You are on page 1of 21

Vibration Mechanisms and Controls of Long-Span Bridges: A Review

Yozo Fujino, Prof.; Dionysius Siringoringo, Res. Asst. Prof.; Bridge and Structure Lab, Civil Engineering Dept., University of Tokyo, Japan. Contact:
DOI: 10.2749/101686613X13439149156886

Dynamic performance is an important consideration in long-span bridge design. Owing to its flexibility and low damping, various types of vibration from different sources of excitation could occur during the lifetime of a long-span bridge. This paper reviews important studies and developments on long-span bridge vibration mechanism and control under wind, seismic,traffic and human-motion excitations. Types of vibration commonly observed on the long-span bridge are discussed from the viewpoint of structure engineering. Discussion for each subject is commenced by describing the vibration mechanism followed by the survey on observed phenomena in many long-span bridges associated with the type of vibration. The paper also describes the engineering solutions adopted as countermeasures for each type of bridge vibration problem. Keywords: long-span bridge vibration, vibration control, wind-induced vibration, cable vibration, seismic-induced vibration

traffic and human-motion excitation. Interested readers are encouraged to refer to literatures cited on each subject for more detailed explanation.

Wind-Induced Vibration of Bridges

Wind is a spatiotemporally varying dynamic and random phenomenon, and therefore its effects on structure also vary in time and space. For a certain time period, wind speed can be defined by the time-averaged (mean) component and the fluctuating component. The total wind force on a bridge is the summation of the mean wind force (also known as the static component), the fluctuating wind force due to turbulence and the motion-induced wind force. The force applied on bridge components cause vibration in three principal directions; drag force causes vibration in along-wind direction, lift force creates vibration in cross-wind direction and the moment force causes vibration in torsional direction. Determination of wind forces is a difficult task as they are sensitively influenced by the turbulence. In designing a bridge, the peak force and displacement in the respective directions are of primary concern. For long-span bridges, wind-induced vibration is the most critical among various types of dynamic excitation. Type and level of vibration depend on bridge structural properties (i.e. mass, stiffness and damping), wind force condition and interaction between wind and bridge. The wind-induced vibration mechanisms in a bridge can be classified on the basis of force excitation into four mechanisms: forced vibration, self-excited vibration, combination of forced and self-excited vibration and random vibration. The bluff body of bridge components such as deck, pylon and cable immersed in a wind flow can move and create a motion that in turn affects the wind flow around the body. Vibration created by this interaction is known as the

Construction of long-span bridges has been very active worldwide in the past few decades. The worlds longest suspension bridge Akashi-Kaikyo in Japan with central span of 1991 m was completed at the end of 20th century, and some of the cable-stayed bridges exceeding 1000 m, such as Stonecutters, Sutong in China and Russky Bridge in Russia with central span of 1104 m were completed in the beginning of 21st century. A number of long-span bridges are now under construction in China1 and Korea,2 and the plans to build super long-span bridges in other parts of the world are also being discussed. As bridge spans get longer and pylons get taller, they become more flexible and prone to vibration. Figure 1 shows how flexible the bridge becomes as the main span increases. Flexible structures tend to vibrate under dynamic loading such as wind, earthquake, vehicle movement, and human motion. Vibration can have several levels of consequence, from a potentially hazardous effect such as causing immediate structural failure to a more subtle
Peer-reviewed by international experts and accepted for publication by SEI Editorial Board Paper received: February 8, 2012 Paper accepted: January 17, 2013

but more extended effect such as structural fatigues. In addition to that, vibration can also affect user safety and comfort and limit bridge serviceability. In the past few decades, extensive research and development has been carried out to understand the mechanisms behind bridge vibration and to reduce undesirable vibration effects through various countermeasures. Results of these research and developments have been adopted in bridge design codes and put into practices by specifying methodologies and guidelines for countermeasures, and by introducing new structural elements or devices as vibration control. In this paper, the authors review important research and developments in long-span bridge vibration with emphasis on mechanism and control. Types of vibration commonly observed on the long-span bridge are discussed from structure engineering viewpoints. Some vibration mechanisms are now well understood, while some others still require further studies to achieve complete understanding. Surveys on the phenomena associated with the type of vibration reported in many long-span bridges are also presented as well as engineering solutions adopted as countermeasures. Owing to space limitation, discussions presented in this paper focus only on bridge vibrations resulting from wind, seismic,


Scientific Paper

Structural Engineering International 3/2013

0,6 1st lateral mode 0,5 1st vertical mode 1st torsional mode 0,4

Natural frequency (Hz)

sections use the original terminologies as described in early literatures on bridge wind-induced vibration while providing other terminologies that have also been used to describe the same phenomena. Aeroelastic Instabilities
Proposed Messina Bridge



AkashiKaikyo Bridge


0 500 1000 1500 2000 2500 3000 3500

Main span length (m)

Fig. 1: Relationship between the lowest natural frequencies and span length obtained from the worlds 40 longest span suspension bridges
Vibration amplitude Torsion ()
6 24

When the structure motion and aerodynamic force interact significantly, the amplitude of self-excited vibration could grow in time with diverging characteristics and induce instability. This phenomenon is known as the aeroelastic instability and there are three types of such phenomenon that can occur in bridge sections: torsional flutter, coupled flutter and galloping. Analytical formulation for studying aeroelastic instabilities originated from aerospace studies, especially in solving the aircraft flutter problem. Early formulation of motion-induced forces on aircraft provided analytical framework for experimenting which was further adopted for analysis of wind-induced vibration in bridge decks. In Theodorsens seminal work,4 which is considered by many as the foundation of flutter analysis, the relationships for unsteady aerodynamic lift force L and moment M acting on a thin airfoil are defined as linear functions of h (lift displacement) and a (twist angle), their time derivatives and the analytical Theodorsens complex circulation function C(K) = F(K) iG(K), which is a function of reduced frequency of oscillation K = w B/U (w : frequency of motion, B: width of the body and U: average wind velocity). This analytical relationship can predict the mean wind speed at which flutter will occur. However, implementations of Theodorsens formulation to bridge decks flutter problem proved unsatisfactory. Attempts to analyze the Tacoma Narrows incident as a flutter problem by applying the Theodorsens aerodynamic formulation to the bridge found that the estimated critical flutter speed was considerably higher than what was experienced.5 Theodorsens circulation function becomes inapplicable as it is derived from a thin airfoil with fully attached flow while most bridge decks are classified as bluff bodies that experience separate flow over significant portions of their surface. Another study6 later suggested that experimentally determined aerodynamic coefficients rather than Theodorsens analytical coefficients could be more useful to determine unsteady aerodynamic forces.
Scientific Paper 249


Torsional flutter 1NT

Vertical 102 (m)


12 1NT (vortex)

1NV (vortex) 0NV (vortex) 2NV (vortex) 3NV (vortex)

0,0 0,3 0,45 0,6 0,75 0,9 1 1,2 1,35 1,5

Wind velocity (m/s)

Fig. 2: Wind-induced vibration type with respect to wind velocity (adopted from aeroelastic full model, wind-tunnel test of the original Tacoma Narrows Bridge,3 NV: vertical bending, NT: torsion)

self-excited vibration and the resulting aerodynamic force is known as motiondependent force. A phenomenon wherein structural motion and aerodynamic force interact significantly is called aeroelastic phenomenon. The aeroelastic phenomena commonly observed in long-span bridges are the galloping, torsional flutter, coupled flutter and vortex-induced vibration (VIV). The occurrence of aeroelastic phenomena depends on wind velocity as schematically shown in Fig. 2. Moreover, wind-induced vibration and its effects on bridge can be classified in terms of stability, type of amplitude response (i.e. vibration with self-limited amplitude or diverging character) and the range of wind velocity where it occurs as illustrated by Fig. 3. In addition to the five types of vibration in this figure, this paper will also discuss wind-induced vibration under

Galloping,torsional flutter,coupled flutter (divergent)

Vibration amplitude (vertical or torsional)

Vortex-induced vibration (self-limited)

Buffeting Buffeting

Wind velocity

Fig. 3: Schematic figure showing wind-induced vibration phenomenon and its effect on bridge

special conditions such as wake interference, rainwind-induced vibration of stay cable, dry-inclined cable galloping and parametric vibration on cablestayed bridge. One would realize that many terminologies have been used differently in describing phenomena in wind-induced vibration. The following

Structural Engineering International 3/2013

Extensive experimental evaluation had been attempted to define unsteady aerodynamic forces of bridge deck without the thin airfoil model assumption. Two types of experiments were conducted: one by directly measuring the aerodynamic force components when the body is in a given specific motion; another by calculating the force indirectly from the induced motion of the body. Probably, the direct method for the measurement of unsteady aerodynamic forces to bridge deck sections was first applied in Ref. [7]. In the experiment, rigid bridge deck models were mechanically driven into a simple harmonic motion with a range of specific frequency and amplitude in two-dimensional (2D) air stream, and the reactions at the model supports were detected. The use of indirect measurement of aerodynamic forces for bridge decks was pioneered in Ref. [8] by detecting the induced response of models in air flow. The indirect measurement generally requires lesscomplicated experimental set-ups than the direct measurement does. This technique is now being widely practiced all over the world and will be described in the following sections. Consider a bridge girder modelled with an N degree-of-freedom under aerodynamic force. The left-hand side of Eq. (1) describes structural properties of the bridge (i.e. mass (M), structural damping (C) and stiffness (K) symmetrical matrices), while the right-hand-side of equation describes the unsteady motion-dependent aerodynamic force (i.e. FV and FD). The term motion-dependent implies that the force contains not only the instantaneous value, but also the effect of previous motion. The forces are defined as a function of frequency (non-dimensional frequency of oscillation K = w/U), structure velocity x (t) and structural displacement x(t):

and potential energy derivation of the system in Eq. (2), the time rate of the total energy can be obtained as9:


Therefore, the increment of total energy in the system of the bridge motion over a period of time T0 can be written as:

dominant than the torsion results in bending flutter. Meanwhile, the torsional flutter resulted when torsional vibration is more dominant than vertical vibration. If the term I2 dominates, the associated stiffness-coupling instability, known as the coupled flutter, is related to phase lag or coupling effect between different vibration modes. Galloping Galloping is defined as single degreeof-freedom large-amplitude aeroelastic oscillation in transverse direction. It is also known as cross-flow galloping, translational galloping, bending flutter or cross-wind galloping. Galloping can be experienced by prisms with certain cross section, D-sections and circular section with some accretion such as in ice-laden cables. The amplitude of oscillation can reach up to ten times or more the cross-sectional dimension of the body. In the galloping phenomenon, the effective angle of attack of wind velocity with respect to the body changes as the body experiences initial motion. On some cross-sectional shapes, this change creates different pressure distributions that enhance the initial motion. Mechanism of galloping can be explained by the quasi-steady aerodynamic theory as explained in Refs [10,11]. Consider a bluff body under self-excited force as shown in Fig. 4 undergoing wind excited vibration in transverse (h) direction with effective angle of attack, a. One can define the equation of motion as:

