Australian Journal of Basic and Applied Sciences, 5(12): 937-945, 2011

ISSN 1991-8178
Corresponding Author: N.A. Akeel, Department of Mechanical & Materials Engineering, Faculty of Engineering and Built
Environment Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia.
E-mail: [email protected]
937
Analysis of Rolling Contact Fatigue Damage Initiation At The Wheel-Rail Interface

N.A. Akeel, Z. Sajuri and A.K. Ariffin

Department of Mechanical & Materials Engineering, Faculty of Engineering and Built Environment
Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia.

Abstract: Rolling contact fatigue (RCF) results in damages on the railhead surface and wheel tread.
The current paper presents the analysis of RCF damage initiation and stress distribution at the wheel-
rail interface at different directions. A three-dimensional elastic frictional finite element model of the
wheel-rail interaction is used to investigate the effect of the applied contact loading force at the
straight, transition, and curved areas of the wheel tread and railhead surface. The interaction between
the left and right wheels is considered. The interface exhibits small damage problems that are solved
via the finite element method (FEM) software code ANSYS 11. The half-space assumption of the
Hertz method is avoided by FEM. The real geometry and the boundary condition of the wheel-rail
interface are accurately shown by the proposed simulation. The numerical results indicate the fatigue
life and equivalent stress at the straight, transition, and curved areas of the rail track. Significant effect
damages from the contact force on the wheel tread and the curve radius of rail track are observed at the
interface.

Key words: Curved track; fatigue life; finite element method; wheel-rail.

INTRODUCTION

The railway track in Malaysia has shown the necessity of an efficient management of railway systems. The
design is aimed toward reducing costs and increasing safety, as well as reliability, of the railway systems. The
wheel-rail interface is one of the most crucial points that must be checked to determine the performance of a
train and consider its safety. However, the calculation of the stress at the wheel-rail interface is dependent on the
static and dynamic loading. The damage in this area is due to the repeated cyclic loading of the wheel. The
damages that affect the mechanisms of fatigue are some of the concerns for the life prediction for a railway
track. Fatigue causes abrupt fractures in the railhead and wheel tread. These failures may cause damage to rails
because of the stress caused by the contact force as stated in the Hertz contact theory (Hosseini Tehrani, 2009).
The stress distribution and fatigue initiation at the wheel-rail interface are calculated by numerical and
analytical methods. The problems that satisfied the assumptions in Hertz contact theory and some practical
problems that do not meet the required assumptions in the aforementioned theory are elucidated in the current
research. The interface is a small area compared with the body dimension and surface curvature of the wheel-
rail contact region (Ringsberg, J.W. 2001; Liu, Y., 2006). A semi-analytical approach is developed for stress
calculation using three-dimensional (3D) finite element analysis but the contact force is considered based on the
Hertz theory. Fine meshing technique at the contact region is used to ensure an efficient and accurate
computation. The stress response of the numerical analysis of the wheel rolling motion on running railhead
surface is used for fatigue life prediction and the equivalent stress (von mises) criterion.
In the wheel-rail rolling contact problem, the interface is near the railhead surface and wheel tread. The
finite element method (FEM) for contact analysis is used to solve the model of such problems (Sraml, M., 2003;
Kyung Su Kim, 2005). The Hertz contact theory is used to calculate the stress response and fatigue damage
problems. The principle stress-strain component in one direction is used for rolling contact fatigue analysis (Liu,
Y., 2005). The two-dimensional (2D) finite element model is used to analyze their proposed model, and a
multiaxial fatigue model is developed for bearing rolling contact fatigue analysis (Howell, M., 1995; Guo, Y.B.,
2004; Fatemi, A., 1988). In the current research, a 3D simulation finite element model for wheel-rail rolling
contact analysis is developed. However, the criterion for multi-axial fatigue of ductile material behavior is
considered in the models. The influences of several parameters are evaluated, and conclusions based on the
present results are obtained. The Goodman approach is used when the mean stress effects are considered. The
method developed in the current study may be used for fatigue life design and service planning of a railroad.

Material Properties:
The material used is a used steel rail track provided by Permanent Way Division, KTMB. The chemical
composition of the rail track material is examined, and the maximum composition of the material
(Cr+Mo+Ni+Cu+V) is limited to be less than 0.35% (Table 1). The mechanical properties of the rail track
material are shown in Table 2. These properties are compared with the standard requirement for rail track
Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
938
material set by KTMB. The tensile properties of the tested rail material complied with the standard property
requirements. Bending strength is not included as one of the parameters to be considered in the standard
property requirements for rail track material set by KTMB. However, this property is one of the most important
parameters in determining crack propagation resistance or fracture resistance of the rail material under Mode I
loading.

