A new model in railvehicles dynamics considering nonlinear suspension

components behavior
H. Sayyaadi

, N. Shokouhi
Department of Mechanical Engineering, Sharif University of Technology, Azadi Avenue, P.O. Box 11365-9567, Tehran, Iran
a r t i c l e i n f o
Article history:
Received 11 June 2008
Received in revised form
11 January 2009
Accepted 14 January 2009
Available online 21 January 2009
Keywords:
Rail vehicle dynamics
Air spring model
Track geometrical irregularities
a b s t r a c t
In this paper, a complete four axle rail vehicle model is addressed with 70 degrees of freedom (DOFs)
including a carbody, two bogies, and four axels. In order to include the effects of the track irregularities
in vehicle dynamics behavior, a simplied track model is proposed and it is validated by some
experimental data and test results. As the performance of the suspension components, especially for air
springs, have signicant effects on railvehicle dynamics and ride comfort of passengers, a complete
nonlinear thermo-dynamical air spring model, which is a combination of two different models, is
introduced and implemented in the complete railvehicle dynamics. By implementing Presthus
formulation [Derivation of air spring model parameters for train simulation. Master dissertation,
Department of Applied Physics and Mechanical Engineering, Division of Fluid Mechanics, LULEA
University, 2002], the thermo-dynamical parameters of air spring are estimated and then they are tuned
based on the experimental data. The results of the complete railvehicle eld tests, show remarkable
agreement between proposed model and test data.
& 2009 Elsevier Ltd. All rights reserved.
1. Introduction
Nowadays, speed-up in technology and its new features bring
higher speed with reliable safety and better ride comfort in rail
transportation industries. Trafc jam in capital and big cities all
around the world, wasting passengers time at the air terminals,
huge mass transportation, and so on brings a good opportunity for
rail industries to attract more and more passengers and cargos
to their services. In addition to safety, the other important factor
for the passengers to decide about their transportation type is
ride comfort. And that is why accessing better ride comfort for
passengers during their trip is essential.
To serve better ride comfort to the passengers, the secondary
suspension of most new EMU and DMU railvehicles is equipped
with air springs. Air springs are very important isolation com-
ponent, which guarantee good ride comfort for the passengers
during their trip. In the most published railvehicle models,
developed by researches [212], the thermo-dynamical effects of
air springs in the railvehicle dynamics are ignored and the
secondary suspension of vehicle is modeled by some simple
springs and dampers models.
In this paper, complete dynamics of one IRICo DMU trailer car
with nonlinear components behavior is addressed. The dynamics
behaviors of all components are validated by some experimental
results. In the proposed model, track rails assumed to be rigid
with viscoelastic bed in vertical and lateral directions [13,14]. In
order to consider the effects of the track irregularities on the
vehicle behavior, track data measured by EM120 machine are
used. In the model of rail vehicle which moves along the straight
line, effects of ballast and sleepers masses on the vehicle
dynamics are ignored. Four contact parameters introduced by
Shabana and Zaazaa [15,16] are used to dene the contact point
between rail and wheel. In order to improve the simulation
performance, a feed-forward neural network (FFNN) is trained
and then it is used to compute the contact point parameters.
Contact forces are calculated based on the Polach theory [17].
Numerical and experimental results are summarized and com-
pared at the end of this article to verify the proposed model and
technique.
2. Vehicle description
IRICo DMU train has four cars; two motor cars at both ends and
two trailer cars in the middle. The schematic diagram for one
complete train is depicted in Fig. 1.
Each car is suspended by two bogies. The side view of two
axle IRICo DMU trailer bogie is shown in Fig. 2. To attain proper
stability and good ride comfort for the passengers, bogies are
equipped with primary and secondary suspension systems. The
secondary suspension has two air springs to suspend the vehicle
body, four vertical and lateral dampers, and two connection links
ARTICLE IN PRESS
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/ijmecsci
International Journal of Mechanical Sciences
0020-7403/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijmecsci.2009.01.003

Corresponding author. Tel.: +982166165682; fax: +982166000021.

