13 Vibration Problems

39

steps. increasing the ampitude of the bridge oscillations, and

foing the effect (Strogatz et al. 2005). Now, the continually rein
Domenon has become a problem that
collective synchronous phe
There are many such accidents.
should be considered in bridge design.
vears of 1825
According to
2000 there happened 39 bridgeincomplete statistics, during he
collapses
with 33 events reported casualties and over 800 deaths. related to the crowd,
with the development ot footbridge
conous new types ot tootbndges, the structures, and continuous appearance of
.ha main problems that inttuencing thecrowd-induced vibration has become one of
w0) The charactenstics of the crowdserviceability
load, its
of footbridges (Chen and Hua
lated control criteria should be considered in vibration effect on man and the
heen aconsensus that dynamuc analysis shouldfootbridge design. Nowadays, it has
be applied in the design of foot
bridges to ensure their serviceability.

L3.5 Bridge Vibration Induced by Running Trains

In recent vears. with durative raise of train speed, steady
growth of traffic density
and continuous increase in vehicle load, the problem of dynamic
interaction of
rain-bridge system becomes more and more important. On the one hand, the
high-speed train will make dynamic impact on bridge structures (Fig. 1.29),
them to vibrate, which directly affects the working status and service lifecausing
of the
bridge. On the other hand, the vibration of bridge will in turn affect the running
stabilitv and safety of the train, which makes the vibration state of bridge structure
become an important index to evaluate the dynamic design parameters for bridge
structure. Thus, it is the actual requirement for engineers to carry out comprehen
sive studies on dynamic interaction of coupled train-bridge system, so as to make
dynamicanalysis and assessment on the dynamic properties of bridge structure and

Fig. 1.29 Thalys train

running through a HSR bridge
Belgium)
 1 Introducüon
40 factor
train-bridge syslem andintlucncing
interaction of
Table 1,7 Dynanic Relatedintluence faclor
Iuhlem Lateral vibration
Vertical vibration
Composition and
running speed
of train: axle Boad and axle
Moving gravity loading on distance of vehicle, etc.
bya
bdge structure induced with
bndge
train nunning on the
car-body, bogies and Masses car-body, bogies and
acertain spoed of
induced Masses of
Impact action on bridge lateral stifness and wheel-sets, vertical sufness and
wheel-sets,
by inertial fones of
car-body. suspension devices
whecl-sets when a damping of suspension devices damping ot
bogies, and
brdge with a
train runs on the
certain spced Vertical deflection ad vibration
Dynamic efect of bridge Lateralvibration of bridge; of bridge:; vertical deformation
defonation and vibration on lateral defornation induced by
temperature difference induced by temperature
nunning vehicle difference and PC beam creep.
and pier foundation setlement
Vertical and cross-level,
Track imegularity, the random Alignment and gauge, geometric and dynamic
dynamic
) excitation source of train-bridgegeometric and iregularities
system vibration when a train imegularities
passes bridge with certain speed
Hunting movenment of vehicle Wave-length and amplitude of
wheels, a periodic random hunting movement
source of self-excited forced
vibration, caused by
wheel-tread taper and wheel-rail
gap
VUnevenness of wheel treads Lateral periodical load Vertical periodical load
Dynamic interaction of coupled Structural form and span of Structural form and span of
train-bridge system, and bridge; composition and axle bridge; composition and axle
resonance appeaing at a certain arrangement of train; properties arrangement of train; properties
train speed of lateral spring and damper; of vertical spring and damper;
lateral and rotational vertical frequencies of train and
frequencies of train and bridge bridge
Curved bridge: lateral and Lateral moving load by
rotational vibration induced by centrifugal force of train
centrifugal forces of moving vehicles
vehicles
Earthquake: the influence of Lateral seismic ground motion Vertical seismic ground motion
earhquake spatial variability on
long bridge, and safety of
running train on bridge during
carthquake
Wind: vibration of coupled Effect of
lateral turbulent wind,Vertical vibration of
train-bridge system induced by dynamic action on bridge of bridge
combined action of wind and induced by wind, and the effect
running train with lateral wind
running rain, and overturn of mean wind
stability of train vehicles pressure, and structural
deformation
wind
induced by mean
Collision loads: lateral or
Magnitude, direction and acting Magnitude, direction and
longitudinal vibration of bridge patterm
induced by collision of vessel, load
of horizontal
colliding pattern of vertical collidingacting
load
vehicle and floating object
|3 Vihration Poblems
4!

