Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
hxia@center.njtu.edu.cn
The objective of this paper is to study the dynamic impacts of moving light-rail trains
on viaduct bridges and the vibrations induced on the surrounding ground. The study is
based on theoretical analysis and field experiment. The theoretical study is performed
in two stages using a 2D interaction model of “train-bridge” system for obtaining the
dynamic loads of moving trains on the bridge piers, and a 2D dynamic model of “pier-
foundation-ground” system for analyzing the vibration responses of the surrounding
ground. In the field experiment, the dynamic responses of bridge piers and the ground
accelerations are measured at different distances under different train speeds.
1. Introduction
With the rapid development of modern industry, vibration effects on residents’
living and working environment in big cities have become more and more serious.
The traffic-induced vibrations that seemed to have been tolerated in the past are
nowadays increasingly being considered as a nuisance. The vibration source, trans-
mission path and control methods are all under study, among which the influences of
light-rail-train borne vibrations on the surrounding environment have also brought
to the attention of engineers and researchers in China and abroad. The generation
mechanism of the vibrations, their propagation properties in the ground, pollution
to the environments, harms to residents and their control measures have been the
focus of recent studies.
Japan is one of the countries where the environmental vibration is serious. In its
“Law of Limiting Vibration”, the traffic induced vibration is specially mentioned,
for which measures should be taken to protect the environments and people’s
health. Fujikake1 proposed a predicting method for the vibration levels of the sur-
rounding environment. Yoshioka2 predicted the environmental vibration induced by
Shinkansen (high speed railways). The Federal Railway of Switzerland (FRS) and
the Experiment Research Institute (ORE) of the International Railway Union (UIC)
jointly performed a project, proposing a way for improving the track structures such
227
June 26, 2002 13:22 WSPC/165-IJSSD 00048
that the influence of the metro trains on the surrounding ground and underground
structures can be reduced.3,4 In Spain,5 Belgium,6 Czech7 and some other coun-
tries, efforts have also been bestowed on this field. The propagating and attenuating
properties of the traffic induced ground vibrations are systemically studied by the
statistics methods using the measured results of different ground soils. Furthermore,
vibration reduction measures are provided by decreasing the vehicle speeds, reduc-
ing vehicle loads and improving the smoothness of road surfaces. Recently, Yang and
Huang studied the elastic waves in visco-elastic half-space generated by various vehi-
cle loads using the finite/infinite element modeling approaches.8,9 Xia et al. studied
the characteristics of traffic induced vibrations and their effects on the environment
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
Today, light-rail traffic systems are under planning in several big cities in China,
owing to its advantages of fast speed, high safety, on schedule in operation and large
transporting capacity. This paper attempts to study the properties of the ground
vibrations induced by light-rail trains via viaduct system using theoretical analysis
and field measurement.
2. Theoretical Study
In this paper, the theoretical study is performed in two stages using two dynamic
models: a 2-dimensional “train-bridge” system model for obtaining the dynamic
loads on piers and a “pier-foundation-ground” system model for analyzing the vi-
bration responses of the surrounding ground.
2*22m
structure is dominantly influenced by the several lowest modes, this approach has
the advantage that an adequate estimation of the dynamic response can be obtained
by considering only a few modes of vibration, thereby reducing the computational
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
efforts required.
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
ϕ1i ϕ2i
Z1i Z2i
bogie Zs
cj
kj
Zwij
beam
lj
L=29.7m
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
movements Zwij and the beam displacements Zb (xij ) satisfy the relation:
X
N
Zwij = Zb (xij ) + Zs (x) = qn φnij + Zs (x) (4)
n=1
where Zs is the vertical displacements of the wheel relative to the rails, represent-
ing the vertical rail irregularities of the track on the bridge. The wheel movements
can then be expressed by the linear composition of the bridge modal coordinates
qn (n = 1, 2, . . . , Nq ).
The dynamic equilibrium equations of the vehicle-bridge system can be obtained
by the composition of the vehicle equations and the bridge equations:
J2 J2 J2
Mi + −
L2 L2 L Z̈1i
J2 J J
− Mi + 12 − 2 Z̈2i
L2
L L
J2 J ϕ̈1i
− 2 J1 + J2
L L
!
