Index

A FINITE ELEMENTS BASED APPROACH
FOR
FRACTURE ANALYSIS OF WELDED JOINTS
IN CONSTRUCTION MACHINERY
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
TANER KARAGÖZ
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF MASTER OF SCIENCE
IN
MECHANICAL ENGINEERING
AUGUST 2007
Approval of the Thesis
“A FINITE ELEMENTS BASED APPROACH FOR FRACTURE ANALYSIS

OF WELDED JOINTS IN CONSTRUCTION MACHINERY”
Submitted by TANER KARAGÖZ in partial fulfillment of the requirements for

the degree of Master of Science in Mechanical Engineering by,
Prof. Dr. Canan Özgen

Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Kemal İder

Head of Department, Mechanical Engineering
Asst. Prof. Dr. Serkan Dağ

Supervisor, Mechanical Engineering, METU
Examining Committee Members:
Prof. Dr. Eres Söylemez (*)

Mechanical Engineering, METU
Asst. Prof. Dr. Serkan Dağ (**)

Prof. Dr. F. Suat Kadıoğlu

Assoc. Prof. Dr. Bora Yıldırım

Mechanical Engineering, Hacettepe University
Kadir Geniş (M.S.)

Hidromek Ltd. Sti.
Date:
(*) Head of Examining Committee

(**) Supervisor
I hereby declare that all information in this document has been
obtained and presented in accordance with academic rules and
ethical conduct. I also declare that, as required by these rules and
conduct, I have fully cited and referenced all material and results that
are not original to this work.
Name, Last name: Taner KARAGÖZ
Signature :
iii
ABSTRACT
A FINITE ELEMENTS BASED APPROACH FOR FRACTURE ANALYSIS

OF WELDED JOINTS IN CONSTRUCTION MACHINERY
Karagöz, Taner
M.S, Department of Mechanical Engineering
Supervisor: Asst. Prof. Dr. Serkan DAĞ
August 2007, 119 pages
This study aims to develop a computer program to perform finite elements

based fracture mechanics analyses of three dimensional surface cracks in
T-welded joints of construction machinery. The geometrical complexity of
the finite elements models and the requirement of large computer
resources for the analyses necessitate the use of shell elements for
general stress distribution optimization. A sub-modeling technique,
together with a shell to solid conversion method, enables the user to
model a local region and analyze it by defining the weld and crack
parameters. It is assumed that the weld material is the same with the
sheet metal material and the surface cracks are considered to occur on
two weld toes and weld root. The surface cracks are assumed to have a
semi elliptical crack front profile. In order to simulate the square-root strain
iv
singularity around the crack front, collapsed 20-node three dimensional
brick elements are utilized. The rest of the local model is modeled by using
20-node three dimensional brick elements. The main objective of this work
is to calculate the mixed mode energy release rates around the crack front
for a sub-model of a global shell model by using J-integral method.
Keywords: Finite elements, sub-modeling, semi elliptical surface crack,

fillet weld, J-integral method
v
ÖZ
İŞ MAKİNALARINDAKİ KAYNAKLI BAĞLANTILARIN SONLU

ELEMANLARA DAYALI BİR YAKLAŞIMLA KIRILMA ANALİZİ
Karagöz, Taner
Yüksek lisans, Makina Mühendisliği Bölümü
Tez Yöneticisi: Y. Doç. Dr. Serkan DAĞ
Ağustos 2007, 119 sayfa
Bu çalışmanın amacı, iş makinelerinin köşe kaynaklı bağlantılarında üç

boyutlu yüzey çatlaklarının sonlu elemanlara dayalı kırılma mekaniği
analizlerini gerçekleştirecek bir bilgisayar programı geliştirmektir. Sonlu
elemanlar modellerinin karmaşık geometrisi ve analizlerin yüksek
kapasiteli bilgisayar gereksinimi genel gerilme dağılımı eniyilemesinde
kabuk elemanların kullanılmasını zorunlu kılar. Bir alt modelleme tekniği,
kabuktan katıya dönüştürücü bir metotla birlikte, kullanıcının tanımlanan
kaynak ve çatlak parametrelerine göre lokal bir bölgeyi modellemesini ve
analizlerini gerçekleştirmesini sağlar. Kaynak malzemesinin sac metal
malzemesiyle aynı olduğu varsayılmış ve yüzey çatlaklarının kaynağın iki
kenarında ve kökünde olduğu düşünülmüştür. Yüzey çatlaklarının yarı
vi
eliptik çatlak yüzü görüntüsüne sahip olduğu varsayılmıştır. Çatlak yüzü
çevresindeki kare kök gerinim tekilliğini benzeştirebilmek için çökertilmiş
20 düğüm noktalı üç boyutlu tuğla elemanlar kullanılmıştır. Lokal modelin
geri kalan kısmı 20 düğüm noktalı üç boyutlu tuğla elemanlar ile
modellenmiştir. Bu çalışmanın asıl amacı, bir global kabuk modelin alt
modelinin çatlak yüzü çevresindeki karışık mod enerji açığa çıkma
oranlarını J-integrali metodunu kullanarak hesaplamaktır.
Anahtar kelimeler: Sonlu elemanlar, alt modelleme, yarı eliptik yüzey

çatlağı, köşe kaynağı, J-integrali metodu
vii
To My Family and My Love Özgül
viii
ACKNOWLEDGMENTS
I wish to express my special thanks to Assistant Prof. Dr. Serkan DAĞ for
his continuous help and guidance throughout the duration of this study.
I would also like to express my appreciation to Kadir GENİŞ, Prof. Dr. Eres
SÖYLEMEZ and Hakan IŞIKTEKİN for their important contributions and
useful discussions during the study.
I would also like to thank my friends Ferhan FIÇICI, C. Can UZER, Erkal
ÖZBAYRAMOĞLU, Murat IŞILDAK and Gençer GENÇOĞLU for their
suggestions and support.
Special thanks to my family for all their love, patience and support during
all my life.
Finally, I wish to express my gratitude to my love M. Özgül ŞAHİNKAYA

who made my life more meaningful for seven years.
This study was supported by Hidromek Ltd. Sti.
ix
TABLE OF CONTENTS
ABSTRACT................................................................................................iv
ÖZ..............................................................................................................vi
DEDICATION........................................................................................... viii
ACKNOWLEDGEMENTS ..........................................................................ix
TABLE OF CONTENTS..............................................................................x
LIST OF TABLES..................................................................................... xiv
LIST OF FIGURES ...................................................................................xv
LIST OF SYMBOLS AND ABBREVIATIONS......................................... xxiv
CHAPTER
1. INTRODUCTION ............................................................................... 1
2. LITERATURE REVIEW...................................................................... 6
3. FRACTURE ANALYSIS IN THREE DIMENSIONS.......................... 17
3.1 Introduction ................................................................................ 17
3.2 Stress Intensity Factor (SIF)....................................................... 18
3.3 Fracture Toughness ................................................................... 21
x
3.4 Two and Three Dimensional Linear Elastic Crack Tip Fields ..... 22
3.5 Energy Release Rate ................................................................. 28
3.6 J-integral .................................................................................... 30
3.7 Numerical Evaluation of the J-integral........................................ 31
4. FINITE ELEMENT MODELING........................................................ 33
4.1 Introduction ................................................................................ 33
4.2 Problem Definition ...................................................................... 34
4.3 Sub-modeling ............................................................................. 36
4.3.1 Global Model ..................................................................... 38
4.3.1.1 Boundary Conditions ............................................. 41
4.3.1.1.1 Bucket Breakout Force ............................ 41
4.3.1.1.2 Arm Breakout Force ................................ 43
4.3.1.1.3 Lateral Force ........................................... 44
4.3.1.1.4 Other Boundary Conditions ..................... 44
4.3.2 Local Region Selection ...................................................... 45
4.3.3 Parameters ........................................................................ 46
xi
4.3.3.1 Weld Parameters ................................................... 46
4.3.3.2 Crack Parameters.................................................. 49
4.3.4 Shell-to-Solid Conversion Method ..................................... 50
4.3.5 Material Properties............................................................. 52
4.3.6 Tyings and Local Model Boundary Conditions................... 53
4.3.7 J-integral Calculation ......................................................... 55
4.4 Verification of Finite Element Modeling and J-integral

Calculation ................................................................................ 56
4.4.1 Verification of Finite Element Modeling............................. 56
4.4.2 Verification of J-integral Calculation.................................. 57
4.4.2.1 Verification of Embedded Circular and Elliptical

Crack Models ....................................................... 57
4.4.2.2 Verification of 3D Inclined Semi Elliptical Surface

Crack Models ....................................................... 63
4.5 Conclusion ................................................................................ 66
5. PARAMETRIC ANALYSES AND NUMERICAL RESULTS.............. 70
5.1 Introduction ............................................................................... 70
xii
5.2 Parametric Analyses ................................................................. 71
5.2.1 Boom Parametric Analyses............................................... 72
5.2.2 Arm Parametric Analyses ................................................. 90
5.3 Numerical Results................................................................... 108
6. CONCLUDING REMARKS ............................................................ 110
REFERENCES ...................................................................................... 113
xiii
LIST OF TABLES
TABLES
Table 4.1 Material Properties of the Steel Used in Global Model
Analyses................................................................................... 53
Table 4.2 Chemical Compositions of the Steel Used in Global Model
Analyses................................................................................... 53
Table 4.3 Details and Comparison for a Penny-shaped Crack Analysis .. 59
Table 4.4 Details for Elliptical Crack Analysis .......................................... 62
Table 4.5 Normalized SIF Comparison for an Elliptical Crack.................. 63
Table 4.6 Details for Noda’s Solution....................................................... 65
Table 4.7 Details for Isida’s Solution........................................................ 66
Table 4.8 Normalized J-integral Value Comparisons for Inclined Semi
Elliptical Surface Cracks .......................................................... 67
Table 4.9 Normalized J-integral Value Comparisons for Inclined Semi
Elliptical Surface Cracks .......................................................... 68
Table 5.1 Parametric Analyses Material Properties and Crack
Parameters............................................................................... 71
Table 5.2 Parametric Analyses Weld Parameters (mm) .......................... 71
xiv
LIST OF FIGURES
FIGURES
Figure 1.1 HMK 102 B Energy Series Backhoe Loader General View....... 2
Figure 1.2 HMK 300 LC Excavator General View ...................................... 3
Figure 3.1 Mode I (Opening Mode) .......................................................... 19
Figure 3.2 Mode II (In-plane Shearing Mode) .......................................... 20
Figure 3.3 Mode III (Out-of-plane Shearing Mode) .................................. 20
Figure 3.4 Fracture Toughness vs. Material Thickness Graph ................ 22
Figure 3.5 Distribution of Stresses in Vicinity of Crack Tip....................... 23
Figure 3.6 A Two-dimensional Crack Configuration................................. 23
Figure 3.7 Three Dimensional Crack Front and the Local Coordinate
System ...................................................................................... 27
Figure 3.8 Definition of the J-integral ....................................................... 31
Figure 4.1 a) Global Model b) Local Model [48] ....................................... 37
Figure 4.2 FEA Model of Excavator Mechanism ...................................... 39
Figure 4.3 Basic Parts of Excavator......................................................... 40
Figure 4.4 Von Mises Stress Map for Shell Outer Layer .......................... 41
Figure 4.5 Bucket Breakout Force Position.............................................. 42
Figure 4.6 Arm Breakout Force (Digging Force) Position ........................ 43
Figure 4.7 Lateral Force on the Bucket.................................................... 45
Figure 4.8 Selection Possibilities on Excavator Boom and Arm............... 46
xv
Figure 4.9 One Sided Weld Parameters .................................................. 47
Figure 4.10 Two Sided Weld Geometry Parameters................................ 48
Figure 4.11 Two Sided Weld Penetration Depth Parameters .................. 48
Figure 4.12 Semi Elliptic Crack Parameters ............................................ 49
Figure 4.13 Crack Inclination Angle Ranges a) Crack at Weld Toe 1
b) Crack at Weld Toe 2 c) Crack at Weld Root ......................... 49
Figure 4.14 Shell-to-Solid Conversion Possibilities a) Shell Model b) Solid
Model Welded from Inner Side c) Solid Model Welded from Outer
Side d) Solid Model Welded from Both sides ............................ 50
Figure 4.15 Semi Elliptic Surface Crack at Weld Toe .............................. 51
Figure 4.16 a) 20-node Isoparametric Brick Element b) Collapsed 20-node
Isoparametric Brick Elements.................................................... 51
Figure 4.17 Collapsed 20-node Isoparametric Brick Elements at Weld Toe
Crack Front Start....................................................................... 52
Figure 4.18 Deformation History Applied to Tying Constrained Nodes at
Local Model Boundaries............................................................ 54
Figure 4.19 Constrained Nodes and the Other (“follower”) Nodes on the
Edge are Forced to Stay in the Plane Defined by the Nodes in
Transverse Direction ................................................................. 55
Figure 4.20 Experimental and Finite Element Analysis Results
Comparison............................................................................... 56
Figure 4.21 Embedded Penny-shaped Crack .......................................... 57
Figure 4.22 General View of Embedded Circular Crack .......................... 59
xvi
Figure 4.23 Close-up View of Embedded Circular Crack......................... 60
Figure 4.24 Parametric Representation of a Point on an Ellipse.............. 61
Figure 4.25 Close-up View of Embedded Elliptic Crack........................... 62
Figure 4.26 An Inclined Semi Elliptical Surface Crack in a Semi-infinite
Body .......................................................................................... 64
Figure 4.27 General View of an Inclined Semi Elliptical Surface Crack Von
Mises Stress Distribution a) Over Deformed Shape of the Crack
Mouth b) Side View of the Loaded Cube Showing the Inclined
Crack......................................................................................... 65
Figure 5.1 Excavator Boom Global Model (Von Mises Stress Outer
Layer) ........................................................................................ 73
Figure 5.2 Excavator Boom Global Model (Von Mises Stress Inner
Layer) ........................................................................................ 73
Figure 5.3 Excavator Boom Model Sheet Metals and the Selected
Region....................................................................................... 74
Figure 5.4 Boom Selected Shell Region and Boom Solid Local Model Von
Mises Stress Map...................................................................... 74
Figure 5.5 Normalized J-integral versus Inclination Angle (ψ) (Crack at
Weld Toe 1, Lateral Force is Included, β =310) ......................... 75
Weld Toe 1, Lateral Force is Included, β=630) .......................... 75
xvii
Weld Toe 1, Lateral Force is Included, β=1170) ........................ 76
Weld Root, Lateral Force is Included, β=310)............................ 80
Weld Root, Lateral Force is Included, β=1170).......................... 81
xviii
Weld Toe 1, Lateral Force is not Included, β=310) ................... 82
Weld Toe 1, Lateral Force is not Included, β=1170).................. 84
xix
Weld Root, Lateral Force is not Included, β=310) ..................... 87
Weld Root, Lateral Force is not Included, β=1170) ................... 89
Figure 5.35 A Fatigue Crack Occurred on HMK 300 LC Excavator Arm
During the Performed Tests in Hidromek Ltd. Sti ...................... 90
Figure 5.36 Excavator Arm Global Model (Von Mises Stress Outer
Layer) ........................................................................................ 91
Figure 5.37 Excavator Arm Global Model (Von Mises Stress Inner
Layer) ........................................................................................ 91
Figure 5.38 Excavator Arm Model Sheet Metals and the Selected
Region....................................................................................... 92
Figure 5.39 Arm Selected Shell Region and Arm Solid Local Model Von
Mises Stress Map...................................................................... 92
xx
xxi
Weld Root, Lateral Force is Included, β=1490)........................ 100
Weld Toe 1, Lateral Force is not Included, β=310) ................. 100
Weld Toe 1, Lateral Force is not Included, β=1170)................ 102
xxii
Weld Root, Lateral Force is not Included, β=1170) ................. 107
Weld Root, Lateral Force is not Included, β=1490) ................. 107
xxiii
LIST OF SYMBOLS AND ABBREVIATIONS
SYMBOLS
m(x,a) : Weight Function

