University of Wollongong

Research Online
Faculty of Engineering and Information Sciences Papers

Faculty of Engineering and Information Sciences

2013

Rolling contact stress states of two contact points

on curved track
Khoa Vo
University of Wollongong, dkv997@uowmail.edu.au

A Kiet Tieu
University of Wollongong, ktieu@uow.edu.au

Hongtao Zhu
University of Wollongong, hongtao@uow.edu.au

Prabuono Buyung Kosasih

University of Wollongong, buyung@uow.edu.au

Publication Details
Vo, K. D., Tieu, A. K., Zhu, H. T. & Kosasih, P. B. (2013). Rolling contact stress states of two contact points on curved track. 10th
World Congress on Railway Research

Research Online is the open access institutional repository for the

University of Wollongong. For further information contact the UOW
Library: research-pubs@uow.edu.au

Rolling contact stress states of two contact points on curved track

Abstract

A three-dimensional finite element model (3D FEM) of wheel/high rail rolling contact on curved track has
been introduced in this article. The elastic-plastic model from FEM code ANSYS/LS-DYNA included a
whole wheel, 680 mm of the high rail and some sub-components of the track (railpads, sleepers, and ballast).
Furthermore, the rail was canted as in practical situation and the influences of angle of attack (AOA) were also
investigated in detail. The obtained results have demonstrated that two-point contact pattern can potentially
appear as the train enters a curved track. The stress states on two contact zones showed that the maximum
stress at the rail gauge was higher than that at the rail head. Besides, the higher AOA could lead to a higher
stress magnitude, especially at the rail gauge where damages appear frequently. This study can provide useful
information for a future research on surface damage mechanism of the rail of a curved track.
Keywords

curved, points, track, two, rolling, states, stress, contact

Publication Details

Vo, K. D., Tieu, A. K., Zhu, H. T. & Kosasih, P. B. (2013). Rolling contact stress states of two contact points on
curved track. 10th World Congress on Railway Research

This conference paper is available at Research Online: http://ro.uow.edu.au/eispapers/3032

ROLLING CONTACT STRESS STATES OF TWO CONTACT POINTS ON CURVED TRACK

1, 2

K. D. Vo

1, 2

, A. K. Tieu

1, 2

, H. T. Zhu

, P. B. Kosasih

Faculty of Engineering, University of Wollongong, Northfields Avenue, Wollongong, NSW 2522, Australia
2
Cooperative Research Centre (CRC) for Rail Innovation, Australia

Abstract: A three-dimensional finite element model (3D FEM) of wheel/high rail rolling contact on
curved track has been introduced in this article. The elastic-plastic model from FEM code ANSYS/LSDYNA included a whole wheel, 680 mm of the high rail and some sub-components of the track
(railpads, sleepers, and ballast). Furthermore, the rail was canted as in practical situation and the
influences of angle of attack (AOA) were also investigated in detail. The obtained results have
demonstrated that two-point contact pattern can potentially appear as the train enters a curved track.
The stress states on two contact zones showed that the maximum stress at the rail gauge was higher
than that at the rail head. Besides, the higher AOA could lead to a higher stress magnitude, especially
at the rail gauge where damages appear frequently. This study can provide useful information for a
future research on surface damage mechanism of the rail of a curved track.
Keywords: contact stress, two-point contact, conformal and non-conformal contact, finite element.
1. Introduction
For wheel/rail contact, the appearance of conformal or non-conformal contact depends on the track
properties, such as curvature of the track and worn status of the wheel/rail profiles, etc [1]. On the
straight track or on the large radius curved track, the non-conformal contact mostly appears on the rail
head. Whereas, the conformal contact between the wheel flange and rail gauge may occur once the
wheel rolls on the curved track, especially on the high rail. The possible contact situations between
the wheel and the rail in curved track are illustrated in Fig. 1:
Low rail: wheel tread-rail head contact
High rail: wheel flange-rail gauge contact