E-mail address: sayyaadi@sharif.edu (H. Sayyaadi).
International Journal of Mechanical Sciences 51 (2009) 222232
to connect the bogie frame to the carbody. The primary
suspension is made of two coil springs, two leaf springs and one
vertical damper at each side of wheel set.
3. Dynamics model
3.1. Track model
In order to include effects due to the track geometrical
irregularities in the railvehicle dynamics, simplied version of
Jin and Wen [13,14] track model is in use here. In the proposed
model, four track irregularities are introduced as bed distur-
bances. As the vehicle is modeled while passing through straight
line, the effects of ballast and sleepers masses in the vehicle
dynamics and coupling effects between left and right ballast
masses are ignored. Track stiffness and damping rates are
according to the work of Jin and Wen [13]. Schematic diagram
of the track model, with four track irregularities, is shown in Fig. 3.
ARTICLE IN PRESS
Fig. 1. Composition of IRICo DMU.
Fig. 2. Side view of IRICo DMU trailer bogie.
Fig. 3. Track model; parameters are based on the work of Jin and Wen [13]. Rail
disturbances: d
left,horizontal
, d
left,vertical
, d
right,horizontal
, and d
right,vertical
.
g(s
1
)
f(s
2
)
Y
rp
Y
w
S
1
Z
rp
Q
Z
w
X
w
S
1
S
2
Q
w
w
r
r
r
S
2
w
Fig. 4. Four contact parameters introduced by Shabana and Zaazaa [15].
Fig. 5. Effective parameters d
L
, d
V
, c
A
, and y
A
for contact point extraction.
Fig. 6. Feed-forward neural network.
Fig. 7. Contact point position in left rail.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 223
3.2. Railwheel contact point and contact forces model
By using four geometrical contact parameters S
r
1
, S
r
2
, S
w
1
,
and S
w
2
, shown in Fig. 4, and the method proposed by Shabana
and Zaazaa [15], contact point between rail and wheel is deter-
mined. According to Shabana, S
r
1
is found out by integrating speed
component of point Q along the rail, shown in Fig. 4, and S
w
2
is
the angle between the Z-axis of coordinate system xed to
the axle and vertical line. The remaining two contact parameters
are calculated by a searching algorithm which guarantees the
perpendicularity of rail reaction force and tangent plane of the
wheel at the contact point. To do this, four geometrical
parameters d
L
, d
R
, y
A
, and c
A
, shown in Fig. 5, are investigated.
By these assumptions and using four contact parameters and
normal reaction force of the rails, creep forces are calculated
based on the Polach theory [17].
For improving the simulation capabilities and also decreasing
the simulation time, a feed-forward neural network is suggested
to compute these two contact parameters as shown in Fig. 6. This
neural network has four inputs d
L
, d
R
, y
A
, and c
A
, two outputs S
r
2
&
S
w
1
and 14 neurons in two hidden layers. The activation function
for hidden layers and output layer is tangent sigmoid and linear,
respectively. The network is trained based on the Levenberg
Marquardt back-propagation method.
Proposed network is trained according to the searching
algorithm results, generated during sinusoidal variations of axle
states. The lateral and vertical axle states changed with 0.01 and
0.003mm amplitude and 0.5 and 0.25Hz frequencies, respec-
tively. At the same time, yawand roll angle of axle varied with 0.41
and 0.21, amplitude and 0.334 and 0.1667Hz frequencies. The
simulation time is selected so that each input signal has at least
two periods in the simulation time. Results are shown in Figs. 7
and 8. Good agreement between network output and searching
algorithm results guarantees the performance of trained network.
Synaptic weights of the trained network are presented in
Appendix 1.
ARTICLE IN PRESS
Fig. 8. Contact point position in left wheel.
4 Leaf Springs
4 Coil Springs
2 Vert. Dampers
CG
car
CG
frame
CG
Axle
4 Leaf Springs
4 Coil Springs
2 Vert. Dampers
CG
Axle
4 Leaf Springs
4 Coil Springs
2 Vert. Dampers
2 Lat Dampers+2 Lat Buffers
Z-Link
2 Air Springs+2 Vert. Dampers
CG
frame
CG
Axle
4 Leaf Springs
4 Coil Springs
2 Vert. Dampers
CG
Axle
2 Lat Dampers+2 Lat Buffers
Z-Link
2 Air Springs+2 Vert. Dampers
Traction Rods
Lat Dampers
Lat Buffers,
Z-Link
CG
car
CG
frame
CG
Axle
Air Spring
Vert. Damper
Air Spring
Vert. Damper
Leaf Springs
Coil Springs
Vert. Damper
Leaf Springs
Coil Springs
Vert. Damper
Fig. 9. Vehicle model with 42 DOFs.
Table 1
Masses and inertia properties of the railvehicle.
Parameters Values Parameters Values Parameters Values
m
axle
1747kg m
frame
2841.3kg m
carbody
33,142kg
I
xx,axle
1098kg m
2
I
xx,frame
1030kg m
2
I
xx,carbody
30,000 kg m
2
I
yy,axle
191kg m
2
I
yy,frame
1054kg m
2
I
yy,carbody
687,231kg m
2
I
zz,axle
1098kg m
2
I
zz,frame
2003kg m
2
I
zz,carbody
687,231kg m
2
K
ex
K
ez
C
z