σ(x) : Stress Distribution
K : Stress Intensity Factor
a : Half Crack Length
b : Crack Depth
KIC : The Critical Mode I Stress Intensity Factor
KC : The Critical Mode I Stress Intensity Factor
B : Specimen Thickness
σX : Normal Stress Component on the Crack Front in x Direction
σY : Normal Stress Component on the Crack Front in y Direction
τXY : Shear Stress Component on the Crack Front
r : Distance from the Crack Tip
θ : Angle from the Crack Plane
σij : The Components of the Stresses
u, v, w : Displacement Components in x, y and z Directions
µ : Shear Modulus
κ : Shear Parameter
v : Poisson’s Ratio
E : Young’s Modulus
KI, KII, KIII : Mode I, II, III Stress Intensity Factors
s : Arc Length of the Crack Front
t : Tangential Direction on the Crack Front
n : Normal Direction on the Crack Front
b : Binormal Direction on the Crack Front
σbb : Normal Stress Component on the Crack Front
xxiv
ubb : Normal Displacement Component on the Crack Front
G : Energy Release Rate
Π : Strain Energy
Gc : Fracture Toughness
GI, GII, GIII : Mode I, II, III Energy Release Rates
J : J-integral
W : Strain Energy Density
T : Kinetic Energy Density
uj : Displacement Vector
Γ : Integration Path
-
J : Converted J-integral
A : Area inside Integration Path
q1 : A Function Introduced in J-integral Conversion
Fd : Bucket Cylinder Force
P : Working Pressure of the Cylinder
Dd : Bucket Cylinder Diameter
Fb : Bucket Breakout Force
c : Perpendicular Distance Bucket Cylinder Axis - Lever Pivot
d : Perpendicular Distance Connecting Link Axis - Lever Pivot
e : Perpendicular Distance Connecting Link Axis - Bucket Pivot
f : Radius Bucket Pivot - Tooth Lip
Fc : Arm Cylinder Force
Dc : Arm Cylinder Diameter
Fa : Arm Breakout Force
g : Perpendicular Distance Arm Cylinder Axis - Arm Pivot
h : Distance Arm Pivot - Tooth Tip
M : Hydraulic Oil Motor-Swing Unit Moment Capacity
F : Lateral Force
Leg1 : Vertical Leg Length of the Weld
Leg2 : Horizontal Leg Length of the Weld
RD : Root Depth
xxv
Leg_1 : Vertical Leg Length of the Inner Weld
Leg_2 : Horizontal Leg Length of the Inner Weld
MCD : Middle Crack Distance
MCL : Middle Crack Length
αx, αy, αz : Rotation About Global x, y and z Axes
KI* : Normalized Mode I Stress Intensity Factor
J* : Normalized J-integral Value
β, φ : Crack Front Position Angles
E (φ) : Elliptical Integral
Ψ : Crack Inclination Angle
JI, JII, JIII : Mode I, II, III J-integral Values
d1, d2, d3 : Cube Dimensions in y, x, z Directions
ABBREVIATIONS
FEA : Finite Element Analysis

FEM : Finite Element Method
BEM : Boundary Element Method
GMAW : Gas Metal Arc Welding
GTAW : Gas Tungsten Arc Welding
FGM : Functionally Graded Material
ASTM : American Society for Testing and Materials
APDL : ANSYS Parametric Design Language
JSSC : Japanese Society of Steel Construction
SAE : Society of Automotive Engineers
SIF : Stress Intensity Factor
CT : Compact Tension
GUI : Graphical User Interface
xxvi
CHAPTER 1
INTRODUCTION
Earth-moving machines are heavy-duty engineering vehicles and are

primarily used for the movement of large quantities of bulk materials, earth,
gravel and broken rock in road building, mining, construction, quarrying,
trenching, demolition, grading, lifting, river dredging and land clearing
applications. The most common types of earth-moving machines are:
backhoe loaders, excavators, bulldozers, loaders, cranes, and graders.
Backhoe loaders (Figure 1.1) are the most common type of earth-moving
machines in the world. Due to their relatively small size and versatility,
backhoe loaders are used for urban engineering applications and small
construction projects (such as building a small house, fixing city roads, small
demolitions, digging holes/excavating etc). The machine is self-propelled,
highly mobile with a mainframe to support and accommodate both rear-
mounted backhoe and front-mounted loader. Backhoe loader consists of two
main mechanisms: backhoe and loader. The Backhoe digs, lifts, swings and
discharges the material while machine is stationary. When used in Loader
mode, the machine loads material into the bucket through forward motion of
the machine and lifts, transports, and discharges the material [1].
1
Figure 1.1 HMK 102 B Energy Series Backhoe Loader General View
Excavator (Figure 1.2), which is also called a 360-degree excavator or

digger, sometimes abbreviated simply to a 360, is an engineering vehicle
consisting of a backhoe and cab mounted on a pivot (a rotating platform)
atop an undercarriage with tracks or wheels [2]. It is a mobile machine which
has an upper structure capable of continuous rotation and which digs,
elevates, swings, and dumps material by action of the boom, the arm, or
telescopic boom and bucket [3]. Excavators come in a wide variety of sizes.
In accordance with their sizes they are used in many roles such as digging of
trenches, holes, foundations, material handling, brush cutting with hydraulic
attachments, demolition, general grading/landscaping, heavy lift, mining, river
dredging.
Often the bucket can be replaced with other tools like a breaker, a grapple or
an auger.
Excavators are usually employed together with loaders and bulldozers.
2
Figure 1.2 HMK 300 LC Excavator General View
Both excavators and backhoe loaders must work reliably in severe and
unpredictable working conditions. While these machines are performing their
required tasks some components of these machines are exposed to repeated
fluctuating stresses, which cause fatigue cracks, especially on digging and
loading components, to occur. Owing to these cracks, the components of the
machines can malfunction, fracture or even cause danger to the life of
people.
Welding is the most dominant joining method during the manufacturing

process of all earth-moving machines and it is the main source of cracks
occurring on the components. The assessment of the welded joints is a major
industrial problem because the welds are the determining factor of expected
life of earth-moving machines. Accordingly, the welded joints are the regions
of weakness in a structure and must be fully understood to improve the
expected life of the earth-moving machines.
3
The crucial problem in assessment of welded joints is the difficulty of defining
weld geometry in a manner, which is sufficiently precise for analysis but
sufficiently simple for industrial use. For this reason, designers have not been
able to take full advantage of the advent of finite element analysis (FEA) and
other numerical methods, which have revolutionized the assessment of
stress concentrations in solid components. In principle a welded joint can be
analyzed using FEA, but in practice a very detailed model is required to
capture the local stresses around the weld bead, making the approach
impractical for real components, especially if they contain many welds. Due
to these reasons, in the previous studies, methods for fatigue assessment of
welds mostly tend to be feature-based and empirical [4] [5] [6] [7] [8] [9] [10]
[11]. In these work, fatigue life predictions were made by using the ‘Paris-
Erdoğan equation’ according to the tests results.
In literature, numerical methods have not been investigated as extensively as

their experimental counterparts. This is possibly due to the complexity in the
formulation of the problems, the requirement of large computer resources for
numerical calculations, and the lack of efficient methods to provide accurate
results. Among the methods used for studying welded joints and calculating
fracture parameters, the finite element method (FEM) together with domain
integrals is commonly used to extract the results for the energy release rate
[12] [13] [14] [15] [16].
In real life applications, the most dominant fatigue cracks evolve into surface
cracks, which often have a semi-elliptical shape. Accordingly, different semi-
elliptical surface cracks were placed into finite element models and boundary
element models by using various software [17] [18] [19] [20]. In these studies,
3-D models were created either manually or parametrically; boundary
conditions were defined either for full or axi-symmetric models. To calculate
the stress intensity factor different techniques are used.
4
In the present study, three-dimensional semi-elliptic surface crack problem in
T-welded joints are examined using a three-dimensional finite element
technique. The crack problem is analyzed by using sub-modeling technique
and the boundary conditions are directly taken from the global model. A shell
to solid conversion technique creates the 3-D solid local models, according to
the selected region. A prepared graphical user interface applies the desired
weld and fracture parameters to automatically formed local model. Then, by
applying J-integral approach to generated 3-D solid models energy release
rate is evaluated for mixed mode loading type. During the study, commercial
finite element analysis software MSC.Mentat-Marc is used.
5
CHAPTER 2
LITERATURE REVIEW
The main objective of this study is to model the three-dimensional semi-

elliptical surface crack at toe and root of fillet welded joints and to calculate
the mixed mode J-integral values around the crack front and performing
structural zooming analysis by using the finite element method (FEM). The
present study focuses only on T-welded joints and it can be considered as
one of the first in the literature dealing with automatically generated three-
dimensional surface cracks at fillet weld toe and root in a structural zooming
analysis.
The accurate calculation of stress intensity factors for 3-D surface and corner
cracks has long been recognized as an important computational problem in
fracture mechanics. Irwin, who first obtained an approximate solution for
surface crack problem in 1962, recognized this. Since the introduction of the
J-integral as a fracture mechanics parameter by Cherepanov [54] and Rice
[55], many numerical solutions have been developed. The application of the
finite element method (FEM) and the boundary element method (BEM) to the
evaluation of the J-integral is well established for two-dimensional problems.
For three-dimensional problems the J-integral has been directly applied to
the finite element method by various workers, however, the evaluation of
surface integrals is cumbersome in FE analyses. This led to the modification
of the J-integral to a domain integral by Nikishkov and Atluri, in which the J-
6
integral is multiplied by a simple function called an "S" function. The method
is known as equivalent domain integral (EDI) and is computationally
appealing as the domain integral is accurately and easily obtained in FE
analysis [21].
Most studies in the literature that deal with surface and corner cracks
concentrate on mode I loading where mode II and mode III stress intensity
factors are zero along the crack front. In practical applications, however,
surface cracks under mixed mode conditions can be encountered frequently.
These flaws may experience mixed mode loading due to mainly three factors:
(1) mixed remote loading, i.e., normal and shear remote loads acting on a
component having perpendicular crack to the normal loading direction, (2)
deflected or inclined crack under normal/uniaxial remote loading, and (3)
mechanical and/or thermal loads combined with arbitrary restraint conditions
[19].
Singh et al. [5] performed experiments to determine the effect of fillet

geometry on fatigue properties of cruciform welded joints in structural steels.
They predicted the fatigue life of AISI 304L cruciform joints failing at the weld
toe using a two stage model. The local stress life method was applied to
calculate the fatigue crack initiation life, whereas the fatigue crack
propagation life was estimated using fracture mechanics concepts. Constant
amplitude fatigue tests with stress ratio R=0 were carried out using a 100 kN
servohydraulic Dartec universal testing machine at a frequency of 30 Hz. An
automatic crack monitoring system based on crack propagation gauges was
used to obtain the propagation data during the fatigue process. In their study
they reported that there are two types of fatigue cracking in fillet welded
joints, namely, (i) root cracking and (ii) toe cracking. According to the test
data they obtained fatigue crack growth curves by using Paris-Erdogan
equation. The results showed that crack growth rate of GMAW with convex
fillets is greater than GTAW joints with concave fillets. Their results show
good correlation with the BS 5400: part 10 design curve.
7
Lahti et al. [23] conducted three-point bending fatigue tests on stainless steel
fillet welds by using two stainless steel grades: ferritic–martensitic EN 1.4003
and austenitic EN 1.4310. The test results obtained were shown to be in
good agreement with suggested fatigue classes in the Eurocode 3 design
standard, derived from fatigue data on structural steels. However, if the size
of the weld was increased, and the failure location could be moved to the
weld toe instead of the weld root, a significant increase in fatigue strength
was observed. Eurocode 3 was found to describe well the fatigue
characteristics of the ‘worst-case’ welds, i.e., welds prone to root failure.
Kainuma et al. [6] performed another experimental study to investigate the

fatigue strength of load-carrying fillet welded cruciform joints with weld root
failures. They used five different weld shapes: an isosceles triangle, two
types of scalene triangles, a concave curvature, and a convex curvature in
their study. They used JSSC average design curve to determine the crack
propagation rate and stress intensity factor. In the study finite element
analysis were also performed to determine the stress intensity factors by
applying energy method to FEM. Based on the experimental and analytical
results, the influence of weld shape, weld size, weld penetration depth, and
plate thickness on fatigue strength was determined, and a numerical
expression for the weld throat thickness was derived to evaluate fatigue
strength.
In the paper by Statnikov et al. [22], the fatigue test was applied to T-weld
joints from steel weldox 420 by means of four-point-bending test method. The
work was intended to obtain the initial data to compare the efficiency of the
post-weld treatment techniques in terms of increasing fatigue strength of
welded joints and develop ultrasonic impact treatment technique that ensures
rather high efficiency of the method. In all tests the cracks were formed at the
weld toe of the specimens. Finally, the fatigue curves for welded joints in the
as-welded and improved conditions were obtained.
8
Fricke et al. [24] studied fillet welded joints in ship hulls experimentally. The
study proposes a simplified approach for the fatigue strength assessment
with respect to root cracking, which is based on a local nominal stress in a
defined area of the weld throat and on common fatigue classes for the
assessment of cruciform joints. Fatigue tests and numerical analyses of local
stresses and crack propagation from the root gap have been performed.
Some tests showed unexpected results in comparison with the calculations.
The study resulted that two different types of cracks are possible, starting
from the weld toe and from the non-welded root gap. The most critical crack
initiation site depends highly on the weld throat thickness and on the actual
axial misalignment. The latter influences more the cracks starting from the
weld toe, which are usually assessed on the basis of the structural hot-spot
stress approach. The approach has been verified by experimental and
numerical (such as BEM and FEM analysis) investigations of two typical
structural configurations.
In the thesis study of Ficici [17] semi– elliptical surface cracks in a test
specimen are modeled. The specimen is declared in the standard ISO/DIS
14345 [25]. The specimen is examined by considering axial and bending
types of loading. All parts of the model including the semi- elliptical surface
crack are generated in the finite element software MSC.MENTAT– MARC
and the crack profile can be placed at the weld toe or at the weld root
depending on the user’s choice. The study uses displacement correlation
technique for computing modes I, II and III stress intensity factors under
mechanical loading. The main goal is to prepare a parametric model with
user interface, which makes all of the stages– including modeling the
specimen, placing the crack, loading, post-processing and computing the
mixed-mode stress intensity factors– automatically.
Gray et al. [26] presented a modification to the quarter-point crack tip

element and employed this element in two-dimensional boundary integral
fracture analysis. They calculated the stress intensity factors with the
9
displacement correlation technique. The obtained results are highly accurate,
and significantly more accurate than with the standard element. The
improvements are especially dramatic for mixed mode problems involving
curved and interacting cracks.
Inan [18] worked on modeling of semi-circular surface cracks in a ceramic

