Sei sulla pagina 1di 8

CABLE OSCILLATION INDUCED BY PARAMETRIC

EXCITATION IN CABLE-STAYED BRIDGES

B. N. Sun1 , Z. G. Wang1 , J. M. Ko2 and Y. Q. Ni2

1
Department of Civil Engineering, Zhejiang University, Hangzhou 310027, P. R. China
2
Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hong Kong

ABSTRACT

In long-span cable-stayed bridges, parametric excitation is very probable due to the presence of many
low frequencies in the deck/tower and in the stay cables. When certain tuning conditions are met, even
considerably small motion of the deck and/or towers may cause dynamic instability and extremely
large vibration amplitudes of the stay cables. In this study, we propose a nonlinear dynamic model for
the simulation and analysis of a kind of parametrically excited vibration of stay cables caused by
support motion in cable-stayed bridges. Based on this model, the oscillation mechanism and dynamic
response characteristics of this kind of vibration are analyzed through numerical computation. It is
noted that the parametrically excited response of stay cables may occur in cable-stayed bridges due to
deck vibration under the condition that the natural frequency of a cable approaches to about half of the
first mode frequency of the bridge deck system. A TMD system installed beneath the bridge deck is
proposed as a possible measure to suppress the cable parametric vibration. Numerical simulation
shows that this kind of vibration can be considerably reduced with the proposed TMD system.

KEYWORDS

Stay cable, cable-stayed bridge, parametric vibration, support excitation, TMD system.

INTRODUCTION

Long-span cable-stayed bridges have become increasingly popular due to their effective use of
materials and their pleasant aesthetics. The construction of cable-stayed bridges is now entering a new
era with main span reaching 1000m (Russell 1999). Increasing bridge span makes the stay cables
prone to vibrations under wind, rain, traffic and seismic loading. The rain-wind-induced cable
vibrations have been observed in a number of cable-stayed bridges worldwide, and have been
extensively studied in the past decade. In the recent years, the cable vibrations with large amplitudes
caused by parametric excitation due to support (deck or tower) motions were also observed in cable-
stayed bridges (Pinto da Costa et al. 1996; Virlogeux 1998; Yamaguchi and Fujino 1998). This kind of
cable oscillations is produced by the deck and/or tower vibrations induced either by vortex shedding
or by buffeting for oblique winds.

The oscillation phenomena and control methods of bridge stay cables excited by the deck/tower
motion are less studied than the rain-wind-induced vibrations. Parametric oscillation and dynamic
stability of suspended cables subjected to support excitation has been studied by Takahashi (1991) and
Perkins (1992). Abdel-Ghaffar and Khalifa (1991) investigated in detail the modal interaction between
the deck-dominated modes and the cable local modes. Fujino et al. (1993) and Warnitchai et al. (1995)
studied theoretically and experimentally the internal resonance between the bridge global modes and
the cable local modes involving linear and quadratic nonlinear couplings. Lilien and Pinto da Costa
(1994), Pinto da Costa et al. (1996) and Takahashi et al. (1997) used a semi-discrete model for the
studies of parametric oscillations of stay cables induced by periodic motions of the deck and/or
towers. Kang and Zhong (1998) proposed a simple mathematical model of MDOF for analyzing the
parametric vibration of stay cables. Zhang and Peil (1999) used a spline finite element model to study
the parametric instability of stay cables under arbitrary dynamic excitation.

The previous studies reveal that when the vibration of the bridge deck or tower falls in certain
frequency ranges, the oscillation of stay cables caused by support motions can become unstable and
exhibit large response amplitude. The most dangerous situations arise when the vibration frequency of
the deck or tower is in the neighborhood of two times the first natural frequency of stay cables. In the
present study, we propose a nonlinear dynamic model for the simulation of cable parametric vibration
induced by bridge deck motion, in which the cable is modeled as a nonlinear string with uniformly
distributed mass and the stiffness effect of bridge deck is represented by a spring. Based on the
proposed model, the large-amplitude oscillation of stay cables excited by support motions is analyzed
by use of the method of multiple scales, and the suppression of cable parametric vibration adopting a
TMD system is studied.

