Computers and Structures 83 (2005) 16271645

www.elsevier.com/locate/compstruc

Three-dimensional dynamic analysis for bridgevehicle

interaction with roadway roughness
a,*
Chul Woo Kim , Mitsuo Kawatani a, Ki Bong Kim b

a
Department of Civil Engineering, Kobe University, 1-1 Rokkodai, Nada, Kobe 657-8501, Japan
b
Department of Civil Engineering, Chung-Ang University, 72-1 Neri, Dedug, Ansung 456-756, Korea

Received 3 March 2004; accepted 21 December 2004

Available online 24 March 2005

Abstract

A three-dimensional means of analysis is proposed for the bridgevehicle interaction to investigate the dynamic
responses of a steel girder bridge and vehicles. The governing equations of motion for a three-dimensional bridgevehi-
cle interaction system taking the roadway surface into account are derived using the Lagrange equation of motion while
the coupled bridgevehicle interaction system is solved using Newmarks b method. A cargo truck, dump truck and steel
girder bridge are considered numerical models and measured roadway roughness proles are used for analyses. The
analytical dynamic wheel loads and acceleration responses of the heavy vehicles and responses of the bridge are com-
pared with data from eld tests to verify the validity of the proposed procedure. The correlation between the analytical
and experimental results is satisfactory.
2005 Elsevier Ltd. All rights reserved.

Keywords: Bridgevehicle interaction; Dynamic wheel load; Field test; Three-dimensional dynamic analysis; Trac-induced vibration
of bridge

1. Introduction which vary over time, have been examined in terms

The problem of vibration in bridge structures due
to moving vehicle loads has been a topic of interest
for over a century. One of the major external dynamic
forces acting on bridges, especially those of short and
medium span, is the moving vehicle load and eects,
*
Corresponding author. Tel.: +81 78 803 6383; fax: +81 78
803 6069.
E-mail addresses: cwkim@kobe-u.ac.jp (C.W. Kim),
m-kawa@kobe-u.ac.jp (M. Kawatani), kbkim@cau.ac.kr
(K.B. Kim).
of impact factors. The impact factor has therefore
been one of the major topics in bridge dynamics
[1,2] since the mid nineteenth century. It is widely
known that the dynamic responses of bridges depend
on the vehicle type, speed, roughness of the pavement,
bridge type, etc. (e.g. Refs. [211]), although the bridge
design codes dene the impact factor as a function of
the bridge span length. Investigations into bridgevehi-
cle interaction have been carried out by many
researchers and several topics relating to this area have
already been fully reviewed by Cantieni in a recent
article [2].
There are two means of investigating the compli-
cated bridgevehicle interaction problem: namely

0045-7949/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compstruc.2004.12.004
1628 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

(a)

(b)

Fig. 1. Idealized heavy trucks. (a) Two-axle vehicle with seven-degrees-of-freedom and (b) three-axle vehicle with eight-degrees-of-
freedom.

Fig. 2. Bridgevehicle interaction system.

experimental and analytical approaches. The experimen-
tal method requires considerable time, facilities and cost
while the analytical approach represents an economical
way to examine the bridgevehicle interaction provided
the validity of this method be veried.
One of the most common analytical approaches to
examine the problems concerning bridgevehicle interac-
tion has been a method using one- or two-dimensional
models of bridges and vehicles [7,9,10,1215]. The two-
dimensional system can provide good analytical results
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645
for investigating the dynamic responses of whole bridge
structures induced by moving vehicles on uneven road-
ways [9,10,13,15]. The three-dimensional bridgevehicle
interaction system is usually adopted to simulate the re-
sponses of local bridge components such as deck slabs
and cross beams [16], and to investigate bridge responses
induced by vehicles with paths that do not follow the
centre line of the bridge. A few three-dimensional ana-
lytical models for the bridgevehicle interaction system
thus have been developed (e.g. Refs. [11,17,18]).
However, most of the previous works to date have fo-
cused on bridge responses, although forces producing
bridge stresses due to moving vehicular loadings are
the results of the vehicles dead weights and the force
of interaction between vehiclepavementbridge. Fur-
thermore, there has been relatively little verication of
analytical results irrespective of their importance in
bridge dynamics, even the case of considering dynamic
vehicular loads.
The major goal of this study is to develop a proce-
dure for determining the dynamic wheel loads of vehicles
as well as the bridge responses thereto, and to verify the
validity of the analytical procedure. To meet the needs
7500
500 6500
250 250 End Section
CL
250
1145
60
1600
1100 2650
G1
G2
G3
5050
: Vehicle path *: Position of right wheel of 7DOF vehicle
(a)
(b) : Support : Observation points
Fig. 3. Steel plate girder bridge. (a) General layout and (b) FE model.
1629
of this study, the governing equations of the bridge
vehicle interaction system are derived using Lagranges
formulation, and the numerical model for the trac-
induced vibration of bridges makes use of the nite ele-
ment method for modal analysis. Newmarks b method
[19] is applied to solve the derived system governing
motion equations. The approach is then validated
through comparison with eld test data.
2. Idealizations and assumptions
2.1. Idealization of vehicle system
The two-axle Ford cargo truck (one front and rear
axle) [2022] is idealized as a vehicle model of oering
seven-degrees-of-freedom (7DOF) (see Fig. 1(a)). The
three-axle Isuzu dump truck (one front and two rear
axles) [16,22] meanwhile is idealized as a vehicle model
with eight-degrees-of-freedom (8DOF) (see Fig. 1(b)).
The vehicle body itself is considered to be rigid and sup-
ported by a set of linear springs and dashpots attached
to each axle.
500
Mid-Section 250 250
75
170
2650 1100 Unit: mm
1400
*(3470)
3200
30300 5050
40400
1630 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

Table 1 of the rear axle of the vehicle model, respectively.

Properties of steel bridge mv11, mv12 and mv22, respectively, indicate the concen-
Mass per unit length (kg/m) 7550 trated mass of the vehicle body, front and rear axles.
Section area of girders (m2) 0.1420 Kvmku and Cvmku are the spring constant and damping
Moment of inertia (m4) 0.2120 coecient of a vehicle; the subscript k is the index to
Torsional constant (m4) 0.0548 indicate the vehicle body and axle (k = 1, 2 indicating
Damping constant 0.0253 the vehicle body and axle, respectively), subscript m is
(for 1st and 2nd modes) the index for the axle/tire positions (if k = 1 then
Fundamental frequency 1st (Bending) 2.33
m = 1, 2 indicating front and rear axle, respectively; if
(Experiment, Hz)
2nd (Torsion) 3.86
k = 2 then m = 1, 2, 3 indicating the tire at the front-axle,
front and rear tires of the rear axle, respectively) and
subscript u is the index for indicating the left and right
In Fig. 1, Z11, Z12, Z22, hx11, hx12, hx22, hy11 and hy22 sides of a vehicle (u = 1, 2 indicating left and right sides,
refer to the bounce of vehicle body, the parallel hop of respectively). The sign is taken to be positive if the defor-
front and rear axle respectively, the rolling of vehicle mation occurs in a downward direction, pitching occurs
body, the axle tramp of the front and rear axles, the from the rear to the front axle and rolling is generated
pitching of vehicle body and the axle windup motion from the right to left side.

