Int J Advanced Design and Manufacturing Technology, Vol. 11/ No.

2/ June - 2018 15

Effect of Burr Grinding on

Fatigue Strength of Steel
Butt-Welded Connections
Abbas Fadaei*
Department of Mechanical Engineering,
Bu-Ali Sina University, Iran
E-mail: [email protected]
*Corresponding author

Annette Betkhoodu
Department of Mechanical Engineering,
Bu-Ali Sina University, Iran
E-mail: [email protected]

Received: 6 January 2018, Revised: 2 February 2018, Accepted: 8 April 2018

Abstract: Among assembling methods, welding is most widely used in various

industries. During the welding operation, material is heated to a temperature above
the melting point and melted material forms a weld pool. This leads to the formation
of tensile residual stresses in the weld toe. In this paper, effect of burr grinding
technique on fatigue strength of butt-welded joint has been evaluated. Burr grinding
is one of the weld geometry modification methods that with removing small crack-
like defects at the weld toe and increasing weld toe radius leads to reduction of stress
concentration factor (SCF) and improvement of fatigue strength of weld. Also, the
finite element simulation was performed by using ABAQUS software and stress
concentration factor was chosen as a criterion. This factor was calculated using
analytical and numerical methods for samples before and after grinding. Burr
grinding procedure on welded samples was done using an electric grinder and
different conical burrs. Burrs with different radii have been selected in order to
provide a better comparison. Fatigue life of samples before and after grinding was
determined by fatigue tests under constant amplitude loading. The results show
43.72 percent improvement in stress concentration factor and 50.61 percent
improvement in fatigue life of samples. In experimental study, the best result belongs
to grinding with a 3 mm tapered burr leading to an improvement of 50.61% in fatigue
life while grinding with 1 mm tapered gives the worst result of 8.88% improvement
in fatigue life.

Keywords: Burr grinding, Fatigue strength, Stress concentration factor butt weld

Reference: Abbas Fadaei and Annette Betkhoodu, “Effect of Burr Grinding on

Fatigue Strength of Steel Butt-Welded Connections”, Int J of Advanced Design and
Manufacturing Technology, Vol. 11/No. 2, 2018, pp. 15-25.

Biographical notes: A. Fadaei received his PhD in Mechanical Engineering from

Bu-Ali Sina University. He is currently Assistant Professor at the Department of
Mechanical Engineering, Bu-Ali Sina University, Hamedan, Iran. His current
research interest includes Stress Analysis and Applied Design. A. Betkhoodu
received her BSc and MSc in Mechanical Engineering from Bu-Ali Sina University,
Iran, in 2012 and 2016, respectively.