(ZrO2) – titanium alloy (Ti-6Al-4V) FGM coating bonded to a homogeneous
titanium alloy substrate under mode I mechanical or thermal loading
conditions. A three dimensional finite element model containing a semi-
circular surface crack is generated using the general purpose finite element
software ANSYS. Mode I stress intensity factor was derived by three–
dimensional displacement correlation technique. The stress intensity factors
are calculated for FGM coating– substrate systems subjected to uniform
tension, bending, fixed-grip tension, three point bending and temperature
gradients. In order to examine the accuracy of the model, calculated stress
intensity factors are compared with those given by Newman and Raju [56] for
various crack dimensions under tension or bending loads. These results
show that calculations of mode I stress intensity factor by means of
displacement correlation technique using finite element analysis was
sufficiently accurate.
In the thesis study of Sabuncuoglu [27], stress intensity factors at the crack
tip for functionally graded materials (FGM’s) were evaluated via the finite
element method in conjunction with the displacement correlation technique. A
parametric modeling code for test specimen given in ASTM E399 was
prepared for mode I stress intensity factor calculations by using ANSYS
software. All the parametric modeling stages were carried out by APDL
codes. Since the cracks are symmetric one forth of the model was formed in
the analysis. In the analysis, 20 nodes brick elements were used in order to
satisfy the strain singularity at the crack front. It is seen that the calculated
values are very close to those given in the studies Kadioglu et al. [28] and
Guo et al. [29]. From these analyses, it can be said that displacement
10
correlation technique is a suitable way of determining the stress intensity
factor for FGM structures.
Ayhan [19] reported mixed mode stress intensity factor solutions for deflected
and inclined surface cracks in finite-thickness plates under uniform tensile
remote loading by using three-dimensional enriched finite elements. The
study demonstrates the convenience of the enriched finite element technique
for these types of problems. Regardless of how a surface or corner crack is
initiated or introduced in a component, accurate prediction of the fracture
conditions, i.e., mixed mode stress intensity factors, is very important to
assess its remaining life. Accordingly, mixed mode stress intensity factor
solutions are generated for semi-circular surface cracks with various
deflection and inclination angles ranging from 00 to 750. The mixed mode
stress intensity factor solutions presented in the paper are obtained using
FRAC3D, a three-dimensional fracture analysis program. It was shown, for
both crack types, that mode I stress intensity factors decrease in magnitude
along the whole crack front as the deflection or inclination angle increases.
Mode II and mode III stress intensity factors, on the other hand, increase
initially as the deflection or inclination angle increases and then decrease for
higher deflection or inclination angles. It was also demonstrated that
decreasing the plates thickness has a magnifying effect on the fracture
parameters, especially on the mode I stress intensity factor. Finally, crack
propagation angles along deflected and inclined crack fronts were shown to
increase in magnitude along the whole crack front with increasing deflection
or inclination angle.
Guo et al. [30] attempted to simplify the stress intensity factor calculation for
integral welded integral structures, which is the current trend in commercial
aircraft manufacture instead of conventional built-up riveted structures. It is
well known that on the conventional T-plate welded joint, many failures are
due to the fatigue cracks initiating and developing from the weld toes where
large stress concentrations are present. The fracture and fatigue analysis are
11
usually very complex and exact solutions for stress intensity factors are not
always available. Therefore, the weight function method is often used
because it enables the stress intensity factors for a variety of loading
conditions to be calculated by simple integration of the weight function m(x,a)
and the stress distribution σ(x) expression. The stress intensity factor weight
function for a single edge crack originating from the T-plate weld toe was
derived from a general weight function form and two reference stress
intensity factors.
The stress intensity factors (K) are obtained using the finite element method.
The finite element analysis was conducted using ABAQUS standard (version
6.4). The model containing a one-dimensional edge crack originating from
the weld toe was analyzed. The sub-modeling technique of the Finite
Element Method was used, so that the mesh at crack tip vicinity could be
refined substantially. The stress intensity factor (K) for Mode I was calculated
from the J-integral which was calculated using the energy domain integral
methodology.
The comparisons showed that the derived weight function can make accurate
predictions for stress intensity factors. The derived weight function is valid for
the relative depth a/t ≤ 0.8. It is also shown that this weight function is
suitable for the stress intensity factor calculation for the cracked laser-welded
padded plate geometries under general loading conditions.
Baumjohann et al. [13] have written a computer code for parametric modeling
of crack geometries. Then, they determined J-integrals of ductile growing
cracks located between two comparative contours by interpolation. The
automatic modeling and a mathematical program processing the finite
element results evaluate the crack growth of the finite element results very
effectively. In their study they used the finite element analysis software
ABAQUS to determine temperature distribution, displacements, stresses and
12
J-integrals. Their study indicated that parametric modeling is very important
for effective SIF calculations.
In their study Lin et al. [15] described a multiple degree of freedom numerical
procedure applicable to the prediction of the fatigue crack growth of surface
cracks in plates under a combined tension and bending load. The procedure
performs a three-dimensional finite element analysis to estimate the stress
intensity factors at a set of points along the crack front, and then calculates
the crack growth increments at these points invoking a fatigue crack growth
relationship. A new crack front is established using a cubic spline
approximation. A remeshing technique developed enables the procedure to
be implemented automatically, and then fatigue crack growth can, therefore,
be predicted in a step-by-step way.
The study displays the sensitivity of stress intensity factor results to crack
shape, the effect of mesh orthogonality and the J-integral path
independence. According to the results the stress intensity factor results are
sensitive to the crack front shape, that the cubic spline approximation gives
more accurate results than the polygonal line approximation. The orthogonal
mesh seems unnecessary for the J-integral but necessary for 1/4-point
displacement method. J-integral path independence is usually maintained but
is lost at the free surface if a slightly non-orthogonal intersection exists
between the free surface and the crack front.
The variation of stress intensity factors along the crack front is estimated
using the 1/4-point crack opening displacement method or the J-integral
method. Based on the results J-integral method gives more reliable results.
The present technique is sufficiently accurate if the crack front is defined by
the cubic spline curve.
In the paper by Hou et al. [31], the finite element method and crack growth
laws in fracture mechanics were combined. The main approach was that the
13
stress intensity factors for general three-dimensional cracks were calculated
by means of the finite element method and the crack growth behavior was
observed by using the crack growth principles in 3-D cases. The computer
code ZENCRACK, which has a direct interface with FE code ABAQUS,
automatically generates the 3D 20-noded crack elements and replaces them
by a group of crack elements to form a desired crack front. The crack front
can be either semi-circular/semi- elliptical or linear within a crack block.
Based upon a cracked FE model the stress intensity factor can be
determined at each node along the crack front using the J-integral method in
ABAQUS.
Courtin et al. [16] aimed to present the test results of several existing
numerical techniques reported in the literature. Both the crack opening
displacement extrapolation method and the J-integral approach are applied in
2D and 3D ABAQUS finite element models. The results obtained by these
various means on CT specimens and cracked round bars are in good
agreement with those found in the literature. From the results obtained it is
indicated that the J-integral method shows some advantages compared to
the displacement extrapolation one. First of all, this method may be applied
automatically with the ABAQUS code. Then, the knowledge of the exact
displacement field in the vicinity of the crack tip is not required, and the use
of singular finite elements is not essential anymore. Besides, non-orthogonal
meshes are without effect on the SIF calculations. The user has just to be
sure that a convergent value is obtained on the different rings. As a
consequence, this approach seems to be particularly suitable to deal with the
fatigue growth of general cracks.
Ranestad et al. [12] described a method for obtaining accurate descriptions

of crack-tip stress-fields in surface cracked welded plates without the need of
large 3D FEA models. The method simply uses the existing shell models in
combination with a plane strain sub-model. The boundary conditions of the
sub-model were taken from the global shell model. When the results were
14
compared to a 3D solid model, the shell models give good predictions of the
J values for homogeneous materials and for weldments with fusion line
crack. The need for the method is because of the expensiveness both in
terms of modeling time and computation time.
Recent studies show that sub-modeling technique is one of the most

powerful techniques when it is necessary to obtain an accurate and detailed
solution in a local region of a large model. In real life conditions, updating the
model for each changing fracture parameters is very time consuming and
unnecessary. That’s why Gerstle et al. [32] [33] performed studies on the
determination of fracture solutions by using sub-modeling technique. They
prepared their own programs for analyzing the fracture problems and get
accurate results by using these programs.
There are various real life applications of sub-modeling technique in

literature. One is performed by Giglio [34] dealing with the analysis of fatigue
damage to upper and lower folding beams on the rear fuselage of a naval
helicopter which may result from flight and folding loads. The study is a FEM-
based analytical approach together with experimental tests. The finite
element model of the helicopter part is created by means of an
ABAQUS/Standard finite element program combined with advanced sub-
modeling techniques. The finite elements calculations were confirmed by
experimental test results. Thus, sub-modeling technique is found to be
reliable for engineering fatigue analysis.
Kitamura et al. [35] applied the sub-modeling technique to ship structure

analysis in which a Bulk Carrier is selected. In their study they used two
types of boundary conditions, the displacement boundary condition and the
stress boundary condition. Applying the displacement boundary condition to
sub-modeling analysis is often used in practice and is a key study in the
literature for the following reasons. First, it is easier to be implemented, and
second, for a raw FEM solution, displacements are generally more accurate
15
than stresses. At the end of their study, they obtained accurate solutions by
using sub-modeling technique.
Another industrial application of sub-modeling technique is carried out by

Larson [36]. In the study a gearbox (2.3 Mdofs) and a control housing (2.9
Mdofs) are investigated. A GUI prepared for adaptive sub-modeling in an
interactive fashion in real time. The study also couples the sub-modeling with
local shape optimization.
As far as the studies mentioned above are considered the sub-modeling

technique is a powerful tool in 3D analysis. In recent studies it is commonly
used and give accurate results. Common commercial finite element analysis
software such as MSC.MENTAT– MARC [37], ABAQUS [38], ANSYS [39]
has sub-modeling technique in them. Therefore, it is easy and inexpensive to
apply this method to large models, where a detailed solution in local region is
desired to obtain.
16
CHAPTER 3
FRACTURE ANALYSIS IN THREE DIMENSIONS
3.1 Introduction
Standard design methods for engineering structures and components under

static loading are usually based on avoiding failure by yielding/plastic
collapse or buckling. The derivation of loading resistance is based on
conventional solid mechanics theories of stress analysis. Conventional
design procedures against fatigue failure are based on experimental results
for particular geometric details and materials. None of these procedures are
capable of allowing for the effects of severe stress concentrations or crack-
like flaws. The presence of such flaws is more or less inevitable to some
extent in practical fabrications.
The modes of failure which are most affected by the presence of crack-like
flaws are fracture and fatigue. The study of the effects of cracks on local
stress and strain fields in the neighborhood of the crack tip and the
consequent effect on failure is the subject of fracture mechanics. The
application of fracture mechanics methods allows analyses to be carried out
to predict the effects of flaws on failure in a wide range of geometries to give
complementary information to that obtained from experimental testing. For
fatigue of welded structures the performance is significantly affected by the
tiny flaws inherent to welding. Fracture mechanics analyses can be very
17
helpful in predicting the effects of geometrical variations on basic fatigue
behavior [40].
Use of crack propagation laws based on stress intensity factor ranges is the
most successful engineering application of fracture mechanics. This chapter
gives a review of the basic concepts of fracture mechanics. In contrast to the
traditional stress-life and strain-life approaches to fatigue, cracks are
assumed to exist in materials and structures within the context of fracture
mechanics. Fracture parameters such as K and J can be used to
characterize the stresses and strains near the crack tips. A fundamental
understanding of fracture mechanics and the limit of using the fracture
parameters is needed for appropriate applications of fracture mechanics to
model fatigue crack propagation.
In this chapter, the concept of Stress Intensity Factor (SIF) is first introduced.
Then, different fracture modes and fracture toughness are discussed. Later,
the expressions of asymptotic crack-tip fields are derived. Afterwards, linear
elastic stress intensity factors and energy release rate are introduced. Finally,
evaluation of J-integral is presented.
3.2 Stress Intensity Factor (SIF)
Irwin (1957) introduced the concept of stress intensity factor, K, as the

parameter, which is providing complete description of the state of stress,
strain and displacement near the tip of a crack caused by a remote load or
residual stresses. It is basically used for representing the strength of the
singularity under different loading conditions. The SIF is proportional to the
applied stress. This relationship is a direct consequence of the linear nature
of the theory of elasticity. Unlike the stress concentration factor, the stress
intensity factor is size dependent, because it contains the crack length as
parameter.
18
If a segment of crack front is considered, it is subjected to three primary
loading modes and their combinations at different loading conditions. A
Cartesian coordinate system is assigned such that the crack front is in the z
direction and idealized planar crack problems, in which the stresses and
strains near the crack tip can be expressed in terms of the in-plane
coordinates x and y only, are considered. As shown in Figure 3.1, the crack
is subject to Mode I, the opening or tensile mode, where the in-plane
stresses and strains are symmetric with respect to the x axis. As shown in
Figure 3.2, the crack is subject to Mode II, the sliding or in-plane shearing
mode, where the stresses and strains are anti-symmetrical with respect to
the x axis. As shown in Figure 3.3, the crack is subject to Mode III, the
tearing or anti-plane shearing mode, where the out-of-plane stresses and
strains are anti-symmetrical with respect to the x axis [41].
Figure 3.1 Mode I (Opening Mode)
19
Figure 3.2 Mode II (In-plane Shearing Mode)
Figure 3.3 Mode III (Out-of-plane Shearing Mode)
Mode I is the most common loading type encountered in engineering design

and in literature there exist many studies related to Mode I loading.
Combinations of these modes are also possible and it is called mixed mode
loading type. The encountered cracks, where high stresses or material
imperfections existing, in industrial applications mostly have mixed mode
loading. Since it is very important to assess the remaining life accurately,
mixed mode loading estimations must be done precisely. This, of course,
requires an accurate and physics based three-dimensional fracture solution
to include all geometrical details and loading conditions in the problem. Even
for surface and corner crack problems involving simple geometry, loading
and boundary conditions, it is still important to apply three-dimensional
20
methods since the corresponding two-dimensional analyses, i.e., plane strain
or plane stress, yield conservative results.
Moreover, in some problems there are truly three dimensional geometrical

features to be considered. In such circumstances, three-dimensional fracture
analysis is also unavoidable to account for the correct geometry and
therefore the correct loading conditions near the crack region.
3.3 Fracture Toughness
As the stress intensity factor reaches a critical value, unstable fracture