z
2
2M
K
vx
C
x
x
2
K
ex
F
fx,max
F
fz,max K
vz
z
1 1
x
Fig. 10. Berg air spring model [27].
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 224
3.3. Vehicle model
This work proposes a new method for studying suspension
components effects on the vehicle dynamics performances.
Accordingly, real behaviors of all components are described in
this section. For this reason, and due to the complexity of
components behaviors, especially for air springs, it is almost
impossible to describe complete vehicle dynamics with a single
set of equations. So, the vehicle is modeled as modular type and
internal forces of each component are calculated, using nonlinear
description functions and system states. Newtonian approach is
implemented for dynamics modeling of different parts and the
interacting forces and moments between them are investigated.
The vehicle model with 7 lumped masses and 42 degrees of
freedom (DOFs) is shown in Fig. 9 and some of its important
specications are tabulated in Table 1.
3.3.1. Air spring model
Air springs, which are made of carbon black lled natural
rubber (CBFNR), have long lifetime and can isolate the vehicle
body from the unpredictable noise, vibration, and disturbances.
ARTICLE IN PRESS
F
X
1
X
1
X
2
F
elastic
F
viscoelastic C = c (|x
1
|+)
X
2
k
F

Fig. 11. One-dimensional Haupt and Sedlan air spring model.

Fig. 12. Proposed modied model for air spring.
Table 2
Air spring parameters value.
Parameters Description Values
l
s
Connecting pipe length 3.2m
A
s
Connecting pipe cross section 0.001359m
2
A
e
Effective area of air spring 0.291m
2
P
0
Initial absolute air spring pressure 3.806bar
P
g
Gauge pressure
r Density of air at P
0
pressure 4.523kg/m
3
V
r0
Reservoir volume 0.04 m
2
V
b0
Air spring volume 0.064m
2
k
t
Total lost coefcient of connection pipes 3.4727
K
auxiliary
Auxiliary spring stiffness in air spring 8234kN/m
M Air mass in the pipes, air bag and air reservoir 198.385kg
Fig. 13. Test rig of air spring at ContiTech Company, Germany.
Fig. 14. Loss angle of air spring M 198.385kg, K
ez
461.629N/mm,
K
vz
351.185N/mm, C
z,1.8
11.508kNs/m
1.8
, k
haupt
96kN/m, and C
haupt
1.52
kNs/m, z 0.00063.
Fig. 15. Dynamic vertical stiffness of the air spring M 198.385kg, K
ez
461.629
N/mm, K
vz
351.185N/mm, C
z,1.8
11.508kNs/m
1.8
, k
haupt
96kN/m, C
haupt

1.52kNs/m, z 0.00063.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 225
The air spring behaviors are so complicated that cannot be
modeled by simple equations. Air spring response is independent
of the excitation frequency [18] and it behaves such as stress
relaxation function [19]. In addition, it has asymmetric hysteresis
loop, which is independent of the excitation frequency [20]. These
behaviors bring difculties to use frictional or columbic descrip-
tion to approximate the air spring characteristics.
A lot of comprehensive researches have been done for real air
springs and CBFNR behaviors identication [2126]. In the latest
3D model, developed by Berg [27], the effects of elasticity, friction
and viscosity of air spring in vertical, lateral and longitudinal
directions are introduced. In this model, stress relaxation is not
represented. The complete explanation about Berg model, shown
in Fig. 10, is presented in [27]. Friction force in the Berg model
which is zero at turning points is
F
friction
/
x x
0