© 2018 IAU, Majlesi Branch

16 Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018

1 INTRODUCTION Pedersen et al. [4] compared three post weld treatments

for improving the fatigue life in the welded joints. They
Welding is a critical technique in the field of connecting studied burr grinding, TIG dressing and ultrasonic
parts in the industry and is one of the production impact (UI) treatments for medium cycle area, i.e.
methods. Its purpose is permanent connection of 10000-500000 cycles, and very high stress ranges. This
engineering materials such as metals, ceramics, study shows that the fatigue resistance of the specimens
polymers and composites to each other. Welding from the three improvement treatments was nearly
defects, welding geometry such as toe radius and the similar, so that the fatigue strength was increased 31%
pattern of tensile residual stresses, have a significant by burr grinding, 33% by UI treatment and 38% by TIG
impact on fatigue strength of welded connections and dressing.
weld failures. During the welding operation, welding Tai et al. [5] considered the fatigue strength
zone experiences high temperature differences improvement of out-of-plane gusset plate specimens by
compared to the surrounding area. Under thermal loads, burr grinding and hammer peening treatments. The
the material expands and the environment is an elderly specimens were tested in high-cycle fatigue condition
inhibition and this leads to excitement of stresses. Since with constant and variable amplitude loading. They
the yield strength decreases with increasing temperature, concluded that the fatigue life can be increased by the
these stresses often exceed yield stress which results in reduction of stress concentration using the burr grinding
development of tensile residual stresses in weld toe. As treatment. Baptista et al. [6] presented the fatigue
a result, welded connections have less fatigue strength in behavior of Duplex S31803 and Austenitic 304L in air
relative to base metal. Weld toe is the main source of and 3% NaCl solution as welded and toe grinding
fatigue cracks due to stress concentration. In this zone, treatment. The results were indicated that the fatigue
impact of stress concentration increases because of the strength of the toe grinding specimens was greater than
presence of defects; hence, fatigue cracks are ready to the as welded specimens. Knight [7] used burr grinding
initiate and develop. and disc grinding methods for improving the fatigue
Recently, improving the fatigue strength of welded strength of the weldment. He found out that in high
connections using different methods has attracted much cycle, fatigue strength improvement using burr grinding
attention. Most of these methods have been developed in of weld toe is about 100% and for disc grinding is about
years between 1960 until early 1970. In general, weld 40%.
fatigue improvement methods can be divided into two Mohr et al. [8] published the results of applying different
main groups [1]: weld geometry modification methods improvement techniques on cross welded samples. They
and residual stress methods. The former removes weld have proved that fatigue improvement of samples due to
toe defects and/or reduces the stress concentration. The grinding is a factor of 2.2. Data have been collected from
latter introduces a compressive stress field in the area researches in United States of America, Canada,
where cracks are likely to initiate. Many parameters England and Norway. Kirkhope et al. [1], [9] used
influence on the performance of different weld fatigue fatigue life improvement methods in ship structures. The
improvement methods such as material of the results had been provided in two different papers; in the
weldments, loading conditions, type of welding process, first one, published data for improving the fatigue life
welding configuration, and the skill and ability of and applications of fatigue life improvement methods to
operator. To improve the repeatability of the methods ship structure were identified. In the second paper,
and to increase the degree of the efficiency of the suitable techniques for ship structures were considered.
methods in the operating conditions, an inter-laboratory It has been shown that burr grinding, TIG dressing and
test program was investigated by Haagensen [2] for hammer peening treatments for ship structure
International Institute of Welding (IIW) in 1995. applications comprise more potential and the fatigue life
Thirteen labs in ten countries took part in this program was increased approximately 120% for all specimens
preoccupied the three universally methods including treated by the toe grinding. Pang [10] analysed the stress
burr grinding, tungsten insert gas (TIG) dressing and in as-welded and weld toe ground fillet weldments. The
hammer peening. weld toe radii were selected 0.5, 1.0, 2.5 and 5.0 mm.
Clegg et al. [3] studied the impact of seven toe The stress distributions through the plate thickness and
improvements on the fatigue resistance in structural steel the stress concentration factor were assessed and the
plate weldments to appoint which one of the methods results of finite element analysis indicated that the toe
possesses the better effect in modifying the fatigue life. grinding leads to increase in the crack propagation
For each of the improvement method, S-N curves were cycles by up to 53%.
obtained by testing the specimens in the bending Huther et al. [11] studied more than 300 sets of data for
condition. Based on the test results, the treatments burr grinding and TIG dressing treatments applied to
including weld toe grinding, hammer peening and butt joints, T-shape joints and longitudinal joints in a
ultrasonic impact produced the better fatigue resistance. non-corrosive environment. Nguyen and Wahab [12]

© 2018 IAU, Majlesi Branch

Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018 17

presented an analytic model to describe the effect of the produce a straightforward-symmetric groove. Generally,
weld geometry parameters and residual stresses on the the burr grinding should reach to minimum depth of 0.5
fatigue resistance of butt weldments for tensile and mm below any evident undercut [2].
bending loading. They deduced that the fatigue
resistance can be increased considerably by reducing the
radius of weld toe.
In this paper, effect of burr grinding technique on fatigue
strength of steel butt-welded connections is investigated.
This technique is based on removing surface material
from the weld toe, increasing the weld toe radii and
reducing the severity of this detail. St 37 is selected as
the material for the welding plates due to its numerous
applications in manufacturing various structures such as
bridges, steel structures and industrial machine
components. This paper presents the fatigue results
obtained from experimental study for two different
conditions (as-welded and weld toe grinded). In order to
investigate the geometry effect of burrs on fatigue life,
three different burrs are selected. Eventually, SCF
results obtained from simulations (numerical method)
are compared with results of analytical method.