(4) Equation (4) implies that the total energy is increasing and the system becomes unstable only when E 0. This unstable condition is defined as flutter, and there are two necessary conditions for the onset of flutter, namely I1 and I2. The term I1 describes the condition associated with diagonal aerodynamic damping matrix and energy due to individual single degreeof-freedom modal motion. The term in I2, on the other hand, is associated with the non-diagonal aerodynamic stiffness matrix and the coupled modal motion, and is referred to as coupled flutter. In the aeroelastic analysis of bridge deck, equation of motion of deck can be described by two degree-offreedom vertical (h) and torsional (a) motion. (5a) (5b) where xh and xa denote the structural damping associated with bending and torsional motion, wh and wa are the natural frequency of bending and torsional mode, respectively; while m and I are the mass and mass moment of inertia, respectively. The aerodynamic forces are described as lift force (Lh) and moment force (Ma) in vertical and torsional direction, respectively. Following description of total energy as previously explained one can distinguish two conditions for aeroelastic instabilities, namely, the single-degree-of-freedom (SDOF) damping-induced instability in either bending or torsion when the term I1 dominates and the stiffness-coupling instability when the term I2 dominates. In the case of SDOF instability, a condition where vertical vibration is more

(6) Defining L = and D = rU2B, total damping of the system in Eq. (6) consists of structural damping and aerodynamic damping, denoted as , where the terms in the bracket are evaluated at a = 0. Considering that structural damping x is positive, the system
F 1 _ C rU2B 2 L 1 _ C 2 D

(1) Equation (1) can be rewritten in the form: (2) where P = C Fv(K) and Q = K FD(K) are the total damping and total stiffness of the system, respectively. Matrices P and Q can be both decomposed into P = P1 + P2 and Q = Q1 + Q2, in which the subscripts 1 and 2 denote the symmetric and skew-symmetric matrices, respectively. Using kinetic
250 Scientific Paper

Urel h U

Fig. 4: Schematic figure of section and wind for quasi-steady analysis of galloping

Structural Engineering International 3/2013

becomes potentially unstable only if the aerodynamic part becomes nega. tive, that is when This galloping threshold is known as the Den Hartog criterion.12 The aerodynamic lift and drag coefficients are obtained experimentally by steady force measurement as a function of angle a, and by following the above criterion, one can obtain the critical wind velocity for the onset of galloping as: (7) From the above derivations, it is clear that galloping is a velocity-dependent phenomenon associated with the occurrence of negative aerodynamic damping. The tendency of a structure to gallop can be evaluated by checking the time-averaged lift and drag coefficients and their signs at a = 0. Negative slope of lift force indicates the tendency for galloping. The propensity of bluff body to gallop depends on sectional characteristics. The section with smaller height-towidth ratio in cross-flow direction has larger tendency to gallop (soft galloping), while the one with larger ratio requires initial perturbation (hard galloping). Furthermore, turbulence also controls the occurrence of galloping, as it changes both the lift and drag force on the bluff body and the separation of flow around the body. It is observed13 that large turbulence intensity causes hard galloping to become soft and soft galloping to become weaker before eventually disappearing. In the foregoing discussion, the critical wind speed for galloping has been predicted according to linear theory. The non-linear quasi-steady aerodynamic theory of galloping was formulized in Ref. [14] using seventh power polynomial approximation to determine the aerodynamic coefficients. It was found that critical velocity is the same as that predicted by the linear theory, only as long as the bifurcation is supercritical. In addition to critical wind velocity, non-linear theory also gives estimate of the amplitude and frequency of galloping. Other studies11,15 have extended the Parkinson and Smith non-linear model14 to continuous elastic structures such as cantilevered towers and proposed the universal response curve that predicts galloping of a given prismatic body shape. Interested readers can find

developments related to galloping mechanisms in other works.13,16 Torsional Flutter Torsional flutter is single degree-offreedom torsional aeroelastic instability. It is also sometimes called torsional galloping, rotational galloping or simply flutter. The mechanism of torsional flutter was initially explained by quasisteady theory in the linear and nonlinear forms in Refs. [17, 18]. Later studies19,20 suggested and validated through series of experiments that the torsional flutter is in fact an unsteady phenomenon. Therefore, unlike galloping, the self-excited force for torsional flutter is an unsteady force and must be described as a function of reduced frequency (K). Torsional flutter occurs when the total damping (structural plus aerodynamic) of the system in torsional motion becomes zero. The unsteady aerodynamic forces are defined in terms of flutter derivatives, which are the function of non-dimensional frequency of oscillation K = w B/U.8,21 The aerodynamic lift force (Lh) and moment force (Ma) are defined as:

mining critical flutter wind speed was by bimodal approach. In this approach, interaction between wind and bridge deck is modelled as a 2D heaving (h) and twisting (a) system. The unsteady aerodynamic forces are described as a function of reduced frequency and the body is assumed to undergo a simple harmonic motion with the same frequency in both heaving and twisting. This type of analysis can be tedious and computationally demanding, especially in the pre-computer days. Therefore, some simplifications were proposed to estimate the flutter speed, among which the most popular one for bridge deck flutter analysis is Selbergs formula22 that was based on theoretical thin airfoil. The use of bimodal approach is not always adequate because often contributions of higher modes are found to be significant. Therefore, mode-by-mode approach is preferred for the flutter analysis. The previous bimodal timedomain flutter analysis8 was extended to multi-mode analysis.21 In the multimode analysis, the onset of flutter is determined by solving the complex eigenvalue problem related to characteristics equation of the system (see,


(8.b) The flutter derivatives, denoted as H* i (K), Ai*(K), for i = 1 to 4 are obtained experimentally in the wind tunnel from sectional test of bridge deck. Using * properties of flutter derivative A2 one can estimate the critical wind velocity as the onset of torsional flutter. The * positive values of A2 represent the tendency for torsional flutter to occur. Coupled Flutter For most bridges, self-excited force produces less significant aerodynamic coupling among modal responses, hence SDOF torsional mode generally dominates the critical flutter speed. However, there is also type of self-excited motion known as coupled flutter or classical flutter that occurs on several degrees of freedom where, the term I2 in Eq. (4) is dominant. In this case, the onset of flutter depends on phase lags or degree of coupling among modes. In the early history of bridges flutter analysis, the common method of deterfor example, procedure in Refs. [23, 24]). Further development includes flutter analysis for three-dimensional (3D) bridge structures, where the unsteady aerodynamic forces are applied directly to a 3D finite element model, or analyzing various 3D vibration modes and assembling them using superposition method. Example of multi-mode 3D flutter analysis of the very long-span Akashi-Kaikyo Bridge is presented in Ref. [25]. The analyses show that six primary modes play significant roles in determining the flutter condition. As the flutter derivatives are functions of reduced frequency, the analysis is basically conducted in the frequency domain. This analysis requires an iterative procedure to obtain the flutter critical wind speed. Time-domain analysis has recently been explored by methods such as indicial function26 and rational function approximation (RFA).27 It has advantages over frequency domain in that it can include non-linear analysis
Scientific Paper 251

Structural Engineering International 3/2013

and provide a platform for flutter control analysis. However, one critical issue in time-domain analysis is the determination of the indicial function from experimentally obtained flutter derivatives, especially for non-streamlined cross section. Occurrences of Aeroelastic Instabilities and Their Countermeasures Galloping can occur on bridge components such as girder, pylon and stay cable. Non-streamlined box girders with relatively small width-to-depth ratio are particularly susceptible to galloping. The flow field under the bridge deck is an important factor to control this phenomenon. Although cases of large amplitude of across-wind galloping of bridge have rarely been observed, countermeasure should be provided in design. In Japan, the galloping of bridge girder is often avoided by means of aerodynamic countermeasures such as adding lower-skirts and horizontal plate on the lower side of box girder cross section (Fig. 5). Examples of such countermeasures are the use of lowerskirts at the corner of Tozaki viaduct girder (part of the Akashi-Ohnaruto line) and horizontal plate on Namihaya

Bridge.28 These countermeasures can reduce wind velocity of the separated flow that originated from the leading edge of girders lower corner, and can thus eliminate the self-excitation force on the girder. Tall and slender pylons are also susceptible to galloping. An example of galloping on the bridge pylon is the damage reported on the single hexagonal shaft of inverted Y-shape pylon of the Lodemann cable-stayed bridge in Germany in 1972 during strong wind excitation of about 40 m/s.31 Pylons of cable-stayed bridges with low transverse beam are also prone to gallop. Pylons of the Higashi-Kobe cablestayed bridge have this characteristic, and galloping during free-standing construction stage was observed in wind tunnel testing.32 In Japan, aerodynamic countermeasures are often applied to prevent galloping of bridge pylons in the form of sectional modification such as corner cut (Tatara Bridge) or by adding appendages: arch-shaped deflector (Katsushika-Harp Bridge), cover plate (Higashi-Kobe Bridge), fairing plates (Megami Bridge).28 Cables in cable-supported bridges commonly have circular cross section and

thus are not prone to gallop. However, accretion of ice on their surface may change the galloping characteristics. Several cases of excessive vibration on bridge cables involving ice accretion have been reported. Among suspension bridges, across-wind and alongwind vibration of the Great Belt East Bridge hanger was observed during 27 to 31 March 2001 with the maximum amplitude of 1,4 m.33 Similarly, violent vibration of stay cables observed at the Oresund cable-stayed bridge was likely to have been caused by snow and sleet accretion on the cable surface.34 A comprehensive list of the studies related to the research on ice accretion galloping of cables is presented in Ref. [35]. Long-span bridge girders are susceptible to torsional flutter. It is generally recognized now that the dramatic collapse of the Tacoma Narrows Bridge was due to torsional flutter (see Fig. 2).3 The weak torsional rigidity and aerodynamically unstable girder cross section are two main factors responsible for the collapse. Long-span bridges with non-streamlined girder and relatively large width-to-depth ratio are susceptible to torsional and coupled flutter. Since the Tacoma Narrows Bridge incident, flutter instability had become the most important issue in the design of long-span bridges. To avoid this instability, two different design approaches have been proposed; one is by streamlining the girder using a box section, this approach was promoted in the United Kingdom in structures such as Severn Bridge and Humber Bridge; and another using the truss-stiffened girder was promoted in the United States, in bridges such as New Tacoma Narrows Bridge, Mackinac Bridge and Verrazano Narrows. Pressure difference between the girder upper and lower surfaces is considered as the main mechanism that induces flutter instability. To reduce such pressure difference, part of the deck floor can be made open using grating, or in case of box girder, by applying the central slot concept.36 Effectiveness of such measures depends on the location and the opening ratio, as explained by experiments in Refs. [37, 38] for open grating system and central slotted system, respectively. Flutter stability of steel box girder can also be improved by various aerodynamic attachments such as fairings, wind nose, deflectors and spoilers.28,29 For truss girder bridge, the use of vertical stabilizer and centre barrier are effective in reducing