(a) (b)

(c) (d)

Fig. 4. Mode shapes and fundamental natural frequencies of steel girder bridge. (a) First bending mode (Analysis: 2.32 Hz,
Experiment: 2.33 Hz), (b) second bending mode (Analysis: 8.79 Hz), (c) rst torsion mode (Analysis: 3.56 Hz, Experiment: 3.86 Hz),
(d) second torsion mode (Analysis: 9.68 Hz).

Table 2
Properties of vehicles
Parameter Two-axle Three-axle
vehicle vehicle
Geometry (m) Tread 2.07 1.80
Distance between front and rear axle 6.20 3.99
Distance of tandem axle 1.20
Distance between front axle and centre of gravity 3.94 2.99
Mass (kg) Sprung mass including payload 14 790 18 500
Steer axle un-sprung mass 650 500
Drive axle un-sprung mass 1070 1450
Spring constant of suspension (kN/m) Front leaf spring 475 1577
Rear leaf spring 1820 4724
Spring constant of tire (kN/m) Front tire 1390 3146
Rear tire 1,170 4724
Damping coecient of suspension (kN s/m) Front left 7.810 11.200
Front right 8.065 11.200
Rear left 3.324 33.420
Rear right 1.649 33.420
Damping coecient of tire (kN s/m) Front tire 13.300
Rear tire 10.000
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1631

10
8 TEST RANGE
6

roughness(cm)
4
2
0
-2
-4 Left Wheel
-6
-8 Right Wheel
-10
0 100 200 300 400 500 600 700 800 900 1000 1100 1200
(a) Distance(m)

10
8 Left Wheel
TEST RANGE
6
Right Wheel
roughness(cm)

4
2
0
-2
-4
-6
-8
-10
0 10 20 30 40 50 60 70
(b) Distance(m)

Fig. 5. Measured roadway proles. (a) Smooth Canadian Road (P1-prole), (b) Umeda Entrance Road (P2-prole).

2.2. Idealization of bridge system
The nite element (FE) method and modal analysis
are adopted as the tools for idealizing bridges for
dynamic response analysis. The lumped mass system
and Rayleigh damping [23] are adopted to form mass
and damping matrices of the bridge model, respec-
tively.
Two types of nite elements are adopted to idealize
members of the bridge super-structure. Beam elements
with six-degrees-of-freedom at each node are used to
idealize girders, cross beams and guard rails of a
bridge. Decks are idealized as a at plate element with
four nodes [24]. To improve the eciency of calcula-
tion, a process known as Guyan reduction is per-
formed [25].
2.3. Idealization of bridgevehicle interaction system
Fig. 2 is the idealized bridgevehicle interaction sys-
tem with the pavement surface in a deformed state
[17]. Z0mu denotes the relative displacement dened by
the dierence between the displacement of bridge and
surface roughness at the contact point of each tire.
The longitudinal position of the tire location xmu is rel-
ative to the bridge entry. The pavement roughness of
the bridge at a tire location is denoted by Zrmu. w(t, xmu)
is the elastic deformation of the bridge at a time t and a
location xmu while the subscripts m and u are the same
indices explained in Section 2.1.
The assumptions considered in this study are: (1)
the bridge follows Hookes law, Naviers hypothesis,
Saint-Venants principle and small-deection theory,
(2) the ground acceleration and support settlement of
the bridge are not taken into account, (3) the bridge
is initially at rest before the vehicle enters the span,
(4) the vehicle speed and path are constant, (5) the
vehicle body is treated as a rigid body: the elastic
behavior of the body is not considered and (6) the sus-
pension springs are assumed to be linearly elastic and
the damping of the suspension system is considered to
be viscous.
3. Equations of motion of bridgevehicle interaction
system
The method known as the Lagrange equation of mo-
tion has been known as one of the most popular meth-
ods to formulate a dynamic system since the French
mathematician Lagrange discovered a relationship pro-
viding a genuinely powerful and exible method to for-
mulate equations of motion for any dynamical system.
Therefore, governing equations of the bridgevehicle
interaction system are derived from the energy method
using the Lagrange equation of motion as shown in
Eq. (1) [7,26].
1632 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

1.0E+8 1.0E+8
1.0E+7 ISO
ISOEstimate
Estimate 1.0E+7 ISO Estimate
ISO Estimate
1.0E+6 (ISO8608,
(ISO 8606,1995)
1996) 1.0E+6 (ISO
(ISO8608,
8606,1995)
1996)
1.0E+5 1.0E+5
1.0E+4 1.0E+4
1.0E+3 1.0E+3
1.0E+2 1.0E+2
1.0E+1 1.0E+1
H H
1.0E+0 G 1.0E+0 G
PSD(cm^2/cycle/m)

PSD(cm^2/cycle/m)
F F
1.0E-1 E 1.0E-1 E
1.0E-2 D 1.0E-2 D
C C
1.0E-3 B 1.0E-3 B
A A
1.0E-4 1.0E-4
1.0E-5 1.0E-5
1.0E-6 1.0E-6
1.0E-7 1.0E-7
1.0E-8 1.0E-8
1.0E-9 1.0E-9
1.0E-10 1.0E-10
1.0E-11 1.0E-11
1.0E-12 1.0E-12
1.0E-13 1.0E-13
Left wheel track Right wheel track
1.0E-14 1.0E-14
1.0E-15 1.0E-15
0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10
(a) Spatial frequency(cycle/m) Spatial frequency(cycle/m)

1.0E+8 1.0E+8
ISO
ISOEstimate
Estimate ISO
ISOEstimate
Estimate
1.0E+7 1.0E+7 (ISO
(ISO
(ISO8608, 1995)
8606, 1996) (ISO8608,
8606,1995)
1996)
1.0E+6 1.0E+6 H
H
1.0E+5 G 1.0E+5 G
F F
1.0E+4 E 1.0E+4 E
D D
1.0E+3 1.0E+3 C
C
1.0E+2 B 1.0E+2 B
PSD(cm^2/cycle/m)

PSD(cm^2/cycle/m)

A A
1.0E+1 1.0E+1
1.0E+0 1.0E+0
1.0E-1 1.0E-1
1.0E-2 1.0E-2
1.0E-3 1.0E-3
1.0E-4 1.0E-4
1.0E-5 1.0E-5
1.0E-6 1.0E-6
1.0E-7 1.0E-7
1.0E-8 Left wheel track 1.0E-8 Right wheel track