occurs.
This critical value of the stress intensity factor is known as the fracture
toughness of the material. The fracture toughness can be considered as the
limiting value of the stress intensity just as the yield stress might be
considered as the limiting value of the applied stress. The fracture toughness
depends on both temperature and the specimen thickness. Mode I plane
strain fracture toughness is denoted as KIC. KC, which is the plane stress
fracture toughness, is used to measure a material's fracture toughness in a
sample that has a thickness that is less than some critical value, B. When the
material's thickness is less than B, and stress is applied, the material is in a
state called plane stress. A material's thickness is related to its fracture
toughness graphically in Figure 3.4. If a stress is applied to a sample with a
thickness greater than B, it is in a state called plane strain [18].
21
Figure 3.4 Fracture Toughness vs. Material Thickness Graph
3.4 Two and Three Dimensional Linear Elastic Crack Tip Fields
The stresses and strains at any point near a crack tip can be derived from the
theory of elasticity. The asymptotic crack-tip stresses and strains for different
modes of loading are known to satisfy a set of fundamental differential
equations resulting from equilibrium, compatibility conditions and physical
properties of the material which constitutes the solid body.
Figure 3.5 and 3.6 show the stress field and polar coordinate system for a
two dimensional crack. Two and three dimensional linear elastic crack tip
fields (stress and displacement relations) and the stress intensity factor
definitions are expressed below for each loading mode.
22
Figure 3.5 Distribution of Stresses in Vicinity of Crack Tip
Figure 3.6 A Two-dimensional Crack Configuration
The distribution of stresses and displacements in small region around a crack

tip given by asymptotic expressions are always the same for any cracked
body. Although the asymptotic expressions are universal, the stress intensity
factor depends on the geometry and the loading conditions. In other words,
23
the stress intensity factor is a function of the size and position of the crack in
the geometry and the applied stress.
For each loading mode (mode I, II and III), two dimensional linear elastic
crack tip fields and definitions of stress intensity factors are cited below. [42]
Mode I Crack:
KI ⎛ θ ⎞⎧⎪ ⎛ θ ⎞ ⎛ 3θ ⎞⎫
σ yy (r ,θ ) = cos⎜⎜ ⎟⎨1 + sin ⎜⎜ ⎟ sin ⎜⎜ ⎟⎬ (3.1)
2πr ⎝ 2 ⎠⎪⎩ ⎝ 2 ⎠ ⎝ 2 ⎠⎭
KI ⎛ θ ⎞⎧⎪ ⎛ θ ⎞ ⎛ 3θ ⎞⎫
σ xx (r ,θ ) = cos⎜⎜ ⎟⎨1 − sin ⎜⎜ ⎟ sin ⎜⎜ ⎟⎬ (3.2)
2πr ⎝ 2 ⎠⎪⎩ ⎝ 2 ⎠ ⎝ 2 ⎠⎭
KI ⎛ θ ⎞ ⎛ θ ⎞ ⎛ 3θ ⎞
σ xy (r ,θ ) = cos⎜⎜ ⎟ sin ⎜⎜ ⎟ sin ⎜⎜ ⎟ (3.3)
2πr ⎝2⎠ ⎝2⎠ ⎝ 2 ⎠
KI r ⎛ θ ⎞⎪⎧ ⎛ θ ⎞⎫
u (r ,θ ) = cos⎜⎜ ⎟⎨κ − 1 + 2 sin 2 ⎜⎜ ⎟⎬ (3.4)
2μ 2π ⎝ 2 ⎠⎪⎩ ⎝ 2 ⎠⎭
KI r ⎛ θ ⎞⎧⎪ ⎛ θ ⎞⎫
v(r ,θ ) = sin ⎜⎜ ⎟⎨κ + 1 − 2 cos 2 ⎜⎜ ⎟⎬ (3.5)
2μ 2π ⎝ 2 ⎠⎪⎩ ⎝ 2 ⎠⎭
where KI is the mode I stress intensity factor and σxx, σyy and σxy are the
stress components (Figure 3.5) at a distance r from the crack tip and at an
angle θ from the crack plane. In Equations 3.4 and 3.5, u and v are the
displacements in x and y directions. µ is the shear modulus and κ is
⎛ 3 −ν ⎞
⎜ ⎟ for plain stress and (3 − 4ν ) for plain strain where ν is the Poisson’s
⎝ 1 +ν ⎠
ratio. The relationship between the shear modulus (µ), Young’s modulus (E)
and Poisson’s ratio (ν) is as follows:
E
μ = (3.6)
2(1 + ν )
24
Definition of the mode I stress intensity factors can be written as:
K I (a ) = lim+ 2π ( x − a )σ yy ( x,0) (3.7)

x→a
K I (− a ) = lim− 2π (− x − a )σ yy (x,0) (3.8)

x→− a
where a is the half of the crack length.
Mode II Crack:
K II ⎛ θ ⎞ ⎛ θ ⎞ ⎛ 3θ ⎞
σ yy (r ,θ ) = sin ⎜ ⎟ cos⎜ ⎟ cos⎜ ⎟ (3.9)
2πr ⎝2⎠ ⎝2⎠ ⎝ 2 ⎠
K II ⎛ θ ⎞⎧ ⎛ θ ⎞ ⎛ 3θ ⎞⎫
σ xx (r ,θ ) = − sin ⎜ ⎟⎨2 + cos⎜ ⎟ cos⎜ ⎟⎬ (3.10)
2πr ⎝ 2 ⎠⎩ ⎝ 2 ⎠ ⎝ 2 ⎠⎭
K II ⎛ θ ⎞⎧ ⎛ θ ⎞ ⎛ 3θ ⎞⎫
σ xy (r ,θ ) = cos⎜ ⎟⎨1 − sin ⎜ ⎟ sin ⎜ ⎟⎬ (3.11)
2πr ⎝ 2 ⎠⎩ ⎝2⎠ ⎝ 2 ⎠⎭
K II r ⎛ θ ⎞⎧ ⎛ θ ⎞⎫
u (r , θ ) = sin ⎜ ⎟⎨κ + 1 + 2 cos 2 ⎜ ⎟⎬ (3.12)
2μ 2π ⎝ 2 ⎠⎩ ⎝ 2 ⎠⎭
K II r ⎛ θ ⎞⎧ ⎛ θ ⎞⎫
v(r , θ ) = − cos⎜ ⎟⎨κ − 1 − 2 sin 2 ⎜ ⎟⎬ (3.13)
2μ 2π ⎝ 2 ⎠⎩ ⎝ 2 ⎠⎭
where KII is the mode II stress intensity factor, which can be defined as:
K II (a ) = lim+ 2π ( x − a )σ xy ( x,0) (3.14)

x→a
K II (− a ) = lim− 2π (− x − a )σ xy ( x,0) (3.15)

x→− a
25
Mode III Crack:
K III ⎛θ ⎞
σ xz (r ,θ ) = − sin ⎜ ⎟ (3.16)
2πr ⎝2⎠
K III ⎛θ ⎞
σ yz (r , θ ) = cos⎜ ⎟ (3.17)
2πr ⎝2⎠
σ xx (r ,θ ) = σ yy (r ,θ ) = σ zz (r ,θ ) = 0 (3.18)
K III r ⎛θ ⎞
w(r ,θ ) = sin ⎜ ⎟ (3.19)
μ 2π ⎝2⎠
u (r , θ ) = v(r , θ ) = 0 (3.20)
where w is the displacement in z direction and KIII is the mode III stress
intensity factor, which can be defined as:
K III (a ) = lim+ 2π ( x − a )σ yz ( x,0 ) (3.21)

x→a
K III (− a ) = lim− 2π (− x − a )σ yz ( x,0) (3.22)

x→− a
In figure 3.7, a three dimensional crack front and a local coordinate system is
shown. The parameter s in this figure is the arc length of the crack front and
t, n, b is a local coordinate system located at point P composed of the
tangential (t), normal (n) and binormal (b) directions, n pointing into the
material side. (r, θ) are the polar coordinates in the normal plane (n, b) [43].
Three dimensional linear elastic crack tip fields are given below:
K I (s ) ⎛ θ ⎞⎡ ⎛ θ ⎞ ⎛ 3θ ⎞⎤
σ nn = cos⎜ ⎟ ⎢1 − sin⎜ ⎟ sin⎜ ⎟⎥
2πr ⎝ 2 ⎠⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
K II (s ) ⎛ θ ⎞⎡ ⎛ θ ⎞ ⎛ 3θ ⎞⎤
− sin ⎜ ⎟ ⎢2 + cos⎜ ⎟ cos⎜ ⎟⎥ (3.23)
2πr ⎝ 2 ⎠⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
26
Figure 3.7 Three Dimensional Crack Front and the Local Coordinate System
K I (s ) ⎛ θ ⎞⎡ ⎛ θ ⎞ ⎛ 3θ ⎞⎤
σ bb = cos⎜ ⎟ ⎢1 + sin ⎜ ⎟ sin ⎜ ⎟⎥
2πr ⎝ 2 ⎠⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
K II (s ) ⎛ θ ⎞ ⎛ θ ⎞ ⎛ 3θ ⎞
+ sin ⎜ ⎟ cos⎜ ⎟ cos⎜ ⎟ (3.24)
2πr ⎝2⎠ ⎝2⎠ ⎝ 2 ⎠
⎡ K I (s )
⎛ θ ⎞ K (s ) ⎛ θ ⎞⎤
σ tt = 2ν ⎢ cos⎜ ⎟ − II sin ⎜ ⎟⎥ (3.25)
⎣ 2πr ⎝2⎠ 2πr ⎝ 2 ⎠⎦
K I (s ) ⎛ θ ⎞ ⎛ θ ⎞ ⎛ 3θ ⎞
σ nb = sin⎜ ⎟ cos⎜ ⎟ cos⎜ ⎟
2πr ⎝2⎠ ⎝2⎠ ⎝ 2 ⎠
K II (s ) ⎛ θ ⎞⎡ ⎛ θ ⎞ ⎛ 3θ ⎞⎤
+ cos⎜ ⎟ ⎢1 − sin ⎜ ⎟ sin ⎜ ⎟⎥ (3.26)
2πr ⎝ 2 ⎠⎣ ⎝ 2 ⎠ ⎝ 2 ⎠⎦
K III (s )⎛θ ⎞
σ nt = − sin⎜ ⎟ (3.27)
2πr ⎝2⎠
27
K III (s )⎛θ ⎞
σ bt = cos⎜ ⎟ (3.28)
2πr ⎝2⎠
1 +ν 2r ⎧ ⎛ θ ⎞⎡ 2 ⎛ θ ⎞⎤
un = ⎨ K I (s ) cos⎜ ⎟ ⎢(1 − 2ν ) + sin ⎜ ⎟⎥
E E⎩ ⎝ 2 ⎠⎣ ⎝ 2 ⎠⎦
⎛ θ ⎞⎡ ⎛ θ ⎞⎤ ⎫
+ K II (s )sin ⎜ ⎟ ⎢2(1 − ν ) + cos 2 ⎜ ⎟⎥ ⎬ (3.29)
⎝ 2 ⎠⎣ ⎝ 2 ⎠⎦ ⎭
1 +ν 2r ⎧ ⎛ θ ⎞⎡ 2 ⎛ θ ⎞⎤
ub = ⎨ K I (s )sin ⎜ ⎟ ⎢2(1 − ν ) − cos ⎜ ⎟⎥
E π ⎩ ⎝ 2 ⎠⎣ ⎝ 2 ⎠⎦
⎛ θ ⎞⎡ ⎛ θ ⎞⎤
− K II (s ) cos⎜ ⎟ ⎢(1 − 2ν ) − sin 2 ⎜ ⎟⎥ (3.30)
⎝ 2 ⎠⎣ ⎝ 2 ⎠⎦
1 +ν 2r ⎛θ ⎞
ut = 2 K III (s )sin⎜ ⎟ (3.31)
E π ⎝2⎠
where KI, KII and KIII are the stress intensity factors and these are defined as:
K I = lim 2πrσ bb (r ,0) (3.32)

r →0
K II = lim 2πrσ nb (r ,0 ) (3.33)

r →0
K III = lim 2πrσ bt (r ,0) (3.34)

r →0
3.5 Energy Release Rate
Linear fracture mechanics presupposes existence of a crack and examines

the conditions under which crack growth occurs. For a crack to propagate,
the rate of elastic energy release should at least be equal the rate of energy
needed for creation of a new crack surface. The energy balance between the
28
strain energy in the structure and the work needed to create a new crack
surface can be expressed using the energy release rate (G) as follows:
G = GC (3.35)
G is defined as:
dΠ
G=− (3.36)
da
where Π is the strain energy and a is the crack area. G depends on the
geometry of the structure and the current loading. Gc is called the fracture
toughness of the material. It is a material property and determined by
experiments. Note that the energy release rate is not a time derivative but a
rate of change in potential energy with crack area. An important feature of
Equation 3.35 is that it can be used as a fracture criterion; a crack starts to
grow when G reaches the critical value Gc.
The connection between the energy release rate and the stress intensity
factors is given by [21]:
1 2 2 1 2
G = G I + G II + G III = ( K I + K II ) + K III (3.37)
E *
2μ
where
E * = E (For plane stress) (3.38)
and
E
E* = (For plane strain) (3.39)
(1 − ν 2 )
29
If only mode I loading type exists, where KII and KIII are equal to zero,
Equation 3.37 becomes:
K2
G= (For plane stress) (3.40)
E
⎛ 1 −ν 2 ⎞ 2
G = ⎜⎜ ⎟⎟ ⋅ K (For plane strain) (3.41)
⎝ E ⎠
3.6 J-integral
The J-integral is similar to G but is more general and is also used for
nonlinear applications. J is equivalent to G when a linear elastic material
model is used.
The J-integral probably offers the best chance to have a single parameter to
relate to the initiation of crack propagation. The J-integral was introduced by
Rice as a path-independent contour integral for the analysis of cracks. As
previously mentioned, it is equivalent to the energy release rate for a linear
elastic material model. It is defined in two dimensions as:
∂u j
J = ∫ ((W + T )n1 − σ ij ni ) dΓ (3.42)
Γ
∂x1
where W is the strain energy density, T is the kinetic energy density, σij is the
stress tensor and uj is the displacement vector. The x1 direction is the same
as the x direction in the local crack tip system in Figure 3.8. The integration
path Γ is a curve surrounding the crack tip, see Figure 3.8. The J-integral is
independent of the path Γ as long as it starts and ends at the two sides of the
crack face and no other singularities are present within the path. This is an
30
important feature for the numerical evaluation since the integral can be
evaluated using results away from the crack tip.
Figure 3.8 Definition of the J-integral
3.7 Numerical Evaluation of the J-integral
The J-integral evaluation in MSC.Marc is based upon the domain integration

method. Since, it is difficult to define the integration path Γ, some
simplifications are done in finite element analysis. In the domain integration
method for two dimensions, the line integral is converted into an area
integration over the area inside the path Γ. This conversion is exact for the
linear elastic case and also for the nonlinear case if the loading is
proportional, that is, if no unloading occurs. By choosing this area as a set of
elements, the integration is straightforward using the finite element solution.
In two dimensions, the converted expression is
31
∂u j δq1
−
J = ∫ (σ ij − Wδ 1i ) dA (3.43)
A
∂x1 δxi
for the simplified case of no thermal strains, body forces or pressure on the
crack faces. A is the area inside Γ and q1 is a function introduced in the
conversion into an area integral. The function q1 can be chosen fairly
general, as long it is equal to one at the crack tip and zero on Γ [37].
This thesis is concerned with semi-elliptical surface cracks arbitrarily inclined

to the free surface of fillet welded joint toes and root subjected to mixed
mode loading condition. Therefore, the estimated J-integral values for fillet
welded joints will allways be for mixed mode loading type. Since it gives
exact solutions for linear elastic case, the J-integral values will be calculated
by using commercial finite element analysis software MSC.Marc.
32
CHAPTER 4
FINITE ELEMENT MODELING
4.1 Introduction
Finite Element Analysis (FEA) is a computer simulation technique used in