Lower skirts

Horizontal plates

Horizontal plates


Open grating


Fairing Vertical stabilizer



Horizontal plate

Guide vanes Fairing Central slot Curved wind flaps Flaps Flaps Fairing Flaps

Spoiler Wind nose Wind nose Deflector Wind spoiler Wind spoiler


Fig. 5: Examples of aerodynamic countermeasure for bridge truss and box girder against galloping, torsional flutter and VIV (after Refs. [2830]). (a) Aerodynamic countermeasures for bridge deck galloping; (b) Aerodynamic countermeasures for bridge deck flutter; (c)Aerodynamic countermeasures for bridge deck VIV 252 Scientific Paper

Structural Engineering International 3/2013

Bridge name Messina42 Akashi-Kaikyo43 Xihoumen44 Great Belt East45 Runyang44 Tsing Ma46

Main span (m) 3300 1991 1650 1624 1490 1377

Girder type Triple box Truss Twin box Single box Single box Single box

Flutter wind speed (m/s) Special utter countermeasures 75 Central slot Centre vertical stabilizer, Open Grating (centre and 84 both sides of girder) 78 Central slot 60 75 Centre vertical stabilizer 74 Central slot

Table 1: Examples of utter critical wind velocity and utter countermeasures on some of worlds longest span bridges4246

the possibility of flutter instabilities. Such countermeasures are adopted in the Akashi-Kaikyo Bridge.28 Figure 5 shows typical deck designs that considered aerodynamic countermeasure for flutter. In the design of long-span bridges, the onset of flutter should be avoided and possibility of its occurrence should be eliminated entirely within the flutter design wind velocity. For this purpose, wind tunnel tests on various girder sections and their aerodynamic attachments should be conducted and the reference wind speed should be carefully defined by considering the design wind speed and wind turbulence effect. The design should ensure that the critical flutter wind speed of the bridge under the worst condition is still higher than the reference wind speed. Table 1 lists flutter critical wind velocity on several long-span bridges and their countermeasures. Interested readers can find developments related to flutter mechanisms and its stabilization of bridge in Refs. [30,39]. Vortex-Induced Vibration (VIV) VIV Mechanism on Bridge Structures VIV is one of major issues in longspan bridge vibration. Although the vibration is self-limited and does not create divergent type of motion, maximum amplitude, fatigue problem and serviceability requirement of the bridge associated with VIV are of major concern. In finding appropriate countermeasures to suppress VIV, generation mechanism should be firstly clarified. Generally, mechanisms of vortex generation can be described as follows. When wind passes a bluff body, unstable shear layers of opposite vorticity are produced, usually referred to as vortex shedding. The vortices create alternating periodic force on the body acting transversely to the flow direction with the frequency fvs. The frequency is known as shedding frequency; and its value depends on wind velocity (U), typical

dimension of bluff body (D) and the Strouhal number (St) (i.e. St = fvsD/U). The body oscillates under influence of the alternating force and the oscillation remains small, unless the shedding frequency becomes closer to the natural frequency of the body. In such a case, a phenomenon known as lockin phenomenon or shedding frequency synchronization will occur. The physics of VIV is very complex and finding a uniformly accepted model to define its complete behaviour remains a challenge. Nevertheless, a number of models have been proposed for engineering applications. The models can be classified based on formulation of fluid force applied on the structure in the lift direction as forced system models, fluidelastic system models and coupled system models.16 In the forced system model, the fluid force depends only on time and is independent of body motion (y), that is, F[t], such as FL(t) 1 rU2BC sin(2P St(u/D)t). In the = __ L 2 fluid-elastic system models, the fluid force does not depend only on time but also on y, denoted as F [y(t),t]. The dependence in y may include all time derivatives, integrals and even time delays. Example of fluid-elastic system models are the oscillator models described in Refs. [40,41], empirical linear and non-linear model.47 In the coupled system models, the fluid force F depends on another variable (q), which depends on body motion y, thus F [q(t), t] such as described in Ref. [48]. Interested readers can find summary of methodologies on VIV in the following works.16,18,49,50 VIV on bridges can be classified into at least three main categories51: oneshear-layer related vortices, two shear layer vortices (including the Karman vortex street and symmetrical vortices) and 3D vortices. The first type is the most commonly observed one on girder of long-span bridges. The 3D vortices type can be generated on the inclined cable of cable-stayed bridge

and on the tip of free-end pylons.52 In bridge design, the maximum response caused by VIV is of particular interest. For this purpose, many formulae based on wind-tunnel test data have been applied in the preliminary design. One example of VIV formulation used in practice in Japan28 is by assuming VIV as self-excited vibration, and using the simplified lift oscillator model, the aerodynamic force is expressed as: (9) The formulation is used to estimate the amplitude of VIV response (y) and the critical damping (x ) that is the amount of damping that should be provided to avoid VIV. The amount of damping can be estimated by x = r UBCL/4wnm. The amplitude of VIV response can be calculated by h = rUB2C/4wnmx, where C, expressed by D/B (ratio of deck height and width) is the self-excited lift force coefficient obtained from wind-tunnel experiment. In addition to wind-tunnel test, vortex response of bridge section can also be estimated by advanced computer simulation using computational fluid dynamics (CFD). Numerical modelling of fluidstructure interaction (FSI) in CFD is attractive for its flexibility in usage over the wind-tunnel tests. The methods could simulate flow and bridge section response under vortex excitation and the results look promising when compared to wind-tunnel tests although it may take sometime before it can completely replace the wind-tunnel test. Examples of application of numerical approaches using CFD to bridge engineering are the discrete vortex method (DVM),53 Reynolds average NavierStokes or RANS-based models54 and large eddy simulations (LES).55 However, CFD is yet to be proven fully effective for fully simulating the detailed small-scale flow features created by the deck components, especially for very streamlined or multiple box girders.
Scientific Paper 253

Structural Engineering International 3/2013

Reported Occurrences of VIV on Bridges and Their Countermeasures VIV mainly occurs on bridge girder and pylon. VIV on cables of cablesupported bridge is also common; but the amplitude is usually small and can be suppressed completely by adding a small amount of damping. VIV on bridge deck is more serious and usually affects the girder vertical or torsional mode. Modern long-span bridges usually use box-girder deck in the trapezoid form. This form has superior aerodynamic instability performance that is higher than critical wind speed for flutter, but may suffer from VIV at low wind speed. Amplitude of VIV depends on wind turbulence, where larger turbulence intensity creates smaller vibration amplitude, shape of girder and structural damping. The occurrences of VIV on the girders of existing bridges have been reported, among others, in the Storebaelt Suspension Bridge,56 Trans-Tokyo Bay Bridge57, Second Severn Crossing Bridge,58 Rio-Niteri Bridge59 and Osteroy Bridge.60 More recently, significant attention has been paid to excessive vertical oscillations resembling VIV that was observed on box girders of the newly built Volgograd Bridge in Russia.61 There has been also reported a case of VIV amplified by interference effect of two parallel decks in Jindo cable-stayed bridge [62] Apart from cable-supported bridges, VIV can also occur on the web members, diagonals and verticals of a longspan truss bridge. Such incident was observed on the vertical truss component of the Commodore Barry Bridge, the longest cantilever truss bridge in the United States.63
Bridge name Trans-Tokyo Bay (multi-girder steel box) [57] Great Belt Suspension bridge (main span) [56] Second Severn Crossing Cable-Stayed Bridge [58] Rio-Niteroi Bridge (multi-girder steel box) [59]

Generally, there are two countermeasures for VIV; one is aerodynamic countermeasure and the other mechanical countermeasure using vibration control such as tuned mass. In some cases, a third option comes into play by changing the structural characteristics such as natural frequency or adding intermediate support and that is also deemed appropriate. To suppress vibration caused by one-shear-layer vortex, one needs to control generation of leading edge vortex caused by body motion. This can be achieved by modification of leading edge of the girder using fairing, wind nose, deflectors, and flap plates. Figure 5 shows example of VIV countermeasures on the bridge girder. While adding aerodynamic appendages can be effective in mitigating VIV of an existing bridge, the added cost of installation and maintenance can be high. Therefore, bridge girder of an optimal shape such that it does not display VIV is preferable, for example by changing the angle of inclined panel in a trapezoid box girder.64 Controlling the pressure difference by gratings that partially open up the deck floor has also been used as a possible stabilising method against VIV of bridge girder. In addition, tuned-mass damper (TMD) can be installed in the box girder to mechanically control the vibration. Table 2 lists several bridges reported to have experienced girder VIV and their countermeasures. The VIV problem encountered on the Trans-Tokyo Bay Bridge and its solution are described here as an example. The Trans-Tokyo Bay Highway Crossing is a combined tunnel with

multiple bridges and sea-crossing highways, has a total length of 11 km, and was completed in 1997. The bridge section includes a ten-span continuous steel box girder with a total length of 1630 m. The two longest spans of this bridge are of 240 m. Significant vibration due to vortex shedding occurred under prevailing winds almost transverse to the bridge axis during the construction stage. The vortex-induced first-mode vibration peaked at a wind velocity around 16 to 17 m/s, with maximum amplitude exceeding 500 mm (Fig. 6).57 Extensive wind-tunnel testing using sectional models as well as 3D models had been conducted in the design stage and the VIV was indeed observed in the experiments. However, higher structural damping and larger turbulence intensity of the wind at the bridge site were also expected. These led to the decision not to install control devices from the beginning; instead, performance of the bridge under wind conditions during the erection was closely monitored. It turned out that the structural damping and the turbulence intensity of the wind were much lower than expected in the design stage and the VIV indeed took place. As countermeasures, TMDs were installed inside the box girder and aerodynamic flaps were added on the deck side before opening the bridge to traffic. These countermeasures significantly reduced the vibration. VIV is also commonly observed on long-span bridge pylons, especially during construction. During the freestanding construction period, the pylons have low natural frequency and their bluff body would normally cause
VIV countermeasures TMD Guide Vanes (main span), TMD (approach span) Baffles TMD TMD Fairings

Country Max span (m) Bridge girder type Japan Denmark United Kingdom 240 1624 456 300 240 Steel box Steel box Steel box Steel twin box

Detected VIV mode 0,34 Hz (first VB) 0,13; 0,209; 0,242 (third, fifth, sixth VB) 0,326 (first VB) 0,32 (first VB)

Brazil United Kessock Cable-Stayed Bridge [29] Kingdom Longs Creek (suspension bridge) [29] Canada Volvograd Bridge (multi-girder steel box) [61] Jindo Bridge (Twin cable-stayed) [62]

Orthotropic deck 0,5 (first VB) I-Section 0,6 (first VB) 0.45 Hz (first VB), 0.57 Hz (second VB) and 0.68 Hz(third VB) 0.438Hz (first VB)

Russia Korea

115 340

steel box Steel Box


Table 2: Long-span bridges reported to have experienced VIV on their girder during their service (VB: Vertical Bending) 254 Scientific Paper