the running qualities of trainvehicles, and to

train and the bridge, deternine the servie reliahility of the
When a train travels on a railway bridye, it will
incvitably cause the bridge to
vibrate. In this case, the bridge bears not only the stalic
load by loading and unloading of thetrain at a ertain load, but also the rOVing
and other dynamic clfects induced by the train and 'spee, and the incrtial forES
vibration of the bridge induced by thee dynamic the bridpe in vibration. The
loads
structure components, reducc their strcngth and tatbility, asdnay cause faligue o!
even alfect the safety
and stability of the train running on bridge when the
loading frequency of the train equals to o approachesvibration
is serious. When the
bridge, he resonance will occur, which will the natural frequency of the
intensify the dynatnic response o
coupled train-bridge system, or even cause an accident. A typical
England in 1847, when the Chester Dee Railway Bridge was broken example was in
excessivc vibration induced by a train passing through it (Lewis 2007). owing w
serious accident provides us with disastrous experience and Jesson, which Every
engineers to constantly inprove the structure design, to adapt the objectivepromotes
Therefore, the dynamic interaction of train-bridye system has been attached laws.great
importance and studied by bridge designers in nany countries.
From the point of view of vehicle dynarmnics, cach railway locomotive or vehícle
is a vibration system composed of a car-body, bogies,
whecel-sets and
spring-damper suspension devices. The running train may produce dynarnic inpact
on the bridge, and cause it to vibrate, while the vibration of the bridge will in turn
affect the running stability and safety of the train. Therefore, the vibrations of the
train and the bridge interact on and interaffect each other, forming a multi-DOFs
(degrees-of-freedoms) complex vibration system.
The dynamic interaction of train-bridge systemn involves various problems,
which are summarized in Table 1.7, with the related influence factors.

14 Research History and Status Quo of Train-Bridge

Coupling Vibration: An Overview
1.4.1 Research on Vibrationof Railway Bridges Under
Moving Train

As early as in 1840s,researchers started to study the dynamic response of railway

bridges. However, as the bridge vibration under moving train loads is a complex
0ssue, many factors should beconsidered in the analysis to get the realistic results.
These factors include the mass of car-body and bogie, the effect of dashpot and
spring on the bogie, the running speed of train, the mass, stiffness and damping of
pier and beam span, the structure and dynamic performance of the track on bridge,
the dynamic interactions between whecl and track, track and beam, etc. as well as
various Yandorn factors such as the ynevenness of wheel rim, the geometric and
42

of track, andthe hunting movement of wheel.

dynamic irregularities
extremely complex. Therefore, previous studies had
oUs approximation
analysis model methods, with grcatlimitations. Only in the recent decales, vint
the wide application of computersand the development of a variety of numerca
solution methods, was promoted the rescarch of this problem. At present, e
dynamic analysisof train-bridgecoupling systemhas become an important TeSearc
of bridge vibration.
topicDuring
inthcthe
ficldlong evolution history of analysis model for train-bridge cOupling
model was developed from the simple inoving force
vibration, the train wheel-spring-mass unil, after thhat to
half-vehicle and then the moving
to spatialfull train model; andthe bridge model was fromthe,
moving mass,model, untilthe
model offul)
constant sectionto the finite element bridge with
the single beam with structures, gradually
foundations, and other accessory
bearings, piers, analysis system.
forming a
complete train-bridge dynamic

1.4.1.1 Researches Abroad

Willis
by neglecting the bridge mass but considering the wheel mass,
In 1849, established the differential equation of
(1849) and Stokes (1 849), respectively,
derived its approximate and exact solu
bridge vibration under moving loads and basic theoretical solution and
tions. Later, Robinson (1887) also obtained the difference between the calculated
conducted the related vibration tests, but the
the bridge mass.
results and those from the tests was large, due to neglecting
carried out three
In the fist half of the 20th century, American engineers
large-scale field tests. In first test, the maximum dynamic responses of the bridge
were measured under various kinds of locomotives passing the plate beam and truss
bridge at ifferent speeds, finding that the hammer effect induced by the unbalanced
wheel weight was the main excitation of bridge vibration, based on which the
concept of "impact factor" was proposed. In the subsequent tests, the resonance
phenomenon was found when the wheel rotation speed of the steam locomotive
matched with the bridge fundamental frequency, and accordingly the concept of
"critical speed" was put forward. The third test was mainly on some small span
bridges, where the difference between steam locomotive and diesel locomotive was
compared, and the vibration attenuating effects of pier elasticity, ballast layers, and
elastic cushions were investigated.
During the same period, researchers in the former Soviet Union caried out a lot
of tests, mainly studied the lateral vibration of bridge induced by locomotive and
other factors, and set up the empirical formulas of the stress coefficient induced by
lateral vibration with respect to train speed. Timoshenko (1922, 1927) established a
bridge model under moving loads by only considering the bridge mass and obtained
its analytical solution, to interpret the resonance
mechanism in theory. By com
bining the theoretical analysis and experimental study and considering the inertial
forces of both locomotive and bridge masses, Inglis
(1934) used the moVinE
periodical forces to solve the motion equation of train-bridge system,which brougnt
L4 Rescarch History and Status Quo ...
43