Table 1: Chemical compositions of rail track material.
C Mn Si P S Cr+Mo+Ni+Cu+V
0.60-0.80 0.83-1.30 0.10-0.50 <0.05 <0.05 <0.35

Table 2: Mechanical properties of the rail material.
Yield stress (MPa) Tensile strength (MPa) Youngs modulus (GPa) Poissons ratio
825 923 200 0.3

Finite Element Method (FEM):
A complete FEM model composed of two wheels and a piece of two rails track is used to analyze the stress
distribution at different curve radii. The simulation is a set of the element size and the quasi-static loads on the
wheel created with a library of ANSYS 11. Force is applied on the wheel which has contact with the rail. Here,
the rail and the wheel tread are defined as two separate profiles. The shape of wheel and rail profiles generated
with a standard mathematical model is described by several keypoints. The Goodman approach in ANSYS 11
via the FEM is used in all simulations to predict the mean stress effects. The detail drawing the wheel and rail
contact model is presented in the previous work (Akeel, N.A., 2011). The 3D geometric model of the wheel is
generated by revolving the 3D curves that describe the profile of the wheel tread. The rail model is created by
extruding railhead profile curves a distance of 600 mm, which is the distance between the sleepers. The two sets
of curves for the new wheels and rails created by the solid model are shown in Figure 1.



Fig. 1: Solid model of wheel/rail contact.

Fine mesh is used in the finite element of the solid model (Figure 2). To predict a reasonable model
configuration, the wheel is spatially oriented relative to the rail according to the quasi-static state calculated by
FEM. A geometric shape, domains, a material model, a value for the coefficient of friction, and knowledge
about the contact forces are required to produce a model of the wheel-rail interaction (Sladkowski, A., 2005).
The material model is based on the stress-strain curves obtained in the previous work (Akeel, N.A., 2011).



Fig. 2: Finite element modeling of wheel/rail contact.


Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
939
Condition Of Wheel-Rail Interface:
The different rail track position (straight, transition, and curved areas) are simulated in the model of a
railway track in Malaysia. The radius of the curved area and the different degree of curvature with every radius
of the curve are shown in Table 3. However, at the transition area with = 25
o
, the right wheel is higher than the
left wheel during motion when the rail track condition shifts from the straight to the curved area (Figure 3).
Therefore, the transition starts from a straight area and changes to the curved area. The curved area is defined as
a shift from the transition area to a straight area as shown Figure 4.

Table 3: Relationship with curve radius and Degree of curvature.
Curve Radius (m) Degree of curvature
100
200
300
400
500
600
700
800
900
1000
12
10
o

8
o

6
o

4
o

4
o

4
o

4
o

4
o

4
o






Fig. 3: Rail track at transition area.



Fig. 4: Crack position on straight, transition and curve radius of rail track.

Numerical Analysis Method:
Crack Initiation Life Damage:
The fatigue crack initiation life prediction and stress distribution under different vertical loading is
calculated using the wheel-rail RCF model. The contact force is assumed at the maximum value based on the
over loading of passengers on the train. The design load of the model is applied on the wheel tread on the
running surface of railhead (Makoto, A.K.A.M.A., 2007). Damage tolerance on the wheel-rail interface is
defined by the following equation:

( )
max
2
FP J
c
o t
A
= + A A
(1)
The constant J is the maximum stress normal to the crack plane,
c A
is the strain range normal to the crack
plane,
t A
is the shear stress range on the crack plane,
A
is the shear strain range on the crack plane, and
FP
max
is the fatigue parameter value defined as the crack plane. The load is chosen as the contact force loading
on the wheel and rail contact areas. The equation for the fatigue crack initiation for the crack plane is as follows
(Kapoor, A., 1994):
Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
940

( )
max
max
max
2
FP J
c
o t
A
= + A A

( )
( ) ( )
2
2 '
2 ' ' 2
b
f f f f
N N b c
E
o
o c = + +
(2)

The crack initiation life is calculated to determine the number of cycles of fatigue life as follows:

r c f
N c c A = /
(3)
where
c
c
is a material constant determined experimentally, and
r
c A
is the equivalent ratcheting strain per
cycle and calculated as

) ( )
(
2
/ 3
r
c c A A + A

(4)

c A

and
A

are calculated as the incremental normal and shear strain acting on the crack plane per load
cycle. The load is the force applied on the wheel tread on the running surface of the railhead at the contact
region. Therefore, this equivalent ratcheting strain per cycle can be used to investigate the crack initiation
corresponding to the material response elucidated through the FEM analysis.