1.0E-9 1.0E-9
0.001 0.01 0.1 1 10 0.001 0.01 0.1 1 10
(b) Spatial frequency(cycle/m) Spatial frequency(cycle/m)

Fig. 6. PSD curves of measured roadway proles. (a) P1-prole and (b) P2-prole.


d oT oT oV oU d 1 _T _ 1 mv11 Z_ 2 J y11 h_ 2 J x11 h_ 2

0 1 T D Mb D 11 y11 x11
dt oq_ i oqi oqi oq_ i 2 2
2 2 2 2
J y22 h_ mv12 Z_ mv22 Z_ J x12 h_
y22 12 22 x12
where, T is kinetic energy of the system; V, potential en- 2
ergy of the system; Ud, dissipation energy of the system; J x22 h_ x22 2
qi, the ith generalized co-ordinate.
The kinetic energy, potential energy including strain 1 1X 3 X 2
V DT Kb D K vm1u R2m1u K vm2u Rm2u
and dissipation energies of the bridgevehicle interaction 2 2 m1 u1
system are expressed in a set of generalized coordinates

Z 0mu 2 2  W mu Z 0mu  3
as follows:
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1633

40 40
Left wheel of front axle Left wheel of front axle

Wheel Load(kN)

Wheel Load(kN)
35 35

30 30

25 25

20 20
0 10 20 30 40 0 10 20 30 40
Time(sec) Time(sec)
40 40
Right wheel of front axle Right wheel of front axle
Wheel Load(kN)

Wheel Load(kN)
35 35

30 30

25 25

20 20
0 10 20 30 40 0 10 20 30 40
Time(sec) Time(sec)
70 70
Left wheel of rear axle Left wheel of rear axle
Wheel Load(kN)

Wheel Load(kN)
60 60

50 50

40 40

30 30
0 10 20 30 40 0 10 20 30 40
Time(sec) Time(sec)
70 70
Right wheel of rear axle Right wheel of rear axle
Wheel Load(kN)

Wheel Load(kN)

60 60

50 50

40 40

30 30
0 10 20 30 40 0 10 20 30 40
(a) Time(sec) (b) Time(sec)

Fig. 7. Dynamic wheel loads of two-axle vehicle (v = 82.73 km/h). (a) Analysis and (b) eld-test.

8 h  i
1 _T _ 1 X 3 X 2
2 >
> 1
g 1
kx1
m m m 1; u 1; 2
Ud D Cb D C vm1u R_ m1u C vm2u R_ m2u >2
< kx v11 v12
2 2 m1 u1 h  i
W mu 1 g 1
kx2 m m m 2; 3; u 1; 2
>
> 4 kx v11 v22

Z_ 0mu 2  4 >
:
0 otherwise
where 7
Z 0mu wt; xmu
Z rmu 5
The superscript dot denotes the dierential in terms of
8 time. J in Eq. (2) and g in Eq. (7) indicate the mass
>
> Z 11

1m kxm hy11

1u ky1 hx11
Z m2
>
>
1u k moment of inertia of the vehicle and its gravity, respec-
>
> ym1 hxm2 _ indicate the displacement
>
> tively. In Eqs. (2)(4), D and D
>
>
>
> m 1; 2; k 1; u 1; 2 and velocity vectors of the bridge, respectively; Mb and
>
>
< Z

1u k h Kb respectively indicate the reduced mass and stiness
12 y2 x12
Rmku 6 matrices [25] of the bridge; Cb, the damping matrix of
> m 1; k 2; u 1;
>
>
> the bridge derived from the assumption of linear relation
>
> m u
> Z 22
1 kx3 hy2

1 ky3 hx22
> between the mass and stiness matrices, which can be
>
>
>
> m 2; 3; k 2; u 1; 2 expressed as Eq. (8) [23]:
>
>
:
0 otherwise Cb pMb qKb 8
1634 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

6 6
4x10 4x10 Left wheel
Left wheel
of front-axle of front-axle
Power spectrum

Power spectrum
6 6
3x10 3x10

6 6
2x10 2x10

6 6
1x10 1x10

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
6 6
4x10 4x10
Right wheel Right wheel
of front-axle

Power spectrum
of front-axle
Power spectrum

6 6
3x10 3x10

6 6
2x10 2x10

6 6
1x10 1x10

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
7 7
5x10 5x10
Left wheel Left wheel
Power spectrum

Power spectrum

7 7
4x10 of rear-axle 4x10 of rear-axle
7 7
3x10 3x10
7 7
2x10 2x10
7 7
1x10 1x10

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
7 7
5x10 5x10
Right wheel Right wheel
Power spectrum
Power spectrum

7 7
4x10 of rear-axle 4x10 of rear-axle
7 7
3x10 3x10

7 7
2x10 2x10

7 7
1x10 1x10

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
(a) (b)

Fig. 8. PSD curves of dynamic wheel loads in Fig. 7 (v = 82.73 km/h). (a) Analysis and (b) eld-test.

where respectively; h1 and h2, damping constants according

to two modes of vibration with natural circular fre-
2x1 x2 h1 x2
h2 x1
p 9 quencies of x1 and x2. In this study, h1 and h2 are as-
x22
x21 sumed to have the same values due to the diculty
involved in estimating the second damping constant
2h2 x2
h1 x1
q 10 h2 [23].
x22
x21 The displacement vector of the bridge is written gen-
In Eqs. (9) and (10), x1 and x2 are the rst and second erally in terms of the normal coordinate as shown in Eq.
natural circular frequencies of the bridge system, (11).
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1635

1 1
0.9 Experiment 0.9 Experiment

Cumulative Distribution

Cumulative Distribution
0.8 Analysis 0.8 Analysis
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
Left wheel of front-axle Right wheel of front-axle
0.1 0.1
0 0
24000 26000 28000 30000 32000 34000 36000 24000 26000 28000 30000 32000 34000 36000
(a) Wheel Load(N) (b) Wheel Load(N)

1 1
0.9 Experiment 0.9 Experiment
Cumulative Distribution

Cumulative Distribution
0.8 Analysis 0.8 Analysis
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2 0.2
Left wheel of rear-axle Right wheel of rear-axle
0.1 0.1
0 0
35000 40000 45000 50000 55000 60000 65000 70000 35000 40000 45000 50000 55000 60000 65000 70000
(c) Wheel load(N) (d) Wheel Load(N)

Fig. 9. Cumulative distributions of dynamic wheel loads of two-axle vehicle in Fig. 7 (v = 82.73 km/h). (a) Left wheel of front-axle, (b)
right wheel of front-axle, (c) left wheel of rear-axle and (d) right wheel of rear-axle.