b x x
0

sign_ x (1)
In which, b is the constant, x the current displacement, and x
0
the displacement at the previous turning point. As it is clear from
ARTICLE IN PRESS
Fig. 16. Lateral hysteresis loop of the air spring K
ey
154kN/m, K
vy
82.26kN/m,
C
y
1.109kNs/m, k
haupt
96kN/m, C
haupt
1.52kNs/m, z 0.00063.
Fig. 17. Test results of IRICo DMU Dampers, done by SACHS Co., Germany.
Table 3
Dampers parameters.
Type of damper Condition a b
Primary vertical dampers
0pj
_
D
damper
jp0:1
40800 0
0:1o
_
D
damper
19163 2163.7
_
D
damper
o0:1
19025 2177.5
Secondary vertical/lateral dampers
0pj
_
D
damper
jp0:05
61375 0
0:05o
_
D
damper
18221 2157.7
_
D
damper
o0:05
14143 2361.6
Stopper
Lateral
(4.5) 171
401
(3.5)
Fig. 18. Lateral buffer position in bogie.
Fig. 19. Lateral buffer stiffness.
Table 4
Haupt and Sedlan model parameters and stiffness coefcients of bushes.
Parameters Values Parameters Values
k
haupt
9kN/mm K
radial
2222.2N/mm
C
haupt
3.8kNs/m K
torsionc
31.25Nm/1
z 0.00063 k
cardanic
45.525Nm/1
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 226
Eq. (1), in this model, displacement at turning points should be
detected and assigned to the x
0
variable, which is not a standard
procedure in the dynamic analysis and causes failure in solving
algorithm.
For simulation of CBFNR behaviors, a model was developed
by Haupt and Sedlan [19], which has elastic and viscoelastic
elements. The viscoelastic element naturally produces asym-
metric stressstrain response and is weakly time-dependent.
These features are not presented in the other models and made it
very powerful to simulate the CBFNR behaviors. The simplied
one-dimensional Haupt and Sedlan model with constant coef-
cients is shown in Fig. 11.
The viscoelastic force in this simplied model is
F
viscoelastic

c_ x
2
j_ x
1
j z
(2)
in which c and z are constant coefcients.
In this research work, the Berg model which is validated by
some experimental data up to 16Hz frequencies [27], is used to
simulate the air spring dynamics. However, because of difculties
comprises from assigning previous turning point displacement
to the variable, the frictional part of this model is replaced by
the simplied viscoelastic model dened by Haupt and Sedlan.
Schematic diagrams of the proposed modied models in vertical
and lateral directions are shown in Fig. 12.
According to Berg and by taking into account of the Haupt and
Sedlan model, differential equations of air spring in vertical,
lateral, and longitudinal directions are proposed as
Vertical direction:
M w
s
K
vz
z w
s
C
zb
j _ w
s
j
b
sign _ w
s
; b 1:8
F
z
p
0
p
a
A
e
K
ez
z K
vz
z w
s

c_ x
2
j_ zj z
(3)
Lateral and longitudinal directions:
F
w
K
ew
wK
eww
y F
visco-elastic;w
K
vw
wu
C
w
_ u K
vw
wu (4)
F
visco-elastic;w

c_ x
2
j _ w
1
j z
w x; y (5)
The parameters in the above equations are estimated according to
the Presthus formulation as given in Eq. (6) [1]:
M l
s
A
s
r
A
e
A
s
V
r0
V
b0
V
r0
_ _
2
K
ez

