projects

In order to validate finite element analysis results with comparing experiment results
we have completed tensile test. Figure 5-6 shows the experiment test piece yielding at
263.4 MPa. The necking started far from the welded area. In the model simulation
found 266.7 MPa stress at the same location where specimen broken. The results
different between experiment and simulation only 1.2%. Considering this minimum
error of the results we are conducting simulations for different type of analysis such as
fatigue life.

Yielding
266.7

Applied force Experiment yield stress Simulation stress

40 kN 263.4 MPa 266.7 MPa
Figure 5-6: Comparison of experiment test with model simulation

Experiment yield stress taken average of five different specimen because the cross
sectional area of the each specimen are not same due to manually prepared of the test
pieces. Simulation stress taken at the same location where yield started although stress
of the model are same everywhere except weld toe.

69
Chapter 5 – Results and Discussion

Weld toe has the maximum stress and minimum number of fatigue life cycles.
However, the weld toe is the concentrated area Figure 5-7 shows the dimension of the
area about 0.25 mm where, we can observe it is very smaller area compare to the cross
sectional area. In real life situation molten weld metal toe will be a bit curve or after
removed some material from toe using grinding operation will be removed most red
coloured area.

Figure 5-7: Zoomed view of stress concentrated area

There are eleven different load combination has been applied to predict one million
cycle fatigue life. Started from ultimate stress load (60 kN) than gradually decrease the
load to get 1 million cycles, the result shows at 88 MPa stress can be used for the
design calculation of this kind of material.

70
 Chapter 6 – Conclusions

6. Chapter 6 – Conclusions
6.1 Chapter overview
This final chapter summarises the achievement of project objectives, identifies future
works and provide recommendations. Definitive project conclusion also find end of
this chapter.

The objective of this project was to validate finite element analysis outcome compare
with experimental investigation. Primary results has been achieved successfully with
some lack of boundary condition input. Initial aims of experiment investigation was
fatigue life of fillet welded joint have not been achieved within the research project
time frame. However tensile test has been done to validate the FEA results.

6.2 Achievement of the project

The ultimate purpose of this project has been to perform finite element analysis of fillet
welded joint and comparison with experimental investigation. In order to achieve the
reasonable outcome study begin with introduction of the project then literature review,
methodology and project planning, model simulation finally result discussion.

Chapter 1 was to understand background of the welded structures and how the
structures fails during operations initiate cracks from the welded area. From
background knowledge define scope of the research and set up expected outcome than
outlined the dissertation.

Chapter 2 of this research has been successfully investigated a broad range of available
literature related to the finite element analysis of welded structure and how welding
operation affected to the materials and how can improved the life of welding joints.

Methodology and project planning developed in the chapter 3, where describe details
drawings of specimen and model development. In the FEA model, mesh element size
also define using iterative process then used boundary condition for analysis.

Chapter 4 detailed the finite element simulation and predicted principal stress and
fatigue life of the model. The stress found for infinite life under cyclic load is 88 MPa
which is 35% of yield stress.

Finally chapter 5 carry out discussion of the results from FEA analysis and experiment
investigation. Appendices has been included end of this report and there are some

71
Chapter 6 – Conclusions

boundary condition haven’t been included to the model which will be discussed in the
future work and recommendation topic.

6.3 Further work and recommendation

Although a significant amount of the original goals have been achieved as part of this
research project, a substantial amount of future work to be completed for the more
accurate results and comparison between experiment investigation and model
simulation. Details recommendations of simulation and experimental investigation
listed below:

 Fatigue testing – Experiment test has been carried out for this project was
tensile test where observed yielding started on parent metal rather than weld
toe area which means welded joint stronger than original material in terms of
tensile force. To investigate fatigue life of the structure, Fatigue testing of
experiment is more relevant to compare with FEA results.
 Model simulation – FEA model will be given result according to input data
and boundary condition. In the model simulation at weld needed to induce
residual stress.
 Mesh element size – Proper mesh element size is the key parameter to get
better results. Mesh method need to be justified between Tetrahedron, Hex
dominant, Sweep and Multizone which will be given more accuracy.