Fig. 1. Possible contact state between the wheel and the low/high rails on a curved track
On the curved track, rail is usually canted and track cant (or superelevation) is also applied by
lowering the inside rail and raising the outside rail by an equal amount. Since a significant
discrepancy of the height is created between the inner and outer canted rails (or low and high rails,
respectively), the contact position and stresses on these two rails are also different. Many research on
rolling contact stresses on straight track [2-5] have been carried out. Some investigations showed the
appearance of multiple-point contact as the wheel flange approaches the rail gauge [6-8]. These
results, however, were still obtained from straight track simulation and the effects of attack angle were
not included. A static finite element model of curved track was proposed by W Daves et al. [9]. This
simulation, nevertheless, was performed on a very short track (70 mm) and the small angle of attack
could not investigate the effect of curved track thoroughly. A simulation of curved track by Aleksander
Sladkowski [10] included the effect of AOA , but the rail cant angle was not considered in the
modelling.
Contact stress is important in the understanding and prediction of the damage behaviour of the
contact bodies [11]. However published research on the stresses on the rail head on the curved track
is very limited. As illustrated in Fig. 1, a non-conformal wheel tread-rail head contact usually appears
on the low rail. This type of contact is similar to the contact on straight track. However, the conformal
wheel flange-rail gauge contact normally occurs on high rail. Such contact location cannot be solved
by the classical theories such as Hertz Theory [12] or Kalkers theories [13-15] because of the
invalidity of half space assumption and elastic assumption of the rail material [16]. Therefore a 3D

elastic-plastic finite element model of rolling contact between the wheel and high rail in a curved track
has been introduced in this paper. Stress states at the contact patches were analysed in detail, in
which angle of attack and rail cant angle were all taken into account. This research will supply a
method and useful datum for future probe on the mechanics of surface damage on curved track.
2. Finite element model
In this work, the finite element models were constructed by ANSYS/LS-DYNA explicit code [17]. The
schematic diagram of the finite element model was shown in Fig. 2. In order to simulate the effects of
the curved track, the rail cant angle and angle of attack denoted by
and , respectively were
applied. The rail was canted 1:20 according to Australian standard [18]. Two cases of AOA were
examined in this study:
o
Case 1: AOA = 1/3
o
Case 2: AOA = 2/3

Fig. 2. Schematic diagram of curved track model on (a) xOz plane, (b) yOz plane, and (c) xOy plane
The Australian standard profiles of both wheel and rail (ANZR1 wheel [19] and 60 kg/m rail [20]) were
used in the current models. The models were mainly built up with the following elements: solid
elements (SOLID164) for wheel/rail, mass element (MASS166) for sleepers, and spring-damping
element (COMBI165) for rail pad and ballast. A non-uniform mesh was created, in which very fine
mesh (1.3 x 1.3 mm) at the contact surfaces and coarse mesh further away from the contact surface,
respectively. It was confirmed in [3] that for engineering applications, the mesh 1.3 x 1.3 mm is fine
enough for wheel/rail contact problem. Different mesh regions in the FEM model are illustrated in Fig.
3.

Fig. 3. 3D mesh of FEM model

The wheelset was modelled as rigid body, and glued with the wheel. Contact between the wheel and
the rail was defined by the automatic contact algorithm and a constant friction coefficient of 0.4 was

applied. The rolling speed was set at 23 rad/s, and initial velocity was also applied to ensure that a
stable state could be achieved faster. A summary of loading conditions is outlined in Table 1 below.
Table 1 Input parameters in FEM models
Parameters
Normal force
Lateral force
Rolling velocity
Cant angle
Angle of Attack (AOA)
COF

Value
116
21
23
2.86
1/3 & 2/3
0.4

Unit
kN
kN
rad/s
degree
degree

3. Results and discussions

Using the finite element models described in section 2, the process of new wheel/high rail rolling
contact in curved track was simulated. FEM models were run on HPC Cluster to minimize the
calculation time. The results for two cases of AOA will be presented in this section.
Contact zone and contact pressure
The results showed that two-point contact appeared when the new wheel travelled on the new rail.
One of the contact patches was placed at the rail head as usual, while the second one was located at
the rail gauge. Under the load conditions described above, the normal contact load on two separate
contact points can be found in Table 2. The total normal contact loads from simulation at two contact
points were almost equal to applied axle load (116 kN). However, the normal load distributions at two
contact areas in two cases are slightly different. In this study, it was found that if the angle of attack
o
o
varies from 1/3 to 2/3 , the normal load at the rail head decreases by a small amount of 3.3 kN while
an increasing of 3.9 kN was found at the rail gauge. This change of load distribution may cause a
variation of the stress state on rail surface.
Table 2 Normal contact load at two contact points (kN)

Case 1: AOA = 1/3

o
Case 2: AOA = 2/3
Difference between case 1 & case 2

Normal load at
wheel tread-rail
head contact
76.6
73.3
-3.3

Normal load at
wheel flange-rail
gauge contact
38.5
42.4
+3.9

Total normal
load
115.1
115.7

The positions of contact zones on the rail surface for different angle of attack are illustrated in Fig. 4. It
should be noted that the distance along the rolling direction between two centres of contact zone in
case 1 is about 7.7 mm, while such distance in case 2 is found wider than case 1 (12.9 mm). The
o
o
increase of AOA from 1/3 to 2/3 has led to a small increase of such distance on the rail surface.