Structural Engineering International 3/2013

Fig. 6: Observed first-mode VIV at the Trans-Tokyo Bay Bridge. Note the disappearance of a car in the red circle (afterRef. [57])

the out-of-plane VIV. Vibration could be excessive and obstruct the construction process, especially for pylons made of low-damped steel, hence vibration control is required. A well-known instance of VIV in tall bridge pylons during construction stage was in 1964, when the towers of the Forth Bridge in United Kingdom vibrated with a maximum amplitude of 1,1 m in 9 to 10 m/s wind.65 VIV was also observed during construction of Akashi-Kaikyo Bridge,66 the Kurushima-Kaikyo Bridges67 and the Storebaelt East Bridge.68 Although most of VIV on

bridge pylons are in cross-wind direction, some along-wind vibrations have also been observed, such as in the Storebaelt East Bridge68 during construction and on the Hakucho Bridge pylons69 after completion. A simple method to control vibration of bridge pylon is by connecting the top of free-standing pylons and heavy sliding blocks with cables.65 More comprehensive control strategies employ aerodynamic countermeasures such as cutting off the corners70 or making slits, and attaching aerodynamic appendages such as deflectors or side-

plates as shown in Fig. 7. Cutting the corner of bridge pylon reduces drag force and cross-wind vibration at low wind speed. This approach was implemented in the Higashi-Kobe Bridge71 and Akashi-Kaikyo Bridge pylon. Mechanical countermeasures such as TMD,72 tuned-liquid damper (TLD), tuned-sloshing damper (TSD) and tuned-liquid column damper (TLCD)73 are also commonly used in recent bridges in Japan. It should be noted that during construction, changes in pylon height lead to changes in natural frequency, hence tuning
Scientific Paper 255

Structural Engineering International 3/2013



Slit 30 a/D 0 20 10 Wind



Side plate


3/18 2h/D 4/18 0,50 0,25 5/18 20 10 6/18 0 30 U/fhD

Effects of section corner cut on pylon aerodynamic performance. (a:length of squared cut section, D:length of original section, fh:heaving natural frequency, U:wind velocity, h:heaving displacement) Critical velocity for pylon galloping increase for a/D=2/18~3/18

Fig. 7: Aerodynamic countermeasures for pylon vibration control using corner cut, slit, deflector and side plate (after Ref. [70, 28])






remain even after bridge completion, and therefore control countermeasure in the form of multiple TMDs is required in the bridge pylons. Figure 8 shows the locations of TMD and HMD. Buffeting Buffeting is a random bridge vibration associated with pressure fluctuations on the bridge due to natural wind turbulence. Natural wind turbulence has vertical and horizontal fluctuation in velocity, thus analysis must take into account the random variation in the angle of attack. Buffeting normally increases monotonically with an increase in wind mean velocity and does not generally lead to catastrophic failure, but is important for evaluation of bridge serviceability. The buffeting forces include lift force in vertical direction, drag force parallel to wind direction, and pitching moment. Concepts of bridge buffeting analysis had its origin from aircraft studies. In frequency domain, buffeting analysis was developed in Ref. 77, inspired by works on thin airfoils [78, 79] by assuming and employing the quasisteady assumption. In this approach, the spectrum of lift force is computed mode-by-mode, neglecting aerodynamic coupling and can be related to span-wise spectrum lift force via the joint acceptance function. Finally, the spectrum of buffeting response is calculated by multiplying the span-wise spectrum lift force with mechanical admittance function. On the basis of the spectrum of buffeting response,






Servo motors

Fig. 8: Vibration control measures at Akashi-Kaikyo Tower (TMD: tuned-mass damper during and after construction is completed, HMD: hybrid-mass damper during construction, after Refs. [28,66])

frequency should be adjusted accordingly. This means several mass dampers of different natural frequencies are needed. To overcome this inefficiency, actively controlled mass dampers have been developed and this technology has been applied.7476 One example of such application is the hybrid-mass damper (HMD) used during construction of the Akashi-Kaikyo Bridge pylons. HMD is a combination of pas256 Scientific Paper

sive TMD and active control actuator. The ability of this device to reduce structural responses relies mainly on the natural motion of the TMD, while the forces from the control actuator are employed to increase the efficiency of the HMD and to increase its robustness to changes in the dynamic characteristics of the structure. In a very long-span bridge, such as AkashiKaikyo Bridge, VIV is expected to

Structural Engineering International 3/2013

one can compute the peak response, which is an important parameter for bridge deck design. The spectra analysis is popular and widely used owing to its computational efficiency and simplicity of analysis when dealing with unsteady aerodynamic forces that are functions of frequency. In addition, it provides the necessary statistical values such as the maximum and mean values that are important for design. However, frequency domain approach has some shortcomings such as: disregarding of aerodynamic coupling effects,80,81 limited to static coefficients that vary linearly with respect to wind angle of attack, and inadequacy in the use of strip assumption to define the span-wise coherence of wind force.82 Recent works have extended buffeting response estimation by spectral method to include the problem of coupled and unsteady aerodynamic damping of multi-modes system.83 Research in time-domain buffeting analysis has been very active recently, and many methodologies have been proposed. They can be classified into four categories: quasi-steady aerodynamics,8486 indicial functions,8789 rational function approximation27,90 and unsteady aerodynamics.91 One critical issue in buffeting analysis has been the determination of aerodynamic admittances. In frequency domain, the admittance functions represent the transfer functions between turbulence component and the sectional forces. When the quasi-steady condition is assumed, the admittance functions become one. In time domain, the relationship between aerodynamic admittances and flutter derivatives associated with vertical motion of the deck are described in Ref. [92], which suggests that the latter should be used to estimate the former. More recently, several methodologies have been proposed to estimate the admittance functions using the flutter derivatives,93 rational functions91 and indicial function.94 Despite current developments, formulation of admittance functions in time-domain modelling and their full realization in wind-tunnel testing remain a challenging task that requires some engineering judgments and are still subject to further research. In long-span bridge design, buffeting must be taken into consideration as deformation under wind buffeting may cause structural fatigue and user discomfort and limit bridge serviceability to vehicles. Buffeting force during

vulnerable construction stages should also be carefully studied as shown in the case of Normandie Bridge.95 Buffeting analysis requires structural parameters, aerodynamic coefficients obtained from wind-tunnel experiments and wind characteristics such as turbulence intensity, wind spectra and wind spatial correlation. In Japanese bridge design code,28 buffeting is taken into consideration not directly as the dynamic response but as the increment of static wind load. However, when fatigue damage, or certain serviceability condition has to be taken into account, the buffeting maximum amplitude against allowable amplitudes should be confirmed. Results of structural monitoring system of long-span bridges have been used to verify buffeting analysis and to improve assumptions made in calculations.96

damping ratio (Fig. 9).98 Large cable oscillations may cause damage to the anchorage and cable fatigue. Cables of cable-supported bridges suffer from various types of vibration including those explained above and they are also well documented in literature (see for example Ref 99 for a review and recommendation by FHWA for stay cable100). In the following section, we shall describe special types of vibration observed only in the cables of long-span bridges (suspension and cable-stay), namely the wake interference vibration, rain- and wind-induced vibration, dry-inclined cable vibration and the indirectly excited cable vibration. Wake Interference Vibration When two circular cylinders placed in staggered arrangements in certain proximity, one windward, producing a wake, and the other leeward within the wake, there is a possibility that the forces in the wake shear flow lead to a coupled instability of the submerged cylinder. Another possibility is that the vortex shedding frequency caused by the windward cylinder coincides with frequency of leeward cylinder creating the wake-induced vortex vibration of the leeward structure. The first phenomenon can result in two types of wake-instability depending upon the mechanism, namely dampingcontrolled mechanism and stiffnesscontrolled mechanism. The former mechanism relates to the galloping condition, in which instability occurs when the leeward motion system has negative aerodynamic damping as described in Den Hartog criterion. The

Cable Vibrations and Control

Cables are essential components of long-span bridges and they are prone to vibration because of their higher flexibility, small mass and low mechanical damping. Main cable of suspension bridge, stay cables of cable-stayed bridge and hangers are generally very flexible compared to other bridge components, and therefore prone to vibrate.97 In long-span cable-stayed bridges, stay cable can be as long as 500 m, resulting in low natural frequencies of about 0,2 to 0,3 Hz in the lowest mode. A survey on inherent damping of stay cables of cable-stayed bridges in Japan shows that structural damping could reach as low as 0,1% critical

Modal damping (log. decrement)

1st Mode


2nd Mode 3rd Mode 4th Mode




Natural frequency (Hz)

Fig. 9: Relation between modal damping and natural frequency of stay cable (after Ref. [98]) Scientific Paper 257

Structural Engineering International 3/2013

damping-control mechanism is velocity dependent and results in single degree-of-freedom motion of the leeward structure and is known as wake galloping.101,102 The later mechanism, also known as wake-flutter, is associated with the coupled flutter condition where oscillation occurs in two degrees of freedom (i.e. vertical and transverse), thus resulting in elliptical orbit motion of the leeward structure.103,104 The wake-flutter condition could occur despite the positive aerodynamic damping condition. Note that some literatures refer to wake-flutter as wake galloping, and this usage is rather inaccurate as galloping and flutter mechanisms are totally different. Vibration of objects caused by interference due to certain arrangements and proximities are very complex in nature. Several mechanisms may take place and this has been studied extensively (see, for example, Ref. [105]). In the case of long-span cable-supported bridges, wake interference has been observed on the stay cables. It is observed in the range of 2 < y/d < 2 and 1 < x/d < 4, where x and y denote the longitudinal and lateral spacing between cables, respectively, and d is cable diameter.106 The oscillation starts at the critical reduced velocity of around 40, which typically corresponds to the wind speed of 5 to 20 m/s, and generally produces an elliptical trajectory with the maximum amplitude less than 3d. The motion is known to be sensitive to the Scruton number (Sc), and becomes hardly recognised at Sc > 50. Stay cables of Yobuko Bridge in Japan107 and the parallel suspender cables of Akashi-Kaikyo Bridge108 were reported to have experienced the wake-flutter. Rain- and-Wind-Induced Vibration Rain- and-wind-induced vibration (RWIV) is probably the most wellknown vibration problem of stay cables. It was first observed by Hikami on the stay cables of the Meiko-Nishi Bridge in Japan.109 The vibration occurs under certain wind velocity range but only during the rains. It occurs only on the inclined cables, and the vibration amplitude that is visible on the vertical plane can reach up to ten times the cable diameter. Vibration frequency involves not only the first mode, but also the higher modes. Characteristics of RWIV are very unique such that the classic cable vibration mechanisms such as VIV and wake interference have been disregarded as the source of
258 Scientific Paper