1
p
0
A
2
e
n=V
b0
V
r0
p
g
dA
e
=dz

1
K
auxiliary
_ _
1
K
vz

1
p
0
A
2
e
n=V
b0
p
g
dA
e
=dz

1
K
auxiliary
_ _
1
K
ez
C
z;b

1
2
rk
t
A
s
A
e
A
s
V
r0
V
b0
V
r0
_ _
1b
; b 2 (6)
ARTICLE IN PRESS
Bogie frame
Axle Box
Axle
Box
Leaf Spring
Bush
k
z,leaf
k
x,leaf
k
x,bush
k
z,bush
k
ty,bush
Fig. 20. Leaf spring schematic diagram in the xz plane.
Table 5
Primary suspension system, mechanical specications.
Parameters Values (N/mm) Parameters Values
K
x,bush
, K
z,bush
45000 K
ty,bush
32Nm/1
K
y,bush
2118 K
tx,bush
, K
tz,bush
1659Nm/1
K
x,leaf spring
44145 K
x,coil spring
365.69N/mm
K
y,leaf spring
2118 K
y,coil spring
365.69N/mm
K
z,leaf spring
71.9618 K
z,coil spring
682.77N/mm
Table 6
Suspension components of rail vehicle model (28 internal DOFs).
Components No. Identier equations Internal DOF
Secondary Vert. damper 2 per bogie
Bi-linear function
Secondary Lat. damper 2 per bogie
Primary Vert. damper 2 per bogie
Air spring 4 per car
Vertical Nonlinear, Berg and Sedlan model x
2
,w
s
Lateral Nonlinear, Berg and Sedlan model x
2
,u
Longitudinal Nonlinear, Berg and Sedlan model x
2
,u
Lateral buffer 4 per car Nonlinear-polynomial, order 4
Bushlink 2 pairs per bogie Nonlinear, Sedlan model x
2
Coil spring 4 per bogie linear in space, based on experimental data
Leaf spring 4 per bogie linear in space, based on analytical calculations
Table 7
Accelerometers specications.
KS77C.100 AS-2TG
Manufacturer Manfred Weber, Germany Kyowa, Japan
Type ICP
s
compatible Strain gauge
Range (g) 760 72
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 227
The exact C
z,1.8
value is calculated based on the Presthus method
and the stiffness K
exw
is approximated by K
exw
0.7(K
ex
h+load) [1].
Numerical values of the above parameters are tabulated in Table 2.
To determine the exact values of the air spring parameters, one
IRICo DMU air spring was tested at ContiTech Company, Germany.
Fig. 13 depicts the air spring test.
Based on the method introduced by Docquier et al. [30] and
according to the test data and simulation results, the loss angle
and vertical dynamics stiffness of air spring were investigated
which are shown in Figs. 14 and 15. As it can be concluded from
these diagrams, good agreement between real test and simulation
results is achieved.
ARTICLE IN PRESS
KS77C.100 KS77C.100
KS77C.100
AS-2TG
KS77C.100 AS-2TG AS-2TG
KS77C.100 KS77C.100
KS77C.100
AS-2TG
Axle 2 Axle 4 Axle 3 Axle 1
Fig. 21. Accelerometers installation position: (a) front bogie, axle 2 and (b) rear bogie, axle 3.
Accelerometers
on rear bogie
Accelerometers
on front bogie
GPS Antenna installed over the roof
Fig. 22. Sensors installation position on the trailer car.
Fig. 23. Tehran-Ghazvin route data: (a) cant variation of Tehran-Ghazvin route and (b) speed and height prole, Tehran-Ghazvin route.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 228
For validation of air spring lateral behavior, hysteresis loop
of air spring is investigated. Results are shown in Fig. 16. Good
agreement between test data and simulation results shows that
the proposed equations can simulate the real behavior of air
spring very well.
3.3.2. Dampers model
IRICo DMU primary and secondary suspension dampers were
tested in Sachs Co., Germany. Results are shown in Fig. 17.
According to the test results, damping rate of each damper is
described by bi-linear function:
F a
_
D
damper
b (7)
The coefcients in Eq. (7) are tabulated in Table 3 for different
conditions and installation locations of dampers.
3.3.3. Lateral buffer model
Lateral displacement of the carbody is restricted by four lateral
buffers, installed on the bogie frame as shown in Fig. 18. Each
lateral buffer has primary compression force equal to 100N.
Forcedisplacement diagram of the lateral buffer is shown in
Fig. 19. An air gap of 17mm between carbody and lateral buffer
lets the carbody moves freely in this range. Lateral displacement
of carbody is restricted over 740mm by two stoppers installed on
the bogie.
According to the experimental data, mechanical behavior of
the lateral buffer is formulated as
F 0; 0oxo17
F 5 10
5
x 17
4
0:0019x 17
3
0:0144x 17
2
; 1:1525x 17 0:100;
17pxp40
_