Fig. 4. Contour distributions of contact pressure on the rail (a) AOA = 1/3 (b) AOA = 2/3

A detailed comparison between contact patches are shown in Table 3. On the whole, there was no
significant change in the dimensions of the contact zones between two cases of AOA (1/3o and 2/3o).
However, the contact dimensions at the rail gauge were quite different to that at the rail head. In
comparison to the dimensions of contact patch at rail head, the longitudinal axis a on the rail gauge is

nearly double that of the rail head, but the lateral axis b has been reduced by half for both two cases
of AOA. This was due to the change of curvatures radius at the contact spots.
Table 3 Comparison between rail head and rail gauge contact patches

AOA = 1/3

AOA = 2/3

Rail head
Rail gauge
Different
Rail head
Rail gauge
Different

Dimension of contact patch (mm)

a
b
a/b
8.75
5.58
1.57
14.6
2.85
5.12
+5.85
-2.73
8.75
5.75
1.52
14.9
2.83
5.26
+6.15
-2.92

Maximum Pressure
(MPa)
920
1028
+108
902
1101
+199

The 3D graphs of pressure distribution in Fig. 5 present a comparison of pressure at the rail gauge
and rail head for various attack angles. It can clearly be seen that the peak of pressure at the rail
gauge was higher than that at the rail head, which means the pressure exerted on the rail gauge was
higher. For instance, maximum pressure at the rail gauge was 1028 MPa, while it was about 920 MPa
o
o
at the rail head in case AOA = 1/3 . With AOA = 2/3 , the maximum pressures at rail gauge and rail
head were 1102 MPa and 902 MPa, respectively. As the AOA was slightly increased, the peak of
contact pressure at rail gauge was also lowered.

@rail gauge

@rail head

(a) Case 1: AOA = 1/3

@rail head

@rail gauge

(b) Case 2: AOA = 2/3

Fig. 5. 3D distribution of contact pressure at the rail gauge and the rail head for two cases of AOA
Contact stresses
The Von-Mises stress distribution at two contact points on rail were plotted in Fig. 6 and Fig.7 below.
As shown in Fig. 6, the maximum Von-Mises stress were found at the centre of contact patch on the

rail gauge in both two cases of AOA. Moreover the maximum value of such stress in case 1 (AOA =
o
1/3 ) is 885 MPa, while it is higher in case 2 (973 MPa). In addition to increasing of normal load at the
rail gauge, the higher angle of attack also resulted in an increase of stress at the rail gauge.

(a) AOA = 1/3

(b) AOA = 2/3

Fig. 6 Von-Mises stress distribution on rail surface

Fig. 7 shows the Von-Mises stress distribution on the cross sections of the wheel and the high rail.
High stress is found not only on the rail gauge but also on the wheel flange. In comparison the stress
at the rail head, the stress generated between the wheel flange and rail gauge was significant larger.
The higher is the Von-Mises stress, the shorter the time damage can occur. It can be speculated that
on curved track, especially as the wheel travels on the high rail, damage is likely to occur at the wheel
flange/rail gauge rather than at the wheel tread/rail head.

(a) AOA = 1/3

(b) AOA = 2/3

Fig. 7. Von-Mises stress distribution on cross section of wheel/rail

4. Conclusions
A 3D dynamic FEM model has been developed to analyse the rolling contact stress states of the high
rail surface in a curved track. The conclusions for current study are as follows:
1) On a curved track, two-point contact case may occur if a wheel travels on a high rail. The first
contact is located at the wheel tread/rail head, and the second one is placed at the wheel
flange/rail gauge.
2) As the angle of attack increases, the normal load distribution on two contact patches can be
changed. In particular such load is reduced at the rail head, whereas it is increased at rail
gauge.
3) The magnitude of Von-Mises stress at rail gauge is larger than this at the rail head, and it is
increased with the increasing of angle of attack. Therefore it can be predicted that the
damage is likely to occur at the rail gauge rather than rail head in curved track.
Acknowledgement
This work was performed as part of CRC for Rail Innovation project no. R3.119. The authors gratefully
acknowledge financial support of CRC for Rail Innovation. Simulations were conducted by using the
HPC Cluster at University of Wollongong, Australia.