Curve Radius of Rail Track:
The curve radius of the railway track is defined as the radius of the approximating circle. The curve radius
is enlarged so the formula for the curve radius R for a horizontal curve can be determined by obtaining the
intended design velocity V, the coefficient of friction, and the allowed superelevation on the curve radius of rail
track.
( )
2
s
R
g e f
u
=
+
(5)

Using this radius, practitioners can determine the degree of curve to check whether it falls within acceptable
standards. The degree of curve, D
a
, can be computed through the following formula:

1746
% %
a
R
D
=
(6)
where
a
D
is the degree of curve along the horizontal curve radius.
The horizontal curve is found at locations where two roadways intersect, providing a gradual transition
between the two paths. The intersection point of the two roads is defined as the Point of Tangent Intersection
(PI). The location of the starting point of the curve is defined as the Point of Curve (PC), whereas that of the
end point is defined as the Point of Tangent (PT). Both PC and PT are at a distance T from PI, where T is
defined as the Tangent Length and can be calculated by determining the central angle of the curve, in degrees.
This angle is equal to the supplement of the interior angle between the two road tangents.

tan
2
T R
A | |
=
|
\ .
(7)
where T is the tangent length (in length units), is the central angle of the curve (in degrees), and R is the curve
radius (in length units).
The distance between PI and the vertex of the curve can be easily calculated using the property of right
triangles with T and R. Considering this distance and subtracting the curve radius R, the external distance E, the
smallest distance between the curve and PI, can be obtained.

1
1
cos
2
E R
| |
|
| =
A | | |
| |
\ . \ .
(8)

Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
941
where E is the external distance (in length units). Similarly, the middle ordinate M can be found and is the
maximum distance between the line drawn between PC, PT, and the curve. This distance falls along the line
between the vertex of the curve and PI.


1 cos
2
M R
| A | | |
=
| |
\ . \ .
(9)

where M is the middle ordinate (in length units), and L is the curve length that can be determined using the
formula for semicircle length.


180
R
L
t A
=
(10)

Similarly, the geometric formula for cord length can be used to calculate C, which represents the cord
length for this curve.


2 sin
2
C R
A | |
=
|
\ .
(11)
RESULTS AND DISCUSSION

The wheel-rail contact model predicts that the maximum value of the von mises stress decreases, in contrast
to that of the fatigue life (Figure 5). However, the rolling contact damage life initiation increases at the straight
area, in contrast to its trends at the transition and curved areas in the wheel-rail contact model. The contour plots
the von mises stresses of static and dynamic loading on the wheel tread and railhead surface. The results
describe the deformation of the stress distribution at the transition and curved areas of wheel-rail interface as the
results of the different curve radii of the rail track as shown in Figures 611.



Fig. 5: Equivalent stresses (von mises) at straight track.

The maximum rolling contact damage at the wheel-rail interface occurs at the corner gauge of the rail track.
The fatigue life of the wheel-rail interface increases as the radius curve increases. However, the value of the
contact loading force increases, in contrast to the fatigue life initiation, at the wheel-rail interface. The effect of
curve radius and angles in wheel-rail contact model on the maximum fatigue life initiation at the right rail track
is plotted in Figure 12.
The maximum value of the von mises stress of the wheel-rail interface decreases with different contact
loading forces from 80 KN to 120 KN and as the sharp radius curve of the rail track increases. The distribution
of the von mises stress in the right rail is higher than that in the left rail with the curve radius of rail track. The
effect of the different curve radii and angles in wheel-rail contact model on the maximum von mises stress in the
right rail is plotted in Figure 13.

Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
942


(a) Front section (b) Left section

Fig. 6: Equivalent stresses (von mises) at transition area.



(a) Front section (b) Left section

Fig. 7: Equivalent stresses (von mises) at curve radius track 100mm.



(a) Front section (b) Left section

Fig. 8: Equivalent stresses (von mises) at curve radius track 200mm.


Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
943


(a) Front section (b) Left section

Fig. 9: Equivalent stresses (von mises) at curve radius track 300mm.



(a) Front section (b) Left section

Fig, 10: Equivalent stresses (von mises) at curve radius track 400mm.



(a) Front section (b) Left section

Fig. 11: Equivalent stresses (von mises) at curve radius track 500mm.


Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
944

Fig. 12: The fatigue life prediction of different area.


Fig. 13: The equivalent stresses (von mises) at different area.

Table 4: The results of stress and fatigue life at straight and transition area.
Load
(KN)
Von mises stress (MPa) Fatigue life (cycle)
Straight area Transition area Straight area Transition area
80
90
100
110
120
235
265
294
324
353
248
279
310
341
372
14314
9604
6934
5165
3947
11964
8169
5898
4393
3357

The fatigue life and von mises stress in wheel-rail contact model at the straight and transition areas of the
rail track are shown in Table 4. However, the fatigue life increases at the straight area in contrast to its behavior
at the transition area. The von mises stress in the wheel-rail contact model increases at the transition area in
contrast to its behavior at the straight area of the rail track. The difference in the contact loading force and the
angle of transition area is approximately
u
=2.5. However, given that the straight area is a flat surface, thus,
u
=0. Therefore, the maximum values of von mises stress at the transition area at the right rail track changes from
the straight to the curved area.