6.4 Conclusion
The ultimate purpose of the project has been achieved with developing techniques of
the finite element analysis of fillet welded joint. The experimental investigation
validate the performance of the FEA analysis results were found 1.2% error on tensile
test. The experiment yield stress was found 263.4 MPa and simulation yield stress at
the same location appears 266.7 MPa. In order to calculate fatigue life of welded joint
used iterative process to define stress at one million cycle. The analysis found 274
MPa stress and 7740 cycle fatigue life applying yielding load. After reduced load at
12kN and found the fatigue life one million cycle where shows 88 MPa stress which
is 35% of yield stress. So that designer can consider 35% of yield strength when design
structure for fluctuating and repeated loading conditions.

72
List of References

List of References
A. M. Kamara, SMaLL 2011, 'A Numerical Investigation into Residual Stress
Characteristics in Laser Deposited Multiple Layer Waspaloy Parts', Manufacturing science
and engineering, viewed 22-05-15,
<[Link]
#FiniteElementModelingofResidualStress>.

A. Thirugnanam, MkaLR 2014, 'Analysis of Stress in Welded Joint in Bending And in

Torsion Using ‘Ansys’', Middle-East Journal of Scientific Research, vol. 20,
<[Link]

Abid, M & Siddique, M 2005, 'Finite-element simulation of tack welds in girth welding of a
pipe-flange joint', Acta Mechanica, vol. 178, no. 1-2, pp. 53-64,
<[Link]

Austmig ES6, WIA, viewed 7/9/2015, <[Link]

metals/view/austmig-es6>.

AWS 2010, Structural Welding Code Steel, American Welding Society, Miami,Florida.

Aygul, M 2012, 'Fatigue Analysis of Welded Structures Using the Finite Element Method ',
Chalmers University of Technology, Sweden.

Brickstad, B & Josefson, BL 1998, 'A parametric study of residual stresses in multi-pass
butt-welded stainless steel pipes', International Journal of Pressure Vessels and Piping, vol.
75, no. 1, pp. 11-25,
<[Link]

Chang, K-H & Lee, C-H 2009, 'Finite element analysis of the residual stresses in T-joint
fillet welds made of similar and dissimilar steels', The International Journal of Advanced
Manufacturing Technology, vol. 41, no. 3-4, pp. 250-8, <[Link]
008-1487-4>.

Deng, D, Liang, W & Murakawa, H 2007, 'Determination of welding deformation in fillet-

welded joint by means of numerical simulation and comparison with experimental
measurements', Journal of Materials Processing Technology, vol. 183, no. 2–3, pp. 219-25,
<[Link]

Dodge, MF, Dong, HB, Gittos, MF & Mobberley, T 2014, 'Fusion zone microstructure
associated with embrittlement of subsea dissimilar joints', in vol. 5.

Gannon, L, Liu, Y, Pegg, N & Smith, M 2010, 'Effect of welding sequence on residual stress
and distortion in flat-bar stiffened plates', Marine Structures, vol. 23, no. 3, pp. 385-404,
<[Link]

Haghani, R 2012, 'Fatigue-Prone Details in Steel Bridges', Buildings (Basel), vol. 2, no. 4,
pp. 456-76, <[Link]

73
List of References

Hale, S 2014, How do I know if my mesh is good enough, CAE Associates, viewed 27/05/15,
<[Link]

Hicks, JG 1987, Welded joint design, 2nd ed. edn, BSP Professional Books, Oxford.