mechanisms. It generally occurs under reduced wind velocity above 20, and therefore cannot be considered as VIV. As the distance between cables are too large and the cross section of a cable with running water is almost identical to that of a dry cable, it cannot be considered as wake interference and section galloping respectively. Hence it is considered as a new type of cable vibration caused by rain and wind. RWIV has been observed in many cable-stayed bridges around the world, such as Tempozan and Aratsu Bridge110 in Japan, Erasmus Bridge in the Netherlands,111 Oresund Bridge in Denmark/Sweden,112 Yangpu Bridge in China and Fred Hartman Bridge in the United States113 to name a few. The vibration has practical consequences that it causes fatigue to the cable and also damage to cable anchorage. Control of this vibration becomes one of the main concerns in long-span cable-stayed bridges. Excitation mechanism of RWIV has been the subject of many studies and understanding the real mechanism is very challenging considering that the phenomenon involves not only wind and cable characteristics, but also flowing liquid on a bluff body. One among the first models that explains this phenomenon was proposed in Ref. [114]. In this model, vibration mechanisms are explained in two steps. First, upper water rivulet on the cable surface are formed as a result of a sensitive equilibrium between gravity, capillary and aerodynamic forces. The water rivulet effectively alters the geometrical cross section of the cable and hence the aerodynamic forces on it. Next, depending on the location and the size, the rivulet tends to give a negative slope of lift curve against the small change in the angle of attack and also significantly reduces the drag force. These combined effects result in the Den Hartog-type galloping instability. Similar approach was taken in a model in Ref. [115], which proposes estimation of critical wind speed and maximum amplitude of rainwind vibration. On the basis of results of numerous wind-tunnel experiments, Ref. 116 describes three mechanisms behind the RWIV, namely, conventional Karman vortex, galloping instability and high-speed vortex excitation. The first mechanism occurs at low wind reduced velocity and its amplitude is relatively small. In the second and third mechanisms, the instability is caused

not only by galloping as mentioned in Refs [109,114], but also by the influence of axial flow generated by the wake and axial vortex in inclined cable (i.e. 3D vortex type). Saito et al.117 conducted a series of wind-tunnel test and proposed a criterion to assign certain damping value to the stay cable so that amplitude of cable vibration can be reduced. Recently, two studies118,119 offered a quite different interpretation. In this, the flow specificity of a yawed cylinder is the main cause of the motion, and the presence of rainwater rivulets enhances the motion. Comparison between full-scale observations and dry-inclined galloping results indicates noticeable similarities between the two instabilities, particularly the effect of the critical Reynolds number regime and associated drag crisis on both instabilities. Despite considerable research activity over the last two decades, the underlying physical mechanisms of the phenomenon are yet to be fully understood. Some common understandings based on current research results can be summarised as follows: (1) Vibrations often occur in flow regimes in which Karman vortex shedding is suppressed. (2) RWIV is related to, but distinct from, dry-inclined galloping, RWIV often occurs in the critical Reynolds number range. However, the precise role of the drag crisis has not yet been fully established. (3) The presence of rivulets can act to increase the likelihood of a galloping-type oscillation, however, the exact role of the size, shape and location of the rivulets and the effect of surface characteristics remain to be fully understood. Dry-Inclined Cable Vibration Vibration with instability characteristics similar to RWIV has been reported on inclined stay cables in tunnel tests during dry (without rain) conditions.100,117,118120 It was shown that if the wind is in an oblique direction against the cable plane, the cable could have similar response characteristics as galloping instability. The instability phenomenon has been observed in several wind-tunnel investigations and it appears to share the characteristics, namely, occurring on dry-inclined cable at specific Reynolds number range and on the few specific geometrical cable positions. Based on wind-tunnel test, Ref. [117] described that unlike the RWIV, unstable condition can occur over a wide range of Scruton numbers, which means the

Structural Engineering International 3/2013

vibration might not be suppressed simply by adding more damping. The work defined the criteria for instability known as Saito instability line. Drygalloping vibration could pose a very serious problem if it takes place as predicted in wind-tunnel tests. However, unlike the RWIV, incidents related to this type of cable motion have rarely been observed in real bridges. Perhaps a clear indication of dry galloping on stay cable was reported only recently in Japan,121 where a 187 m length stay cable showed intense vibration with amplitude up to 1,5 m at the wind speed of 18 m/s. The large vibration damaged the viscous-damper at the cable-end and deck connection as well as part of edge faring installed at the bridge girder edge. Recently, a series of wind-tunnel tests were conducted by FHWA to investigate the phenomenon more closely.100 Using the results of the experiments, the instability criteria was redefined such that large-amplitude unstable conditions could occur only within limited range of Scruton number (i.e. <3). The common Scruton number of stay cable is usually larger than 3 unless damping is extremely low, such as less than 0,2%. Excitation mechanism of this instability has been centred on the role of the air flow along the cable axis and the critical Reynolds number range. According to Ref. [122] the axial flow has a significant role in dry-inclined cable vibration. Using artificial axial flow in the experiment, it was shown that the presence of axial flow could induce negative slope of the lift force, and galloping of dry-inclined cable would occur. Through a series of windtunnel tests Refs [123,124] showed

that the critical onset condition for the divergent motion can be predicted quite accurately by the Den Hartog criterion. Although, they also noted that effect of axial flow on the leeward side of the inclined cable in terms of intensity and frequency should be taken into account. Other authors have pointed out that instability clearly depends on the critical Reynolds number range.125 Further, another study126 showed that the main reason of inclined cable galloping is the difference in lift force due to the Reynolds number, that is the term CL/Re. The critical Reynolds number is sensitive to the presence of surface roughness, flow turbulence, motion of the cable and orientation of the body to the mean flow direction. Support-Excited Vibration of Stay Cables Cable vibrations are usually considered local, in that the anchorage points at girder and pylon are fixed. On the other hand, girderpylon vibrations are considered global, as the whole bridge span vibrates, while the cables do not vibrate locally but behave like massless elastic tendons. Vibration mechanisms in which the cables are directly excited by external force have been discussed so far. In addition to this, motion of the girder and pylons can also induce vibration of cables through the supports or anchorages. This type of excitation is known as support-excited excitation or indirect excitation. When the respective natural frequencies of local and global vibrations become nearly equal to each other, strong coupling of local and global vibrations may occur. In case of long-span cable-stayed bridge (Fig. 10), natural frequency of the stay

cables in local horizontal vibration (fy) can be almost equal to the natural frequency of the girder global horizontal vibration (fh), a condition that leads to a linear interaction known as parametric vibration. As the cable is much lighter compared to girder and pylons, its vibration can easily be of large amplitude such that the geometrical non-linearity of the cable cannot be ignored. This creates the non-linear interaction. The linear internal resonance occurs when the natural frequency ratio of the local cable motion to the girderor pylon-dominant global mode is 1:1. In the forced-excitation experiment of cable-stayed bridges, cable local vibration is often excited and this is mostly due to the linear internal resonance. Parametrically excited stay cables were reported in Ben-Ahin Bridge,127 Burlington Bridge and the Second Severn Crossing Bridge.128 In addition to that, the frequency of girder global horizontal vibration (fh), can be nearly half the natural frequency of the girder vertical vibration (fg), which creates a condition known as auto-parametric interaction. The condition where the ratio of fy:fh:fg = 1:1:2, is very likely in long-span, multistay-cable bridges, and it may occur simultaneously. A pedestrian bridge129 near Tokyo is reported to have satisfied the above ratio between stay cables and girders, and interaction between cable local vibration and bridge global vibration has been reported. Autoparametric vibration of cables has also been observed in Hitsuishijima Bridge (Fig. 11) and Tatara Bridge during the forced vibration testing [28]. Mechanisms and estimation of the maximum amplitude of vibration inter-

S Ca bl e
y (s




d Gir


Local cable horizontal mode

Global girder horizontal mode

Global girder vertical mode

Fig. 10: Local and global vibration modes of stay cable in a cable-stayed bridge (after Ref [130])

Structural Engineering International 3/2013

Scientific Paper



790 420


sag, bending stiffness of cable and the flexibility of damper supports on the optimal performance of the dampers. Vibration mitigation usually utilizes supplemental damping devices such as high-damping rubber,137 viscous dampers,134136 and linear138 and nonlinear dampers.139 In addition to that, recently the magnetorheological (MR) damper technique has been implemented to cable-stayed bridges in China, for example on the Dongting Bridge140 and Shandong Binzhou Bridge.141 Interested readers can find summary of technologies for vibration control of cables on long-span bridges in the following works (see Refs. [28,99,100,142]).

28 HB 1 P

25 HB 2 P

29 HB 3 P

28 HB 4 P

m/s2 0,7

0,7 m/s2 0,7

(a) Girder vibration : 2 Hz

Seismic-Induced Vibration of Bridges

(b) Tower vibration : 2 Hz mm



10 (c) Cable vibration : 2 Hz

Fig. 11: Parametric vibration observed on Hitsuishijima cable-stayed bridge during forced vibration test

Seismic-induced vibration is a forced vibration due to inertia force produced by the bridge mass and ground motion. Structures with long natural period such as long-span bridges have less acceleration during an earthquake, but experience large displacement. Less acceleration implies less inertia force on the superstructure but large displacement means certain countermeasure to prevent pounding between girder and approaching span may be required. In the design of superstructure of a long-span bridge, wind load is generally dominant; however, if large earthquake produces unacceptable inelastic response, then seismic load may become a significant factor. Design of sub-structures and pylons are generally governed by seismic load, while wind load governs design of the girder. From an economic point of view, it is necessary to identify acceptable levels of damage to a long-span bridge and to establish verification procedures for seismic performance against large earthquakes. Analysis of seismic loads on long-span bridges can be conducted in pseudodynamic or static approach and in fully dynamic approach depending on the complexity of the bridge. Pseudodynamic analysis is a simplified method and can be divided into equivalent-static seismic method, multimode spectral method and response spectrum method. Dynamic analysis is more realistic as it includes 3D structure, geometric and material non-linear effect, multiple-support excitation, soilstructure interaction and spatial variability of ground motion. For a

action between pylon, girder and cables in the parametric and auto-parametric excitations are explained in some of the studies.130,131,132 Cable vibration can be very large if damping is very low. However, the possibility of having such a vibration due to the gust response of the actual bridges during service may not be high because of the positive aerodynamic damping of the cables. Control of Cable Vibration Cable vibration can be controlled by aerodynamic, structural and mechanical countermeasures. For RWIV, the aerodynamic countermeasures aim at breaking up formation of the upper water rivulet on the cable surface. This can be achieved by whirling a helical wire on the cable surface, adding dimples to the cable surface, or using axially protuberated surface (Fig. 12). To suppress wake interference vibration, structural countermeasures by installing spacer and cable crossties between cables are commonly applied in an attempt to increase cable stiffness.
260 Scientific Paper

This, however, may not be effective for very long cables. Currently, mechanical countermeasure is the most widely used method. The main purpose of this approach is to increase cable damping by installing various types of dampers near the anchorage points or between cables. The relative motion between the cables and between cable and girder generate damping force.99,100,133 This method, however, has a limited performance and adequate damping force cannot be provided for extremely long stay cables. The damper has customarily been designed by neglecting several influencing factors, one of which is the cable flexural rigidity.133,134 It can be anticipated that the small flexural rigidity possessed by real cable systems affects performance of the damper as it is usually installed near the anchorage. As a result, additional damping given to the cable may not be as high as designed. Recent studies135,136 on the theoretical aspects have provided deeper understanding on the effects of