Table 3
DLC values and dominant frequencies of dynamic wheel loads of two-axle vehicle
Vehicle speed (km/h) Wheel DLC Dominant frequency (Hz)
Experiment Analysis Experiment Analysis
64.96 Front left 0.0309 0.0328 1.93 1.96
Front right 0.0319 0.0337 1.93 1.96
Rear left 0.0444 0.0511 2.53 2.58
Rear right 0.0471 0.0531 2.53 2.58
75.68 Front left 0.0342 0.0394 1.88 1.93
Front right 0.0357 0.0371 1.88 1.93
Rear left 0.0579 0.0572 2.46 2.43
Rear right 0.0582 0.0609 2.46 2.43
82.73 Front left 0.0384 0.0399 1.98 1.66
Front right 0.0397 0.0377 1.98 1.66
Rear left 0.0632 0.0598 2.31 2.48
Rear right 0.0649 0.0601 2.31 2.48

X
n
wt; xmu WTmu D 12
D /j aj Ua 11
j1
where, Wmu is the distribution vector delivering wheel
where, U and a are the modal matrix and the generalized loads through a plate element to each node of the ele-
displacement vector of the bridge, respectively. ment, and represented as Eq. (13).
The displacement of the bridge w(t, xmu) can be ob-
tained using a combination of the displacement and dis- Wmu 0; . . . ; 0; Wk;mu t; Wk1;mu t; Wk2;mu t;
tribution vectors as shown in Eq. (12). Wk3;mu t; . . . ; 0 13
1636 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

6 6
Bounce motion at center of front axle Bounce motion at center of front axle

Acc.(*100 Gal)

Acc.(*100 Gal)
4 4
2 2
0 0
-2 -2
-4 -4
-6 -6
3 4 5 6 7 8 9 10 3 4 5 6 7 8 9 10
Time(sec) Time(sec)
6 6
Bounce motion at center of rear axle Bounce motion at center of rear axle
4 4
Acc.(*100 Gal)

Acc.(*100 Gal)
2 2
0 0
-2 -2
-4 -4
-6 -6
3 4 5 6 7 8 9 10 3 4 5 6 7 8 9 10
Time(sec) Time(sec)
6 6
Axle-hop motion of front axle Axle-hop motion of front axle
4
Acc.(*100 Gal)

Acc.(*100 Gal)
4
2 2
0 0
-2 -2
-4 -4
-6 -6
3 4 5 6 7 8 9 10 3 4 5 6 7 8 9 10
Time(sec) Time(sec)
6 6
Axle-hop motion of rear axle Axle-hop motion of rear axle
Acc.(*100 Gal)

Acc.(*100 Gal)

4 4
2 2
0 0
-2 -2
-4 -4
-6 -6
4 5 6 7 8 9 10 4 5 6 7 8 9 10
Time(sec) Time(sec)
(a) (b)

Fig. 10. Acceleration time histories of three-axle vehicle (v = 16.8 km/h). (a) Analysis and (b) eld-test.

The nal formulation of the governing die- X

2

rential equations for the bridgevehicle interaction
C v12u WT1u Ua_ K v12u WT1u Ua mv12 Z 12
u1
system is obtained from the relationships in Eqs. 2 X
X 2
(1)(13):

1k C v1ku R_ 1ku K v1ku R1ku
k1 u1
X
3 X
2 X 2

a Cb a_ Kb a UT
Mb Wmu C vm2u WTmu Ua_
C v12u Z_ r1u K v12u Z r1u 16
m1 u1 u1

X
3 X
2
X
3 X
2
K vm2u WTmu Ua
UT Wmu C vm2u R_ m2u
C vm2u WTmu Ua_ K vm2u WTmu Ua mv22 Z 22
m1 u1
m2 u1
K vm2u Rm2u 3 X
X 2 X
2

1k C vmku R_ mku K vmku Rmku
X
3 X
2
UT Wmu W mu C vm2u Z_ rmu K vm2u Z rmu m2 k1 u1

m1 u1 3 X
X 2

C vm2u Z_ rmu K vm2u Z rmu 17
14 m2 u1

2 X
X 2
J x11
hx11

1u ky1 C vm1u R_ m1u K vm1u Rm1u 0
2 X
X 2
m1 u1
mv11 Z 11 C vm1u R_ m1u K vm1u Rm1u 0 15
m1 u1
18
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1637

2.0 2.0
Bounce motion Bounce motion
at center at center
1.5 of front axle of front axle
Power Spectrum
1.5

Power Spectrum
1.0 1.0

0.5 0.5

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
6.0 6.0
Bounce motion Bounce motion
at center at center
Power Spectrum

of rear axle of rear axle

Power Spectrum
4.0 4.0

2.0 2.0

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
0.3 0.3
Axle-hop motion Axle-hop motion
of front axle of front axle
Power Spectrum
Power Spectrum

0.2 0.2

0.1 0.1

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
1.2 1.2
Axle-hop motion Axle-hop motion
1.0 of rear axle 1.0 of rear axle
Power Spectrum

Power Spectrum

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0 0
0.1 1 10 100 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
(a) (b)

Fig. 11. PSD curves of accelerations in Fig. 10 (v = 16.8 km/h). (a) Analysis and (b) eld-test.

X
2 X
3 X
2

1u ky2 C v12u WT1u Ua_ K v12u WT1u Ua J x12 hx12
1u ky3 C vm2u WTmu Ua_ K vm2u WTmu Ua J x22
hx22
u1 m2 u1

X
2 X
2 3 X
X 2 X
2


1k
1u ky2 C v1ku R_ 1ku K v1ku R1ku

1k
1u ky3 C vmku R_ mku K vmku Rmku
k1 u1 m2 k1 u1

X
2 3 X
X 2

1u ky2 C v12u Z_ r1u K v12u Z r1u 19
1u ky3 C vm2u Z_ rmu K vm2u Z rmu 20
u1 m2 u1
1638 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

2 X
X 2 The equations of the bridgevehicle interaction are rep-
J y11 hy11

1m kxm C vm1u R_ m1u K vm1u Rm1u 0 resented as following a matrix formation by combining
m1 u1
Eqs. (14)(23):
21
Cs W
Ms W _ Ks W Fs 24
X
3 X
2


1m kx3 C vm2u WTmu Ua_ K vm2u WTmu Ua where, Ms, Cs and Ks indicate the mass, damping and
m2 u1
stiness matrices of the bridgevehicle interaction sys-
3 X
X 2 W,
tem, respectively. The W, _ W and Fs respectively refer
J y22 hy22
1m kx3 C vm2u R_ m2u K vm2u Rm2u to the acceleration, velocity, displacement and force vec-
m2 u1
tors of the system. Details concerning the system matri-
3 X
X 2 ces are described in the Appendix section.