Conclusion:
The effect of the von mises stress and the fatigue damage life initiation is demonstrated by the wheel-rail
contact model. The results of the investigations in the current study can be summarized as follows:
0 200 400 600 800 1000 1200
10
2
10
3
10
4
10
5
Curve radius (m)
F
a
t
i
g
u
e

l
i
f
e

(
C
y
l
c
e
s
)


80KN
90KN
100KN
110KN
120KN
0 200 400 600 800 1000 1200
0
100
200
300
400
500
600
700
800
Curve radius (m)
V
o
n

m
i
s
e
s

s
t
r
e
s
s

(
M
P
a
)


80KN
90KN
100KN
110KN
120KN
Aust. J. Basic & Appl. Sci., 5(12): 937-945, 2011
945
1. An increasing contact loading force and decreasing curve radius results in more damage in the wheel-rail
interface.
2. The von mises stress decreases at the straight area in the wheel-rail contact model compared with those at
the transition and curved areas of the rail track.
3. The maximum value of the von mises stress is 660 MPa at the wheel-rail interface with an approximately
100 m curve radius as a result from an applied force load of approximately 120 KN.
4. The effect of fatigue damage life initiation increases at the straight area but decreases at the transition and
curved areas in the wheel-rail contact model.
5. The maximum value of fatigue life prediction at the straight area is approximately 14314 cycles.

ACKNOWLEDGMENTS

The authors would like to thank Keretapi Tanah Melayu Berhad (KTMB) for supplying the rail track
materials and to Ministry of Science, Technology and Innovation (MOSTI) for sponsoring this work.

REFERENCES

Akeel, N.A., M.A. Aziman, Zainuddin Sajuri, Ahmad Kamal Ariffin, A.W. Ikhsan, 2011. Identification of
Damages and Stress Analysis of Rail/Wheel Rolling Contact Region. Key Engineering Materials, pp: 462-463.
Fatemi, A., D.F. Socie, 1988. Multiaxial fatigue damage including out-of-phase loading. Fatigue Fract Eng
Mater Struct., 11: 149-65.
Guo, Y.B., M.E. Barkey, 2004. Modeling of rolling contact fatigue for hard-machined components with
process induced residual stress. Int. J. Fatigue, 26(6): 605-13.
Hosseini Tehrani, P., M. Saket, 2009. Fatigue crack initiation life prediction of railroad, Journal of Physics:
Conference Series 181. 012038, doi:10.1088/1742-6596/181/1/012038.
Howell, M., G.T. Hahn, C.A. Rubin, D.L. McDowell, 1995. Finite element analysis of rolling contact for
nonlinear kinematic hardening bearing steel. ASME J Tribol., 117: 729-36.
http://www.wikibooks.org/Fundamentals of Transportation/ Horizontal Curves. on 26 May 2010.
Kapoor, A., 1994. ''A re-evaluation of the life to rupture of ductile metals by cyclic plastic strain'', Fatigue
and Fracture of Engineering Materials and Structures, 17: 201-219.
Kyung Su Kim, Dong In Cho, Jae Wook Ahn, Seung Bok Choi, 2005. An Experimental Study on Fatigue
Crack Propagation under Cyclic Loading with Multiple Overloads. Key Engineering Materials, 297-300, pp:
2495-2500.
Liu, Y., B. Stratman, S. Mahadevan, 2006. Fatigue crack initiation life prediction of railroad wheels. Int J
Fatigue, 28(7): 747-56.
Liu, Y., S. Mahadevan, 2005. Multiaxial high-cycle fatigue criterion and life prediction for metals. Int J
Fatigue, 7(7): 790-800.
Makoto, A.K.A.M.A., 2007. Development of Finite Element Model for Analysis of Rolling Contact
Fatigue Cracks in Wheel/Rail Systems. Quarterly Report of RTRI, 48(1): 8-14.
Ringsberg, J.W., 2001. Life Prediction of Rolling Contact Fatigue Crack Initiation, Int. J. Fatigue, 23:
575-586.
Sladkowski, A., M. Sitarz, 2005. Analysis of wheel-rail interaction using FE Software, Wear, 258:
1217-1223.
Sraml, M., J. Flasker, I. Potrc, 2003. Numerical procedure for predicting the rolling contact fatigue crack
initiation. Int J Fatigue, 25(7): 585-95.