Professional Documents
Culture Documents
To cite this article: W.W. Guo , H. Xia & N. Zhang (2013) Dynamic responses of Tsing Ma Bridge and
running safety of trains subjected to Typhoon York, International Journal of Rail Transportation,
1:3, 181-192, DOI: 10.1080/21650349.2013.808417
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Downloaded by [University of Kent] at 09:10 24 November 2014
International Journal of Rail Transportation, 2013
Vol. 1, No. 3, 181–192, http://dx.doi.org/10.1080/21650349.2013.808417
1. Introduction
The dynamic behaviours of bridge structures and the running safety of train vehicles
under wind actions are always important issues that should be considered in a railway
bridge design. Due to wind actions, the bridge will experience considerable deformations
and vibrations. When a train runs on the bridge, the deflections and vibrations of the
bridge may be further exaggerated. The large deflections and vibrations of the bridge may
in turn affect the running safety of train vehicles and the riding comfort of passengers. In
some cases, train vehicles may overturn when they are running on a long-span bridge
exposed to strong winds. Therefore, the understanding of dynamic behaviours and the
prediction of dynamic responses of long-span suspension bridges under both high winds
and running trains becomes an important task.
Before the end of the twentieth century, however, most of the studies focused on three
separate research fields: (1) wind effects on a long-span bridge without considering the
effects of running trains [1–5]; (2) the dynamic analysis of coupled train and bridge
system excluding wind effects [6–13] and (3) the running safety of trains in high cross
winds on the ground rather than on the bridge [14]. Since the beginning of this century,
some studies have come into consideration of the dynamic interactions among the wind
action, running trains and bridge structures. Xu et al. [15] presented a framework for
predicting the dynamic response of a long suspension bridge to high winds and running
trains based on the vehicle–bridge dynamic interaction and then extended their work to
sion bridges to high winds and running trains. The field measurement data during this
event are thus analysed and used to verify, to some extent, the finite element-based
framework developed by the writers in the time domain for predicting dynamic response
of coupled train–bridge systems subjected to wind actions. Furthermore, this article
researches the critical conditions for determining whether the rail traffic on the Tsing
Ma Bridge should be closed temporarily to ensure running safety of the train.
2. Theoretical background
The dynamic analysis model of the wind–train–bridge system consists of a wind model, a
train model and a bridge model.
pffiffiffiffiffiffiffiffiffiffiffiffiffi
where Ω ¼ 2ðΔωÞ; Nf is the total number of frequency interval Δω in the spectrum;
j = 1, 2,…, n; n is the total number of wind velocity simulation points; Sw(ω) are the
alongwind and upward wind auto-spectra, respectively; φmk is a random variable uni-
formly distributed between 0 and 2π; Gjm(ω) is the correlation matrix between two
different wind velocity points and can be expressed as
International Journal of Rail Transportation 183
8
< 0; 1j<mn
Gjm ðωÞ ¼ C jjmj p
; m ¼ 1; m j n (2)
: jjmj ffiffiffiffiffiffiffiffiffiffiffiffiffiffi2
C 1C ; 2mjn
λωD
C ¼ exp (3)
2πU
m
ωmk ¼ ðk 1ÞΔω þ Δω ; k ¼ 1 ; 2 ; Nf (4)
n1
where λ is the dimensionless factor between 7 and 10, U is the mean wind velocity at the
Downloaded by [University of Kent] at 09:10 24 November 2014
elevation of the main deck and D is the horizontal distance between the wind simulation
points.
The wind spectra for the digital simulation of stochastic wind velocity field are fitted
using nonlinear least-squares method with the following objective function [19]:
a
nSj ðnÞ ¼ cm m =s
2
(5)
ð1 þ bn1=m Þ
10
u (m/s)
0
−5
−10
0 100 200 300 400 500 600
Time (s)
(a) Alongwind
8
4
v (m/s)
0
Downloaded by [University of Kent] at 09:10 24 November 2014
−4
−8
0 100 200 300 400 500 600
Time (s)
(b) Upward
Figure 1. Time histories of simulated fluctuating wind components of bridge at mid-main span
(U = 20 m/s).