2 BURR GRINDING

The intensity of the stress concentration in weld toe

leads to cracking in weldment. The existence of little
crack-like flaws increases the stress concentration effect
and as a result, flaws extend to depths to a few tenths of
a millimetre. Fatigue cracks easily initiate at these flaws
and slowly propagate. These flaws can be eliminated or
decreased by size using the fatigue life improvement
methods and results in delaying the crack initiation or
extending the fatigue life. In addition, local stress
concentration can be reduced to produce smooth edge
across the plate and the weld face [2].
Today, grinding is the most important way to create a
smooth surface which can be used in various industries
such as aerospace, automotive, transportation, medical
devices and electronics which needs high-level quality.
The primary aim of the grinding is to remove or reduce Fig. 1 The burr grinding technique: (a): Side view and
(b): Top view [2]
size of the weld toe flaws from which fatigue cracks
propagate. At the same time, it aims to reduce the local
stress concentration effect of the weld profile by
smoothly blending the transition between the plate and 3 EXPERIMENTS
the weld face. This process is implemented by mounting
tool bits in grinding devices. The tool bits are normally 3.1. Welding Specimens
burrs which are classified on the basis of materials, Two St-37 plates with dimension of 300 mm in length,
hardness grades, tool shapes and grit sizes. The burr 125 mm in width and 8 mm in thickness were butt
grinding treatment is demonstrated in “Fig. 1ˮ. As it is welded by the Gas Metal Arc Welding (GMAW) method
shown, burr is positioned on the weld toe and the angle using CO2 as protective gas and E7018 electrodes. The
between the tool axis and the horizon should be cross section of weld was selected according to AWS
considered 45-60°. Besides, the angle of tool axis D1.1 [13]. The test and Standard: BS EN 10025-2 (2004)
relative to the direction of movement is selected 30-45° Grade S235 JR (RSt37-2)-1.0038 chemical composition
and the tool is squeezed or dragged along the weld toe. and mechanical properties of the material are given in
The treatment should be conducted so that the tool can Table 1 and Table 2, respectively.

© 2018 IAU, Majlesi Branch

18 Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018

Table 1 The chemical composition of St37 steel

Element Test, % Standard, % R6
C 0.050 Max. 0.190
Mn 0.380 Max. 1.500
P 0.013 Max. 0.045
40 28
S 0.008 Max. 0.045
Cu 0.070 Max. 0.600
79 79

Table 2 The mechanical properties of St37 steel

Ultimate Yield Elongation
strength, MPa strength,Mpa %
Test 1 373 286 41.0 8
Test 2 375 284 42.5
Standard 360-510 >235 >26 250

Fig. 3 Test specimen designed for fatigue test

(dimensions in mm)
3.2. Fatigue Testing
Fatigue testing was performed under constant amplitude 3.3. Weld Toe Burr Grinding
loading by using a 200 KN fatigue testing machine at Burr grinding process was performed by an electric
room temperature and no heating of the specimens grinder with rotational speed 27000 rpm and power of
occurred. Low cycle fatigue tests using the equipment 500 watts. To investigate the effect of burr grinding on
shown in Fig. 2, were implemented. The stress ratio (R) fatigue life of welded joints, Tapered burrs in hardness
and frequency were kept at 0.125 and 3Hz, respectively. grade M with three different radii were selected. This
The final specimen geometry is shown in Fig. 3. grade of burrs is manufactured from a mix of dark red
and white aluminium oxide in a vitrified bond. This grit
combination leads to a good balance between stock
removal rate and tool life. The hardness grade M is the
most universally used bond for steel and cast steel.
Fig. 4 gives an overview of the tapered burrs [14].

d D

Fig. 4 Dimensions of Tapered burr [14]

Grinding process was conducted by using three different

sizes of burrs in order to allow a better comparison. In
weld toe burr grinding only the weld toe is machined to
remove weld toe defects and reduce the weld toe angle
which results in a decrease in the weld toe stress
concentration. Therefore, material is removed to a depth
of 1mm with respect to the thickness of specimen.
Details of burrs used in burr grinding are listed in
Table 3 and in Fig. 5, a view of them is shown. In
Fig. 2 Fatigue testing machine Fig. 6, a grinded specimen is represented.