MODEL DEVELOPMENT

Figure 1 shows the cable-deck model proposed for analysis of the cable parametric vibration caused
by deck motions. The stay cable is modeled as a nonlinear string with the initial length L, and uniform
mass density ρA per unit length. The geometric nonlinearity of the cable is introduced so that the
dynamic interaction between the cable and deck can be accounted for. The cable possesses an axial
elastic stiffness EA, but the weight and bending stiffness of the cable are omitted. The static tension
force T0 of the cable represents the action of weight and elastic deformation of the bridge deck with
the mass m and bending stiffness k. Only the motion of the cable in the plane of static equilibrium
configuration is considered in the present work.

The relationship between the transverse displacement x 1 at position z and the tension force T(t) of the
cable can be obtained from the standard string theory as

∂ 2 x1 ∂ 2 x1
ρA = T (1)
∂t 2 ∂z 2

For the taut string, the transverse displacement x 1 (z, t) can be expressed by use of the separation of
variables as
x1 ( z , t ) = ∑ X i (t )ϕi ( z ) (2)
z

L x1

m
kc x2
k
xc mc

Figure 1: Dynamic model of stay cable

where ϕi(z) is the ith mode shape of the string vibration and Xi(t) is the ith generalized coordinates.
For the sake of simplicity, only the first mode ϕ(z) = cos(πz/L) is taken into account, namely,
x1 = X 1 (t ) cos(πz / L ) (3)

Substituting Eqn. 3 into Eqn. 1 yields


X&& 1 + (π 2 T / L2 ρA) X 1 = 0 (4)

The total extension of the string includes the following three parts: (1) the initial extension z0 due to
tension force T0 ; (2) boundary extension X2 induced by the movement of mass m; (3) extension δ
during the elastic deformation. Therefore the tension force T can be expressed as
T = ( AE / L)[ z 0 + X 2 + δ ] (5)

For stay cables in cable-stayed bridges, the sag influence can be ignored because the ratio of sag to
cable length is very small (i.e., ∂x1 / ∂z << 1 ). By using Eqn. 3, the extension δ induced by elastic
deformation can be obtained as
∂x L 2 ∂x1 2 π 2 X 12
δ = ∫ ( ds − dz ) = ∫−LL22 {dz[1 + ( 1 ) 2 ]1 / 2 − dz} ≈ ∫0 ( ) dz = (6)
∂z ∂z 4L
Therefore the tension force T in Eqn. 5 can be expressed as
AEz 0 AE π 2 X 12
T = To + Td = + [X 2 + ] (7)
L L 4L

By substituting Eqn. 7 into Eqn. 4, we have


X π 2 X 12
X&&1 + ω12 [1 + 2 + ] X1 = 0 (8)
z0 4 Lz 0
where ω12 = π 2 T0 / L2 ρA .
The differential motion equation of mass m can be expressed as
Td + mX&& 2 + kX 2 = 0 (9)

By combining Eqn. 8 and Eqn. 9, the nonlinear governing equations of motion of the cable-deck
system are derived as
X&&1 + ω12 (1 + X 2 / z 0 + KX 12 ) X 1 = 0 (10a)
AE
X&& 2 + ω22 X 2 + Kz 0 X 12 = 0 (10b)
mL
where K = π /( 4Lz 0 ) and ω 2 = AE /( mL) + k / m .
2 2

Because of the nonlinear terms in Eqn. 10, the dynamic responses X1 (t) and X2 (t) are coupled strongly.
Although the structural parameters of the system do not contain time t explicitly, the vibration of mass
m will cause the cable tension force to vary with time t because the cable sustains a restoring force
depending on X2 . That is, the cable is excited by the bridge deck through parametric vibration. In this
way, the transverse vibration of the cable and the vertical vibration of the deck (mass m) are coupled
with each other. According to the internal resonance criteria of second-order nonlinear systems, the
vertical vibration of the cable will exhibit large-amplitude responses when ω1 ≈ 0.5ω2 . In this case, the
coupled vibration of the cable and the mass m performs as parametric resonance and the cable
vibration will be unstable.