X
4 X 8
lj2 X8
lj lj2 X4 X 8
lj2
cj + cj − η + 2 cj − ηlj cj + cj
L2 L L L
j=1 j=5 j=5 j=1 j=5 Ż1i
! 8 2 8
X 8
lj2 X X
− lj lj
η−
lj
+ η + 2 cj 1+η cj cj Ż2i
L L L L
j=5 j=5 j=5
ϕ̇1i
X 4 X 8
l 2 X8
l X 8
− ηlj cj +
j
c η−
j
cj 2
lj cj
L j L
j=1 j=5 j=5 j=1
June 26, 2002 13:22 WSPC/165-IJSSD 00048
!
X
4 X8
lj2 X8
lj lj2 X4 X 8
lj2
kj + kj − η + 2 kj − ηlj kj + k
L2 L L L j
j=1 j=5 j=5 j=1 j=5 Z1i
! 8 2 8
X 8
l 2 X X
− lj j lj
η−
lj
+ η + 2 kj 1+η kj kj Z2i
L L L L
j=5
ϕ1i
j=5 j=5
X 4 X lj
8 2 X8
lj X 8
− ηlj cj + kj η− kj lj2 kj
L L
j=1 j=5 j=5 j=1
4
X n X8
lj n
φij (kj qn + cj q̇n ) − η φij (kj qn + cj q̇n )
L
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
j=1 j=5
N X
X 8
lj
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
n
= 1−η φij (kj qn + cj q̇n )
L
n=1
j=5
X
8
n
− ηl φ (k
j ij j n q + c q̇
j n )
j=1
X4 X 8
lj
[kj Zs (xij ) + cj Żs (xij )] − η [kj Zs (xij ) + cj Żs (xij )]
L
j=1 j=5
X 8
lj
+ 1−η [kj Zs (xij ) + cj Żs (xij )]
L
j=5
X
8
−
ηl [k
j j Z s (x ij ) + cj Żs (x ij )]
j=1
(
2
X
Nv X
8 X
N
q̈n +2ξn ωn q̇n +ωn qn = φn
ij − φm
ij [mij q̈m +ci q̇m +ki qm ]
i=1 j=1 m=1
X
4 X
8
lj
+ (kj Z1i +cj Ż1i )− η (kj Z1i +cj Ż1i )
L
j=1 j=5
8
X
lj
+ 1+η (kj Z2i +cj Ż2i )
L
j=5
X
8 X
8
− ηlj (kj ϕ1i +cj ϕ̇1i )− [mij Z̈s (xij )+cj Żs (xij )]+kj Zs (xij )]
j=1 j=1
)
1
+ g(Mi +mij ) (i = 1, 2, . . . , Nv , n = 1, 2, . . . , Nq ) . (5)
8
Since the coefficients “φ” are changing continuously when the train moves over
the bridge, the dynamic equations of the vehicle-bridge system should be recog-
nized as a set of 2-order linear simultaneous differential equations with time-varying
coefficients. The equations are solved by the Newmark β step by step integration
June 26, 2002 13:22 WSPC/165-IJSSD 00048
algorithm with β = 1/4. When the generalized coordinates qn (t) is obtained, the
reaction Rr (t) of support r can be determined by
X
N
Rr (t) = qn (t)φnr · 4Sr · Er /Hr (6)
n=1
where φnr is the value of the nth modal function of the beam at support r, and Sr ,
Er and Hr are cross area, elastic modulus and thickness of the rubber bearings.
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
By the vehicle-bridge system model, the whole histories of a train running over the
bridge is simulated by computer. Two spans are taken as the calculation part, in
which six modes are considered for each span. The train speed is 70 km/h; the
integration time interval is 0.005 second.
Figure 3 shows that the calculated time history of the dynamic reaction force
acting on a pier, from which the dynamic behaviour of the acting force induced by
the passing train can be appreciated.