1m kx3 C vm2u Z_ rmu K vm2u Z rmu 22 By eliminating elements of the column and row re-
m2 u1
lated to hy22 in the system matrices and replacing 1/4
where, Mb , Cb and Kb in Eq. (14) refer to normalized of Eq. (7) to 1/2, the Eq. (24) can be applied to the
mass, damping and stiness matrices, respectively: bridgevehicle interaction problem exposed to the two-
Mb UT Mb U; Cb UT Cb U; Kb UT Kb U. axle vehicle (7DOF vehicle). Neglecting the matrices
The dynamic wheel loads at each tire of the vehicle and variables related to the bridge response in Eq. (24)
are estimated based on the following formula: allows for the dynamic response analysis for a vehicle
operating on the rigid surface.
P mu t W mu C vm2u R_ m2u
WTmu Ua_
Z_ rmu  The dynamic equation for the vehicle-bridge interac-
tion is a non-stationary dynamic problem since the coef-
K vm2u Rm2u
WTmu Ua
Z rmu ; m 1; 2; 3
cient matrices of the equations vary according to the
23 vehicle position. Thus, the simultaneous dierential

Table 4
Dominant frequencies of three-axle vehicle
Roadway condition Vehicle speed (km/h) Position Bounce motion Axle-hop motion
Experiment Analysis Experiment Analysis
Case 1 16.8 Front 3.47 4.00 22.90 17.33
Rear 4.15 4.00 14.11 20.40
21.7 Front 3.59 3.13 17.39 19.14
Rear 4.33 3.52 14.23 18.95
31.5 Front 3.28 3.52 14.11 15.23
Rear 4.21 3.52 14.17 20.20
Case 2 23.2 Front 4.02 3.32 17.76 20.51
Rear 4.39 3.52 14.36 19.53
30.8 Front 3.47 3.52 17.08 19.92
Rear 4.46 3.71 14.23 19.34
Case 3 14.9 Front 3.96 3.61 23.64 13.57
Rear 4.27 3.52 13.55 19.04
30.0 Front 3.53 3.32 14.17 20.31
Rear 4.4 3.91 13.99 18.95
Case 4 17.3 Front 4.15 3.52 17.02 13.48
Rear 4.21 3.52 14.91 14.45
23.9 Front 3.28 3.52 16.46 21.09
Rear 4.27 3.71 13.74 19.92
30.9 Front 4.27 3.32 14.23 19.53
Rear 4.33 3.71 13.61 19.53
Case 1: P2-prole with no bump at expansion joint of entrance.
Case 2: P2-prole with 5 mm bump (plank with 5 mm thickness) at expansion joint of entrance.
Case 3: P2-prole with 10 mm bump (plank with 10 mm thickness) at expansion joint of entrance.
Case 4: P2-prole with 15 mm bump (plank with 15 mm thickness) at expansion joint of entrance.
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1639

1E-4 1E-4
Analysis Analysis
8E-5 Experiment 8E-5 Experiment
6E-5 v=16.8km/hr 6E-5 v=16.8km/hr

strain

strain
4E-5 4E-5
2E-5 2E-5
0E+0 0E+0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Time(sec) Time(sec)

1E-4 1E-4
Analysis Analysis
8E-5 Experiment 8E-5 Experiment
6E-5 v=21.7km/hr 6E-5 v=21.7km/hr

strain
strain

4E-5 4E-5

2E-5 2E-5

0E+0 0E+0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
Time(sec) Time(sec)

1E-4 1E-4
Analysis Analysis
8E-5 Experiment 8E-5 Experiment
strain

strain

6E-5 v=31.5km/hr 6E-5 v=31.5km/hr

4E-5 4E-5
2E-5 2E-5
0E+0 0E+0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
(a) Time(sec) (b) Time(sec)

Fig. 12. Strain responses of bridge due to three-axle vehicle running on P2-prole. (a) External girder G1 at span centre and (b)
internal girder G2 at span centre.

equations involved in the bridgevehicle interaction sys-
tem are solved using Newmark b as a direct integration
method [19].
4. Numerical examples
4.1. Numerical models
The bridge studied in this paper is a steel plate girder
bridge with a span measuring 40.4 m in length. It is com-
posed of three girders and reinforced concrete (RC)
decks. Fig. 3 shows the general layout and FE model
of the bridge and details of the bridge model are summa-
rized in Table 1.
Mode shapes from the eigen value analysis are shown
in Fig. 4 with natural frequencies taken from analysis
and experiment. The rst natural frequencies for bend-
ing and torsion obtained through analysis represent a
reasonable match for those obtained through experi-
mental results. It assists the validity of bridge modeling
for the dynamic response analysis.
Table 2 shows details of the vehicle models. The
spring constants and damping coecients of the two-
axle vehicle are estimated in DIVINE (Dynamic Interac-
tion VehicleINfrastructure Experiment) project devel-
oped by the OECD [21]. Part of the project was
coordinated by the TNO (Road-Vehicles Research Insti-
tute of Netherlands Organization for Applied Scientic
Research) working under the auspices of the DIVINE
joint research program [20]. In eld tests, vehicle speeds
were measured up to 64.96 km/h, 75.68 km/h and
82.73 km/h respectively with sample rates of 100 Hz
used during the experiments.
For the three-axle vehicle, the spring constants and
damping coecients of suspensions in Table 2 are mea-
sured during the on-site test at Umeda entrance bridge
in Osaka, Japan [16,22,27]. For tires, on the other hand,
the test has not been performed; meaning empirically as-
sumed values of the truck manufacturer are used in the
analysis. The vehicle speeds during the experiment were
measured as 14.017.3 km/h, 21.723.9 km/h and 30.0
31.5 km/h. The sample rates were 100 Hz during the
experiments.
Measured roadway proles used in analyses are
shown in Fig. 5(a) [20,21] and Fig. 5(b) [16,22,27]. The
proles are designated as P1-prole and P2-prole
respectively in this paper. The roadway proles were
1640 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

5.0E-11 5.0E-11
External girder G1 External girder G1
at span centre at span centre
4.0E-11 4.0E-11
Power spectrum

Power spectrum
3.0E-11 3.0E-11

2.0E-11 2.0E-11

1.0E-11 1.0E-11

0.0E+0 0.0E+0
0.01 0.1 1 10 100 0.01 0.1 1 10 100
Frequency(Hz) Frequency(Hz)

6.0E-12 6.0E-12
Internal girder G2 Internal girder G2
5.0E-12 at span centre 5.0E-12 at span centre
Power spectrum

4.0E-12 4.0E-12

3.0E-12 Power spectrum 3.0E-12

2.0E-12 2.0E-12

1.0E-12 1.0E-12

0.0E+0 0.0E+0
0.01 0.1 1 10 100 0.01 0.1 1 10 100
Frequency(Hz) Frequency(Hz)
(a) (b)

Fig. 13. PSD curves of bridge responses in Fig. 12 (v = 21.7 km/h). (a) Analysis and (b) eld-test.