ANALYSIS OF PARAMETRIC VIBRATION

Numerical analysis has been accomplished to reveal the oscillation mechanism and dynamic response
characteristics of cable parametric vibration induced by deck motion. In the numerical computation,
the parameters of the cable-deck system are taken as same as those given in Kang and Zhong (1998):
mass per unit length of cable ρA = 0.25kg/m; cable length L = 40m; elastic stiffness of cable EA =
16000N; static tension force T0 = 101N. Assuming the mass of bridge deck per unit length is m =
100kg; the stiffness of deck k = 600N/m, and initial perturbations are X1 = 0.001m and X2 = 0.1m
respectively. In this case, the frequency ratio of the cable to the bridge deck is ω1 /ω2 = 0.4992. By
solving Eqn. 10 using the method of multiple scales (Nayfeh and Mook 1979), the free vibration
responses of the cable and the deck, X1 (t) and X2 (t), are obtained as shown in Figure 2.
The calculation results show that the vibrations of the cable and the bridge deck are coupled
inseparably. The cable oscillation excited by the deck motion becomes unstable when ω1 ≈ 0.5ω2 . It is

(m) (m)
1

0.75 0.1

0.5
0.05
0.25

0 0
-0.25

-0.5 -0.05

-0.75
-0.1

0 50 100 150 200 250 300 t (sec) 0 50 100 150 200 250 300 t (sec)
(a) at the cable (b) at the deck
Figure 2: Time histories of displacement responses (T0 = 101N, m = 100kg, k = 600N/m)
found that the response amplitude of the cable increases drastically from the initial perturbation
0.001m to the maximum response peak 0.8m. It is clear from Figure 2 that the system vibrations
exhibit obvious beating characteristic. The time instant at which the response of the cable reaches
its maximum is exactly the time instant at minimum response of the deck. The energy of the dynamic
system is transferred between the bridge deck and the cable. By comparing the analysis results of the
present study with those given in Kang and Zhong (1998), the revealed beating phenomenon is
similar but the beating frequency ratio ω1 /ω2 is less than their result. This is because in their
study the cable was modeled by a lump mass located at the mid-span of the cable, resulting in the
values of ω1 and ω2 different from those obtained in the present study. Based on the cable linear
theory (Irvine 1981), the natural frequency ωs of a linear elastic cable is governed by

tan( Ωs / 2) = (Ωs / 2) − ( 4 / λ2 )(Ω s / 2) 3 (11)


where Ω s = ω s L T0 ρA .

The value of parameter λ2 is very small for a flat-sag cable. By ignoring the λ2 term in Eqn. 11, the
expression of the cable natural frequency can be obtained as

ω s = ( s π / L) T0 / ρA ( s = 1, 2, L, n ) (12)

The first mode frequency derived from Eqn. 12 is exactly same as the ω1 used in the present study.
However, the first frequency obtained from the simplified model in Kang and Zhong (1998) is less
than that calculated from Eqn. 12. This explains why the value of the beating frequency ratio in
the present study is different from the value given in Kang and Zhong (1998).

(m) (m)
1
0.1
0.75

0.5
0.05
0.25

0 0

-0.25
-0.05
-0.5

-0.75
-0.1

0 50 100 150 200 250 300 t (sec) 0 50 100 150 200 250 300 t (sec)
(a) at the cable (b) at the deck
Figure 3: Time histories of displacement responses (T0 = 101N, m = 50kg, k = 100N/m)

(m) (m)
1
0.1
0.75

0.5
0.05
0.25

0 0

-0.25
-0.05
-0.5

-0.75
-0.1

0 50 100 150 200 250 300 t (sec) 0 50 100 150 200 250 300 t (sec)
(a) at the cable (b) at the deck
Figure 4: Time histories of displacement responses (T0 = 201N, m = 100kg, k = 1600N/m)
The analysis results of Kang and Zhong (1998) indicated that when the ratio of the cable frequency to
the deck frequency keeps a constant value, the mass ratio of cable to deck would influence the cable
dynamic response significantly. The present study reveals that the mass ratio affects the response
amplitude of the cable, but does not influence the beating frequency. Figure 3 shows the cable
and deck responses in the case when the deck mass m = 50kg, the deck stiffness k = 100N/m, and the
static tension force T0 remains same as before (T0 = 101N, so the frequency ratio keeps unvaried). It is
seen that the dynamic response amplitudes of the cable are decreased in comparison with the previous
case but the beating frequency is not changed. Also it is found that the larger the mass ratio of
deck to cable is, the larger the vibration amplitude of the cable.
It is revealed that change in the static tension force of the cable will alter the beating frequency of
the parametric vibration, but does not influence the dynamic response amplitude of the cable. Figure 4
shows the cable and deck responses when the deck mass m = 100kg and the static tension force is
increased to be T0 = 201N. In this case, the deck stiffness is adjusted to be k = 1600N/m so that the
frequency ratio of cable to deck keeps the same as before. It is seen from Figure 4 that the dynamic
response amplitude of the cable does not change much, but the beating frequency is significantly
altered. Also, it is found that the larger the static tension force T0 is, the lower the beating
frequency of the parametric vibration system.