V = constant
vehicle, respectively; ywijl, zwijl and θwijl are the 3 DOFs of the lth wheel in the jth bogie of
the ith vehicle, respectively. The connections between the car body and a bogie are
represented by two linear springs and two viscous dashpots of the same properties in
either the horizontal direction or the vertical direction, as well as the connections between
h h
bogies and wheelsets. The stiffness and damping coefficients are denoted as k2ij , k1ij , ch2ij ,
ch1ij in the horizontal direction and k2ij
v v
, k1ij , cv2ij , cv1ij in the vertical direction.
206.0 206.0
B D E F G H
Ma Wan Tsing Yi
Island Island
4.318 2.914
2.35
13.4 7.1
41.0
taken into account in the computation in order to include the effects of both the global
deformation of the bridge and the local deformation of the structural components support-
ing the tracks. The modal damping ratios in the lateral direction, the vertical direction and
the torsional direction are taken as 1.0%, 1.0% and 0.5%, respectively.
where M, K and C are mass, stiffness and damping matrices; X, X_ and X € are displace-
ment, velocity and acceleration vectors, with the subscripts v and b representing the
vehicle and the bridge, respectively; Fv and Fb are interaction forces between the vehicles
and the bridge; Fbf and Fse are modal buffeting force vector and the self-excited force
vector. Details of the matrices in Equation (6) can be found in [23]. Equation (6) is a
second-order linear nonhomogeneous differential equation with time-varying coefficients,
which is solved using the Newmark implicit integral algorithm with β = 1/4 in this study.
50
−25
−50
0 30 60 90 120 150 180
Time (s)
(a) Measured
50
Lat Acc (mm/s2)
25
0
Downloaded by [University of Kent] at 09:10 24 November 2014
−25
−50
0 30 60 90 120 150 180
Time (s)
(b) Computed
Figure 6. Comparison of measured and computed deck lateral acceleration time histories at
Section D.
300
Vert Acc (mm/s2)
150
−150
−300
0 30 60 90 120 150 180
Time (s)
(a) Measured
300
Vert Acc (mm/s2)
150
−150
−300
0 30 60 90 120 150 180
Time (s)
(b) Computed
Figure 7. Comparison of measured and computed deck vertical acceleration time histories at
Section G.
Section G is located at mid-main span. The measured time histories have gone through a
lowpass filter of a 1.2 Hz cutoff frequency. According to the measurements, the train runs
on the bridge at a constant speed of 99 km/h. The mean wind velocity is 18.9 m/s, and the
turbulent intensity is 14.9% in the horizontal direction and 7.9% in the vertical direction.
188 W.W. Guo et al.
100
50
−200
Downloaded by [University of Kent] at 09:10 24 November 2014
Measured
−400 Computed
−600
0 20 40 60 80 100 120 140 160 180
Time (s)
(b) Section G
Figure 8. Comparison of measured and computed deck vertical displacement time histories.
The mean wind speed is almost perpendicular to the bridge alignment. It can be seen that
the two sets of data are in good accordance.
Figure 8 illustrates the measured and computed time histories of deck deflection
responses at Sections B and G. Section B is located at 1/2 of Ma Wan side span. To
capture the main effects of moving load of the train on the bridge, both the measured and
computed time histories of vertical displacement response go through a lowpass filter of a
0.05 Hz upper frequency.
It can be clearly seen that the moving single train causes a single peak vertical
displacement response at a section when the train passes over this section. The moving
train induces larger peak displacement response at the main span than at the side span.
The relative discrepancies of the two sets of results are 27% at Section B and 77% at
Section G. Although the computed peak vertical displacements are greater than the
measured ones, the numerical computation does predict the effect of a moving train on
the bridge.
In this study, no measurement data of the running trains are available and only the
computed results are presented for illustrative purposes. Figure 9 shows the time histories
of lateral and vertical acceleration responses of the vehicle on the bridge.
There are three important indices currently adopted in Chinese railways in the
evaluation of the running safety of train vehicles under wind actions. The first index is
the derailment factor Q/P, the ratio of lateral force Q acting on the wheelset to the total
vertical force P acting on the same wheelset. The total vertical force is the sum of the self-
weight of the vehicle per wheelset Ps and the dynamic vertical force Pd on the wheelset.