engineering analysis. It uses a numerical technique called the finite element
method (FEM). The finite element method (FEM) is a computer technique for
solving partial differential equations. By using this technique the stresses,
deflections and reaction forces etc in objects can be estimated. The
technique involves dividing the object into relatively small elements whose
individual behavior is easily calculated. The behavior of all of the small
elements is then put together to estimate the stresses and deflections in the
entire object.
In finite element analysis, the design is discretized or subdivided into a series

of elements that are connected by nodes. Material properties and element
properties are specified to represent the physical properties of the model.
Boundary conditions and applied loads are then defined to represent the
operating environment for which the design is to be subjected. Finite element
analysis is a simulation tool that enables engineers to simulate the behavior
of a structure [44].
33
In general, there are three phases in finite element analysis:
• Pre-processing (defining the finite element model and

environmental factors to be applied to it)
• Analysis solver (solution of finite element model)
• Post-processing of results (using visualization tools)
Finite element analysis is an important part of the overall design process,

serving to verify or validate a design prior to its manufacture. However it is
important that the results from finite element analysis are always examined
carefully. The credibility of the results should always be checked by carrying
out experiments whenever possible.
4.2 Problem Definition
There are a large number of commercial computer programs available today

that are specially designed for creating and analyzing finite element models.
Many of these computer programs are highly sophisticated in their ability to
generate a model, solve it and finally analyze the results.
Modeling is the most time consuming and human controlled phase of finite
element analysis. Modeling can be defined as the simulation of a physical
structure or physical process by means of a substitute analytical or numerical
construct. It is not simply preparing a mesh of nodes and elements. Modeling
require that the physical action of the problem be understood well enough to
choose suitable kinds of elements, and enough of them, to represent the
physical action adequately. It requires to avoid badly shaped elements and
too large to represent important variations of the field quantity (displacement
field or stress field). At the other extreme it is desired to avoid waste of
analyst time and computer resources associated with over-refinement, that
34
is, using many more elements than needed to adequately represent the field
and its gradient [45].
The optimization of earthmoving machinery attachments requires lots of

analyses to be performed. However, it takes too much time to model the
critical parts (such as boom and arm etc) of the machines. Therefore, few
pre-processors are prepared for modeling the earthmoving machinery
attachments automatically. One of these pre-processors is described in the
thesis study performed by Yener [47]. In the thesis study of Yener, finite
element analyses of excavator boom are carried out automatically. A
prepared GUI lets the user set the parameters of the boom geometry and
performs the analyses by using MSC.Marc behind it. The program defines
the geometric and material properties, boundary conditions and other related
parameters automatically. However, while modeling the boom geometry
quadrilateral thick shell elements are used. This is due to the reason that the
main aim of the study is to optimize the general stress distribution of the
boom and lots of analyses must be performed for the optimization, which
means that the solution time must be as low as possible. Thus, the prepared
programs for earthmoving machinery components do not involve welded
joints on them. However, according to the performed tests the failures
occurring on the attachments are mainly due to the defects on welded joints
and they must be well examined to obtain the optimum attachment geometry.
In this thesis, a prepared GUI will be used to convert the user-selected local
shell region to solid region with the desired weld shape and to implement a
semi elliptical surface crack into the weld toe and root at different inclination
angles. The program will automatically take the geometric and material
properties from the global model and it will use the sub-modeling technique
for boundary condition application. Although the prepared program is a
general use program for box cross section corner joints, only the boom and
arm of the HMK 300 LC excavator will be studied in the scope of this study.
35
4.3 Sub-modeling
If the local stress field is of interest this requires a detailed finite element
model in order to give sufficiently accurate results. In a large or geometrically
complex structure it is often not possible to model whole structure with such
accuracy, since this would involve a great deal of modeling work and result in
a model that was difficult to work with. In such cases it is often a good
solution to analyze the local stresses using a sub-model.
The sub-modeling approach involves creating a relatively coarse model to

calculate the global stresses in the structure. In this model the weld can
either be totally ignored or just be modeled roughly. Once the stresses have
been calculated in the global model, a local model (sub-model) is constructed
to model the area in question separately and give the degree of accuracy that
is needed to resolve the stresses correctly. An example of this is shown in
Figure 4.1. The deformations that are obtained from calculations using the
global finite element model are then used as boundary conditions for the
local model. This allows the local stresses at the weld to be calculated with a
good degree of accuracy without taking up excessive computer time.
A typical structural zooming analysis contains two steps:
1. Global run to obtain a post file containing global results.

2. Local run to define kinematic boundary conditions in the local
model and to obtain refined results in the local model.
This procedure can be repeated as many times as desired. Any local

analysis can be the global analysis of next level refinement.
36
Figure 4.1 a) Global Model b) Local Model [48]
Aside from the obvious benefits of giving more accurate results in a region of
the model, the sub-modeling technique has other advantages:
• It reduces, or even eliminates, the need for complicated transition

regions in finite element models.
• It enables the user to experiment with different designs for the
region of interest without requiring a complete re-analysis of the
entire model
• It provides adequate mesh refinement
37
The sub-modeling technique used in this study is a little bit different from
standard sub-modeling techniques used in commercial finite elements
software.
The steps included in this study are as follows:
1. Automatic model formation of boom and arm geometry by using

prepared GUI in Hidromek Ltd. Sti. and analysis of global model
having quadrilateral thick shell elements as element type.
2. Selection of the interested local region.
3. Setting the weld and crack implementation parameters in GUI
prepared for this study.
4. Conversion of local quadrilateral shell meshed model to 20-noded
hex meshed solid model.
5. Application of set weld parameters to solid model.
6. Implementation of semi elliptical surface cracks either to one of the
weld toe or to weld root.
7. Transfer of material properties from global model.
8. Definition of kinematic boundary condition application points.
9. Application of specially prepared tyings to shell-to-solid transition
region.
10. Analysis of the sub-model.
The steps mentioned above will be explained in detail below.
4.3.1 Global Model
The first step of sub-modeling requires a complete analysis of global model.

In this work, the boom and arm of HMK 300 LC excavator will be used as the
global model. The global models will be prepared by using GUI prepared in
Hidromek Ltd. Sti. It is assumed that material behavior is linear and strains
are small. Therefore, linear elastic analysis will be carried out. Sheet metal
38
parts of the boom and arm are modeled by using quadrilateral thick shell
elements. In the analysis only the boom and arm are realistically modeled.
Hydraulic cylinders, arm, bucket and connecting linkage are modeled using
beam elements (Figure 4.2).
The most critical boundary condition of an excavator is assumed to be when

the machine is subjected to maximum digging force at maximum digging
force position (Figure 4.6). The maximum digging force calculation of an
excavator is defined in standard SAE J1179 [49].
Maximum digging force of an excavator can be calculated in two ways. First

when the bucket cylinder is active and the second when the arm cylinder is
active, see Figure 4.3.
Figure 4.2 FEA Model of Excavator Mechanism
39
In this study, arm breakout force is taken to be the maximum digging force as
the global model boundary condition. In order to see the effect of lateral loads
on cracks, the global analyses are done by including the lateral loads and not
including them. Figure 4.4 shows an example of global model analysis result
including the lateral loads in boundary conditions.
Figure 4.3 Basic Parts of Excavator
40
Figure 4.4 Von Mises Stress Map for Shell Outer Layer
4.3.1.1 Boundary Conditions
4.3.1.1.1 Bucket Breakout Force
The bucket breakout force is the available force at the tip of the teeth created
by the bucket cylinder. Maximum breakout force is reached when the
available tooth force reaches its maximum.
41
Figure 4.5 Bucket Breakout Force Position
2
Dd
Fd = P. (4.1)
4
Fd .c.e
Fb = (4.2)
d. f
Fd – Bucket cylinder force

P – Working pressure of bucket cylinder
Dd – Bucket cylinder diameter
Fb – Bucket breakout force
c – Perpendicular distance bucket cylinder axis - lever pivot
d – Perpendicular distance connecting link axis - lever pivot
e – Perpendicular distance connecting link axis - bucket pivot
f – Radius bucket pivot - tooth lip
42
4.3.1.1.2 Arm Breakout Force
The digging force is the available force at the tip of the bucket teeth created
by the arm cylinder(s). Maximum digging force is calculated with dimension
“g” at its maximum and with the bucket in a position calculated for maximum
bucket breakout force.
Figure 4.6 Arm Breakout Force (Digging Force) Position
2
Dc
Fc = P. (4.3)
4
Fc .g
Fa = (4.4)
h
43
Fc – Arm cylinder force
P – Working pressure of arm cylinder
Dc – Arm cylinder diameter
Fa – Arm breakout force
g – Perpendicular distance arm cylinder axis - arm pivot
h – Distance arm pivot - tooth tip
4.3.1.1.3 Lateral Force
Upper part of an excavator is capable of full rotation about the center of the
chassis by means of a hydraulic system. This system is capable of applying a
moment ‘M’ limited by the capacity of hydraulic oil motor-swing unit. This
moment causes a lateral force ‘F’ at the tip of the bucket (Figure 4.7).
Therefore, in finite element analysis, this lateral force must be included.
However, to see the effect of lateral force on cracks, analysis will be divided
into two groups: the analysis including lateral force and not including lateral
force.
4.3.1.1.4 Other Boundary Conditions
In finite element analysis, pivot joints of boom and boom cylinder on the
chassis will be fixed in all directions except joint rotation direction (Figure
4.2).
To learn more about the global model analysis the thesis written by Yener
[47] can be examined.
44
Figure 4.7 Lateral Force on the Bucket
4.3.2 Local Region Selection
Local region selection is of great importance for examining the cracks in

critical regions. There are some limitations while selecting the local region on
global model. First, this study includes only the corner joints of box cross
section, therefore the selected region must be on the corner. Second, the
selected region junction nodes should follow a linear path on the global
model. Third, thickness transition regions should not be selected as local
model regions. Although the prepared computer code is for general use, this
study will only be interested in excavator boom and arm. Figure 4.8 shows
45
the selection possibilities (thick black lines) according to the specified
limitations mentioned above.
Figure 4.8 Selection Possibilities on Excavator Boom and Arm
4.3.3 Parameters
The parameters of local models are divided into two groups: weld parameters
and crack parameters. In the following, the parameters are explained in
detail.
4.3.3.1 Weld Parameters
In the analyses, the weld shapes are always assumed to be in triangular,

either isosceles or scalene, shape. In GUI, the set weld parameters differ
with respect to the weld application side selection of the user. If the user
46
wants to apply one sided weld to created solid models the parameters that
defines the weld geometry are (Figure 4.9):
Leg1 – Vertical leg length of the weld

Leg2 – Horizontal leg length of the weld
RD – Root Depth (determines the penetration depth of the weld)
Figure 4.9 One Sided Weld Parameters
If the user wants to apply the weld to both sides of the solid model, the
parameters are as follows (Figure 4.10, Figure 4.11):
Leg1 – Vertical leg length of the outer weld

Leg2 – Horizontal leg length of the outer weld
Leg_1 – Vertical leg length of the inner weld
Leg_2 – Horizontal leg length of the inner weld
47
MCD – Middle Crack Distance (determines the penetration depth of the weld)
MCL – Middle Crack Length (determines the penetration depth of the weld)
Figure 4.10 Two Sided Weld Geometry Parameters
Figure 4.11 Two Sided Weld Penetration Depth Parameters
48
4.3.3.2 Crack Parameters
In real life applications the most dominant fatigue cracks are semi elliptical
surface cracks. A semi elliptic surface crack can be defined by two
parameters (Figure 4.12):
a – Half surface crack length

b – Crack depth
If half surface crack length (a) is equal to the crack depth (b) the cracks
shape becomes an exact circle. In present work semi elliptic surface cracks
are placed into toes and root of local solid models. The analyses are done for
various inclination angles ranging from 00 to 900 at weld toe and 00 to 1800 at
weld root (Figure 4.13).
Figure 4.12 Semi Elliptic Crack Parameters
Figure 4.13 Crack Inclination Angle Ranges a) Crack at Weld Toe 1 b) Crack
at Weld Toe 2 c) Crack at Weld Root
49
4.3.4 Shell-to-Solid Conversion Method
As mentioned before, sheet metal parts of the global model are modeled by
using quadrilateral thick shell elements. Shell-to-Solid conversion method
simply forms a crack block and positions it to the specified location. Then, the
weld shape is formed and shell model is converted to solid model (Figure
4.14). The local solid model consists of 20-node isoparametric three
dimensional brick elements (Figure 4.15). In fracture analysis, in order to
simulate the square-root strain singularity around the crack front, collapsed
20-node isoparametric three dimensional brick elements are utilized. These
elements are depicted in Figure 4.16 [43]. The representation of collapsed
20-node isoparametric elements at weld to crack front start can be seen in
Figure 4.17.
Figure 4.14 Shell-to-Solid Conversion Possibilities a) Shell Model b) Solid

Model Welded from Inner Side c) Solid Model Welded from Outer Side d)
Solid Model Welded from Both Sides
50
Figure 4.15 Semi Elliptic Surface Crack at Weld Toe
Figure 4.16 a) 20-node Isoparametric Brick Element b) Collapsed 20-node

Isoparametric Brick Elements
51
Figure 4.17 Collapsed 20-node Isoparametric Brick Elements at Weld Toe
Crack Front Start
4.3.5 Material Properties
The computer code is prepared such that it takes all material properties from
the global model (Table 4.1, Table 4.2). In addition, the material properties of
the welded connections are assumed to be the same with the global model.
52
Table 4.1 Material Properties of the Steel Used in Global Model Analyses
Yield Tensile
Material Elongation Poisson's Density
Strength Strength E (MPa)
Type (%) Ratio (kg/m3)
(MPa) (MPa)
St-52 355 510 21 210000 0.3 7850
Table 4.2 Chemical Compositions of the Steel Used in Global Model

Analyses
Material
C mass (%) Si mass (%) Mn mass (%) P mass (%) S mass (%)
Type
St-52 0.22 0.55 1.60 0.035 0.035
4.3.6 Tyings and Local Model Boundary Conditions
The sheet metal parts of the global model are composed of quadrilateral thick
shell elements. These elements’ nodes have six degrees of freedom, which
are global displacements and rotations in three dimensions.
Degrees of freedom per node:
1 = u = global (Cartesian) x-displacement

2 = v = global (Cartesian) y-displacement
3 = w = global (Cartesian) z-displacement
4= x = rotation about global x-axis
5= y = rotation about global y-axis
6= z = rotation about global z-axis
53
However, 20-node brick elements’ nodes have only global displacements (u,
v, w) as degrees of freedom. Therefore, if there exists a transition from shell
to solid (or vice versa), a proper tying type must be used not to loose the
deformation history of the global model nodes. In this work, a similar method
to the method used in the study of Ranestad et al. [12] is adopted. In this
method, the sub-model is loaded with the deformation history taken from the
global shell model. The deformation history of the nodes is applied to
constrained nodes (Mid-surface nodes) (Figure 4.18). The other (“follower”)
nodes on the edge are forced to stay in the plane defined by the nodes in
transverse direction (Figure 4.19). This method gives sufficiently precise
values far from the boundary condition application points. Therefore, special
attention must be paid not to implement cracks very close to the boundaries.
This property is also parameterized in prepared computer code and it can be
set according to the size of selected region of interest.
Figure 4.18 Deformation History Applied to Tying Constrained Nodes at Local

Model Boundaries
54
Figure 4.19 Constrained Nodes and the Other (“follower”) Nodes on the Edge
are Forced to Stay in the Plane Defined by the Nodes in Transverse Direction
4.3.7 J-integral Calculation
As mentioned in section 3.7, J-integral values are calculated in MSC.Marc.