© 2018 IAU, Majlesi Branch

Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018 19

4 FINITE ELEMENT ANALYSIS

A
4.1. Welding Simulation
Welding simulation is a thermo-mechanical analysis. In
finite element method, there are two distinct approaches:
direct method (coupled) and indirect (uncoupled). In this
paper, the thermal and mechanical analyses are only
sequentially coupled, so the heat effect of the plastic
deformation in the mechanical field is neglected during
the thermal analysis. As a result, first a thermal analysis
is conducted to calculate the transient temperature field
used as input data for the second step. The material
B properties are considered to be temperature dependent
but microstructural phase transformation is not
considered in order to simplify the analysis [15]. In both
temperature dependent thermal and mechanical
analyses, the thermo-physical and mechanical properties
of the base metal were used in the computations. The
material properties of St 37-2 as the function of
temperature, including thermal conductivity, specific
heat, yield stress, elastic modulus, Poisson's ratio and
thermal expansion coefficient were extracted from [16].
C The density is considered 7870 Kg ⁄m3 and temperature
independent.
In this research, The Goldak double-ellipsoid heat
source model was adopted to calculate the volumetric
heat flux distribution as the input heat into the welding
pool. The heat source distribution as shown in Fig. 7,
combines two different ellipses, i.e. one in the front
quadrant of the heat source and the other in the rear
quadrant of it. The heat input is modelled by the user
subroutine DFLUX which is written as FORTRAN
Fig. 5 Burrs used in the burr grinding
code.
The power densities of the double-ellipsoid heat source,
Table 3 Dimensions of burrs used for grinding sampling qf(x,y,z) and qr(x,y,z) describing the heat flux distribution
(dimensions mm) inside the front and rear quadrant of the heat source, can
Case Sd L1 L2 D d be expressed as following [17]:
A 6 18 40 14 2
B 6 25 40 10 5
C 6 40 40 20 6

Fig. 7 The Goldak double-ellipsoid heat source model

Fig. 6 Grinded specimen [17]

© 2018 IAU, Majlesi Branch

20 Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018
𝑥2 𝑦2
−3( ) −3( ) 2 ⁄𝑐 2 ) 𝑄𝑟 = 𝜀𝜎0 (𝑇𝑎 4 − 𝑇04 ) (6)
𝑞𝑓 = (6√3𝑄𝑓𝑓 ⁄𝑎𝑏𝑐𝑓 𝜋√𝜋)𝑒 𝑎2 𝑒 𝑏2 𝑒 −3(𝑧 𝑓 (1)

𝑥2 𝑦2
Where ε, emissivity coefficient, is assumed 0.77 and σ0
𝑞𝑟 = (6√3𝑄𝑓𝑟 ⁄𝑎𝑏𝑐𝑟
−3( ) −3( ) 2 2
𝜋√𝜋)𝑒 𝑎2 𝑒 𝑏2 𝑒 −3(𝑧 ⁄𝑐𝑟 ) (2) is the Stefan-Boltzmann constant (=5.670 × 10-8 W⁄
(m2K4)) [8].
In mentioned equations, a is the weld width along the
tangent direction x, b is the weld penetration depth along 4.2. The Weld Joint Geometry and Welding
the arc direction y, cf and cr are the front and rear weld Parameters
pool length in the weld path direction z, and ff and fr are Two St-37 plates with dimension of 300 mm in length,
some dimensionless factors. In order to calculate a and 125 mm in width and 8 mm in thickness were butt
b, the cross section of weld was sanded using different welded. It was assumed that the base metal and the weld
grades of sandpapers in metallographic laboratory to metal have the same thermal and mechanical properties.
observe its elliptical distribution. Two parameters cf and Fig. 8 illustrates the mesh configuration of the weld plate
cr were determined using the following relationships in the thermal and mechanical analyses. Due to high
[17]: temperature and the stress gradient near the weld line, a
relatively finer mesh was provided and the element size
𝑐𝑓 ∈ (0.5𝑎, 𝑎) & 𝑐𝑟 ∈ (2𝑎, 4𝑎) (3) was increased progressively with distance from the weld
line. The element birth and death technique was used to
simulate the weld passes and the filler metal deposition
Two dimensionless factors ff and fr were obtained from
into the weld pool. In thermal analysis, 5200 elements of
the following equations [17]:
DC3D8 (eight-node hexahedral elements) were used for
3D meshing and the same number of elements of C3D8R
𝑓𝑓 = 2𝑐𝑓 ⁄(𝑐𝑓 + 𝑐𝑟 ) & 𝑓𝑟 = 2𝑐𝑟 ⁄(𝑐𝑓 + 𝑐𝑟 ) (4)
were employed in mechanical analysis. Welding
parameters are given in Table 5.
The calculated parameters of the heat flux distribution
are listed in Table 4.
Welding direction
Table 4 The parameters of the heat flux distribution
(dimensions include a, b, cf and cr are mm)
Parameter a b cf cr ff fr
Value 6 5 6 18 0.5 1.5

Two kinds of the boundary conditions including

convection and radiation were considered in the thermal
analysis [11]:

(a)- Heat transfer by convection: Heat transfer by

convection occurs in liquids or gases caused by
movement of molecules. The Newton’s law describes
convection with [17]:
Fig. 8 The mesh configuration of the weld plate
𝑄𝑐 = ℎ𝑐 (𝑇𝑎 − 𝑇0 ) (5)
Table 5 The parameters of welding
Where Ta is the surface temperature of plate; T0 is the Parameter/ Current Voltage Speed efficiency
ambience temperature; and hc is the heat transfer Pass (A) (V) (mm/s) (%)
coefficient due to convection. The heat transfer First 180 20 3.0 70
coefficient is dependent on the flow conditions, the Second 180 20 1.5 70
surface and temperature differences between the
material and the surroundings and was assumed 15 W⁄ 4.3. Burr Grinding Simulation
(m2K) [17]. Three different parameters including weld toe radius,
grinding depth, and flank angle influence burr grinding
(b)- Heat transfer by radiation: According to the analysis. In this paper, the effect of weld toe radius has
Stephan-Boltzmann’s law the heat flow density of a gray been investigated and grinding depth and flank angle
body caused by radiation is proportional to the fourth were assumed 1 mm and 45°, respectively. In burr
power of the temperature differences to its surroundings grinding simulation, just a small fillet at the intersection
[17]: of weld and base metal was developed resulting in

© 2018 IAU, Majlesi Branch

Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018 21

reduction of stress concentration factor. Fig. 9 illustrates 4.4. Calculation of Stress Concentration Factor
weld toe profile after burr grinding. In this paper, the purpose of simulations is finding stress
concentration factor by which the effect of burr grinding
is investigated. To find out stress concentration factor,
both analytical and numerical methods have been
employed. In numerical method, after finding stress
Flank angle distribution using finite element method, stress
Weld toe radius concentration factor is defined as a ratio between local
Grinding depth maximum stress in the longitudinal direction and
nominal stress [6]:

𝐾𝑡 = 𝜎𝑚𝑎𝑥 (𝑙𝑜𝑛𝑔𝑖𝑡𝑢𝑑𝑖𝑛𝑎𝑙)/𝜎𝑛𝑜𝑚𝑖𝑛𝑎𝑙 (7)

Where Kt is the stress concentration factor.

In order to calculate Kt, the stress analysis including two
Fig. 9 Weld toe profile after burr grinding [10]
steps was performed using finite element method. At the
first step, stress field due to burr grinding was defined
Since this process is carried out at a temperature slightly and then at the second step, axial loading was subjected
higher than ambient temperature and amount of material to the end of one side, while displacement at the other
removal is also very low, creating a new model was not side was considered zero. This analysis was carried out
required and a new mechanical analysis was submitted before and after burr grinding and for each case stress
by changing geometry of the same model. Therefore, concentration factor was obtained.
residual stress field due to welding of plates was In analytical method, formulae proposed by Gill et al.
transferred to the grinding model which is the fatigue [18] for prediction of stress concentration factor of butt
test sample and then geometrical changes were applied. welded joints were used. As it can be seen, the stress
Fig. 10 shows the grinding model. Simulation of burr concentration factor is obtained by multiplying the
grinding was carried out at three different weld toe radii parameters of weld geometry:
and stress distribution of each was obtained. Different
modes of grinding simulation are represented in Table 6. 𝐹(𝐻 ⁄𝑇) = 1 − 𝑒𝑥𝑝[−10.4 × (𝐻 ⁄𝑇 )0.7 ] (8)
. 3]
𝐹(W⁄𝑇) = 1 − exp[−4.92 × (W⁄𝑇) (9)

{1 − exp[−4.10 × (𝑅 ⁄𝑇)0. × (𝜋⁄180) × 𝜃]}

𝐹(𝜃) =
{1 − exp[−4.10 × (𝑅 ⁄𝑇)0. × (𝜋⁄2)]}
(10)
𝐾𝑡 = 1 + 0.322 × 𝐹(𝐻 ⁄𝑇 ) × 𝐹(𝑊 ⁄𝑇 ) × 𝐹(𝜃)
× (𝑇⁄𝑅 )0.5
(11)

Where H is height of weld bead, T is plate thickness, W

is width of weld bead, R is weld toe radius, θ is flank
angle, and Kt is stress concentration factor.
Fig. 10 Grinding model

Table 6 Different modes of grinding simulation 5 RESULTS AND DISCUSSION

Case Weld toe radius Grinding Flank Angle
(mm) depth (degree) 5.1. Results of Welding Simulation
(mm) Fig. 11 shows the temperature field of welded plates
A 1.0 1.0 45 after 100s from welding start in second pass, when the
B 2.5 1.0 45 heat source (electrical arc) was in the middle of the plate
C 3.0 1.0 45 at a distance of 150mm from the edge of plate. Elliptical
distribution of Goldak heat source can be seen in Fig. 11.