Structural Engineering International 3/2013

Cable cross-tie system


Helical stripes

ated with the inertial force in the case of fixed base conditions while the quasistatic (or pseudo-static) displacements are due to the different motions of the supports.143 Examples of the study on the multiple-support excitation and spatial variability of the ground motion of long-span bridges are given in Ref. [144] for suspension bridge and in Ref. [145] for cable-stayed bridge. Effect of multiple-support excitation can be also considered in design using response spectrum method by taking into account not only the cross-modal correlation but also the cross-support correlation.146 Type of connections between pylon and girder affect characteristics of seismic loads on long-span bridges. Generally, shock transmission units (STU) are used as the connection between the pylons and the deck. These devices are installed in the longitudinal direction to allow for expansion of the deck due to temperature changes. Under dynamic loads, these devices are extremely stiff and behave as rigid links. Various combinations of connections such as fixed, movable, elastically constrained and unsupported or rigidly connected with pylons are possible. Fixed or rigid type of connection will reduce the displacement of the deck but transfer larger inertia force of superstructure to the base of the tower during seismic excitation, thus increasing the base shear and moment force on the pylons. On the contrary, movable or floating type connection will result in smaller inertia force but requires special measure to prevent excessive deck displacement. The common approach is to compromise between displacements of the deck, shear forces and moments at the base of the pylons. Table 3 shows an example of how connection types affect forces on the pylon and girder displacements of Higashi-Kobe cable-stayed bridge.71 In this case, the all moveable connection type was selected in the final design and countermeasure to limit the effect of excessive girder displacement was provided. Alternatively, we can use energy dissipation devices such as active or semi-active control systems between the bridge deck and pylons. Passive energy dissipation devices can be designed effectively for a specified ground motion but they may be less effective for ground motions with different characteristics. Effectiveness of seismic control on long-span bridges such as cable-stayed bridge has been a subject of extensive research (see for example, Ref. [147]).
Scientific Paper 261


Indented cable surface

Axially protuberated cable surface

Fig. 12: Aerodynamic countermeasures for cable vibration control at (a) Yobuko Bridge (Cross-tie), (b) Akashi-Kaikyo Bridge (helical stripes), (c) Higashi-Kobe Bridge (axially protuberated surface) and (d) Tatara Bridge (indented surface)

Connection between pier/pylon Frequency Force at the pylon base Displacement and girder (Hz) 103 kN, kN m Girder (mm) Two connection fixed (MFFM) 0,7 M = 608; N = 90; S = 24 200 All connections fixed (FFFF) 0,79 M = 609; N = 85; S = 24 180 Elastic spring connections (MSSM) 0,33 M = 308; N = 88; S = 10 370 Moveable connections (MMMM) 0,12 M = 155; N = 90; S = 2 560 One connection fixed (MFMM) 0,45 M = 602; N = 97; S = 23 220
Load combination: dead load + live load during earthquake + earthquake load (M: moveable, F: fixed).

Table 3: Examples of how type of pier/pylon-to-girder connections affect the force and displacement (Higashi-Kobe Bridge, after Ref [71])

long-span bridge, multiple-support excitation and spatial variability of the ground motion become significant factors in the seismic response. Spatial variability of ground motions is possible considering that input forces from multiple supports separated by few hundred meters can be different depending on the local site condition. The contributing factors to spatial variability of ground motions are dynamic

properties of surrounding soils, wave passage effects such as propagation speed, distance between supports and effects of incoherence in the seismic wave. The common approach for including multiple-support excitation and spatial variability of the ground motion in the analysis of long-span bridges is by decomposing the total structural response into dynamic component and pseudo-static component. The dynamic displacements are associ-

Structural Engineering International 3/2013

Generally, designs of long-span cablestayed bridges in Japan employ more flexible type of connections that produce less seismic force while controlling displacement with special devices such as148: short tension links (Yokohama-Bay Bridge149), elastic constraint (Nishi Meiko Bridge), large springs (Hitsushi/Iwakuro Bridge) and long-vertical suspended cable (Higashi-Kobe Bridge). In the RionAntirion Bridge,150 a different strategy was adopted. In this bridge, wind and static loads are transferred to the pier in transverse direction using a stiff metallic fuse device. Parallel to this system, a number of viscous dampers are placed and designed to dissipate large amounts of energy after the fuses yield in the event of large earthquake. Although many long-span bridges are built in seismically active regions, only few are known to have been damaged by earthquakes. Among them are: the Higashi-Kobe cable-stayed bridge that suffered from end-link failure during the 1995 Hanshin Kobe Earthquake, Japan; eastern span of the San FranciscoOakland Bay Bridge, a cantilever truss bridge whose upper truss collapsed during the 1989 Loma Prieta earthquake; Ji Ji Da cablestayed bridge in Taiwan that suffered from damage on the main pylon and the main girder during the 1999 ChiChi earthquake; and Shipshaw cablestayed bridge in Canada during the 1993 Saguenay earthquake. Recently, many long-span cable-supported bridges in seismically active regions have been instrumented with vibration sensors to investigate structural responses during large earthquakes. Seismic responses of these bridges were analyzed during large earthquakes such as during the 1987 Whittier Narrows and the 1994 Northridge earthquake for Vincent Thomas Suspension Bridge151, the 1995 Hyogo-Nambu earthquake for Higashi-Kobe Bridge152 and the recent 2011 Great East Japan earthquake for Yokohama-Bay Bridge.153 They have enhanced our understanding in the nonlinearity aspects of long-span bridge response,151,152 soilstructure interaction152 and performance of structural connections during earthquakes.153,154

bridges load-bearing capacity. In addition to gravitational vehicle load, vehicle movement can introduce load amplification commonly known as dynamic amplification factor (DAF). The DAF depends on three main factors, namely, bridge structure (i.e. type of bridge, natural frequency, support and joint condition), vehicle characteristics (i.e. weight, speed and dynamic characteristics) and pavement conditions (i.e surface roughness and transverse path). DAF has been studied comprehensively and described in various bridge design codes, although consensus of its value and methods remain a topic of discussion. A comprehensive review of research and design code of the DAF is provided in Ref. [155]. Traffic-induced vibration may create local stress fluctuation on the bridge. Large stress fluctuation due to traffic may induce structural fatigue at certain structural components, especially steel bridges that may reduce the expected service life of the bridge. Fatigue of steel bridge components has been extensively investigated; Ref. [156], for example, presents a comprehensive list of studies on this subject. Vehiclebridge interaction is an important factor in analysis of traffic-induced bridge vibration. In the early studies on this topic, the interaction problem was simplified using moving-force and moving-mass model on the simple bridge model to obtain fundamental understanding.157,158 Nowadays, with the advancement of digital computer, more realistic vehicle and bridge models can be realised and the effect of interaction can be better understood.159 The effect of vehiclebridge interaction is especially important for railway bridges, where large kinetic energy carried by trains at high speeds may interact significantly with a bridge and even resonate with it under certain conditions. Vertical vibration caused by train speed will induce excessive vibration in the columns of the viaducts supporting the rails. The lateral vibration is of primary concern with regard to the safety of the passengers to prevent derailment. Resonance between a supporting bridge and high-speed train may occur under certain train speed and bogies spacing. And these conditions are generally found in the short to medium railway bridges having span length between 10 and 40 m. For long-span bridges, there are at least three important issues in the analysis

of trainbridge interaction, in addition to the problems mentioned above. They are train runability, safety and passenger comfort.160 Train runability concerns with the vertical slopes and horizontal planes of the railroadon which the train can operate safely. For the long-span bridges, train runability may be affected by bridge deformation not only because of gravitational load but strong wind and seismic action as well, and these conditions must be fully investigated during design. Safety issues concerning possible derailment during extreme events such as strong wind and earthquake and their effects on bridge and train have to be looked into carefully.

Human-Induced Vibration of Pedestrian Bridges

Human-induced vibration is a problem commonly observed on pedestrian bridge, although, recently some of vehicle bridges too are reported to have experienced human-induced vibration during events such as a marathon or a rally. The topic is included in this paper as it is considered as an important vibration problem that should be addressed while designing a long-span pedestrian bridge. There have been many reports of human-induced vibration in pedestrian bridges in the past, but the problem has attracted considerable attention from bridge engineers and researchers after the infamous incident of Millennium Bridge in London in year 2000. Several studies161163 have presented a comprehensive review on reported cases of human-induced vibration on pedestrian bridges and recent research and developments on the remedies. The problem has attracted particular interest from research communities in European countries as pedestrian bridges are very popular in this region. In general, human-induced vibration is a serviceability issue rather than a safety concern, and it becomes an issue mainly because human beings are very sensitive to vibration. However, if vibration occurs very frequently, fatigue may become a concern regarding service lifetime of the bridge. Human-induced vibration of pedestrian bridge may occur in vertical and lateral directions. There are two issues in humanbridge interaction concerning pedestrian bridge, one being the interaction between human motion and the bridge, and the other is

Traffic-Induced Vibration of Bridges

The main concern related to trafficinduced vibration is its effect on the
262 Scientific Paper

Structural Engineering International 3/2013

synchronization of movement among pedestrians as input force to the bridge vibration. Regarding the first issue, it is well known that human walking produces vertical periodic force that excites vertical motion of the girders. Frequency of this force is about 2 Hz (two steps per second), and design code in Japan for instance specifies that the natural frequency of the vertical modes of pedestrian bridge should be outside of the 1,7 to 2,5 Hz range to avoid the resonance vibration of the girder.164 In addition to vertical vibration, human walking also produces horizontal periodic force, because human beings use two legs in walking and its frequency is around 1 Hz. The lateral force of human walking can potentially excite lateral motion of the girder if the natural frequency of the lateral modes is close to 1 Hz. Synchronization of Pedestrian Movement Synchronization of movement among pedestrians can occur in vertical and lateral directions as they are passing the bridge. While synchronization in vertical direction is much less likely, synchronization in lateral direction can occur and becomes more pronounced with larger vibration. One of the first studies that investigate such vibration is the lateral vibration observed on T-Park Bridge in Japan during congested passage of pedestrians.129 Analysis of recorded bridge and human vibration motion through video of pedestrians head motions revealed that pedestrians who noticed the lateral sway of the bridge attempted to re-establish his/her balance by moving his/her body in the opposite direction. By doing so, he or she exerted load on the pavement and enhanced structural vibration. The frequency of human walking tends to synchronize and becomes lock-in to the frequency of bridge lateral vibration. This excites the bridge in a resonant manner, resulting in a large lateral girder vibration with amplitude of several centimeters. This phenomenon is mentioned in Ref. 165 as subconscious adjustment of individual pedestrian to any vibration of the pavement that leads to feedback and synchronization. Similar phenomenon was observed in Solferino Bridge in Paris and Millennium Bridge in London during its opening day in year 2000.166 In the case of T-Park Bridge,129 human walking is modelled as periodic force