Hobbacher, AF 2009, 'The new IIW recommendations for fatigue assessment of welded
joints and components – A comprehensive code recently updated', International Journal of
Fatigue, vol. 31, no. 1, pp. 50-8,
<[Link]

Joshi, S & Price, JWH 2009, 'A comparative study on application of design codes for
prediction of fatigue life of a mining dragline cluster', Engineering Failure Analysis, vol. 16,
no. 5, pp. 1562-9, <[Link]

Kirkhope, KJ, Bell, R, Caron, L, Basu, RI & Ma, KT 1999, 'Weld detail fatigue life
improvement techniques. Part 1: review', Marine Structures, vol. 12, no. 6, pp. 447-74,
<[Link]

Lassen, BT, Darcis, P & Recho, N 2006, 'Fatigue Behavior of Welded Joints
Part 2: Physical modelling of the fatigue process', Welding journal, vol. 26, viewed
1/6/2015, <[Link]

Leap Australia, 2014, LEAP Australia Pty Ltd, viewed 7-5-2015,

<[Link]

Lindgren, L-E 2001, 'FINITE ELEMENT MODELING AND SIMULATION OF

WELDING. PART 2: IMPROVED MATERIAL MODELING', Journal of Thermal
Stresses, vol. 24, no. 3, pp. 195-231, viewed 2015/05/06,
<[Link]

Maddox, SJ 2002, Fatigue strength of welded structures, Second edn, The Welding Institure,
Milton Keynes England.

Moaveni, S 2007, Finite element analysis : theory and application with ANSYS, 3rd ed. edn,
Pearson Prentice Hall, Upper Saddle River NJ.

Morris, JL 1955, Welding processes and procedures, Prentice Hall, New York.

PadmaKumari, V 2013, 'Finite Element Analysis of EBW Welded Joint Using Sysweld',
International Journal of engineering technology and advance engineering, vol. 3, no. 2,
viewed 11-04-15, <[Link]

Perić, M, Tonković, Z, Rodić, A, Surjak, M, Garašić, I, Boras, I & Švaić, S 2014, 'Numerical
analysis and experimental investigation of welding residual stresses and distortions in a T-

74
List of References

joint fillet weld', Materials & Design, vol. 53, no. 0, pp. 1052-63,
<[Link]

Pisarski, HG, Tkach, Y & Quintana, M 2004, 'Evaluation of weld metal strength mismatch in
X100 pipeline girth welds', <[Link]
papers/evaluation-of-weld-metal-strength-mismatch-in-x100-pipeline-girth-welds/>.

Qi, DHJ 2006, Finite Element Analysis, viewed 24/04/2015,

<[Link]

Qingyu, S, Anli, L, Haiyan, Z & Aiping, W 2002, 'Development and application of the
adaptive mesh technique in the three-dimensional numerical simulation of the welding
process', Journal of Materials Processing Technology, vol. 121, no. 2–3, pp. 167-72,
<[Link]

Rabih Kamal Kassab1, HC, Ngan Van Lê1, Jacques Lanteigne2 and Marc Thomas1 2012,
'Experimental and Finite Element Analysis of a T-Joint Welding', Journal of Mechanics
Engineering and Automation, viewed 3/5/2015,
<[Link]
>.

Richards, KG 1969, Fatigue strength of welded structures, Welding Institute, Cambridge.

Rother, K & Rudolph, J 2011, 'Fatigue assessment of welded structures: practical aspects for
stress analysis and fatigue assessment', Fatigue & Fracture of Engineering Materials &
Structures, vol. 34, no. 3, pp. 177-204, <[Link]
2695.2010.01506.x>.