© 2018 IAU, Majlesi Branch

22 Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018

180

160

Longitudinal residual stress(MPa)

140

120

100

80

60

40

20

0
0 50 100 150 200 250 300 350

Along the weld(mm)

(a)
Fig. 11 Temperature field of 100s after the start of welding
60
in second pass
40
Temperature history for a node in welding direction at a
Longitudinal residual stress(MPa)
distance of 150mm from the edge of plate is shown in 20
Fig. 12. It was evident from Fig. 11 that the temperature
around the torch reached 1750°C. The thermal 0
distribution attained via the Goldak thermal model,
indicated that the highest temperature was achieved at -20

the torch location. Then, the temperature gradually drops

as the cooling starts and finally reaches the steady state. -40

-60
0 20 40 60 80 100 120 140

Distance from weld line(mm)

2000
(b)
1800
Fig. 13 Longitudinal residual stress distribution along the:
1600 (a): welding direction and (b): transverse direction
1400
5.2. Results of Weld Toe Burr Grinding Simulation
Temperature(°C)

1200
Burr grinding method was simulated in ABAQUS by
1000
applying geometrical changes at weld toe region. For
800 this purpose, three different cases of burr grinding with
600 three different weld toe radius were analysed. The major
400
aim of simulations is calculating stress concentration
factor before and after burr grinding. Stress
200
concentration factor for longitudinal stress and in
0
0 500 1000 1500 2000 2500 3000
transverse direction was selected as a criterion for
comparison. Longitudinal residual stress before and
Time(s)
after grinding has been shown in Fig. 14.
Fig. 14 also indicates that distance between weld line
Fig. 12 Temperature history at z=150 and peak of each graph in which maximum longitudinal
residual stress occurs, increases by applying burr
Fig. 13 shows residual stress distribution along the grinding. Thus, it is possible to conclude that weld toe
transverse direction and welding direction. In transverse grinding which increases the weld toe radii, decreases
direction, residual stress around the weld line is tensile the local stress levels leading to reduction of stress
due to the thermal gradient of welding and as the concentration factor and an improved fatigue life. Stress
distance from weld line to edge of the plate increases, it concentration factors before and after grinding are
changes to compressive residual stress. shown in Fig. 15.

© 2018 IAU, Majlesi Branch

Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018 23

60 Table 7 Weld bead geometry and stress concentration factor

Before grinding with respect to weld parameters
A
40 B Parameter/ F(H/T) F(W/T) F(θ) Kt
Longitudinal residual stress(MPa)

C
Case
20
Before
0.980 0.980 0.964 2.790
grinding
0
A 0.980 0.980 0.977 1.850
-20 B 0.980 0.980 0.983 1.540
C 0.980 0.980 0.984 1.490
-40

Table 8 Comparison between results of analytical and

-60
0 20 40 60 80 100 120 140
numerical methods
Distance from weld line(mm) Kt Finite Difference
Kt Analytical
Fig. 14 Longitudinal residual stress in transverse direction Case element percentage
method
before and after grinding method
Before
2.63 2.79 5.73
3.0
grinding
A 1.73 1.85 6.48
2.5 B 1.51 1.54 1.94
C 1.48 1.49 0.67
Stress concentration factor