the research of train-bridge dynamicsystem to a new

into account the masses of a single wheel and bridge,level. Subsequently, by taking
the Fourier series expansionof wheel inetia force Sehallenkamp (1937) studied
and bridge deflection and derived
the analytical solution by solving the
relationship among the series coeficients.
Muchnikov (1953)carTied out accurate analysis on the above models by the integral
equation method.
In addition, Carter (1926), as the
founder of the rolling contact theory, skillfully
analyzed and presented the relationship between the longitudinal and the tangential
wheel-rail forces with the creep rate, according to the Hertz theory
half-space theory. On the basis of Carter's research, Johnson (1958)andproposed
the elastic
the
pure creep roll1ng and pure spin rolling contact theory, and generalized it to the
three-dimensional rolling contact. In the 1970s, Kalker (1973, 1979) eventually
proposed the three-dimensional linear creep theory for wheel-rail rolling contact,
which has been widely used in the wheel-rail relationship analysis.
In this
stage, although there was still a large gap between the analysis and the
actual results, the moving sprung-mass model was developed, being the prototype
of the modern train-bridge analysis model, and some basic concepts such as
impact
coefficient and resonance etfect were established. All these research laid a solid
foundation for the subsequent development of the modern train-bridge coupling
vibration theory.
Entering the 1960s, with the aid of the structural FEM and the progress of
computational technique, the research on vibration problem of train-bridge system
broke through the traditional methods, stepping into the new period of system
research. During this period, along with the emergence and development of
high-speed railways, the neds for the dynamic analysis of train-bridge coupling
system became more urgent, and meanwhile, a large number of railway construc
tions made the field tests more convenient. Under this background, through the
systematic and in-depth efforts of researchers in various countries,the research on
train-bridge coupling vibration, from mechanical model, excitation source simula
tion to analysis method and calculation technique, all achieved qualitative devel
opments and obtained a lot of results in model tests and field experiments.
In the United States, Chu et al. (1979) took the lead in the research on spatial
vibration of train-bridge system. They proposed the vehicle model with 11DOFs by
connecting the rigid bodies of car-body, bogies, and wheel-sets with the vertical
suspensions and established the train-bridge system spatial analysis model by
simplifying the truss mass into the nodes. Wiriyachai (1980) investigated the impact
coefficient and the fatigue damage of truss bridge with rigid, hinged and semi-rigid
nodes. Bhatti (1982) analyzed the spatial dynamic responses of the train-bridge
system with the 21 DOFs' model considering the vehicle suspension nonlinearity.
Later, Wang et al. (1991) improved Bhatti's vehicle model to 23 DOFs to get a
more realistic simulation movement of wheel. Vu-Quoc and Olsson (1989) estab
Iished the nonlinear differential equations for the vehicle and flexible structure
system, using the Runge-Kutta method and the linear multi-step integral method, to
predict and modify the unconditional stability of the train-bridge interaction
problem.
 Japan. Matsuura (1974, 1976) studied the dynamic responses of ahalf- \ntroi cion
irviethidk
In
model passing the bridge, and
based on the results, proposedthe deflection
suspension vehicle
the bridge.
DOFs, Theythe
deduced established
also motion atwo-layer
equation analyzed fohex
by the energy method and model
composition
tntiuence of several factors (vehicle parameters, traun and unNing
Speed. bidge span, trackiregularity, etc.) on the bridge dynamic coeficients.
when the train speed reached 300 km/h, the bridge
esults showed
coeticient that
could appear abnormal value, and when the loading period induced by
rgularly aranged axle loads was an integer times the natural frequency of the dynani,
beam, the bridge resonance would occur. Tanabe et al. (1987) established the
tour-axis passenger vehicle model of 31 DOFs on Shinkansen in Japan and used the

synthesis technology ofto solve the problem.