_
(8)
3.3.4. Connection link model
The carbody is linked to the bogie frame by two connection
links through center pivot device. Each link has a rubber bush
at each end. Static stiffness of the rubber bush in radial, torsional,
and cardanic movements are presented according to the tests
done by GMT Co., Germany. Because these rubber bushes are
made from CBFNR, for exact modeling of connection link behavior
in the vehicle dynamics, the simplied one-dimensional Haupt
and Sedlan model, described in Fig. 11, is used for radial
movement of the bushes with the following equation:
F
visco-elastic;radial

c_ x
2
j_ r
1
j z
(9)
The Haupt and Sedlan parameters, tabulated in Table 4, are
according to the Allen results [28].
ARTICLE IN PRESS
Fig. 24. Axles 2 and 3 accelerationsvalidation of track model: (a) FFT of lateral accelerationaxle 2; (b) FFT of vertical accelerationaxle 2; (c) FFT of lateral
accelerationaxle 3; and (d) FFT of vertical accelerationaxle 3.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 229
3.3.5. Primary suspension, stiffness coefcients model
Four leaf springs in each axle, direct the wheel-set along the
curves. Each leaf spring has one rubber bush at each end.
Schematic diagram of the leaf spring in the xz plane is shown
in Fig. 20. The same conguration is developed for the other
planes.
Considering boundary condition of leaf spring at two ends, the
stiffness of leaf spring in bending is determined as follows:
k
1
leaf vertical

1
EI
L
4
k
ty;bush
4Lk
ty;bush
EI

L
3
3
_ _
(10)
Based on the manufacturer test results, coil spring stiffness is
approximated linearly. By using Eq. (10), stiffness coefcients
of the primary suspension are calculated. Numerical value of
primary suspension parameters are tabulated in Table 5.
3.4. Complete vehicle model
All components model used in the railvehicle dynamics with
related internal DOFs are listed in Table 6. It can be seen that
suspension components add up 28 additional internal DOFs to the
ARTICLE IN PRESS
Fig. 25. Validation of complete vehicle model: (a) FFT of carbody lateral acceleration; (b) FFT of carbody vertical acceleration; (c) FFT of front bogie frame lateral
acceleration; (d) FFT of front bogie frame vertical acceleration; (e) FFT of rear bogie frame lateral acceleration; and (f) FFT of rear bogie frame vertical acceleration.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 230
model. Whereas the vehicle masses have totally 42 DOFs, the
complete railvehicle will be a model of 70 DOFs.
4. Model simulation and test results
Dynamics test of the IRICo DMU was performed in Tehran-
Ghazvin route to validate the performance of the proposed
dynamics model. VBOX III GPS, manufactured by Racelogic, UK,
is used to measure the exact position and speed of the train.
Measured speed is monitored and recorded by one of the 16
channels of CRONOS PL data acquisition system manufactured by
imc, Australia, which was connected to the portable computer.
The accelerations of the trailer car masses are recorded by 15
remaining channels by means of accelerometers with the
specications presented in Table 7. These accelerometers are
installed on two axles, two bogie frames and on the oor of the
carbody inside the car according to Figs. 21 and 22.
Cant variations of the track, recorded by EM120 machine,
vehicle speed and height prole of the line are shown in Fig. 23.
Referring to this gure, it is perceived that cant parameter in the
interval between 110 and 115km does not have any signicant
variation that means the track line is almost straight without
signicant curvature. Constant traveling speed and smooth height
prole make this section of the test route suitable for validation of
the proposed vehicle model.
As track data and geometrical irregularities recorded by EM120
machine are not up-to-date and were recorded 25 months before
the test, by using track data, poor agreement between test and
simulation results is achieved. For this reason, the vehicle model
is studied in two different phases. In the rst phase, the model
was executed by considering old track irregularities and axle
accelerations are investigated to validate track model. In the
second phase, measured axle accelerations and their integrals in
time domain are used as reference states to validate complete
vehicle model.
As in the vehicletrack system, track has the highest stiffness,
the high-frequency responses of the vehicle are related to the
track behavior. To validate dynamic model of track system,
responses of the vehicle in frequency domain are investigated
[29]. Fig. 24 shows the fast Fourier transform (FFT) of measured
signals and simulation results for 100m track length. It can be
seen that in the frequency range between 20 and 60Hz, good
agreement between test and simulation results is achieved.
For validation of the proposed vehicle model, the measured
vertical and lateral acceleration of axles within 200m of the track
with their integrals are used as reference inputs. Whereas the
accelerometers are not installed on the axle nos. 1 and 4, the
accelerations of axles 2 and 3 are used for axles 1 and 4 with a
constant lead/lag equal to axle base. FFT of test and simulation
results are shown in Fig. 25. Because each car is equipped with an
auxiliary power unit (APU) which works with an internal
combustion engine, there is a peak response in carbody accelera-
tions signals at 50Hz frequency. To simulate the effect of APU on
the system, a sinusoidal force with relevant amplitude and
frequency is added to the vehicle model as a disturbance.
It can be seen that the vehicle test results and the proposed
vehicle model have almost the same behavior. Little deviations in
the diagrams can be judged by the following items:
1. The reference axle states only cover vertical and lateral
movement of the axles. During validation, other DOFs are
constrained constant.
2. Only accelerations of axle nos. 2 and 3 are recorded. The other
two axle states are generated based on the measured
accelerations.
3. Measured accelerations contain some sources of errors. By
integrating accelerations, errors of speed and position signals
have been accumulated.
5. Conclusion
This research work proposes a new model for studding
inuences of suspension system components behavior on rail
vehicle dynamics. For studding the behavior and performance of
suspension components, complete trackvehicle model with 70
DOFs is addressed as a modular type. In this new model, behavior
of each component is dened and validated using real test data
from eld experiments. Complete nonlinear air spring model, with
taking into consideration of thermo-dynamical effects, is devel-
oped and model coefcients are tuned based on the real test data.
This model can be easily used in dynamic modeling of air springs.
For validation of the proposed trackvehicle model, dynamics test
of the vehicle was carried out. Comparison of the results show
good agreement between proposed model and test results that
says this new model can be used for simulation of the vehicle
performances very well and then it is a good model for further
applications such as improvement in ride quality and comfort
index, passive and/or active control and so on.
Acknowledgments
Authors acknowledge Sharif University of Technology, grateful
for the excessive support from Irankhodro Rail Transport
Industries Co., and also like to express their sincere thanks to
Mr. M.S. Ghorashi for his help in managing eld tests.
ARTICLE IN PRESS
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 231
ARTICLE IN PRESS
Appendix 1. Neural network synaptic weights
y f W x b; f
hidden layers
x
2
1 e
2x
1; f
output layer
x x
W
1