The second index is the offload factor ΔP/P, the ratio of the vertical force difference ΔP to
the total vertical force P acting on the wheelset with ΔP = Ps – Pd. The third index is the
overturn factor D induced by the direct wind actions on the car body. Since the train runs
International Journal of Rail Transportation 189
0.4
−0.2
−0.4
0 30 60 90 120 150 180
Time (s)
(a) Lateral
0.4
Max Vert Acc (m/s2)
0.2
0
Downloaded by [University of Kent] at 09:10 24 November 2014
−0.2
−0.4
0 30 60 90 120 150 180
Time (s)
(b) Vertical
inside the Tsing Ma Bridge deck, only the first two indices are hereinafter considered. The
allowable derail factor Q/P and the offloading factor ΔP/P specified in the Chinese design
guideline are 0.8 and 0.6, respectively.
Displayed in Figure 10(a) and (b) are, respectively, the distributions of the maximum
derailment factor and offload factor of the vehicle wheel versus the train speed when the
train runs through the bridge under several different wind velocities. It can be seen that
these two factors increase significantly with the mean wind velocity, whereas fluctuations
appear with the train speed.
By taking the mean wind velocity and the train speed as the main parameters that
control the running safety of train vehicles, the threshold speed of train vehicles travelling
on the bridge under turbulent winds can be determined according to the aforementioned
safety evaluation indices in the following way:
(1) By keeping the mean wind velocity constant at each stage, the dynamic responses
of the train vehicles are calculated by changing the train speed. The critical train
speeds at which the two indices exceed the allowances are obtained.
(2) By changing the mean wind velocity, the two indices of the train vehicles are
calculated at different train speeds.
(3) By plotting the calculated results under all wind velocities and the corresponding
train speeds in a same coordinate system, the relationships between the mean
wind velocity and the critical train speed can thus be calculated.
The threshold curve of wind velocity for ensuring the running safety of the train in the
bridge deck is then calculated, as shown in Figure 11, from which the allowable train
speed at different wind velocities can be determined. It can be seen that with the increase
190 W.W. Guo et al.
0.7
0.6
0.5
0.4
0.3
0.2
U = 0 m/s U = 10 m/s
0.1
U = 15 m/s U = 20 m/s
0
60 80 100 120 140 160 180 200
Train speed (km/h)
(b) Offload factor
Figure 10. Maximum derailment factor and offload factor versus train speed.
250
200
Train speed (km/h)
Dangerous area
150
100
50 Safe area
0
16 18 20 22 24 26 28 30
Mean wind velocity (m/s)
Figure 11. Critical train speed for running safety versus mean wind velocity.
of mean wind velocity, the critical train speed drops off remarkably. As the mean wind
velocity reaches 30 m/s, the critical train speed is around 2 km/h, indicating that the rail
traffic on the Tsing Ma Bridge should be closed to ensure running safety of the train. This
method can be applied to other railway bridges to determine their critical train speed
curves by taking into account the mean wind velocity on site and the dynamic properties
of the bridges.
International Journal of Rail Transportation 191
4. Conclusions
The framework for performing dynamic analysis of coupled train and long-span suspen-
sion bridge systems under cross winds has been established and then applied to the Tsing
Ma suspension bridge. The bridge responses are computed and compared with the
measured responses. The train responses are also computed and their performance is
evaluated. The major results are summarised as follows:
(1) The computed acceleration and displacement responses of the bridge deck agree
well in general with the measured ones.
(2) The threshold curve of wind velocity and train speed for ensuring the running
safety of the train in the Tsing Ma Bridge is proposed according to the two safety
evaluation indices in Chinese railway design code.
(3) The critical train speed drops off remarkably with the increase of mean wind
Downloaded by [University of Kent] at 09:10 24 November 2014
velocity. As the mean wind velocity reaches 30 m/s, the critical train speed is
around 2 km/h, indicating that the rail traffic on the Tsing Ma Bridge should be
closed.
(4) The numerical results demonstrate that the proposed framework and the asso-
ciated computer programme can efficiently predict the dynamic response of
coupled train–bridge system in turbulent winds with reasonable computation
efforts.