4.2. Verication of numerical results

Y static,max. 4.2.1. Dynamic wheel loads of two-axle vehicle

To verify the validity of the analytical dynamic
Static response wheel loads of the two-axle cargo truck running on
Dynamic response
DIF=1+ Y/Y static.max. the rigid roadway, time histories of wheel loads, the
Y dynamic load coecient (DLC) [20,21] and dominant
frequencies of wheel loads for each tire are compared
to the eld test data. The DLC is dened as the ratio
one cycle
of the root mean square (RMS) dynamic wheel loads
Fig. 14. Denition of DIF. against the mean wheel load where the RMS dynamic
wheel load represents the standard deviation of the
probability distribution. The DLC is hence the wheel
load coecient of variation. Since the DLC value
sampled every 0.05 m and the state of the roadway pro- has a limited role in validating the model, the cumula-
les can be categorized as very good (Class A) according tive distribution function is also adopted as a function
to ISO estimate [28] based on the vehicle ride comfort to analyze and validate the model. The measured dy-
compared to the power spectral density (PSD) curves namic wheel load of the two-axle vehicle running on
(see Fig. 6). the P1-prole is taken from the DIVINE project
Under analysis, the vehicle speeds recorded are then [20,21].
used to compare analytical and experimental results. A Figs. 7 and 8 show, respectively, time histories and
time step of 0.001 s for bridge and vehicle models and PSD curves of dynamic wheel loads of the two-axle
b of 0.25 are used to obtain a stable and accurate solu- vehicle at v = 82.73 km/h, and positive correlation
tion. The solution can be obtained within the relative is observed between analytical and experimental
margin of error of less than 0.001. results.
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1641

Table 5
DIF values of steel bridge
Roadway condition Vehicle speed (km/h) External girder G1 (at span Internal girder G2 (at span
centre) centre)
Experiment Analysis Experiment Analysis
Case 1 16.8 1.20 1.15 1.11 1.05
21.7 1.11 1.06 1.04 1.04
31.5 1.09 1.09 1.10 1.09
Case 2 14.0 1.13 1.15 1.21 1.06
23.2 1.11 1.06 1.09 1.04
30.8 1.08 1.08 1.07 1.13
Case 3 14.9 1.13 1.15 1.09 1.06
23.9 1.07 1.04 1.09 1.12
30.0 1.10 1.12 1.06 1.06
Case 4 17.3 1.20 1.15 1.07 1.07
23.9 1.10 1.06 1.09 1.12
30.9 1.09 1.12
Case 1: P2-prole with no bump at expansion joint of entrance.
Case 2: P2-prole with 5 mm bump (plank with 5 mm thickness) at expansion joint of entrance.
Case 3: P2-prole with 10 mm bump (plank with 10 mm thickness) at expansion joint of entrance.
Case 4: P2-prole with 15 mm bump (plank with 15 mm thickness) at expansion joint of entrance.

DLC value (two-axle, Wheel load) prole, which can resonant with a vehicle system, to
Frequency (two-axle, Wheel load) change in accordance to the vehicle speed.
Frequency (three-axle, Bounce motion)
Frequency (three-axle, Axle-hop motion)
4.2.2. Acceleration responses of three-axle vehicle
DIF of bridge
1.6 The experiment for the three-axle vehicle was carried
out as a part of a eld test for the Umeda entrance
1.4
bridge located on the Hanshin Expressway in Osaka,
Analysis/Experiment

1.2 Japan [16,22,27]. The acceleration responses in relation

1 to bounce and axle-hop motions are taken from acceler-
0.8
0.6
0.4
0 5
Fig. 15. Ratio of analytical and experimental results.
To assess the validity of the analytical dynamic wheel
loads quantitatively, cumulative distributions, DLC
values and dominant frequencies of the wheel loads
are compared with experimental ones. Cumulative dis-
tributions of the dynamic wheel load in Fig. 7 are shown
in Fig. 9, and the DLC values and dominant frequencies
are summarized in Table 3. The correlation between the
analysis and experiment shown in Fig. 9 and Table 3
demonstrates that the model is capable of simulating
the dynamic wheel loads within an error margin of
under 10% on average. One thing worth noting in Table
3 is that dominant frequencies of the dynamic wheel
loads vary according to speed. One of the reasons for
such dominant frequency variation may be the propen-
sity for the dominant space frequency of the roadway
ometers equipped on the vehicle body and each axle.
The roadway prole measured under the path of the
three-axle vehicle is the P2-prole (see Fig. 5(b)).
Acceleration responses of the bounce and the axle-
10 15 20 25
Case No.
hop motion of the vehicle are shown in Fig. 10. It is clear
that the quality of the correlation between the experi-
mental and analytical results is considered quite accept-
able for the bounce motion. Fig. 11 shows a positive
correlation of PSD curves for the bounce motion com-
paring experiment and analysis. On the other hand,
poorly assumed spring constant and damping coecient
related to tires leads the axle-hop motion which some-
what diers from the experimental result. The dominant
frequencies of the bounce and axle-hop motion of the
three-axle vehicle are summarized in Table 4.
4.2.3. Bridge responses
Fig. 12 shows typical strain responses taken from
analysis and eld tests at the span centre of the main
bridge girders due to the passage of the three-axle vehi-
cle with P2-prole on the bridge. The PSD curves of the
strain responses are shown in Fig. 13. G1 and G2 in the
Figs. 12 and 13 indicate the external and internal girders,
1642 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