2.0

As it can be observed from Table 8, stress concentration

1.5
factor results of both methods are in good agreement
with each other and the difference between them is
1.0
acceptable. Moreover, it can be noticed that in both
0.5
methods increasing of weld toe radius results in
reduction of stress concentration factor.
0.0
before burr grinding A B C 5.4. Experimental Results
Modelling cases based on weld toe radius The fatigue cycles of samples were obtained according
Fig. 15 Stress concentration factors before and after to fatigue tests results. Overall, two samples before and
grinding
five samples after burr grinding were tested. After burr
grinding, fatigue test was repeated for each radius once.
After modelling weld toe burr grinding with different
Table 9 shows the results of fatigue tests.
weld toe radii, a significant reduction of stress
concentration factor was observed. This suggests that
Table 9 Fatigue tests results
increasing weld toe radius and smoothing the transition
Before or after Fatigue life
area between weld bead and base plate leads to reduction Case
grinding (cycles)
of stress concentration factor. Before grinding Test 28357
Before grinding Re- test 30159
5.3. Analytical Results of Stress Concentration
After grinding A 30875
Factor
After grinding B 34356
In this section, analytical method has been applied in
order to obtain stress concentration factor. Therefore, After grinding Re-B 36544
equation (11) has been used to calculate stress After grinding C 39076
concentration factor before and after grinding and Table After grinding Re-C 42710
7 shows weld bead geometry and stress concentration
factor with respect to weld parameters for all modelling As it is noticed from Table 9, burr grinding with
cases in which dimensions are H=2 mm, T=8 mm, different radii leads to improvement of fatigue life of
W=6.7 mm, R=0.22 mm and θ=61.4º. In order to provide welded samples which is due to weld geometry
a better comparison between analytical and numerical improvement by burr grinding. In burr grinding for cases
results, Table 8 has been prepared based on stress A, B and C, radius of 1, 2.5 and 3mm was selected,
concentration factors obtained from both methods respectively. It is evident that by increasing radius of
before and after burr grinding and the difference burr grinding, a better improvement in fatigue life is
percentage between two methods was calculated. obtained. This indicates that by increasing the weld toe

© 2018 IAU, Majlesi Branch

24 Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018

radii and consequently reducing the severity of the  It is possible to conclude that weld toe grinding
detail, stress concentration factor decreases. For a better which increases the weld toe radii, decreases the local
comparison, the experimental results are plotted in Fig. stress levels leading to reduction of stress concentration
16. Table 10 shows percentage of fatigue life factor and an improved fatigue life.
improvement for specimens after burr grinding  By comparing numerical and analytical
compared to the test and the re-test results of as-welded methods, it is observed that stress concentration factor
specimens. results of both methods are in good agreement with each
other and the difference between them is acceptable.
44000  In numerical method, the best result is obtained
42000
for a weld toe radius of 3 mm which leads to a decrease
of 43.72% in stress concentration factor compared to
40000
before grinding.
Number of failure cycles

38000  In analytical method, the best result is obtained

36000
for a weld toe radius of 3 mm which leads to a decrease
of 46.59% in stress concentration factor compared to
34000 before grinding.
32000  Experimental results have proved that weld toe
burr grinding treatment increases significantly the
30000
fatigue life of welded specimens.
28000  A better improvement in fatigue life of welded
before burr grinding A B repeat of B C repeat of C
specimens is observed by increasing the radius of burrs.
Specimens
Fig. 16 Number of failure cycles versus different cases of  In experimental study, the best result belongs to
burr grinding grinding with a 3 mm tapered burr (Case C) leading to
an improvement of 50.61% in fatigue life while grinding
Table 10 Fatigue improvement percentage of specimens due with 1 mm tapered burr (Case A) gives the worst result
to burr grinding relative to the test and the re-test results of of 8.88% improvement in fatigue life.
as-welded specimens
Fatigue improvement
Relative to Case
percentage
The test A 8.9 7 NOMENCLATURE
The test B 21.2
Q Total heat input rate (W)
The test Re-B 28.9
A Weld width (mm)
The test C 37.8 B Weld penetration depth (mm)
The test Re-C 50.6 H Height of weld bead (mm)
The re-test A 2.4 T Plate thickness (mm)
The re-test B 13.9 W Width of weld bead (mm)
The re-test Re-B 21.2 R Weld toe radius (mm)
The re-test C 29.6 qf Heat flux distribution inside the front quadrant
The re-test Re-C 41.6 of the heat source (W⁄m)
qr Heat flux distribution inside the rear quadrant
of the heat source (W⁄m)
ff Fraction of heat in the front quadrant
6 CONCLUSIONS fr Fraction of heat in the rear quadrant
cf Front weld pool length (mm)
Burr grinding treatment has been investigated as fatigue
cr Rear weld pool length (mm)
strength improvement method. In order to validate the
Qc Heat transfer by convection (W⁄m2)
improvement effect on fatigue strength, fatigue tests
hc Convection coefficient (W⁄(m2K)
were carried out. The fatigue and finite element analyses
Ta Surface temperature of plate (K)
results suggested that weld toe treatment by burr
T0 Ambient temperature (K)
grinding can improve the fatigue life. The main findings
Qr Heat transfer by Radiation (W⁄m2)
are summarized as follows:
Kt Stress concentration factor
 By comparing the simulated weld toe grinding
Ε Emissivity coefficient
specimens with the simulated as-welded specimens, it is σ0 Stefan-Boltzmann constant (W⁄(m2K4)
clear that weld toe grinding treatment is very successful θ Flank angle (degree)
and leads to reduction of stress concentration factor.