Seok, C-S, Suh, M-W & Park, J-H 1999, 'Investigation of welding residual stress of high
tensile steel by finite element method and experiment', KSME International Journal, vol. 13,
no. 12, pp. 879-85, <[Link]

Shen, WY & Clayton, P 1996, 'Fatigue of fillet welded A515 steel', Engineering Fracture
Mechanics, vol. 53, no. 6, pp. 1007-16,
<[Link]

SIEMENS 2015, Finite Element Analysis, Siemens, viewed 12/04/15,

<[Link]

Sulaiman, M, Manurung, Y, Haruman, E, Rahim, M, Redza, M, Lidam, R, Abas, S, Tham, G

& Chau, C 2011, 'Simulation and experimental study on distortion of butt and T-joints using
WELD PLANNER', Journal of Mechanical Science and Technology, vol. 25, no. 10, pp.
2641-6, <[Link]

Syahroni, N & Hidayat, MIP 2012, 3D Finite Element Simulation of T-Joint Fillet Weld:
Effect of Various Welding Sequences on the Residual Stresses and Distortions.

75
List of References

Teng, T-L, Fung, C-P, Chang, P-H & Yang, W-C 2001, 'Analysis of residual stresses and
distortions in T-joint fillet welds', International Journal of Pressure Vessels and Piping, vol.
78, no. 8, pp. 523-38.

TWI Fatigue testinng, 2015, viewed 15/09/15, <[Link]

knowledge/job-knowledge/fatigue-testing-078/>.

WANG Rui, RS, SERIZAWA Hisashi, MURAKAWA Hidekazu, Jianxun ZHANG 2008,
'Numerical and Experimental Investigations on Welding Deformation', vol. 37, viewed
28/05/15, <[Link]

Wang, T, Hopperstad, OS, Lademo, OG & Larsen, PK 2007, 'Finite element analysis of
welded beam-to-column joints in aluminium alloy EN AW 6082 T6', Finite Elements in
Analysis and Design, vol. 44, no. 1–2, pp. 1-16,
<[Link]

Welded Connections, viewed 14/8/2015,

<[Link]

Welding : theory and practice, 1990, North-Holland, Amsterdam.

Welding fundamentals and processes, 2011, ASM International, 10/6/2015,

<[Link]

Welding Inspection Cswip, 2010, viewed 2/6/15,

<[Link]

Woolling, P & Carrouge, D 2002, 'Heat affected zone microstructures in supermartensitic

stainless steels', <[Link]
affected-zone-microstructures-in-supermartensitic-stainless-steels-october-2002/>.

Worksafe, 2015, Government of western Australia, viewed 1/6/15,

<[Link]

Yang, X, Zou, L & Deng, W 2015, 'Fatigue life prediction for welding components based on
hybrid intelligent technique', Materials Science and Engineering: A, vol. 642, pp. 253-61,
<[Link]

Zhang, Y-H, Smith, S, Wei, L, Johnston, C & Stacey, A 2013, 'Measurement and modelling
of residual stresses in offshore circumferential welds', in vol. 3.

76
List of Appendices

List of Appendices
A - Project specification
University of Southern Queensland
FACULTY OF ENGINEERING AND SURVEYING
ENG4111/ENG4112 Research Project
PROJECT SPECIFICATION

FOR: RAFIQUL ISLAM MOHAMMED

FINITE ELEMENTS ANALYSIS OF FILLET WELDED

TOPIC:
JOINT

SUPERVISOR: CHRIS SNOOK

PROJECT AIM:
The aim of the project is to develop techniques for
simplifying finite element analysis such as stress analysis
and dynamic properties of structures under vibration
excitation on non-complete joint penetrated (Fillet) welding.

SPONSORSHIP: University of Southern Queensland

PROGRAMME: Issue A, 18th March 2015

 Research the back ground information of finite element analysis of welded

joint.
 Research on behaviour of welded joints under cyclic load.

 Learn how to use ANSYS to analyse fillet welding joints with specific
parameter, including ANSYS fatigue analysis.

 Design and analyse the model structures with software (Creo/ANSYS) to

compare between complete penetration joints and non-complete penetration
joints.

 Fabricate the T-joint specimen in workshop.

 Complete physical testing of the specimen (using INSTRON).

As time permits:

 Develop techniques for simplifying finite element analysis of welding joints.