respectively (see Fig. 3(a)). The trends, maximum ampli- frequencies for wheel loads of the two-axle vehicle
tudes and overall responses match quite well. are scattered within 88110% and 84110% of the
The dynamic increment factor (DIF) [29] is chosen as experimental results, respectively.
a parameter for the quantitative investigation of bridge 2. For the three-axle vehicle, the dominant frequencies
responses. The DIF is dened as the ratio of absolute of the bounce motion correlate positively with exper-
maximal dierence between the dynamic and static re- imental resultsrepresenting 80120% of these lat-
sponses to the maximum static response during one ter. On the other hand, for the axle-hop motion, a
major period of the dynamic response including the less accurate correlation between the analytical and
maximum static response (see Fig. 14) [29]. The DY in experimental results than those of bounce motion is
Fig. 14 is expressed as DY = jYdynamic
Ystaticjmax. DIF observed, namely 60145% of the experimental
values of the bridge are summarized in Table 5, which results, because of the dierence of the accuracy in
also shows a positive correlation between experiment assuming the spring constant of tires.
and analysis despite the poor assumption of tire-related 3. The analytical DIF values of the bridge display a
properties. positive correlation with the experimental results
despite the tire stiness being poorly estimated: i.e.
4.2.4. Summary the values are distributed at 93% and 105% of the
The summarized results for the relationship between experimental results. It encourages the use of
experiment and analysis are plotted in Fig. 15, which assumed properties for tires during the simulation,
illustrates the ratio of analytical to experimental results. where the analysis is focused on the dynamic
It is observed that the analytic DLC values of the two- responses of medium span bridges with similar fre-
axle vehicle wheel loads are distributed between 88% quency characteristics to the bounce motion of
and 110% of the experimental ones. The dominant fre- vehicles.
quencies of wheel loads of the two-axle vehicle taken 4. The eld test data indicates that analytical method to
from analysis represent between 84% and 110% of the be accurate for predicting the dynamic wheel load
experimental results. On the other hand, the analytical and response of highway bridges induced by moving
frequencies of the bounce motion and the axle-hop mo- vehicles.
tion of the three-axle vehicle range from 80% to 120%
and 60% to 145% of the experimental results, respec-
tively. The damping coecient and spring constant
for each tire of the three-axle vehicle used in analysis Appendix A
are not measured but taken as empirically assumed val-
ues, and the ratios between the analytical and experi- The Eq. (24) can be expressed as follows:
mental results of the three-axle vehicle are thus " # " #
  
scattered more widely than those of the two-axle vehi- Mb 0
a Cb Cbv a_
cle. For DIF values of the bridge, analytical results are
Symm: Mv
D Symm: Cv D_
distributed between 93% and 105% of the experimental " 
#  ( )
ones. Kb Kbv a
f b
A:1
The eld test data indicates that the analytical method Symm: Kv D fv
is accurate for predicting the dynamic wheel load and
response of highway bridges induced by moving where,
vehicles. 2 3
M b1  0
6 .. .. .. 7
6
Mb 4 7 A:2
. . . 5
5. Conclusions
Symm:  M bn
In this study, the simultaneous dierential equations 2 3
of motion to simulate the dynamic responses of the mv11
6 mv12 7
bridgevehicle interaction system are derived using the 6 7
6 7
Lagrange equation of motion from Hamiltons principle, 6 mv22 0 7
6 7
and FE method for modal analysis is adopted. The 6 J x11 7
6 7
major conclusions that can be drawn from the analysis Mv 6 7
6 J x12 7
6 7
and experiment are as follows: 6 Symm: J y22 7
6 7
6 7
4 J y11 5
1. Analytical dynamic wheel loads provide good results
J x22
when the properties of the vehicle are well estimated
or tested. The analytical DLC values and dominant A:3
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645 1643
2 3
P3 P 2
T
6 C b1 U1 Wmu C vm2u WTmu U1 0 7
6 m1 u1 7
 6 7
Cb 6
6
..
.
7
7 A:4
6 7
4 P
3 P
2 5
Symm: C bn UTn Wmu C vm2u WTmu Un
m1 u1

2 3
0 C bv12 C bv13 0 C bv15 C bv16 0 C bv18
6. .. .. .. .. .. .. .. 7
Cbv 6
4 .. 7 A:5
. . . . . . . 5
0 C bvn2 C bvn3 0 C bvn5 C bvn6 0 C bvn8

X
2 3 X
X 2 X
2
C bvj2
C v12u WT1u Uj ; C bvj3
C vmnu WTmu Uj ; C bvj5
1u ky2 C v12u WT1u Uj
u1 m2 u1 u1

3 X
X 2 3 X
X 2
C bvj6
1u ky3 C vm2u WTmu Uj ; C bvj8

1m kx3 C vm2u WTmu Uj ; j 1n
m2 u1 m2 u1

" #
C 11
v C 12
v
Cv A:6
C 21
v C 22
v

2 3
P
2 P
2 P
2 P
2 P
2 P
2
u
6 C vm1u
C v11u
C v21u

1 ky1 C vm1u 7
6 m1 u1 u1 u1 m1 u1 7
6 7
6 P
2 P
2 P2 7
6 C v1ku 0 u

1 ky1 C v11u 7
6 7
6 k1 u1 u1 7
C11
v 6
 7
6 P
2 P
3 P2 7
6 C v21u C vm2u
1u ky1 C v21u 7
6 7
6 u1 m2 u1 7
6 7
4 P2 P 2
2 5
Symm: ky1 C vm1u
m1 u1

2 3
P
2
u P
2
u
2 P
P 2
m
6
1 ky2 C v11u
1 ky3 C v21u

1 kxm C vm1u 0 7
6 u1 u1 m1 u1 7
6 7
6 P2 P 2 P
2 7
6

1u ky2 C v1ku 0
kx1 C v11u 0 7
6 7
12 6 k1 u1 u1 7
Cv 6
 7
6 P
2 P
3 P
2 P P
3 2 7
6 0

1u ky3 C v21u C vm2u kx2 C v21u
1 kx3 C vm2u 7
m
6 7
6 u1 m2 u1 m2 u1 7
6 7
4 P 2 P
2 P
2 P
2 5

ky1 ky2 C v11u
ky1 ky3 C v21u
1m
1u kxm ky1 C vm1u 0
u1 u1 m1 u1

2 3
2 P
P 2 P
2
6 k2y2 C v1ku 0
1u kx1 ky2 C v11u 0 7
6 k1 u1 u1 7
6 7
6 P
2 P
3 P
2 P3 P 2 7
6 k2y3 C v21u C vm2u

1u ky3 kx2 C v21u

1m
1u kx3 ky3 C vm2u 7
6 7
22 6 u1 m2 u1 m2 u1 7
Cv 6 7
6 P
2 P
2 7
6 k2xm C vm1u 0 7
6 7
6 m1 u1 7
6 7
4 P3 P
2
2 5
Symm: kx3 C vm2u
m2 u1

C21 12
v Cv

A similar procedure can be employed to derive each element in Kb , Kbv and Kv matrices.
1644 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645

The displacement and force vectors shown in Eq. [13] Green MF, Cebon D. Dynamic interaction between
(A.1) are dened as; heavy vehicles and highway bridges. Comput Struct
  1997;62(2):25364.
a1 ; . . . ; an ; Z 11 ; Z 12 ; Z 22 ; hx11 ;
fa; Dg [14] Pesterev AV, Bergman LA. Response of elastic continuum
hx12 ; hx22 ; hy11 ; hy22 carrying moving linear oscillator. J Eng Mech ASCE
A:7 1997;123(8):87884.
( ) [15] Yang YB, Lin BH. Vehiclebridge interaction analysis by
X
3 X
2 dynamic condensation method. J Struct Eng ASCE

f b UT Wmu W mu K vm2u Z rmu C vm2u Z_ rmu 1995;121(11):163643.
m1 u1 [16] Kawatani M, Kim CW. Eects of gap at expansion joint
A:8 on trac-induced vibration of highway bridge. In: Pro-
ceedings of developments in short and medium span bridge
( engineering98. Calgary, Canada: 1998. p. 28594.
X
2
fv 0;
K v21u Z r1u C v21u Z_ r1u ; [17] Mulcathy NL. Bridge response with tractortrailer vehicle
u1 loading. Earthq Eng Struct Dyn 1983;11:64965.
3 X
X 2 X
2 [18] Huang D, Wang TL. Vibration of highway steel bridges with