© 2018 IAU, Majlesi Branch

Int J Advanced Design and Manufacturing Technology, Vol. 11/ No. 2/ June - 2018 25

Techniques: Part 2- Application to Ship Structures,

REFERENCES
Marine Structures, Vol. 12, No. 6, 1999, pp. 477-
496.
[1] Kirkhope, K. J., Bell, R., Caron, L., Basu, R. I. and [10] Pang, H. L. J., Analysis of Weld Toe Radius Effects
Ma, K. T., Weld Detail Fatigue Life Improvement
on Fatigue Weld Toe Cracks, International Journal
Techniques: Part 1- Review, Marine Structures, Vol. of Pressure Vessel and Piping, Vol. 58, No. 2, 1994,
12, No. 6, 1999, pp. 447- 474. pp. 171-177.
[2] Haagensen, P. J., Maddox, S. J., Recommendations
[11] Huther, I., et al., The Influence of Improvement
on Methods for Improving the Fatigue Strength of Techniques on Welded Joint Fatigue Strength, IIW
Welded Joints, International Institute of Welding, Document XIII-1562-94, 1994.
Commission XIII-1815-00, 2013.
[12] Nguyen, T. N., Wahab, M. A., The Effect of Weld
[3] Clegg, R. E., Mcleod, A. J. and Ruddell, W., Effect Geometry and Residual Stresses on the Fatigue of
of Toe Treatment on the Fatigue Resistance of Welded Joints under Combined Loading, Journal of
Structural Steel Welds, Australasian Welding
Materials Processing Technology, Vol. 77, No. 1-3,
Journal, Vol. 58, No.3, 2013, pp. 34-41. 1998, pp. 201-208.
[4] Pedersen, M. M., Mouritsen, O. O., Hansen, M. R., [13] American Welding Society AWS-D1 Committee on
Andersen, J. G. and Wenderby, J., Comparison of
Structural Welding, Structural Welding Code- Steel,
Post Weld Treatment of High Strength Steel Welded 22th ed. American National Standards Institute,
Joints in Medium Cycle Fatigue, International 2010.
Institute of Welding, XIII-2272-09, 2010.
[14] PFERD Tool Selection Manual, 22th ed., Section
[5] Tai, M., Miki, C., Improvement Effects of Fatigue 203, Germany, August Rüggeberg GmbH & Co.
Strength by Burr Grinding and Hammer Peening KG, 2014.
under Variable Amplitude Loading, Welding in the
[15] Fanous, I. F. Z., Younan, M. and Wifi, A., 3D Finite
world, Vol. 56, No. 7, 2012, pp. 109-117. Element Modeling of the Welding Process Using
[6] Baptista, R., Infante, V., and Branco, C. M., Study of Element Birth and Element Movement Techniques,
the Fatigue Behavior in Welded Joints of Stainless
International Journal of Pressure Vessel Technology,
Steels Treated by Weld Toe Grinding and Subjected Vol. 125, No. 2, 2003, pp. 144-150.
to Salt Water Corrosion, International Journal of [16] Barsoum, Z., Barsoum, I., Residual Stress Effects on
Fatigue, Vol. 30, No. 3, 2008, pp. 453-462.
Fatigue Life of Welded Structures Using LEFM,
[7] Knight, J. W., Improving the Fatigue Strength of Engineering Failure Analysis, Vol. 16, No. 1, 2009,
Fillet Welded Joints by Grinding and Peening, pp. 449-467.
Welding Research International, Vol. 9, No. 6, 1978,
[17] Goldak, J., Akhlaghi, M., Computational Welding
pp. 519-540. Mechanics, New York, Springer, 2005.
[8] Mohr, W. C., Tsai, C. and Tso, C. M., Fatigue [18] Gill, J., Singh, J., Effect of Welding Speed and Heat
Strength of Welds with Profile and Post-Weld
Input Rate on Stress Concentration Factor of Butt
Improvements, Proceedings of fourth International Welded Joint, International Journal of Advanced
Conference Offshore Mechanics and Arctic Engineering Research and Studies 2012, Vol. 1, No.
Engineering, Copenhagen, 1995.
3, 2012, pp. 98-100.
[9] Kirkhope, K. J., Bell, R., Caron, L., Basu, R. I. and
Ma, K. T., Weld Detail Fatigue Life Improvement

© 2018 IAU, Majlesi Branch