The principal frequency of the force wave is approximately 2.6 Hz. It is just the
loading frequency f = V /La = 19.44/7.425 ≈ 2.6 Hz of the train (with the average
interval of bogies La = 7.425 m and the train speed V = 70/3.6 = 19.44 m/s),
which indicates that the loading frequency of the train plays an important role on
the forces.
Table 3 lists the comparison of the static and dynamic reactions and their dy-
namic factors. Since the two support reactions on a pier are in different phases
(with the interval 1.10 m between the two supports on one pier, the average bogie
distance 7.425 m, the phase difference between the two support reactions is about
170◦ when V = 70 km/h, approximately in inverse phases), the dynamic loading
(resultant reaction 1 + 2) of the train on the pier is much smaller than the algebraic
sum of the two individual supports.
F(kN)
600
400
200
t(sec)
0
0 1 2 3 4 5 6 7 8 9
• The foundation and the surrounding soils are in close contact during the whole
course of response.
• Since the train induced vibrations of the system are very small, the system is
assumed to be elastic.
Dynamic loading
Pier
Medium
sand
Piles
H grit
L
The time histories of responses at other points are almost the same, the only dif-
ference is the magnitudes. Figure 6 shows the maximum acceleration and velocity
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
responses of the ground surface versus the distance to the pier on shallow founda-
tion and pile foundation. As can be seen, the simulation range of the FEM model
should be large enough to reduce the boundary damping and reflection effects.
The FEM network for pile foundation is the same as that for shallow foundation.
The pile length is 13 m. The shapes of the vibration curves are almost the same as
those for shallow foundation. One can see from these figures that the ground vibra-
tions are much smaller for pile foundation than for shallow foundation. Compared
with the results of shallow foundation, the calculated acceleration and velocity are
only 2.51/4.02 = 1/1.6 and 0.24/1.13 = 1/4.7, respectively.
Acceleration (mm/s/s)
10
t(sec)
0
1 2 3 7 8 9
4 5 6
-10
(a)
Velocity (mm/s)
3.0
1.5
t(sec)
0.0
1 2 8 9
3 7
4 5 6
-1.5
-3.0
(b)
Displacement (mm)
4.0
3.0
2.0
1.0
t(sec)
0.0
0 1 2 3 7 8 9
4 5 6
(c)
4.5 1.2
4.0 Shallow 72dB Shallow
1.0
Ground acceleration (gal)
3.5 Pile Pile
1.5 0.4
1.0 60dB
0.2
0.5 54dB
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
0.0 0.0
0 4 8 12 16 20 0 4 8 12 16 20
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
3. Experimental Study
The experiment was carried out at the Daqinghe Bridge site. The purpose of this
experiment was to study the vibration level and the distribution characteristics
of the train induced ground vibration at different distances under different train
speeds.
^ _ a b c e g i j l m n o p q
! " # % ' (
H I K L
) * + , - . / 0 1
M N P R S U V X Y Z [ \ ]
2 3 4 6 8 9 ; < = > ? A B C D E F G
r s t u w x z { | } ~
different distances under different train speeds were obtained from the measure-
ment, as listed in Table 5.
The vibration levels are calculated on the basis of G = 20 lg(a) + 60, where G
is the vibration level, in dB; a is the measured acceleration, in cm/s2 .
9:
6 78
3 4
/01
, -.
% & ( )
! # $
(a)
78
4 56
1 2
-./
# $ & '
* +,
! "
(b)
becomes greater. There are 7–8 dB difference between the maximum and average
vibration levels of the pier under the train speed difference of 20 km/h.
67
3 54
0 1
-./
+ ,
! # $
& '(
(a)
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
56
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
2 34
/ 0
,-.
* +
! " # $
% &'
(b)
110
100 V=80
Ground vibration level (dB)
V=75
V=70
90
V=65
80 V=60
70
60
50
D(m)
40
-10
0
10 20 30 40 50 60 70
is 100.5 dB, occurring at the foot of the pier (0 m). The measured data show 37–
43 dB decrease for the vibration levels in the distances from 0 to 60 m at different
test train speeds. The distribution of the average vibration levels at different train
speeds versus the distance D to the vibration source shows the same trend as the
maximum ones, with a maximum average value of 96.1 dB.