K vm2u Z rmu C vm2u Z_ rmu ; 0;
1u ky2 longitudinal grades. Comput Struct 1998;69(2): 23545.
m2 u1 u1 [19] Newmark NM. A method of computation for structural
3 X
X 2
 K v12u Z r1u C v12u Z_ r1u ;
1u ky3 K vm2u Z rmu dynamics. J Eng Mech Div ASCE 1970;96:593620.
m2 u1
[20] Hoogvelt RBJ, Ruijs PAJ. OECD-IR6 DIVINE Element 4.
X3 X 2 Computer simulation of heavy vehicle dynamic wheel loads.
C vm2u Z_ rmu ; 0;

1m kx3 K vm2u Z rmu TNO Report 97.OR.016.1/H/PR, TNO, Delft, The Neth-
) m2 u1 erlands: 1997.
[21] Technical Report, DSTI/DOT/RTR/IR6(98)1/FINAL,
C vm2u Z_ rmu A:9 OECD; 1998.
[22] Kim CW, Kawatani M. A comparative study on dynamic
wheel loads of multi-axle vehicle and bridge responses. In:
Proceedings of DETC01 ASME 2001 design engineering
References technical conference and computers and information in
engineering conference. Pittsburgh, USA: 2001 (CD-ROM).
[1] Timoshenko SP. History of strength of materials. Dover; [23] Agabein ME. The eect of various damping assumptions on
1982. p. 17381. the dynamic response of structures. Bulletin of international
[2] Cantieni R. Dynamic behavior of highway bridges under institute of seismology and earthquake engineering, vol.
the passage of heavy vehicles. EMPA Report No. 220; 8. Building Research Institute of Japan; 1971. p. 21736.
1992. [24] Bathe KJ. Finite element procedures in engineering anal-
[3] The AASHO road test: Report 4-Bridge Research, Special ysis. Prentice-Hall; 1982.
Report 61D: National Academy of Science-National [25] Guyan RJ. Reduction of stiness and mass matrices.
Research Council; 1962. AIAA J 1965;3(2):380.
[4] Walker WH, Veletsos AS. Response of simple-span high- [26] Hurty WC, Rubinstein MF. Dynamics of struc-
way bridges to moving vehicles. Bulletin 486: University of tures. Prentice-Hall; 1960. p. 90103.
Illinois, Engineering Experiment Station, Urbana IL: 1966. [27] Technical report-Inspections on Umeda entrance bridge
[5] Nieto-Ramirez JA, Veletsos AS. Response of three-span structures. Hanshin Expressway Public Corporation
continuous highway bridges to moving vehicles. Bulletin (HEPC), 1992 (in Japanese).
489: University of Illinois, Engineering Experiment Sta- [28] ISO 8608. Mechanical vibrationRoad surface proles
tion, Urbana IL: 1966. Reporting of measured data. ISO; 1995.
[6] Hwang ES, Nowak AS. Simulation of dynamic load for [29] Kawatani M, Nishiyama S, Yamada Y. Dynamic response
bridges. J Struct Eng ASCE 1991;117:141334. analysis of highway bridges under moving vehicle. Techn
[7] Hutton SG, Cheung YK. Dynamic response of single span Rep Osaka Univ 1993;43(2137):10918.
highway bridges. Earthq Eng Struct Dyn 1979;7:54353.
[8] Billing JR. Dynamic loading and testing of bridges in
Ontario. Can J Civ Eng CSCE 1984;11:83343. Chul Woo Kim was working in the
[9] Honda H, Kobori T, Yamada Y. Dynamic factor of Department of Civil Engineering,
highway steel girder bridges. IABSE Proc 1986;P-98:5775. Chung-Ang University, Korea, from
[10] Inbahathan MJ, Wieland M. Bridge vibrations due to 1984 to 1988. He obtained his M.Sc.
vehicle moving over rough surface. J Struct Eng ASCE degree in 1990 and Ph.D. degree in
1987;113(9):19942008. 1997 from the Graduate School of the
[11] Kou JW, DeWolf JT. Vibrational behavior of continuous same University. He was a Visiting
span highway bridgeInuencing variables. J Struct Eng Researcher in the Department of Civil
ASCE 1997;123(3):33344. Engineering, Osaka University Japan,
[12] Biggs JM. Introduction to structural dynamics. McGraw- during 19961997. He worked as a
Hill; 1982. p. 31527. part-time lecturer during 19971998 in
 C.W. Kim et al. / Computers and Structures 83 (2005) 16271645
the Department of Civil Engineering, College of Engineering,
Chung-Ang University, Korea. He worked as a lecturer during
19982003 in the Department of Civil and Urban Engineering,
Pohang College, Korea. He was an invited foreign researcher
during 20022003, Department of Civil Engineering, Kobe
University, Japan. He obtained his Doctorate in Engineering in
2003 from the Graduate School of Science and Technology,
Kobe University, Japan and at present, he is a COE researcher
in the same University.
Mitsuo Kawatani was working in the
Department of Civil Engineering,
Osaka University, during 19681972.
He was working in the Department of
Civil Engineering, Graduate School of
Engineering, Osaka University, Japan,
during 19721974. He was a Research
Associate in the Department of Civil
Engineering, Osaka University, Japan,
during 19741986. He was a Visiting
Researcher in the Columbia University
in the city of New York, USA, during 19851986. He was a
lecturer in the Department of Civil Engineering, Osaka Uni-
1645
versity, Japan, during 19861988. He was an Associate Pro-
fessor in the Department of Civil Engineering, Osaka
University, Japan, during 19881999. He was presently working
as a Professor in the Department of Civil Engineering, Kobe
University, Japan.
Ki Bong Kim was working in the
Department of Civil Engineering,
Seoul National University, Korea,
during 19691978. He obtained his
M.Sc. degree in 1980 from the Gradu-
ate School of the same university. He
obtained his Ph.D. degree from the
Graduate School, The George Wash-
ington University, USA. He was a
researcher from January 1986 to
August 1986 in The George Washing-
ton University, USA. He was an Assitant Professor during
19871991 and an Associate Professor during 19911996 in the
Department of Civil Engineering, Chung-Ang University,
Korea. He was pres ently working as a professor in the same
University.