CONTROL OF PARAMETRIC VIBRATION

A damping system can be installed on the cable or bridge deck to reduce the cable dynamic response
caused by the parametric vibration. Because the cable parametric vibration is induced by the bridge
deck motion, we propose here a cable vibration suppression measure by installing a tuned mass
damper (TMD) under the deck as shown in Figure 1. The governing equation of motion of the cable-
deck-TMD system cab be written as
X&&1 + ω12 (1 + X 2 / z 0 + KX 12 ) X 1 = 0 (13a)
AE k c
X&& 2 + ω 22 X 2 + Kz 0 X 12 + c ( X 2 − X c ) + ( X& 2 − X& c ) = 0 (13b)
mL m m
c &
X&& c + ωc2 ( X c− X 2 ) + ( X c − X& 2 ) = 0 (13c)
mc
where mc, k c and c are the mass, stiffness and damping coefficient of the TMD system respectively, as
shown in Figure 1; and ω c2 = k c mc .

Figure 5 shows the cable and deck responses of the cable-deck-TMD system when the damping
coefficient of the TMD is taken as zero (c = 0). In this example the TMD mass and stiffness are taken
to be mc = 0.5kg (equal to 1/2000 deck mass) and k c = 5N/m respectively, and the initial displacement
of the TMD is set as zero. This results in ωc ≈ ω2 . It is observed from Figure 5 that the dynamic
response amplitude of the cable can be reduced considerably when the ratio of TMD stiffness k c to
mass mc is adjusted properly to meet ωc ≈ ω2 . Compared with Figure 2, the cable response amplitude
is reduced to about 0.003m from the original 0.8m, while the vibration amplitude of the deck has not
been magnified. Therefore, the TMD is efficient for reducing the cable parametric vibration. The
beating frequency of the system increases greatly when the TMD is added.
Figure 6 illustrates the cable and deck responses of the cable-deck-TMD system when the damping
coefficient of the TMD is taken as c = 0.05N⋅s/m. The other parameters of the TMD are same as
(m) (m)

0.004 0.1

0.002 0.05

0 0

-0.002
-0.05

-0.004
-0.1
0 20 40 60 80 100 120 140 t (sec) 0 20 40 60 80 100 120 140 t (sec)
(a) at the cable (b) at the deck
Figure 5: Time histories of displacement responses with TMD (mc = 0.5kg, k c = 5N/m, c = 0)

(m) (m)
0.004
0.1
0.003

0.002
0.05
0.001

0 0
-0.001

-0.002 -0.05

-0.003
-0.1

0 20 40 60 80 100 120 140 t (sec) 0 20 40 60 80 100 120 140 t (sec)


(a) at the cable (b) at the deck
Figure 6: Time histories of displacement responses with TMD (mc = 0.5kg, k c = 5N/m, c = 0.05N⋅s/m)

before. It is seen from Figure 6 that the TMD system with damping can effectively reduce the
dynamic responses of both the cable and the bridge deck. The quasi-steady state response of the cable
in the case of c = 0.05 N⋅s/m is only about half of the response amplitude in the case of c = 0. From
the viewpoint of energy, the TMD absorbs the majority of vibration energy in the bridge deck and
therefore the vibration energy transferred from the deck to the cable is reduced.