Examiner/Co-examiner____________________________________

77
List of Appendices

B - Resource analysis
In order to complete the project various resources are required. This project divide in
to two part, one is to prepare experiment and conduct physical test, another is to design
and simulation with software using computer.

The resource that will required to complete the experiment:

 Access to metal fabrication workshop

 Grade 250 mild steel, size 400mmX300mmX10mm, one piece
 Grade 250 mild steel, size 400mmX150mmX10mm, two pieces
 Welding (MIG) machine and welding wire (ES609S)
 Band saw
 Angle grinder
 Personal protective equipment (PPE)
 Access to material testing lab (Z104.1)
 Test machine operator personnel

The resource that will be required to complete simulation:

 A standard computer capable to run simulation software and producing data

 Access to the internet, journals, articles, websites and newspaper to collect
theoretical information and perform literature review
 Access to library data-base for previous work which were published
 Access to standards
 ANSYS software for design and simulations
 Access to ANSYS official tutorials to learn operating and useful commands
 Access to Z block computer laboratories

Critical resource:

Learning ANSYS program is the critical service in this project. In the simulation most
important data input is to applying welding parameter on the geometry. Knowing this
operation is the most critical.

Cost involves
Material cost: $100, manpower cost: $100 X 4 = $400, overhead cost: $100 there for
total cos involves = $600

78
C - Risk management
Risk assessment is a process for developing knowledge and understanding about
hazards and risks so that good decisions can be taken about controlling them. This
project required use of welding machine, argon mixture with CO2 gas, high power
cable and band saw in laboratory Z4. For testing use of 810 material testing system in
a testing lab Z104.1.

To carry out a basic risk assessment, follow these four steps (Worksafe 2015);

 Gather information about each hazard that can identify.

 Workout likelihood of an accident or incident occurring and consider how
many people are likely to involve.
 Assess the consequence of hazard for example, fatal, suffer major injuries,
suffer minor injuries or negligible.
 Rate the risk.

79
Risk assessment for location Z4

Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)
Material falling off Person who Up to 5 Feet, lower Low
involve material body
handling
Categories Short term controls Long term controls Completion details
Employer: USQ
Use trolley or lifting
Material handling Avoid bending and twisting of body Faculty of Health Engineering and
equipment
Science
Welding on plate Prepared by : Rafiqul Mohammed
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)
Sharp edges of machine parts Person who Up to 5 Fingers, hands Low
involve material and feet.
cutting
Categories Short term controls Long term controls Completion details
Employer: USQ
Material handling Do not keep fingers close to the machine blade Fixed machine guard Faculty of Health Engineering and
Science
Welding on plate PPE must worn Prepared by : Rafiqul Mohammed
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
80
 Number at Parts of
Description of hazards People at risk Risk level
risk (1-5) body
Electric shock Person who Up to 5 All Medium
involve Welding
Categories Short term controls Long term controls Completion details
Inspect welding machine Employer: USQ
Material handling PPE must worn, do not weld if place is wet and cable regularly to Faculty of Health Engineering and
make sure no damage Science
If damage during
Welding on plate operation report to the Prepared by : Rafiqul Mohammed
supervisor
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
Number at Parts of
Description of hazards People at risk Risk level
risk (1-5) body
Welding fume and gases Person who Up to 5 Lung, skin, Low
involve Welding kidneys
Categories Short term controls Long term controls Completion details
Employer: USQ
PPE must be worn and fume mask (welding mask) Local exhaust
Material handling Faculty of Health Engineering and
need to use during welding (ventilation)
Science
Adequate air flow
Welding on plate (ventilation) should be in Prepared by : Rafiqul Mohammed
the workplace
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
81
 Number at Parts of
Description of hazards People at risk Risk level
risk (1-5) body
Radiations Person who Up to 5 Eyes and Low
involve Welding skin
and surrounding
people
Categories Short term controls Long term controls Completion details
Employer: USQ
Skin protection, Eye protection Non-flammable Welding bay covered with
Material handling Faculty of Health Engineering and
clothing welding screens
Science
Welding on plate PPE must be worn Prepared by : Rafiqul Mohammed
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
Number at Parts of
Description of hazards People at risk Risk level
risk (1-5) body
Fire All people, Up to 5 All body Low
whoever in the
workshop
Categories Short term controls Long term controls Completion details
Safe working practice, Employer: USQ
Material handling Use fire extinguisher and fire hoses Flammable goods never Faculty of Health Engineering and
allow in the welding bay Science
Inspect welding machine
Welding on plate Checked welded area after half an hour Prepared by : Rafiqul Mohammed
regularly before start work
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
82
 Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)
Heat during welding Person who Up to 5 Whole body Low
involve Welding
Categories Short term controls Long term controls Completion details
Employer: USQ
Use proper clothing’s, stay away from welding spark Local exhaust and
Material handling Faculty of Health Engineering and
as much as possible adequate air flow
Science
Welding on plate Prepared by : Rafiqul Mohammed
Cutting off welded
Assented to by: Chris Snook
plate
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:
Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)
Compressed gas cylinder explosion Everybody in the Up to 5 All Low
workshop
Categories Short term controls Long term controls Completion details
Case around cylinders Employer: USQ
Material handling Positioned gas cylinder upright preventing access or Faculty of Health Engineering and
accidental contact Science
Welding on plate Use proper valve to control the gases Prepared by : Rafiqul Mohammed
Cutting off welded ensure all gas cylinder are kept in the cool area and
Assented to by: Chris Snook
plate away from potential heat source
Position: Discipline group leader
Workshop PPE
Mechanical engineering
Signature:
Date:

83
Risk management chart for Z104.1 testing lab

Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)

Electric shock Person undertaking Up to 5 Whole body Low

the test

Categories Short term controls Long term controls Completion details

Employer: USQ
Check all the cables connected to the machine for Operate machine
Material handling Faculty of Health Engineering and
inspection tags whether it is up to date according to specifications
Science

Clamping test piece Do not over loaded Prepared by : Rafiqul Mohammed

Operating test
Follow instructions Assented to by: Chris Snook
machine

Position: Discipline group leader

Workshop PPE
Mechanical engineering

Signature:

Date:

84
 Number at
Description of hazards People at risk Parts of body Risk level
risk (1-5)

Flying metal chips Person undertaking Up to 5 Any parts of Low

the test body

Categories Short term controls Long term controls Completion details

Employer: USQ
PPE must be worn such as full covered shows, safety
Material handling Use proper machine guards Faculty of Health Engineering and
goggles and air plugs.
Science

Clamping test piece Know the location of emergency switch Machine up to date Prepared by : Rafiqul Mohammed

Operating test
Do not touch specimen during loading Assented to by: Chris Snook
machine

Position: Discipline group leader

Workshop PPE
Mechanical engineering

Signature:

Date:

85
D - Project timeline
Table 6-1: project timeline

Project proposal: Completed

Project specification: Completed

Literature review: Partially completed, further need to review until completing the project.

Learn ANSYS program: Introduction has been completed with YouTube help. Found access of
ANSYS software at USQ lab with help from supervisor (Mr. Chris Snook). Still unable to access
official tutorial due to registration problem, this learning will be carried out until September, in the
meantime some simulation will be progressed.

Writing preliminary report: First preliminary report submitted 4th June 2015, feedback received
with lots of comments. Revision 2 of this report will submitted by 17th July 2015.

Model simulation: Model has been finalised at this stage now updating knowledge to applying
welding parameter to the model.

Prepare experiment: Request has been sent to the workshop to fabrication of specimen and
expected completion time will be end of July.

Experiment test: Once specimen done then booked for the test. The test will required one day to
complete. According to timeline by August can be done.

Thesis writing: Thesis writing in progress, continue writing until the day of submission.

86