Mechanism: Viewed from above



1,0 s

1 Hz 1 Hz

1 Hz

Fig. 13: T-Park Pedestrian Bridge near Tokyo experienced lateral synchronized humaninduced vibration and the two excitation mechanisms (after Ref. [129])

on the bridge, caused by N number of people with specific pedestrian lateral force (p), walking frequency (wp) and random phase (fi): (10) The bridge lateral vibration was thought to be caused mostly by direct linear resonance (Fig. 13) between lateral force of human walking and the girder lateral mode. The simple forced-excitation model of human walking can explain qualitatively the occurrence of lateral motion but cannot explain the large lateral vibration observed on the bridge. The model was unable to explain explicitly the growth process of lateral vibration, but synchronization process of human walking is assumed to increase the human force and to have caused the bridge resonance. In a subsequent research, a two-leg rigid rocking model was proposed in Ref. [167] to describe pedestrian walking. Lateral force is calculated by this model and interaction between pedestrian walking and girder lateral vibration is considered under assumption that phase lag exists between human motion and walking force. The results of the calculation agree well with experiments of human walking on a lateral shaking table. More recently, especially after the Millennium Bridge incident, vari-

ous models have been proposed to describe the pedestrian loading and synchronization mechanism. Some models assumed that synchronization is the cause of lateral vibration, such as in Refs. [168, 169]. The models use parameters whose values are fitted with experimental data. The case of Millennium Bridge lateral vibration169 was used to model the human walking force on the bridge as forced-excitation mechanism by separating the natural movement of the centre mass on a stationary pavement x(t) and movement of the pavement y(t), and examined the influence of time lag () between people reaction and the bridge movement, and some constants a and b. The human walking load force is modelled as: (11) Furthermore, the study investigates the effect of people-to-bridge mass ratio on the stability of bridge vibration and evaluates the amount of required damping to suppress the vibration. The pedestrian force and synchronization process are described using a model adopted from mathematical biology in Refs. [170, 171]. In this model, each pedestrian (i) exerts alternating sideways force F(t) defined as: , where G is the maximum force and the phase increase by 2 during a full sideways walking cycle. In turn, pedestrian walking pattern is
Scientific Paper 263

Structural Engineering International 3/2013

altered by the bridges movement according to: (12) Using experimental results of Millennium Bridge, the model can estimate the critical number of pedestrians (N) on which the large lateral motion and the pedestrians synchronized walking occur. Using the biomechanics field study, Ref. [172] modelled the lateral excitation of the bridge as the pedestrians balancing force. The model was able to provide reasonable representation of the lateral pedestrian force applied to the bridge in the absence of bridge motion, but could not match accurately the numerical results from full-scale measurement when the bridge motion was included. In addition to deterministic models, some researchers have proposed to model pedestrian loads using stochastic approach. The work in Ref. [173] for example, represents pedestrian load model through power spectrum density measured in the absence of lateral motion. Pedestrianbridge interaction is then considered through the velocity and acceleration proportional loads. A related work174 uses a general multiharmonic force model in a probabilistic framework that considers the intra- and inter-subject variability in the walking force, and the force model was successfully verified in a simulation case. Experimental investigations of pedestrian bridge vibration in a controlled (laboratory) condition and in a real (full-scale) bridge have been conducted in recent years to clarify the load-governing mechanism that is still somewhat disputed. The increasing interest in pedestrian bridge vibration is testified by the increasing number of publications on this topic and the establishment of an international conference in 2002 named Footbridge that was specifically designated for this issue. Countermeasures for Pedestrian Bridge Vibration To avoid excessive vibration, pedestrian bridge is usually designed following two principles: one by conducting dynamic analysis and confirming the response against serviceability requirement, and the other one by avoiding designing the bridge within the frequency range of human walking. The two principles have been adopted in
264 Scientific Paper

design guidelines of pedestrian bridges in several countries. For the lateral synchronization problem, while the governing mechanism that generates the load is still somewhat debatable and remains a subject of research, there is one thing that is commonly agreed upon, that is the amplitude of lateral vibration and the occurrence of input force synchronization depending on the number of pedestrians crossing the bridge. Therefore, limiting the number of people could be one way to avoid excessive lateral vibration. However, for longer span pedestrian bridge this may be impractical, hence suppressing vibration by increasing damping is a more reasonable option. Analytical models can estimate the amount of damping required to suppress the vibration.169 In the case of the cable-stayed pedestrian bridge near Tokyo, TLDs were used to control the lateral vibration, while in the Millennium Bridge TMDs and viscous dampers were used.175 Design codes for footbridge have taken the effect of lateral vibration and synchronization into account; the French design guideline for footbridges,176 for example, defines the specific acceleration criterion as the transition between random and synchronization. On the basis of the results of Millennium Bridge study, the FIB design guideline for footbridges177 specifies the necessary damping factor and critical number of people for prohibiting lateral synchronization. More recently, there has been a collective effort by European researchers178 to provide a more comprehensive guideline for design of pedestrian bridges against human-induced vibration. The guideline is expected to improve serviceability and mitigate the human-induced vibration problem of pedestrian bridges.

ing a suitable cure for the problem. Through decades of research, we now understand the mechanisms behind most of vibration problems commonly observed in long-span bridges. Vibration problems whose mechanisms are yet to be fully understood are still present which require further studies. Damping, for example, is a very important component in the vibration of long-span bridges, yet its quantification remains challenging. Structural damping is expected to be smaller for longer span bridges; therefore, mechanisms to increase damping will become more important for future design. Countermeasures against vibration problems have been developed on the basis of the present understanding of vibration mechanisms. Performance improvement for these engineering countermeasures is still a subject for further study, especially for bridge with longer spans and taller pylons. One particular aspect of the countermeasures is the use of control technology for long-span bridge vibration problems. While passive control strategies have been widely applied, active control strategies have yet to reach a mature state to be directly applied to solve long-span bridge vibration issues, which provides both challenges and opportunities for bridge researchers and engineers in the future. Recent advancements in computer technology have provided us with strong simulation tools to investigate vibration problems. CFD technique, for example, has been developed and used extensively as a supplement in recent years. With rapid development, CFD is expected to become a strong tool, although it may take some time to become a complete replacement for wind-tunnel testing because of the high complexity of the geometry of the bridge section and the 3D nature of the bridge. For seismic-induced vibration analysis, the use of computer simulation improves our understanding of the behaviour of bridge components under specific earthquakes, which is essential in seismic performance-based design. New developments and technologies in sensing and monitoring have been applied in modern long-span bridges. Such monitoring systems have provided data that is useful for improving our understanding of bridge vibration problems.179 Furthermore, monitoring data will be essential for validation of structural models and for providing

Conclusion and Future Outlook

Vibration mechanisms and controls are critical factors in the design and construction of long-span bridges. In this paper, we reviewed the vibration mechanisms commonly observed in long-span bridges, especially cable-supported ones, which were due to wind, seismic forces, and traffic and also their control methods. In addition, we also discussed the problem of humaninduced vibration of pedestrian bridges. Understanding vibration mechanisms is the most important step in find-

Structural Engineering International 3/2013

feedback on the performance of current vibration countermeasures. Newly observed vibration phenomenons69,154 are, in authors experience, the results of continuous monitoring of longspan bridges. One direction in bridge structural health monitoring is toward realising a low-cost, smart, continuous and reliable monitoring system using wireless sensor network. The objective is to combine bridge engineering knowledge with developments in sensor technology and network/information management in order to provide a solution that is a robust and significantly cost-effective alternative to traditional wired monitoring systems.180

[12] Den Hartog JP. Mechanical Vibrations, 4th edn. McGraw-Hill: New York, 1956. [13] Parkinson GV. Phenomena and Modeling of Flow-Induced Vibrations of Bluff Bodies. Prog. Aerospace Sci. 1989; 26: 169224. [14] Parkinson GV, Smith JD. The square prism as an aeroelastic non-linear oscillator. Quart. J. Mech. Appl. Math. 1964; 17: 225239. [15] Novak M. Aeroelastic galloping of prismatic bodies. J. Eng. Mech. ASCE 1969; 96: 115142. [16] Paidoussis MP, Price SJ, de Langre E. FluidStructure Interactions Cross-Flow-Induced Instabilities. Cambridge University Press, Cambridge, UK, 2011. [17] Blevins RD, Iwan WD. The galloping response of a two-degree-of-freedom system. J. Appl. Mech. 1974; 41: 11131118. [18] Blevins RD. Flow-induced Vibration, 2nd edn. Van Nostrand Reinhold: New York, 1990. [19] Nakamura Y, Mizota T. Torsional flutter of rectangular prisms. J. Eng. Mech. ASCE 1975; 101: 125142. [20] Nakamura Y. On the aerodynamic mechanism of torsional flutter of bluff structures. J. Sound Vib. 1979; 67: 163177. [21] Scanlan RH. The action of flexible bridges under wind, I: flutter theory. J. Sound Vib. 1978; 60(2): 187-199. [22] Selberg A. Oscillation and aerodynamic stability of suspension bridge. ACTA Polytech. Scand. C 1961; 13: 164. [23] Jain A, Jones NP, Scanlan RH. Coupled flutter and buffeting analysis of long-span bridges. J. Eng. Mech. ASCE 1996; 122(7): 716725. [24] Miyata T, Yamada H. Coupled flutter estimate of a suspension bridge J. Wind Eng. Ind. Aerodyn. 1990; 33(1): 341348. [25] Katsuchi H, Jones NP, Scanlan RH. Multimode coupled flutter and buffeting analysis of the Akashi-Kaikyo Bridge. J. Struc. Eng. ASCE 1999; 125: 6070. [26] Scanlan RH. Role of indicial functions in buffeting analysis of bridges. J. Struct. Eng. 1984; 110(7): 14331446. [27] Wilde K, Fujino Y, Masukawa J. Time domain modeling of bridge deck flutter. J. Struct. Mech, Earthq. Eng. Japan Soc. Civil Eng (JSCE) 1996; 13(2): 93104. [28] Fujino Y, Kimura K,Tanaka H. Wind Resistant Design of Bridges in Japan: Developments and Practices. Springer, Tokyo, Japan 2012. [29] Wardlaw RL. The improvement of aerodynamic performance. In Aerodynamics of Large Bridges, Proceedings of 1st International Symposium on Aerodynamics of Large Bridge, Copenhagen, Denmark, 1992. [30] Simiu E, Miyata T. Design of Buildings and Bridges for Wind. John Wiley and Sons, Hoboken, NJ, USA, 2006. [31] Mahrenholtz O, Bardowicks H. Aeroelastic problems at masts and chimneys. J. Wind Eng. Ind. Aerodyn. 1979; 4: 261272. [32] Pheinsusom P, Fujino Y, Ito M. Galloping of tower-like structure with two closely-spaced natural frequencies. J. Wind Eng. Ind. Aerodyn. 1989; 32(12): 189198.