June 26, 2002 13:22 WSPC/165-IJSSD 00048
110
100 D=0m
90 D=20m
80
70 D=40m
60
D=60m
50
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
V(km/h)
40
by UNIVERSITY OF QUEENSLAND on 03/13/15. For personal use only.
55 60 65 70 75 80 85
4. Conclusions
The following conclusions can be drawn from the theoretical study and field exper-
iment of the example considered in this paper:
• The train induced ground responses obtained from the theoretical analysis show
the same trends as those from the field measurement, which partly proved the
reliability of the analytical model and the computer simulation method.
• The ground vibration induced by light railway trains, as well as by ordinary
trains, attenuates with the distance to the bridge pier (point vibration source).
• The ground vibration induced by running trains at places near a bridge increases
with the train speed. When the speed of the test train changes from 60 km/h
to 80 km/h, the maximum vibration level of the ground near the pier foot is
80.5 dB–100.5 dB.
• The type of bridge foundations of light rail systems has important influence on
the ground vibration. Using pile foundations can greatly reduce the responses
induced by moving train loads.
June 26, 2002 13:22 WSPC/165-IJSSD 00048
• For the light rail system where rubber bearings are adopted as in the example
considered, the maximum velocities at ground surface 5.2 m from the pier are
0.63 cm/s and 0.129 cm/s for shallow foundation and pile foundation, respec-
tively, see Fig. 6.
• The ground vibration around the light-rail bridge system is mainly induced by
the dynamic impacts of trains as moving loads. In the design of bridge system,
special attention should be paid to avoid resonance.
• The ground vibration induced by light rail trains and viaduct system is much
smaller than that induced by trains (with heavier loads) in ordinary railways,
compare the results of Fig. 6 and Fig. 11.
Int. J. Str. Stab. Dyn. 2002.02:227-240. Downloaded from www.worldscientific.com
materials often leads to safer results. The example and the corresponding results,
therefore, can only be regarded as a case study. Further study should be carried
out for different types of field conditions.
Acknowledgments
This study is supported by the Doctoral Research Foundation of the Education
Ministry of China (Number 2000000407).
References
1. T. A. Fujikake, “A prediction method for the propagation of ground vibration from
railway trains,” J. Sound Vibration 111(2), pp. 289–297, 1986.
2. O. Yoshioka, “Basic characteristics of Shinkansen-induced ground vibration and its
reduction measures,” Proc. Wave 2000, pp. 219–239, 2000.
3. E. Tassily, “Interaction dynamique voie/roue: Modeles existants et perspectives de
recherche,” Revue Generale des Chemins de Fer, 107, pp. 23–30, 1988.
4. ORE: Question D151 DT217 Measures against Structure Borne Noise and Vibrations
Technical Report, Office for Research and Experiments of International Union of
Railways, Utrecht, NL, 1989.
5. G. Volberg, “Propagation of ground vibration near railway tracks,” J. Sound Vibration
87(2), 371–376, 1983.
6. G. Degrande and G. Lombaert, “High-speed train induced free field vibrations: In
situ measurements and numerical modelling,” Proc. Wave 2000, pp. 29–42, 2000.
7. T. M. Dawn and C. G. Stanworth, “Ground vibrations from passing trains,” J. Sound
Vibration, 66(3), pp. 355–362, 1979.
8. H. H. Huang and Y. B. Yang, “Elastic waves in visco-elastic half-space generated by
various vehicle load,” J. Soil Dynam. Earthq. Eng. 21, pp. 1–17.
9. Y. B. Yang and H. H. Huang, “A 2.5D finite/infinite element approach for mod-
elling visco-elastic bodies subjected to moving loads,” Int. J. Numer. Meth. Eng. 51,
pp. 1317–1336, 2001.
10. H. Xia, N. Zhang and J. J. Qu, “Traffic-induced vibrations and their influences on sur-
rounding environments,” Proc. Int. Conf. Mod. Composite Con., University of Delft,
Netherlands, pp. 123–130, 2000.
11. H. Xia, “Characteristics of traffic induced vibrations and their effects on environ-
ments,” Proc. TIVC’2001, pp. 83–90, 2001.