Referenes
[1]
R. Lewis and U. Olofsson, Wheel-rail interface handbook. Oxford: CRC Press, 2009.
[2]
X. Zhao, Z. L. C. Esveld, and R. Dollevoet, "The Dynamic Stress State of the Wheel-Rail
Contact," Proceedings of the 2nd IASME/WSEAS Int. Conf. on Continuum Mechanics, pp.
127-133, 2007.
[3]
X. Zhao and Z. Li, "The solution of frictional wheelrail rolling contact with a 3D transient finite
element model: Validation and error analysis," Wear, vol. 271, pp. 444-452, 2011.
[4]
Z. Wen, L. Wu, W. Li, X. Jin, and M. Zhu, "Three-dimensional elasticplastic stress analysis
of wheelrail rolling contact," Wear, vol. 271, pp. 426-436, 2011.
[5]
M. A. Arslan and O. Kayaba, "3-D RailWheel contact analysis using FEA," Advances in
Engineering Software, vol. 45, pp. 325-331, 2012.
[6]
T. Telliskivi, U. Olofsson, U. Sellgren, and P. Kruse, "A tool and a method for FE analysis of
wheel and rail interaction," Proceedings of the ANSYS Conference in Pittsburgh (on CD),
Pennsylvania, 2000.
[7]
T. Telliskivi and U. Olofsson, "Contact mechanics analysis of measured wheel-rail profiles
using the finite element method," Proceedings of the Institution of Mechanical Engineers, Part
F: Journal of Rail and Rapid Transit, vol. 215, pp. 65-72, March 1, 2001 2001.
[8]
J. Zhang, S. Sun, and X. Jin, "Numerical Simulation of Two-Point Contact Between Wheel
and Rail," Acta Mechanica Solida Sinica, vol. 22, pp. 352-359, 2009.
[9]
W. Daves, B. P. Singh, and K. Six, "Rolling contact fatigue of wheels - finite element
modelling," In: P.Mutton, B. Thomas (Eds) Proceedings of the 7th International Conference
on Contact Mechanics andWear of Rail/Wheel Systems (CM2006), Brisbane, Australia, pp.
447-453, 2006.
[10]
A. Sladkowski and M. Sitarz, "Analysis of wheelrail interaction using FE software," Wear, vol.
258, pp. 1217-1223, 2005.
[11]
U. Zerbst, R. Lundn, K. O. Edel, and R. A. Smith, "Introduction to the damage tolerance
behaviour of railway rails a review," Engineering Fracture Mechanics, vol. 76, pp. 25632601, 2009.
[12]
W. Yan and F. Fischer, "Applicability of the Hertz contact theory to rail-wheel contact
problems," Archive of Applied Mechanics, vol. 70, pp. 255-268, 2000.
[13]
J. J. Kalker, "Simplified theory of rolling contact " Department of Mathematics, Delft University
of Technology 1973.
[14]
J. J. Kalker, "A strip theory for rolling with slip and spin, Proc. Koninkl. Ned. Akad.
Wetenschap. Ser. B," pp. 1062, 1967.
[15]
K. E. Zaazaa and A. L. Schwab, "Review of Joost Kalker's Wheel-Rail Contact Theories and
Their Implementation in Multibody Codes," ASME Conference Proceedings, vol. 2009, pp.
1889-1900, 2009.
[16]
K. L. Johnson, Contact mechanics. New York: Cambridge University Press, 1985.
[17]
Ansys Theory Reference for the Mechanical APDL and Mechanical Applications: ANSYS,
Inc., 2009.
[18]
"ESC 230 Sleeper and track support," ed: RailCorp Engineering Standard, 2012.
[19]
"ESR 0001-G RSU APPENDIX G DRAWINGS," ed: Engineering Standard Rolling Stock, Jun
2010.
[20]
"AS 1085.1-2002/Amdt 1-2005 : Railway track materials - Steel rails," ed: Standards
Australia, 2005.