0:4879 143:29 153:58 296:96
440:24 1618:3 922:66 1334:3
31:414 61:038 58:806 286:63
24:646 3:7075 68:557 168:59
90:293 204:3 141:32 326:46
25:573 141:03 86:023 164:47
74:14 5:2036 304:34 42:909
19:799 90:972 249:49 199:13
29:727 80:828 67:982 241:81
_

_
_

_
; b
1

1:7844
343:02
21:85
18:173
71:443
16:336
58:029
11:211
20:332
_

_
_

_
W
2

3:8078 0:34866 81:023 13:621 0:1302 28:59 1:6923 10:078 107:8
23:264 6:9369 340:7 27:582 6:3036 210:98 10:203 1:4993 556:53
3:8083 0:34666 80:939 13:612 0:12951 28:561 1:6881 10:077 107:69
1:464 0:20806 51:506 3:1561 2:0031 23:906 3:4124 3:1543 76:946
0:76366 0:01884 9:4441 0:49465 0:0555 6:2466 0:25575 0:15104 16:544
_

_
_

_
; b
2

4:2292
2:5291
4:2332
1:0352
0:5185
_

_
_

_
W
3

7:73 0:003244 7:7259 0:022746 0:00677
7:5261 0:003135 7:5231 0:021734 0:020275
_ _
; b
3