CONCLUSIONS

The following conclusions are drawn from the numerical analyses explored in the present study:

1. In cable-stayed bridges, if the ratio of natural frequencies of cable to bridge deck falls into certain
range, serious parametric vibrations of the cables with large amplitude may occur;

2. The parametric vibrations of cables can exhibit obvious beating characteristic. The cable
dynamic response is dependent on the vibration level of the bridge deck, and the frequency and
mass ratios of cable to deck, but independent of initial perturbation. The beating frequencies
relate to tension force of the cable. The larger the static tension force is, the lower the beating
frequency. The matching frequency exciting cable parametric vibration in deck motion also relates
to the vibration amplitude of the deck and the mass ratio of cable to deck;

3. Installation of a TMD beneath the bridge deck can effectively suppress the deck-motion-induced
cable parametric vibration when appropriate TMD parameters are determined. When no damping
included, the TMD can considerably reduce the cable response amplitude and while keeping the
deck vibration level being unvaried. When the TMD system contains damping as well, the dynamic
responses of both the cable and the bridge deck can be significantly reduced and the beating
phenomenon is weakened.

ACKNOWLEDGMENTS

The funding support from the Hong Kong Polytechnic University to the first author when he worked
as a visiting scholar at the Hong Kong Polytechnic University is gratefully acknowledged. The third
and fourth authors are supported by a grant from the Research Grants Council of the Hong Kong
Special Administrative Region, China (Project No. PolyU 5045/00E).

References

Abdel-Ghaffar A.M. and Khalifa M.A. (1991). Importance of Cable Vibration in Dynamics of Cable-
Stayed Bridges. ASCE Journal of Engineering Mechanics 117, 2571-2589.
Fujino Y., Warnitchai P. and Pacheco B.M. (1993). An Experimental and Analytical Study of
Autoparametric Resonance in a 3DOF Model of Cable-Stayed-Beam. Nonlinear Dynamics 4, 111-
138.
Irvine H.M. (1981). Cable Structures. The MIT Press, Cambridge, Massachusetts.
Kang Z. and Zhong W.X. (1998). Numerical Study on Parametric Resonance of Cable in Cable
Stayed Bridge. China Civil Engineering Journal 31:4,14-22.
Lilien J.L. and Pinto da Costa A. (1994). Vibration Amplitudes Caused by Parametric Excitation of
Cable Stayed Structures. Journal of Sound and Vibration 174, 69-90.
Nayfeh A.H and Mook D.T. (1979). Nonlinear Oscillations. Wiley-Interscience, New York.
Perkins N.C. (1992). Modal Interactions in the Non-Linear Response of Elastic Cables under
Parametric/External Excitation. International Journal of Non-Linear Mechanics 27, 233-250.
Pinto da Costa A., Martins J.A.C., Branco F. and Lillien J.L. (1996). Oscillations of Bridge Stay
Cables Induced by Periodic Motions of Deck and/or Towers. ASCE Journal of Engineering
Mechanics 122, 613-622.
Russell H. (1999). Hong Kong Bids for Cable-Stayed Bridge Record. Bridge Design and Engineering,
Second Quarter, p.7.
Takahashi K. (1991). Dynamic Stability of Cables Subjected to an Axial Periodic Load. Journal of
Sound and Vibration 144, 323-330.
Takahashi K., Herath C.R. and Hanada H. (1997). Nonlinear Dynamic Response of Cable Subjected
to Support Excitation. Proceedings of the Asia-Pacific Vibration Conference 97, Kyongju, Korea,
Vol. 2, 940-944.
Virlogeux M. (1998). Cable Vibrations in Cable-Stayed Bridges. Bridge Aerodynamics, A. Larsen and
S. Esdahl (eds.), A.A. Balkema, Rotterdam, Netherlands, 213-233.
Warnitchai P., Fujino Y. and Susumpow T. (1995). A Non-Linear Dynamic Model for Cables and its
Application to a Cable-Structure System. Journal of Sound and Vibration 187, 695-712.
Yamaguchi H. and Fujino Y. (1998). Stayed Cable Dynamics and its Vibration Control. Bridge
Aerodynamics, A. Larsen and S. Esdahl (eds.), A.A. Balkema, Rotterdam, Netherlands, 235-253.
Zhang Q.-L. and Peil U. (1999). Dynamic Behaviours of Cables in Parametrically Unstable Zones.
Computers and Structures 73, 437-443.

Potrebbero piacerti anche