Acknowledgements
This work was supported by the National Key Basic Research Program of China (973 Program)
under grant 2013CB036203, the Natural Science Foundation of China (50838006) and the
Fundamental Research Funds for the Central Universities (2012JBM075). The authors express
their sincere acknowledgement to Professor Y.L. Xu for his guidance when W.W. Guo and N.
Zhang were working in the Hong Kong Polytechnic University.
References
[1] Diana G. Wind effects on the dynamic behavior of suspension bridge. Milano: International
Report; 1986.
[2] Scanlan RH, Jones NP. Aeroelastic analysis of cable-stayed bridges. J Struct Eng. 1990;116(2):
279–297.
[3] Simiu E, Scanlan RH. Wind effects on structures. New York, NY: John Wiley & Sons; 1996.
[4] Strømmen E, Hjorth-Hansen E, Hansen SO, Jakobsen JB. Aerodynamic investigations for the
tender design concepts of the Øresund cable-stayed bridge. J Wind Eng Ind Aerod.
1999;80:351–372.
[5] Cao YH, Xiang HF, Zhou Y. Simulation of stochastic wind velocity field on long-span bridges.
J Eng Mech. 2000;126(1):1–6.
[6] Chu KH, Garg VK, Wiriyachi A. Dynamic interaction of railway train and bridges. Veh Syst
Dyn. 1980;9(4):207–236.
[7] Bhatti MH. Vertical and lateral dynamic response of railway bridges due to nonlinear vehicle
and track irregularities [PhD dissertation]. Chicago, IL: Illinois Institute of Technology; 1982.
[8] Tanabe M, Yamada Y. Model method for interaction of train and bridge. Comput Struct.
1987;27(1):119–127.
[9] Diana G, Cheli F. Dynamic interaction of railway systems with large bridges. Veh Syst Dyn.
1989;18(1–3):71–106.
[10] Bogaert V. Dynamic response of trains crossing large span double-track bridges. J
Constructional Steel Res. 1993;24(1):57–74.
[11] Frýba L. Dynamics of railway bridges. London: Thomas Telford; 1996.
192 W.W. Guo et al.
[12] Yang YB, Yau JD. Vehicle-bridge interaction element for dynamic analysis. J Struct Eng.
1997;123(11):1512–1518.
[13] Xia H, Xu YL, Chan THT. Dynamic interaction of long suspension bridges with running
trains. J Sound Vib. 2000;237(2):263–280.
[14] Baker CJ. Ground vehicles in high cross winds – part II: unsteady aerodynamic forces. J Fluids
Struct. 1991;5:91–111.
[15] Xu YL, Xia H. Dynamic response of suspension bridge to high wind and running train. J
Bridge Eng. 2003;8:46–55.
[16] Xu YL, Zhang N, Xia H. Vibration of coupled train and cable-stayed bridge system in cross
wind. Eng Struct. 2004;26:1389–1406.
[17] Li YL, Qiang SZ, Liao HL, Xu YL. Dynamics of wind-rail vehicle-bridge systems. J Wind
Eng Ind Aerod. 2005;93:483–507.
[18] Xia H, Guo WW, Zhang N, Sun GJ. Dynamic analysis of a train-bridge system under wind
action. Comput Struct. 2008;86:1845–1855.
[19] Xu YL, Zhu LD. Buffeting response of long-span cable-supported bridges under skew winds –
part II: case study. J Sound Vib. 2005;281(3–5):675–697.
Downloaded by [University of Kent] at 09:10 24 November 2014
[20] Lin YK, Yang JN. Multimode bridge response to wind excitations. J Eng Mech. 1983;109(2):
586–603.
[21] Guo WW, Xia H, Xu YL. Running safety analysis of a train on the Tsing Ma Bridge under
turbulent winds. Earthq Eng Eng Vib. 2010;9(3):307–318.
[22] Xu YL, Ko JM, Zhang WS. Vibration studies of Tsing Ma suspension bridge. J Bridge Eng.
1997;2(4):149–156.
[23] Guo WW, Xu YL, Xia H, Zhang WS, Shum KM. Dynamic response of suspension bridge to
typhoon and trains – part II: numerical results. J Struct Eng. 2007;133(1):12–21.