The method used in the program is exact for linear cases. Therefore, it can
be used for linear analyses of excavator boom and arm. However, to
calculate the J-integral some conditions must be satisfied. First, a crack
shape must be modeled at the interested location. Second, crack front must
be modeled by using quarter-point crack tip elements (Figure 4.16, Figure
4.17). Third, crack front nodes must be defined for J-integral calculation.
Crack front nodes are the nodes on the red path shown in Figure 4.15. After
satisfying all of the conditions, the analysis of sub-modeled region must be
done. The J-integral values of crack front nodes are given in the output file of
the performed analysis.
55
4.4 Verification of Finite Element Modeling and J-integral Calculation
4.4.1 Verification of Finite Element Modeling
Verification of finite element modeling requires tests to be performed. The

performed tests must also have the same boundary conditions with finite
element analyses. Finite elements results at some locations must be
compared with the test results to verify the model. According to the
experimental stress analysis results of HMK 300 LC excavator, maximum
deviation from finite element analysis results is 15% (Figure 4.20). Thus, the
global model analysis gives sufficiently accurate results.
160
140
Von Mises Stress Values [MPa]
120
100
80
60
40
T
e
s
t
R
e
s
u
l
t
s
A
n
a
l
y
s
i
s
R
e
s
u
l
t
s
20
0
0 5 10 15 20
G
a
u
g
e
N
u
m
b
e
r
Figure 4.20 Experimental and Finite Element Analysis Results Comparison
56
4.4.2 Verification of J-integral Calculation
To verify the calculated J-integral values, the results in literature are

compared with the results obtained in this thesis.
Firstly, embedded circular and elliptical crack solutions of Sneddon [50] and
Irwin [51] are investigated. The results are compared and the models are
verified.
Secondly, 3D inclined semi elliptical surface cracks in a semi-infinite body
under tension are examined. In this context, the solutions of Noda et al. [52]
and Isida et al. [53] are compared with the results found by J-integral method.
4.4.2.1 Verification of Embedded Circular and Elliptical Crack Models
Sneddon [50] treated the problem of a circular crack of radius a (penny-

shaped crack) embedded in an infinite solid subjected to uniform tension
(Figure 4.21). He arrived at:
2
K= a (4.5)
Figure 4.21 Embedded Penny-shaped Crack
57
If plane strain assumption is assumed to exist, Mode-I stress intensity factor
can be calculated by using J-integral values as follows:
E
KI = 2
J (4.6)
1
In the analyses, it is seen that the size of the collapsed 20-node brick
elements and the size of the solid cube has a vital effect on the solutions.
Therefore, to obtain more accurate results the edge of the solid cube is
increased up to forty times the circular crack radius (a) and the size of the
collapsed brick elements surrounding the crack front is decreased up to one
over twenty-five times the circular crack radius. Since the model is
symmetric, in finite element analysis half of the model is used to calculate the
J-integral values (Figure 4.22, Figure 4.23). Calculated J-integral values are
all the same at the crack front nodes. To be able to compare the present
results with the bibliographical ones, dimensionless SIF and J-integral are
introduced
* KI
KI = (4.7)
a
where KI is the real SIF, the applied stress and a the crack radius or crack
depth and
J
J* = 2
(4.8)
( a / E)
where J is the real J-integral value, the applied stress, a the crack radius
or crack depth and E modulus of elasticity (Table 4.3).
58
Table 4.3 Details and Comparison for a Penny-shaped Crack Analysis
J-integral Present Sneddon’s

Poisson’s a Study Solution Difference
E (MPa) Value
Ratio ( ) (mm) (MPa) * * (%)
( J*) ( KI ) ( KI )
210000 0.3 50 100 0.3684 0.6362 0.6366 0.0596
Figure 4.22 General View of Embedded Circular Crack
59
Figure 4.23 Close-up View of Embedded Circular Crack
A solution for an embedded elliptical flaw not being available, Irwin [51]
derived a useful expression on the basis of the stress field around the crack
front of an elliptical crack in an infinite body in remote tension. The calculated
SIF values showed that they vary around the crack front. The results of
Irwin’s analysis is:
1
a a2 4
K= sin 2 + 2 cos 2 (4.9)
E( ) c
where a, c, are as shown in Figure 4.24, and E ( ) is the elliptical integral:
60
2 12
a2
E( )= 1 1 2 sin 2
d (4.10)
0 c
Figure 4.24 Parametric Representation of a Point on an Ellipse
In Irwin’s solution the maximum value of SIF is obtained on minor axis, and
the minimum value of the SIF is obtained on major axis of the elliptical crack.
In the analysis, embedded elliptical crack dimensions are 125 mm as major
axis and 50 mm as minor axis. The solution of eq. (4.10) is calculated by
using MATLAB for the specified major and minor axis values. The details of
the analysis and the normalized SIF comparison of Irwin’s solution and the
present study can be seen in Table 4.4 and Table 4.5.
To decrease the size effect of the solid cube and collapsed brick elements
the edge of the solid cube is increased up to sixty times the elliptical crack
minor axis (a) and the size of the collapsed brick elements surrounding the
61
crack front is decreased up to one over twenty-five times the elliptical crack
minor axis (Figure 4.25). Since the model is symmetric, in finite element
analysis half of the model is used to calculate the J-integral values.
Figure 4.25 Close-up View of Embedded Elliptic Crack
Table 4.4 Details for Elliptical Crack Analysis
Poisson’s
E (MPa) a (mm) c (mm) E( )
Ratio ( ) (MPa)
210000 0.3 50 125 100 1.1507
62
Table 4.5 Normalized SIF Comparison for an Elliptical Crack
Irwin's Difference
(degree) Present Study
Solution (%)
0.00 0.5523 0.5496 0.4886
5.64 0.5594 0.5565 0.5284
10.96 0.5773 0.5740 0.5725
16.27 0.6034 0.5991 0.6991
21.68 0.6340 0.6291 0.7633
27.14 0.6662 0.6610 0.7800
32.54 0.6982 0.6924 0.8332
37.89 0.7287 0.7222 0.8969
43.26 0.7568 0.7499 0.9110
48.15 0.7822 0.7730 1.1706
54.14 0.8045 0.7982 0.7829
59.98 0.8236 0.8192 0.5291
65.31 0.8393 0.8353 0.4796
70.62 0.8513 0.8482 0.3641
75.94 0.8601 0.8581 0.2415
80.24 0.8655 0.8637 0.2020
90.00 0.8678 0.8690 -0.1379
4.4.2.2 Verification of 3D Inclined Semi Elliptical Surface Crack Models
In the study of Noda et al. [52], a singular integral equation method is applied
to calculate the stress intensity factor along crack front of a 3D inclined semi-
elliptical surface crack in a semi-infinite body under tension. The results show
that their method yields smooth variations of mixed modes stress intensity
factors along the crack front accurately for various geometrical conditions. In
their study one of the parameters examined is the effect of inclination angle
( ) on SIF (Figure 4.26, Figure 4.27). In this manner, they calculated mode I,
II and III stress intensity factors at different inclination angles. Their results
showed that, it is difficult to obtain accurate results for large inclination angles
63
(ψ > 45 degrees), because the effect of free surface on the results is
complicated.
The study of Noda et al. gives SIF values at each mode for different
inclination angles (Table 4.8). In this study, to make a comparison, the given
SIF values will be converted to normalized mixed mode J-integral values
through eq. (4.11) [21] and eq. (4.8):
1 1
J = J I + J II + J III = ( K I + K II ) +
2 2 2
K III (4.11)
E *
2μ
where E* is defined in eq. (3.39) and μ is defined in eq. (3.6). The details of
the analysis can be seen in Table 4.6.
Figure 4.26 An Inclined Semi Elliptical Surface Crack in a Semi-infinite Body
64
Figure 4.27 General View of an Inclined Semi Elliptical Surface Crack Von
Mises Stress Distribution a) Over Deformed Shape of the Crack Mouth b)
Side View of the Loaded Cube Showing the Inclined Crack
Table 4.6 Details for Noda’s Solution
Poisson’s d1 d2 d3
E (MPa) a (mm) a/c
Ratio ( ) (MPa) (mm) (mm) (mm)
210000 0.3 1 0.5 1 20 10 20
65
Isida et al. [53] also studied semi-elliptical surface cracks arbitrarily inclined
to the free surface of a semi-infinite solid subjected to tension. They
calculated the SIF at the maximum depth point ( = 900) of the crack front at
different inclination angles. The calculated SIF values are converted to
normalized J-integral values by using eq. (4.11) and eq. (4.8). The details
and the comparison of their study and the present study are given in Table
4.7 and in Table 4.9.
Table 4.7 Details for Isida’s Solution
Poisson’s
E (MPa) a (mm) (degree)
Ratio ( ) (MPa)
210000 0.3 1 90 1
4.5 Conclusion
As can be seen from the comparison tables 4.3, 4.5, 4.8 and 4.9, computed
stress intensity factors and J-integral values are approximately the same with
referred values. However, the results differ near the free surface where the
inclination angle is large. This is a commonly encountered problem as can be
seen in literature. From these results, it can be said that the J-integral
method adopted in this study gives sufficiently accurate results and can be
used to calculate the mixed mode J-integral values around crack front of
inclined semi-elliptical surface cracks at weld toe and root of the parts.
66
Table 4.8 Normalized J-integral Value Comparisons for Inclined Semi Elliptical Surface Cracks
67
Table 4.9 Normalized J-integral Value Comparisons for Inclined Semi Elliptical Surface Cracks
68
Table 4.9 (continued)
69
CHAPTER 5
PARAMETRIC ANALYSES AND NUMERICAL RESULTS
5.1 Introduction
This chapter presents the parametric analysis and obtained results by using
J-integral method. Although the prepared program is a general use program,
this study is mainly focused on the most critical parts (boom and arm) of
HMK 300 LC excavator. In this work, each part is examined in two groups.
First, global model boundary conditions include the lateral force; and the
second, global model boundary conditions do not include the lateral force. In
each group, the models are examined in three sub-groups: local solid model
welded from outer side, local solid model welded from inner side and local
solid model welded from both sides. For this purpose, semi-elliptical surface
cracks are placed to the weld toe and weld root of the prepared models for a
range of inclination angles ( ). For both sides welded case the root cracks
are implemented to the outer side of the root. Eventually, the mixed mode J-
integral values are found along the crack front at different crack front position
angles ( ).
The boundary conditions of the sub-models are directly taken from the global
model. The global model used in this study is verified by the experiments
carried out in Hidromek Ltd. Sti. (Section 4.4.1).
70
In case studies, 252 sub-models are formed for different parameters. The
calculated mixed mode J-integral values showing the effect of inclined semi
elliptical surface cracks on weld toes and weld root of different welded
connections are given in Figures 5.5-5.34 and in Figures 5.40-5.69. The
obtained results are non-dimensionalized by using eq. (4.7) and eq. (4.8).
Since the sub-model applied stresses are variable around the boundaries of
the sub-models, a constant applied stress value (8) is used during the
normalization. The details of the analyses are given in Table 5.1. and Table
5.2.
Table 5.1 Parametric Analyses Material Properties and Crack Parameters
Poisson’s
E (MPa) a (mm) c (mm)
Ratio ( ) (MPa)
210000 0.3 0.5 1 100
Table 5.2 Parametric Analyses Weld Parameters (mm)
Root Depth Middle Crack Middle Crack

Leg1 Leg2 Leg_1 Leg_2
(RD) Distance (MCD) Length (MCL)
11 6 6 6 6 3 3
5.2 Parametric Analyses
The boom and arm of the excavator are the most critical parts of all structure.
Since almost all of the earth-moving machineries are exposed to repeated
71
fluctuating stresses, fatigue cracks occur especially on boom and arm of the
machines due to high stresses on these parts. The cracks appear generally
on welded connections of sheet metals. Thus, optimization of these parts is a
very important procedure for reducing the stresses that the welds are
exposed to. However, the shape optimization is not enough alone without
proper weld parameters selection. Therefore, this study will examine the weld
parameters of an optimized excavator boom and arm in terms of fracture
mechanics. In all of the analyses three cases: welding from outside, welding
from inside and welding from both sides will be compared to each other. In
the following, the details of the boom and arm parametric analyses will be
explained.
5.2.1 Boom Parametric Analyses
A typical structural zooming (sub-modeling) analysis starts with the global

model analysis. This is due to the reason that the displacement type
boundary conditions are taken from the global post file. In Figure 5.1 and in
Figure 5.2 the stress map of the excavator boom global shell model and the
interested regions are shown. The interested region must be at the corner of
the box section and must not contain any thickness transition region. Figure
5.3 shows the different sheet metal thicknesses with different colors and the
position of the selected region.
When the selection is made, the interested region is in shell form and must
be transformed to solid local model. Shell-to-solid conversion procedure, the
boundary conditions, weld and crack parameters are explained in detail in
Chapter 4. An example to a completed solid local model analysis is given in
Figure 5.4. In Figures 5.5-5.34 the normalized J-integral values versus
inclination angle ( ) for five crack front position angles ( ) are presented.
72
Figure 5.1 Excavator Boom Global Model (Von Mises Stress Outer Layer)
Figure 5.2 Excavator Boom Global Model (Von Mises Stress Inner Layer)
73
Figure 5.3 Excavator Boom Model Sheet Metals and the Selected Region
Figure 5.4 Boom Selected Shell Region and Boom Solid Local Model Von
Mises Stress Map
74
7
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
2
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
4
.
0
0
E
-
2
J*
3
.
0
0
E
-
2
2
.
0
0
E
-
2
1
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]
Figure 5.5 Normalized J-integral versus Inclination Angle ( )

(Crack at Weld Toe 1, Lateral Force is Included, =310)
7
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
2
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
4
.
0
0
E
-
2
J*
3
.
0
0
E
-
2
2
.
0
0
E
-
2
1
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

75
7
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
2
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
4
.
0
0
E
-
2
J*
3
.
0
0
E
-
2
2
.
0
0
E
-
2
1
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

7
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
2
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
4
.
0
0
E
-
2
J*
3
.
0
0
E
-
2
2
.
0
0
E
-
2
1
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

76
7
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
2
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
4
.
0
0
E
-
2
J*
3
.
0
0
E
-
2
2
.
0
0
E
-
2
1
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
2
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
2
6
.
0
0
E
-
2
J*
4
.
0
0
E
-
2
2
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

77
1
.
2
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
2
6
.
0
0
E
-
2
J*
4
.
0
0
E
-
2
2
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
2
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
2
6
.
0
0
E
-
2
J*
4
.
0
0
E
-
2
2
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

78
1
.
2
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
2
6
.
0
0
E
-
2
J*
4
.
0
0
E
-
2
2
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
2
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
2
6
.
0
0
E
-
2
J*
4
.
0
0
E
-
2
2
.
0
0
E
-
2
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

79
3
.
0
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
2
.
5
0
E
-
1
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
1
1
.
5
0
E
-
1
J*
1
.
0
0
E
-
1
5
.
0
0
E
-
2
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