0:014172
0:00269
_ _
References
[1] Presthus M. Derivation of air spring model parameters for train simulation.
Master dissertation, Department of Applied Physics and Mechanical En-
gineering, Division of Fluid Mechanics, LULEA University, 2002.
[2] Sun YQ, Dhanasekar M. A dynamic model for the vertical interaction of the
rail track and wagon system. International Journal of Solids and Structures
2002;39:133759.
[3] Lei X, Noda NA. Analyses of dynamic response of vehicle and track coupling
systemwith random irregularity of track vertical prole. Journal of Sound and
Vibration 2002;258:14765.
[4] Hou K, Kalousek J, Dong R. A dynamic model for an asymmetrical vehicle/
track system. Journal of Sound and Vibration 2003;267:591604.
[5] Popp K, Kaiser I, Kruse H. System dynamics of railway vehicles and track.
Archive of Applied Mechanics 2003;72:94961.
[6] Tanabe M, et al. Computational model of a Shinkansen train running on the
railway structure and the industrial applications. Journal of Materials
Processing Technology 2003;140:70510.
[7] Durali M, Bahabadi MMJ. Investigation of train dynamics in passing through
curves using a full model. In: Rail conference, Proceedings of the 2004 ASME/
IEEE joint, p. 838.
[8] Li Y, Qiang Sh, et al. Dynamics of windrail vehiclebridge systems. Journal of
Wind Engineering and Industrial Aerodynamics 2005;93:483507.
[9] Baeza L, Roda A, Nielsen JCO. Railway vehicle/track interaction analysis using
a modal substructuring approach. Journal of Sound and Vibration 2006;293:
11224.
[10] Li P, Goodall R, et al. Estimation of railway vehicle suspension parameters for
condition monitoring. Control Engineering Practice 2007;15:4355.
[11] Zhang N, Xia H, et al. Vehiclebridge interaction analysis under high-speed
trains. Journal of Sound and Vibration 2008;309:40725.
[12] Banerjee N, et al. Bond graph modeling of a railway truck on curved track.
Simulation Modeling Practice and Theory. Available online from 2008.
[13] Jin XS, Wen ZF. Effect of passenger car curving on rail corrugation at a curved
track. Journal of Wear 2006;260:61933.
[14] Jin XS, Wen ZF. Effect of discrete track support by sleepers on rail corrugation
at a curved track. Journal of Sound and Vibration. Available online from 2008.
[15] Shabana AA, Zaazaa KhE. Development of elastic force model for wheel/rail
contact problems. Journal of Sound and Vibration 2004;269:295325.
[16] Shabana AA, Zaazaa KhE. Railroad vehicle dynamics, a computational
approach. Boca Raton, FL: CRC Press; London: Taylor & Francis, 2008.
[17] Polach O. A fast wheelrail forces calculation computer code. Journal of
Vehicle System Dynamics 1999(Suppl. 33):72839.
[18] Boast D, Fellows S, Hale M. Effects of temperature, frequency and amplitude
on the dynamic properties of elastomers. In: Conference paper, AVON
Automotive 2002; 03.
[19] Haupt P, Sedlan K. Viscoplasticity of elastomeric materials: experimental
facts and constitutive modeling. Archive of Applied Mechanics 2001;71:
89109.
[20] Berg M. A non-linear rubber spring model for rail vehicle dynamics analysis.
Journal of Vehicle System Dynamics 1998;30:197212.
[21] Turner DM. A triboelastic model for the mechanical behaviour of rubber.
Journal of Plastics and Rubber Processing and Applications 1988;9:197201.
[22] Coveney VA, Johnson DE, Turner DM. A triboelastic model for the cyclic
mechanical behaviour of lled vulcanizates. Journal of Rubber Chemistry and
Technology 1995;68:66070.
[23] Coveney VA, Johnson DE. Rate-dependent modelling of a highly lled
vucanizate. Journal of Rubber Chemistry and Technology 2000;73(4):56577.
[24] Bergstrom JS, Boyce MC. Large strain time-dependent behavior of lled
elastomers. Journal of Mechanics of Materials 2000;32:62744.
[25] Miehe C, Keck J. Superimposed nite elasticviscoelasticplastoelastic stress
response with damage in lled rubbery polymers. Experiments, modelling
and algorithmic implementation. Journal of Mechanics and Physics of Solids
2000;48:32365.
[26] Lion A. A constitutive model for carbon black lled rubber: experimental
investigation and mathematical representation. Journal of Continuum
Mechanics and Thermodynamics 1996;8:15369.
[27] Berg M. A three-dimensional airspring model with friction and orice
damping. Journal of Vehicle System Dynamics 1999;33:52839.
[28] Allen P. Models for the dynamic simulation of tank track components. PhD
dissertation, Craneld University, Academic Year 20052006.
[29] Lombaert G, Degrande G, et al. The experimental validation of a numerical
model for the prediction of railway induced vibrations. Journal of Sound and
Vibration 2006;297:51235.
[30] Docquier N, Fisette P, Jeanmart H. Multiphysic modelling of railway vehicles
equipped with pneumatic suspensions. Journal of Vehicle System Dynamics
2007;45(6):50524.
H. Sayyaadi, N. Shokouhi / International Journal of Mechanical Sciences 51 (2009) 222232 232