FE-and-modal
In Europe, with the development HSR constructioninthe 1970s, Frýba (1972,
1976) established the vertical vibration analysis model for train-bridge system,
including the moving load, moving mass, moving wheel-spring-mass unit, and
considerine
vehicle element with car-body and two-layer suspensions. By
iegularity of bridge deck, he systematically compared the effect of differen
parameters, such as bridge dynamic properties, vehicle characteristics, train speed,
imegularities and wheel-rail forces, on dynamic response of the bridoe
track method to reduce the analytical Do
(1985, 1991)adopted the modal coordinate
general vehicle element composed
of the tbridge, and then derived the system, and presented the an
time-varying, asymmetric matrices for train-bridge of
lyical expression and numerical
solution dynamic responses for
simply-supported beam under moving loads.
research of
Since the 1980s, scholars from various countries have extended the
train-bridge coupling vibration tomore application scopes. Diana and Cheli (1980)
established the vehicle model of 23 DOFs considering the wheel-rail interaction
and analyzed the dynamic responses of a long-span suspension bridge under wind
loads. Green and Cebon (1994, 1997)obtained the dynamic responsesof the bridge
through FFT transform and by convolution of the vehicle load and bridge modes in
the frequency domain.
The European Railway Research Institute (ERRI) conducted a series of resear
ches on the bridgevibration caused by high-speed trains, and the results were used
to guide the dynamic design of HSR bridge (ERRI D214 Commitee 1999). The
D214 commitee of ERRI introduced some new concepts such as the "tran ii
gerprint" and the "bridge dynamic influence line," and classified the faclos
affecting the amplification of bridge dynamicresponses into two categories. One 1
from the train (arrangement of wheel-axles): the other is from the bridge (frequeny
and damping of structure). They also presented the harmonic decompostu
method to calculate the maximum dynamic displacements and accelerations ofthe
bridge. Although the harmonic decomposition method was only applied of the
upper limits
simply-supported beam with uniform section, it can give the while avoidthe
dynamic response of the bridge nder the resonance train speed Majkaand
step-by-step integration of dynamic equations in the time domain.
Hartnett(2008, 2009) studied in detail the inffuences of various parameters, su
1.4 Research History and Status Quo ...
45

bridge frequency, mass, damping and span, on the dynamic

span simply-supported bridge under diferent train speeds. responses of the single
De Roeck et al. (2000) performed
systematic research on the bridge damage
jdentification. By continuous measurement on a bridge for a whole year, they found
that temperature was the most important
the boundary conditions as well as the environmental factor, as it could change
elastic modulus of
variationof vibration signais induced by bridge damage is mainlytheduebridge,
to
and the
the decline
of local stiffness. Therefore, it is necessary to
eliminate the interference of envi
ronment factors when using the bridge dynamic responses as
Liu et al. (2014a, b) analyzed the evaluation indices.
dynamic responses of
multi-span
simply-supported viaducts under the passage of train, considering the weak cou
pling effect between bridge and track due to ballast and rails, and
found that fairly
ideal analysis results of bridge dynamic behavior could be obtained by
the boundary conditions of a single span beam model. optimizing
Rebelo et al. (2008) carried out dynamic tests on several medium and small
span
ballasted bridges in Austria, and discovered that friction forces between ballast and
deck had big influence on the peak acceleration of bridge.
Rigueiro et al. (2010) studied the effect of various ballasted tracks, and found
that track structure worked as a filter to the high frequency vibration of elevated
railway bridges, which would reduce the maximum acceleration of the deck.
Kaliyaperumal et al. (2011)established the finite element model of askewed steel
bridge, using combined plate and beam elements. The calculated results matched
well with the experience values after the model being updated by the measured
characteristic parameters. It was proved that the global bridge model established
with combined plate and beam elements could achieve good dynamic responses of
the bridge, and can be used to estimate the average stress range of fatigue damage.
With the development of research in this field, some shortages in the analysis
method and the simplification assumptions are gradually recognized, where, for
example, the track structure was simplified to have the same deformation with the
beam by ignoring its individual changes, and the wheel-rail relationship was mostly
considered by force equilibrium and displacement compatibility, while lacked
substantive physical description on real wheel-rail contact state.
In recent years, the computer simulation based on complex vehicle model and
spatial bridge model has become a necessary tool for train-bridge coupling vibra
tion research. Using more detailed wheel-rail relationship to further perfect the
vehicle model and adopting more efficient calculation method to objectively sim
ulate the real dynamic response of train-bridge system has become the common
goal for all international scholars.

1.4.1.2 Researches in China

In China, studies on bridge vibration subjected to moving train loads started early in
1970s. Professor Li GH proposed the flexure-torsion theory of trussed beam, using
the finite difference method to calculate the spatial free vibrations of several