(Crack at Weld Root, Lateral Force is Included, =310)
2
.
5
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
1
1
.
5
0
E
-
1
J*
1
.
0
0
E
-
1
5
.
0
0
E
-
2
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

80
2
.
5
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
1
1
.
5
0
E
-
1
J*
1
.
0
0
E
-
1
5
.
0
0
E
-
2
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

2
.
5
0
E
-
1
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
1
1
.
5
0
E
-
1
J*
1
.
0
0
E
-
1
5
.
0
0
E
-
2
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

81
2
.
5
0
E
-
1
2
.
0
0
E
-
1
1
.
5
0
E
-
1
O
u
t
s
i
d
e
J*
I
n
s
i
d
e
1
.
0
0
E
-
1
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
2
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

3
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
3
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
2
.
5
0
E
-
3
2
.
0
0
E
-
3
J*
1
.
5
0
E
-
3
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

(Crack at Weld Toe 1, Lateral Force is not Included, =310)
82
4
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
3
.
5
0
E
-
3
B
o
t
h
S
i
d
e
s
3
.
0
0
E
-
3
2
.
5
0
E
-
3
2
.
0
0
E
-
3
J*
1
.
5
0
E
-
3
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

4
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
3
.
5
0
E
-
3
B
o
t
h
S
i
d
e
s
3
.
0
0
E
-
3
2
.
5
0
E
-
3
2
.
0
0
E
-
3
J*
1
.
5
0
E
-
3
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

83
4
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
3
.
5
0
E
-
3
B
o
t
h
S
i
d
e
s
3
.
0
0
E
-
3
2
.
5
0
E
-
3
2
.
0
0
E
-
3
J*
1
.
5
0
E
-
3
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

3
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
3
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
2
.
5
0
E
-
3
2
.
0
0
E
-
3
J*
1
.
5
0
E
-
3
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

84
6
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
5
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
4
.
0
0
E
-
3
3
.
0
0
E
-
3
J*
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

7
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
3
4
.
0
0
E
-
3
J*
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

85
7
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
3
4
.
0
0
E
-
3
J*
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

7
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
3
4
.
0
0
E
-
3
J*
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

86
7
.
0
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
6
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
5
.
0
0
E
-
3
4
.
0
0
E
-
3
J*
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
8
0
E
-
2
1
.
6
0
E
-
2
1
.
4
0
E
-
2
1
.
2
0
E
-
2
O
u
t
s
i
d
e
1
.
0
0
E
-
2
I
n
s
i
d
e
J*
8
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

(Crack at Weld Root, Lateral Force is not Included, =310)
87
1
.
6
0
E
-
2
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
8
.
0
0
E
-
3
J*
B
o
t
h
S
i
d
e
s
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

1
.
6
0
E
-
2
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
8
.
0
0
E
-
3
J*
B
o
t
h
S
i
d
e
s
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

88
1
.
6
0
E
-
2
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
O
u
t
s
i
d
e
I
n
s
i
d
e
8
.
0
0
E
-
3
J*
B
o
t
h
S
i
d
e
s
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

2
.
0
0
E
-
2
1
.
8
0
E
-
2
1
.
6
0
E
-
2
1
.
4
0
E
-
2
1
.
2
0
E
-
2
O
u
t
s
i
d
e
1
.
0
0
E
-
2
J*
I
n
s
i
d
e
8
.
0
0
E
-
3
B
o
t
h
S
i
d
e
s
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

89
5.2.2 Arm Parametric Analyses
The excavator arm is one of the main components of digging mechanism. It

is also exposed to repeated fluctuations, which cause cracks to occur. Figure
5.35 shows a crack encountered during the fatigue tests performed in
Hidromek Ltd. Sti. Therefore, this location can be a good starting point for
fracture mechanics analyses of excavator arm. The global shell model
analysis of the arm will again be the initial step in structural zooming analysis
(Figure 5.36, 5.37). The selected region on global shell model is shown in
Figure 5.38. Notice that the selected region is at the corner of the box section
and does not contain any thickness transition region. An example to the
generated solid local model analysis is given in Figure 5.39. The resulting
normalized J- integral values versus inclination angle ( ) for five crack front
position angles ( ) are given in Figures 5.40-5.69.
Figure 5.35 A Fatigue Crack Occurred on HMK 300 LC Excavator Arm

During the Performed Tests in Hidromek Ltd. Sti.
90
Figure 5.36 Excavator Arm Global Model (Von Mises Stress Outer Layer)
Figure 5.37 Excavator Arm Global Model (Von Mises Stress Inner Layer)
91
Figure 5.38 Excavator Arm Model Sheet Metals and the Selected Region
Figure 5.39 Arm Selected Shell Region and Arm Solid Local Model Von
Mises Stress Map
92
9
.
0
0
E
-
3
O
u
t
s
i
d
e
8
.
0
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
J*
4
.
0
0
E
-
3
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
0
0
E
-
2
O
u
t
s
i
d
e
9
.
0
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
3
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
J*
4
.
0
0
E
-
3
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

93
1
.
0
0
E
-
2
O
u
t
s
i
d
e
9
.
0
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
3
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
J*
4
.
0
0
E
-
3
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
0
0
E
-
2
O
u
t
s
i
d
e
9
.
0
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
8
.
0
0
E
-
3
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
J*
4
.
0
0
E
-
3
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

94
9
.
0
0
E
-
3
O
u
t
s
i
d
e
8
.
0
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
J*
4
.
0
0
E
-
3
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
8
0
E
-
2
O
u
t
s
i
d
e
1
.
6
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
J*
8
.
0
0
E
-
3
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

95
1
.
8
0
E
-
2
O
u
t
s
i
d
e
1
.
6
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
J*
8
.
0
0
E
-
3
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
8
0
E
-
2
O
u
t
s
i
d
e
1
.
6
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
J*
8
.
0
0
E
-
3
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

96
1
.
8
0
E
-
2
O
u
t
s
i
d
e
1
.
6
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
J*
8
.
0
0
E
-
3
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
8
0
E
-
2
O
u
t
s
i
d
e
1
.
6
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
2
1
.
2
0
E
-
2
1
.
0
0
E
-
2
J*
8
.
0
0
E
-
3
6
.
0
0
E
-
3
4
.
0
0
E
-
3
2
.
0
0
E
-
3
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

97
5
.
0
0
E
-
2
O
u
t
s
i
d
e
4
.
5
0
E
-
2
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
4
.
0
0
E
-
2
3
.
5
0
E
-
2
3
.
0
0
E
-
2
2
.
5
0
E
-
2
J*
2
.
0
0
E
-
2
1
.
5
0
E
-
2
1
.
0
0
E
-
2
5
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

4
.
0
0
E
-
2
3
.
5
0
E
-
2
3
.
0
0
E
-
2
2
.
5
0
E
-
2
O
u
t
s
i
d
e
2
.
0
0
E
-
2
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
5
0
E
-
2
1
.
0
0
E
-
2
5
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

98
4
.
0
0
E
-
2
3
.
5
0
E
-
2
3
.
0
0
E
-
2
2
.
5
0
E
-
2
O
u
t
s
i
d
e
2
.
0
0
E
-
2
I
n
s
i
d
e
J*
B
o
t
h
S
i
d
e
s
1
.
5
0
E
-
2
1
.
0
0
E
-
2
5
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

4
.
0
0
E
-
2
3
.
5
0
E
-
2
3
.
0
0
E
-
2
2
.
5
0
E
-
2
O
u
t
s
i
d
e
2
.
0
0
E
-
2
I
n
s
i
d
e
J*
B
o
t
h
S
i
d
e
s
1
.
5
0
E
-
2
1
.
0
0
E
-
2
5
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

99
5
.
0
0
E
-
2
4
.
5
0
E
-
2
4
.
0
0
E
-
2
3
.
5
0
E
-
2
3
.
0
0
E
-
2
O
u
t
s
i
d
e
2
.
5
0
E
-
2
I
n
s
i
d
e
J*
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
2
1
.
5
0
E
-
2
1
.
0
0
E
-
2
5
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

1
.
4
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
2
0
E
-
3
B
o
t
h
S
i
d
e
s
1
.
0
0
E
-
3
8
.
0
0
E
-
4
J*
6
.
0
0
E
-
4
4
.
0
0
E
-
4
2
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

100
1
.
6
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
4
0
E
-
3
B
o
t
h
S
i
d
e
s
1
.
2
0
E
-
3
1
.
0
0
E
-
3
8
.
0
0
E
-
4
J*
6
.
0
0
E
-
4
4
.
0
0
E
-
4
2
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
8
0
E
-
3
O
u
t
s
i
d
e
1
.
6
0
E
-
3
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
1
.
4
0
E
-
3
1
.
2
0
E
-
3
1
.
0
0
E
-
3
J*
8
.
0
0
E
-
4
6
.
0
0
E
-
4
4
.
0
0
E
-
4
2
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

101
1
.
6
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
4
0
E
-
3
B
o
t
h
S
i
d
e
s
1
.
2
0
E
-
3
1
.
0
0
E
-
3
8
.
0
0
E
-
4
J*
6
.
0
0
E
-
4
4
.
0
0
E
-
4
2
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

1
.
6
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
1
.
4
0
E
-
3
B
o
t
h
S
i
d
e
s
1
.
2
0
E
-
3
1
.
0
0
E
-
3
8
.
0
0
E
-
4
J*
6
.
0
0
E
-
4
4
.
0
0
E
-
4
2
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

102
2
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
5
0
E
-
3
J*
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

2
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
5
0
E
-
3
J*
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

103
2
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
5
0
E
-
3
J*
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

2
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
5
0
E
-
3
J*
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

104
2
.
5
0
E
-
3
O
u
t
s
i
d
e
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
5
0
E
-
3
J*
1
.
0
0
E
-
3
5
.
0
0
E
-
4
0
.
0
0
E
+
0
0
1
0
2
0
3
0
4
0
5
0
6
0
7
0
8
0
9
0
[degree]

8
.
0
0
E
-
3
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
O
u
t
s
i
d
e
4
.
0
0
E
-
3
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

105
6
.
0
0
E
-
3
5
.
0
0
E
-
3
4
.
0
0
E
-
3
O
u
t
s
i
d
e
3
.
0
0
E
-
3
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

6
.
0
0
E
-
3
5
.
0
0
E
-
3
4
.
0
0
E
-
3
O
u
t
s
i
d
e
3
.
0
0
E
-
3
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

106
6
.
0
0
E
-
3
5
.
0
0
E
-
3
4
.
0
0
E
-
3
O
u
t
s
i
d
e
3
.
0
0
E
-
3
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

8
.
0
0
E
-
3
7
.
0
0
E
-
3
6
.
0
0
E
-
3
5
.
0
0
E
-
3
O
u
t
s
i
d
e
4
.
0
0
E
-
3
J*
I
n
s
i
d
e
B
o
t
h
S
i
d
e
s
3
.
0
0
E
-
3
2
.
0
0
E
-
3
1
.
0
0
E
-
3
0
.
0
0
E
+
0
0
2
0
4
0
6
0
8
0
1
0
0
1
2
0
1
4
0
1
6
0
1
8
0
[degree]

107
5.3 Numerical Results
The obtained results give useful knowledge about the welding location (one-
sided, two-sided) and geometry and probable failure locations of boom and
arm welded connections. Since the performed analyses are divided into two
groups as boom and arm analysis, it will be reasonable to examine the
results group by group.
Boom and arm analyses can be investigated in two groups: analyses

including lateral force in boundary conditions and analyses not including
lateral force in boundary conditions. A vital result attained in the analyses is
the effect of lateral force on J-integral. It can easily be seen that the
corresponding J-integral values in Figures 5.5-5.19 are much more higher
than the values in Figures 5.20-5.34, which means that the lateral force has
an important increasing effect on J-integral. Thus, if a machine is exposed to
lateral force too much, the expected life of the digging mechanism may
decrease dramatically. Therefore, if the total life of the digging mechanism is
going to be estimated, the fraction of the lateral force in loading spectrum is
very important and must be carefully taken into consideration.
If boom analyses are examined (Figures 5.5-5.34), it is seen that if the

welding is made from outside the probable failure location is the root of the
weld, if the welding is made from inside the probable failure location is toe 2
of the weld and if the welding is made from both sides the probable failure
location is toe 1 of the weld. The tendency of toe 1 weld failure of both sides
welded connections shifts to the root failures if the lateral force is excluded
from boundary conditions. If the three welding locations (outside, inside and
both sides) are compared to each other, in all boom analyses both sides weld
gives the best and outside weld gives the worst J-integral values. Therefore,
both sides weld is more suitable than the other two alternatives at the
specified assumptions, made in previous chapters.
108
Arm analyses (Figures 5.40-5.69) show also similar trend with boom
analyses. The probable failure locations for the three weld locations and the
shift to root failures for both sides weld applications are the same. The
ordering of the weld location is also the same. Thus, the results indicate that
the weld from both sides is the best and the weld from outside is the worst
alternative among them.
A significant result is that the risk of failure on the selected region of the
boom is higher than the selected region of the arm in spite of the high
stresses on the latter case.
109
CHAPTER 6
CONCLUDING REMARKS
In the present study, three-dimensional semi-elliptic surface crack problem in

T-welded joints are examined using a three-dimensional finite element sub-
modeling technique. An interface, which lets the user define the related weld
and crack parameters, has been developed during the study. The developed
interface enables the creation of three dimensional finite element solid model
of a selected region by using a shell-to-solid conversion method. The
prepared models are analyzed by using commercial finite element analysis
software MSC.Mentat-Marc. The linear analyses results give the energy
release rate for mixed mode loading type. In the analyses J-integral approach
is used and the local solid models are prepared in this manner. The local
solid models consist of 20-node and collapsed 20-node isoparametric three
dimensional brick elements. The failure locations of a typical weld are
assumed be the two weld toes and the weld root. It is also assumed that
these locations have semi elliptical surface cracks, which are inclined for a
range of inclination angles (ψ).
The accuracy of the adopted method is analyzed in two steps: global model
verification and local model verification. Since this study is a sub-modeling
study, which takes the kinematic boundary conditions directly from the global
model, the global model analysis must be verified first. The performed
experiments in Hidromek Ltd. Sti. show that the global model analyses
results and experiment results fit in well. Therefore, the global model is
110
assumed to be sufficiently precise for the local model analyses. To verify the
local models, the solutions of Sneddon [50], Irwin [51], Noda et al [52]. and
Isida et al. [53] are compared with the results found by J-integral method.
According to the obtained results, the J-integral method is sufficiently reliable
unless the inclination angle is above 45o or the interested node is at free
surface.
In case study analyses, two different regions one is on boom and the other is
on arm of HMK 300 LC excavator are analyzed in terms of fracture
mechanics. Each region is investigated either including the lateral force or
not including the lateral force in boundary conditions. In all groups, three weld
location cases namely welding from outside, welding from inside and welding
from both side are studied for different inclination angles (ψ). In toe regions
the range of inclination angle is 0o - 90o and in root region the range of
inclination angle is 0o - 180o. Thus, the resulting graphs are formed in that
way.
The computations show that welding from both side is the best and welding
from outside is the worst alternative among the three welding location cases
for the indicated global model and for the specified assumptions, which are
made in previous chapters.
Fatigue failure of welded connections is the main criteria determining the total
life of the construction machinery. This study aims to analyze the different
weldments in construction machineries in terms of fracture mechanics. The
method used in this study is applicable to linearly varying box cross section
corners. By improving the method, curvilinear parts of the box cross sections
and sheet metal thickness transition regions can also be examined. Thus,
any region on any parts can be investigated by fracture mechanics point of
view. The study can also be improved by including the total life assessments
of the weldments. In this manner, a crack map for all parts of the construction
machinery can be defined for the given loading spectrum. In addition to
111
these, the effect of heat treatment and the weld material can also be taken
into consideration in subsequent studies. After completing all studies, the
models must be verified by controlled experiments as future work.
112
REFERENCES
[1] E. A. Avallone, T. Baumaister iii, Marks’ Standard Handbook for

Mechanical Engineers 10th edition, McGRAW-HILL International
editions, pp 10.25
[2] “Wikipedia The Free Encyclopedia”

http://en.wikipedia.org/wiki/Excavator, Last accessed 13.05.2007.
[3] “SAE J1057 SEP88: Identification Terminology of Earthmoving

Machines”
[4] Taylor D., Barrett N., Lucano G., “Some New Methods for Predicting
Fatigue in Welded Joints”, International Journal of Fatigue, Vol.24,
509-518, 2002.
[5] Singh P. J., Guha B., Achar D. R. G., “Fatigue Life Prediction Using
Two Stage Model for AISI 304L Cruciform Joints, with Different Fillet
Geometry, Failing at Toe”, Science and Technology of Welding and
Joining, Vol.8, 69-75, 2003.
[6] Kainuma S., Mori T., “A Fatigue Strength Evaluation Method for Load-
Carrying Fillet Welded Cruciform Joints”, International Journal of
Fatigue, Vol.28, 864-872, 2006.
113
[7] Darcis P., Santarosa D., Recho N., Lassen T., “A Fracture Mechanics
Approach for the Crack Growth in Welded Joints with Reference to BS
7910”, The 15th European Conference of Fracture, 2004.
[8] Sidhom N., Laamouri A., Fathallah R., Braham C., Lieurade H. P.,
“Fatigue Strength Improvement of 5083 H11 Al-alloy T-welded Joints
by Shot Peening: Experimental Characterization and Predictive
Approach”, International Journal of Fatigue, Vol.27, 729-745, 2005.
[9] Ferreira J., Castiglioni C. A., Calado L., Agatino M. R., “Low Cycle
Fatigue Strength Assessment of Cruciform Welded Joints”, Journal of
Construction Steel Research, Vol.47, 223-244, 1998.
[10] Haagensen P. J., Statnikov E. S., Martinez L. L., “Introductory Fatigue

Tests on Welded Joints in High Strength Steel and Aliminium
Improved by Various Methods Including Ultrasonic Impact Treatment
(UIT)”, International Institute of Welding, IIW Doc. XIII- 1748-98, 1998.
[11] Singh P. J., Guha B., Achar D. R. G., “Fatigue Life Prediction of GTA
Welded AISI 304L Cruciform Joints with Lack of Penetration Using
Local-Stress Approach”, Engineering Failure Analysis, Vol.10, 25-36,
2003.
[12] Ranestad Q., Zhang Z. L., Thaulow C., “An Engineering Method for
Constraint Based Fracture Assessment of Welded Structural
Components with Surface Cracks”, Engineering Fracture Mechanics,
Vol.63, 653-674, 1999.
[13] Baumjohann F., Kröning J., “A Practical Method for Computation of

Ductile Crack Growth by Means of Finite Elements and Parametric
3D-modelling”, Nuclear Engineering and Design, Vol.190, 197-206,
1999.
114
[14] Nykänen T., Marquis G., Björk T., “Fatigue Analysis of Non-load-
carrying Fillet Welded Cruciform Joints”, Engineering Fracture
Mechanics, Vol.74, 399-415, 2007.
[15] Lin X.B., Smith R.A., “Finite Element Modelling of Fatigue Crack
Growth of Surface Cracked Plates Part I: The Numerical Technique”,
Engineering Fracture Mechanics, Vol.63, 503-522, 1999.
[16] Courtin S., Gardin C., Bezine G., Hamouda H. B. H., “Advantages of
the J-integral Approach for Calculating stress Intensity Factors When
Using the Commercial Finite Element Software ABAQUS”,
Engineering Fracture Mechanics, Vol.72, 2174-2185, 2005.
[17] Ficici F., “Three Dimensional Fracture Analysis of Fillet Welds”, Ms.
Thesis Submitted to Middle East Technical University, Turkey, 2007.
[18] Inan O., “Three Dimensional Fracture Analysis of FGM Coatings”, Ms.
Thesis Submitted to Middle East Technical University, Turkey, 2004.
[19] Ayhan A. O., “Mixed Mode Stress Intensity Factors for Deflected and
Inclined Surface Cracks in Finite-Thickness Plates”, Engineering
Fracture Mechanics, Vol.71, 1059-1079, 2004.
[20] dell’Erba D. N., Aliabadi M. H., “BEM Analysis of Fracture Probşems in

Three-dimensional Thermoelasticity Using J-integral”, International
Journal of Solids and Structures, Vol.38, 4609-4630, 2001.
[21] Rigby R. H., Aliabadi M. H., “Decomposition of the Mixed-Mode J-

integral- Revisited”, International Journal of Solids and Structures,
Vol.35, 2073-2099, 1998.
115
[22] Statnikov E. S., Muktepavel V. O., Trufyakov V. I., Mikheev P. P.,
Kuzmenko A. Z., Blomqvist A., “Efficiency Evaluation of Ultrasonic
Impact Treatment (UIT) of Welded Joints in Weldox 420 Steel in
accordance with the IIW Program”, IIW Doc. XIII-1817-00, 2000.
[23] Lahti K. E., Hanninen H., Niemi E., “Nominal Stress Range Fatigue of
Stainless Steel Fillet Welds — the Effect of Weld Size”, Journal of
Constructional Steel Research, Vol.54, 161-172, 2000.
[24] Fricke W., Doerk O., “Simplified Approach to Fatigue Strength

Assessment of Fillet-Welded Attachment Ends”, International Journal
of Fatigue, Vol.28, 141-150, 2006.
[25] “ISO/DIS 14345: Fatigue testing of welded components”, 2001
[26] Gray L. J., Phan A. V., Paulino G. H., Kaplan T., “Improved Quarter-
Point Crack Tip Element”, Engineering Fracture Mechanics, Vol.70,
269-283, 2003.
[27] Sabuncuoglu B., “Fatigue Crack Growth Analysis Models for FGM’s”,
Ms. Thesis Submitted to Middle East Technical University, Turkey,
2006.
[28] Kadioglu S., Dag S., Yahsi S., “Crack Problem for a Functionally
Graded Layer on an Elastic Foundation”, International Journal of
Fracture, Vol.47, 533-545, 1991.
[29] Guo L.C., Wu L.Z., Zeng T., Ma L., “Mode I Crack Problem for a
Functionally Graded Orthotropic Strip”, European Journal of
Mechanics A/Solids, Vol. 23, 219-234, 2004.
116
[30] Guo K., Bell R., Wang X., “The Stress Intensity Factor Solutions for
Edge Cracks in a Padded Plate Geometry under General Loading
Conditions”, International Journal of Fatigue, Vol.29, 481-488, 2007.
[31] Hou J., Goldstraw M., Maan S., Knop M., “An Evaluation of 3D Crack
Growth Using ZENCRACK”, DSTO-TR-1158, 1-28, 2001.
[32] Gerstle W., “Structural Zooming Research and Development of an

Interactive Computer Graphical Interface for Stress Analysis of
Cracks", NASA/American Society for Engineering Education (ASEE)
Summer Faculty Fellowship Program 1989, Tiwari, S. N. (Compiler),
NASA Contractor Report 181894, pp. 67-68, Sept., 1989.
[33] Gerstle, W., Martha, L., Ingraffea, A., "Finite and Boundary Element
Modeling of Crack Propagation in Two and Three Dimensions,"
Engineering with Computers, Vol. 2, pp. 167-183, 1987.
[34] Giglio M., “FEM Submodelling Fatigue Analysis of a Complex

Helicopter Component”, International Journal of Fatigue, Vol.21, 445-
455, 1999.
[35] Kitamura M., Ohtsubo H., Akiyama A., Bandoh H., “Submodeling
Analysis of Ship Structure with Superconvergent Patch Recovery
Method”, International Journal of Offshore and Polar Engineering,
Vol.13, 2003.
[36] Larson M. G., “Adaptive Submodeling”, Computational Mathematics,

Chalmers
[37] MSC.MARC Volume A: Theory and User Information Version 2005,

Vol.A, Chapter 5, 123-127.
117
[38] ABAQUS Analysis User's Manual v6.5-1: Part III Analysis: Section 7.3
[39] ANSYS Advanced Analysis Techniques Guide Release 11.0, Chapter

9, 229-239.
[40] “Lecture 12.10: Basics of Fracture Mechanics”, ESDEB WG12,

http://www.kuleuven.ac.be/bwk/materials/Teaching/master/wg12/l1000
.htm#SEC_1, Last accessed 18.05.2007.
[41] Lee Y., Pan J., Hathaway R. B., Barkey M. E., “Fatigue Testing and
Analysis (Theory and Practice)”, Elsevier, 2005.
[42] “ME 583 Engineering Fracture Mechanics”

http://www.me.metu.edu.tr/me583, Last accessed 21.05.2007.
[43] Yildirim B., Dag S., Erdogan F., “Three Dimensional Fracture Analysis
of FGM Coatings Under Thermo-mechanical Loading”, International
Journal of Fracture, Vol.132, 369-395, 2005.
[44] Sharma A. K., “Kinematics and Finite Element Analysis of Excavator

Attachments”, Ms. Thesis Submitted to Nirma University of Science &
Technology, Ahmedabad, 2005.
[45] Cook R. D., “Finite Element Modeling for Stress Analysis”, John Wiley
& Sons, Inc., 1995.
[46] Eriksson A., Lignell A. M., Olsson C., Spennare H., “Weld Evaluation
Using FEM”, Industrilitteratur AB., 2003.
[47] Yener M., “Design of a Computer Interface for Automatic Finite

Element Analysis of an Excavator Boom”, Ms. Thesis Submitted to
Middle East Technical University, Turkey, 2005.
118
[48] Makinen H., “Global and Local FE Models in the Fatigue Analysis of
Ships”, Ms. Thesis Submitted to VTT Technical Research Centre of
Finland, Finland, 2005.
[49] SAE J1179, “Hydraulic Excavator and Backhoe Digging Forces”, SAE
Handbook, Vol.3, 40.355, 1995.
[50] Broek D., “Elementary Engineering Fracture Mechanics”, Martinus

Nijhoff Publishers, 1984
[51] Sanford R.J., “Principles of Fracture Mechanics”, Prentice Hall, 2003
[52] Noda N., Kobayashi K., Yagishita M., “Variation of mixed modes
stress intensity factors of an inclined semi-elliptical surface crack”,
International Journal of Fracture, Vol.100, 207-225, 1999
[53] Isida M., Tokumoto A., Noguchi H., “Oblique Semi-elliptical Surface
Crack in Semi-infinite Solid Subjected to Tension”, Engineering
Fracture Mechanics, Vol.36, 889-892, 1990.
[54] Cherepanov G.P., “The Propagation of Cracks in a Continuous

Medium”, Journal of Applied Mathematics and Mechanics, Vol. 31,
503-512, 1967
[55] Rice J.R., “A Path Independent Integral and the Approximate Analysis
of Strain Concentration by Notches and Cracks”, Journal of Applied
Mechanics Transactions of the ASME, Vol. 35, 379-386, 1968
[56] Newman J.C., Raju I.S., “An Empirical Stress Intensity Factor
Equation for the Surface Crack”, Engineering Fracture Mechanics, Vol.
15, 185-192, 1981
119

Index

Uploaded by

Copyright:

Available Formats

Index

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Index

Uploaded by

Copyright:

Available Formats

A FINITE ELEMENTS BASED APPROACH

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

“A FINITE ELEMENTS BASED APPROACH FOR FRACTURE ANALYSIS

Submitted by TANER KARAGÖZ in partial fulfillment of the requirements for

Prof. Dr. Canan Özgen

Prof. Dr. Kemal İder

Asst. Prof. Dr. Serkan Dağ

Examining Committee Members:

Prof. Dr. Eres Söylemez (*)

Asst. Prof. Dr. Serkan Dağ (**)

Prof. Dr. F. Suat Kadıoğlu

Assoc. Prof. Dr. Bora Yıldırım

Kadir Geniş (M.S.)

(*) Head of Examining Committee

Name, Last name: Taner KARAGÖZ

A FINITE ELEMENTS BASED APPROACH FOR FRACTURE ANALYSIS

August 2007, 119 pages

This study aims to develop a computer program to perform finite elements

Keywords: Finite elements, sub-modeling, semi elliptical surface crack,

İŞ MAKİNALARINDAKİ KAYNAKLI BAĞLANTILARIN SONLU

Ağustos 2007, 119 sayfa

Bu çalışmanın amacı, iş makinelerinin köşe kaynaklı bağlantılarında üç

Anahtar kelimeler: Sonlu elemanlar, alt modelleme, yarı eliptik yüzey

Finally, I wish to express my gratitude to my love M. Özgül ŞAHİNKAYA

This study was supported by Hidromek Ltd. Sti.

LIST OF TABLES..................................................................................... xiv

LIST OF FIGURES ...................................................................................xv

LIST OF SYMBOLS AND ABBREVIATIONS......................................... xxiv

3. FRACTURE ANALYSIS IN THREE DIMENSIONS.......................... 17

3.1 Introduction ................................................................................ 17

3.2 Stress Intensity Factor (SIF)....................................................... 18

3.3 Fracture Toughness ................................................................... 21

3.5 Energy Release Rate ................................................................. 28

3.6 J-integral .................................................................................... 30

3.7 Numerical Evaluation of the J-integral........................................ 31

4. FINITE ELEMENT MODELING........................................................ 33

4.1 Introduction ................................................................................ 33

4.2 Problem Definition ...................................................................... 34

4.3 Sub-modeling ............................................................................. 36

4.3.1 Global Model ..................................................................... 38

4.3.1.1 Boundary Conditions ............................................. 41

4.3.1.1.1 Bucket Breakout Force ............................ 41

4.3.1.1.2 Arm Breakout Force ................................ 43

4.3.1.1.3 Lateral Force ........................................... 44

4.3.1.1.4 Other Boundary Conditions ..................... 44

4.3.2 Local Region Selection ...................................................... 45

4.3.3 Parameters ........................................................................ 46

4.3.3.2 Crack Parameters.................................................. 49

4.3.4 Shell-to-Solid Conversion Method ..................................... 50

4.3.5 Material Properties............................................................. 52

4.3.6 Tyings and Local Model Boundary Conditions................... 53

4.3.7 J-integral Calculation ......................................................... 55

4.4 Verification of Finite Element Modeling and J-integral

4.4.1 Verification of Finite Element Modeling............................. 56

4.4.2 Verification of J-integral Calculation.................................. 57

4.4.2.1 Verification of Embedded Circular and Elliptical

4.4.2.2 Verification of 3D Inclined Semi Elliptical Surface

4.5 Conclusion ................................................................................ 66

5. PARAMETRIC ANALYSES AND NUMERICAL RESULTS.............. 70

5.1 Introduction ............................................................................... 70