[33] Gjelstrup H, Georgakis C, Larsen A. A preliminary investigation of the hanger vibrations on the Great Belt East Bridge. Proceedings of 7th International Symposium on Cable Dynamics, Vienna Austria, 2007. [34] Svensson B, Emanuelsson L, Svensson E. resund BridgeCable systemVibration incidents, mechanisms and alleviating measures. Proceedings of 4th International Cable Supported Bridge Operators Conference, Copenhagen, Denmark, 2004; 99108. [35] Lilien JL. Galloping of overhead electrical lines, mechanisms, wind tunnel experimentsfield measurements. Proceedings of the 2nd International Seminar on Cable Dynamics (ISCD), Tokyo, 1997; 3748. [36] Sato H, Kusuhara S, Ogi KI, Matsufuji H. Aerodynamic characteristics of super long-span bridges with slotted box girder. J. Wind Eng. Ind. Aerodyn. 2000; 88(2): 297306. [37] Matsumoto M, Nakajima N, Taniwaki Y, Shijo R. Grating Effect on Flutter Instability. J. Wind Eng. Ind. Aerodyn. 2001; 89: 14871497. [38] Sato H, Hirahara N, Fumoto K, Hirano S, Kusuhara S. Full aeroelastic model test of a super long-span bridge with slotted box girder. J. Wind Eng. Ind. Aerodyn. 2002; 90(12): 20232032. [39] Matsumoto M, Mizuno K, Okubo K, Ito Y, Matsumiya H. Flutter instability and recent development in stabilization of structures. J. Wind Eng. Ind. Aerodyn., 2007; 95: 888907. [40] Griffin OM and Ramberg SE, Vortex shedding from a cylinder vibrating in line with an incident flow. J. Fluid Mechs. 1976; 75: 257271. [41] Skop RA and Griffin OM, A model for the vortex-excited resonant response of bluff cylinders. J. Sound and Vib. 1973; 27: 225233. [42] Larsen A, Diana G, Poulin S, Zasso A. Aerodynamics of the Messina Bridge Deck Progetto Definitivo. Proceedings of IABSEIASS, London, UK, 2011. [43] Miyata T, Yamaguchi K. Aerodynamics of wind effects on the Akashi Kaikyo Bridge. J. Wind Eng. Ind. Aerodyn, 1993; 48(23): 287315. [44] Ge YJ, Xiang HF, Recent development of bridge aerodynamics in China. J. Wind Eng. Ind. Aerodyn. 2008; 96(67): 736768.. [45] Larsen A. Aerodynamic aspects of the final design of the 1624m suspension bridge across the Great Belt. J. Wind Eng. Ind. Aerodyn. 1993; 48: 261285. [46] Zhu LD, Xu YL, Xiang HF. Tsing Ma Bridge deck under skew winds. J. Wind Eng. Ind. Aerodyn. 2002; 90(7): 781837 [47] Simiu E, Scanlan RH. Wind Effects on Structures, 2nd edn. John Wiley & Sons, Hoboken, NJ, USA 1986. [48] Hartlen RT, Currie IG. Lift-oscillator model of vortex-induced vibration. J. Eng. Mech. ASCE 1970; 96: 577591. [49] Sarpkaya T. A critical review of the intrinsic nature of vortex-induced vibrations. J. Fluids Struct. 2004; 19(4): 389447. [50] Griffin OM. Vortex shedding from bluff bodies in a shear flowa review. ASME J. Fluids Eng. 1985; 107: 298306. [51] Matsumoto M. Vortex shedding of bluff bodies: a review. J. Fluids Struct. 1999; 13: 791811.

As a closing note, we would like to acknowledge many other insightful papers, apart from numerous works quoted here, that have helped lay the foundation for todays understanding of long-span bridge vibration.

[1] Feng M. Chinas major bridges. In IABSE Symposium Report, IABSE Workshop, Shanghai 2009; 124. [2] Koh HM, Choo JF. Recent major bridges in Korea. In IABSE Symposium Report, IABSE Workshop, Shanghai 2009; 88104. [3] Farquharson FB. Aerodynamics Stability of Suspension Bridges with Special Reference to the Tacoma Narrows Bridge. Parts I, III and IV. Bulletin No. 116, 19491954 University of Washington Engineering Experiment Station, Seattle,WA, USA. [4] Theodorsen T. General Theory of Aerodynamic Instability and the Mechanism of Flutter, NACA R496, National Aeronautics and Space Administration, 1934. [5] Bleich F. Dynamic instability of truss-stiffened suspension bridges under wind action. J. Eng. Mech. ASCE 1948; 74(8): 12691314. [6] Pugsley A. Some experimental work on model suspension bridges. Struct. Eng. 1949; 27 (8): 327348. [7] Ukeguchi M, Sakata H, and Nishitani H. An investigation of aeroelastic instability of suspension bridges. Proceedings of Symposium on Suspension Bridges, Lisbon, 1966. [8] Scanlan RH, Tomko JJ, Airfoil and bridge deck flutter derivatives. J. Eng. Mech. ASCE 1971; 97(6): 17171737. [9] Nakamura Y, Yoshimura T. Binary flutter of suspension bridge deck section. J. Eng. Mech. ASCE 1976; EM4: 685700. [10] Parkinson GV. Mathematical models of flow-induced vibrations of bluff bodies. In FlowInduced Structural Vibrations, Naudascher E (ed.). Springer-Verlag: Berlin, 1974; 81127. [11] Novak M. Galloping oscillations of prismatic structures. J. Eng. Mech. ASCE 1972; 98: 2746.

Structural Engineering International 3/2013

Scientific Paper


[52] Kitagawa T, Fujino Y, Kimura K. An experimental study on vortex-induced vibration of circular cylinder tower at a high wind speed. J. Wind Eng. Ind. Aerodyn. 1997; 6971: 731744. [53] Larsen A, Walther JH. Discrete vortex simulation of flow around five generic bridge deck sections. J. Wind Eng. Ind. Aerodyn. 1998; 7778: 591602. [54] Shimada K, Ishihara T. Predictability of unsteady two-dimensional model on the aerodynamic instabilities of some rectangular prisms. J. Fluid Struct. 2012; 28, 2039. [55] Sarwar MW, Ishihara T. Numerical study on suppression of vortex-induced vibrations of box girder bridge section by aerodynamic counter measures. J. Wind Eng. Ind. Aerodyn. 2010; 98(12): 701711. [56] Larsen A, Esdahl S, Andersen JE, Vejrum T. Storebaelt suspension bridge-vortex shedding excitation and mitigation by guide vanes J. Wind Eng. Ind. Aerodyn., 2000; 88, 283296. [57] Fujino Y, Yoshida Y. Wind-induced vibration and control of Trans-Tokyo Bay Crossing Bridge. J. Struct. Eng. ASCE 2002; 128: 10121025. [58] Macdonald JHG, Irwin PA, Fetcher MS. Vortex-induced vibrations of the second severn crossing cable-stayed bridgefull-scale and wind tunnel measurements. Proc. ICE Struct. Build. 2002; 152(2): 123134. [59] Battista R, Pfeil MS. Reduction of vortexinduced oscillations of RioNiteri bridge by dynamic control devices. J. Wind Eng. Ind. Aerodyn. 2000; 84(3): 273288. [60] Larsen A, Poulin S. Vortex-shedding excitation of box-girder bridges and mitigation. Struct. Eng. Int. 2005; 4: 258263. [61] Weber F, Ma slanka M. Frequency and damping adaptation of a TMD with controlled MR damper. Smart Mater. Struct. 2012; 21(5): 055011. [62] Seo JW, Kim HK, Park J, Kim KT and Kim GN, Interference effect on vortex-induced vibration in a parallel twin cable-stayed bridge. J. Wind Eng. Ind. Aerodyn. 2013; 116: 720. [63] Fox GF. Wind-induced vibration of truss members of the Commodore Barry Bridge. Proceedings of 12th US-Japan Cooperative Program in Natural Resources, 1980; 460484. [64] Larsen A, Wall A. Shaping of bridge box girders to avoid vortex shedding response. J. Wind Eng. Ind. Aerodyn. 2012; 104106: 159165. [65] Scruton C. An Introduction to Wind Effect on Structures. Oxford University Press: Oxford, 1981. [66] Hata K, Tatsumi M. Vibration control of the main towers of the Akashi Kaikyo Bridge. Proc IABSE Symposium Kobe; 1998, 5762.. [67] Suzuki S, Sasaki M, Yamaguchi K. Countermeasures to suppress vibration of the towers in the Kurushima Kaikyo Bridge in the erection stage (in Japanese). Honshi Technical Report BOEA 1996; 20(80): 2739. [68] Larose GL, Zasso A, Melelli S, Casanova D. Field measurements of the wind-induced response of a 254 m high free-standing bridge pylon. J. Wind Eng. Ind. Aerodyn. 1998; 7476: 891902.

[69] Siringoringo DM, Fujino Y. Observed along-wind vibration of a long-span suspension bridge tower. J. Wind Eng. Ind. Aerodyn. 2012; 103: 107121. [70] Shiraishi N, Matsumoto M, Shirato H and Ishizaki H. On Aerodynamic stability effects for bluff rectangular cylinders by their corner-cut. J. Wind Eng. Ind. Aerodyn. 1988; 28: 371380. [71] Yamada Y, Shiraishi N, Toki K, Earthquakeresistant and wind-resistant design of the Higashi-Kobe Bridge. Cable-Stayed Bridges Recent Developments and their Futures, Ito M, Fujino Y, Miyata T, Narita N. (eds). Elsevier, Amsterdam, The Netherlands, 1991; 397416. [72] Fujino Y. Countermeasures of vibration of civil engineering structures (in Japanese). J. Wind Eng. JAWE 1990; 44: 5369. [73] Fujino Y, Sun L, Pacheco BM, Chaiseri P. Tuned liquid damper (TLD) for suppressing horizontal motion of structures. J. Eng. Mech. ASCE 1992; 118(10): 20172030. [74] Tanida K. Progress in the application of active vibration control technologies to longspan bridges in Japan. Prog. Struct. Eng. Mater. 2002; 4(4): 363371. [75] Spencer BF, Nagarajaiah S Jr. State of the art of structural control. J. Struct. Eng. 2003; 129(7): 845885. [76] Preumont A, Seto K. Active Control of Structures. John Wiley & Sons, Hoboken, NJ, USA, 2008. [77] Davenport AG. Buffeting of a suspension bridge by storm winds. Proc. ASCE J. Struct. Div. ST3 1962; 88: 233268. [78] Sears WR. Operational methods in the theory of airfoils in non-uniform motion. J. Franklin Inst. 1940; 230(1): 95111. [79] Liepmann HW. Extension of the statistical approach to buffeting and gust response of wings of finite span. J. Aeronaut. Sci. 1955; 22(3): 197. [80] Scanlan RH. The action of flexible bridges under wind, I: buffeting theory. J. Sound Vib 1978; 60(2): 201211. [81] Jain A, Jones NP, Scanlan RH. Coupled flutter and buffeting analysis of long-span bridges. J. Struct. Eng. 1996; 122(7): 716725. [82] Larose GL, Tanaka H, Gimsing NJ, Dyrbye C. Direct measurements of buffeting wind forces on bridge decks. J. Wind Eng. Ind. Aerodyn. 1998; 74: 809818. [83] Katsuchi H, Jones NP, Scanlan RH, Akiyama H. Multi-mode flutter and buffeting analysis of the Akashi-Kaikyo Bridge. J. Wind Eng. Ind. Aerodyn. 1998; 7778: 431441. [84] Miyata T, Yamada H, Boonyapinyo V, Santos JC. Analytical investigation on the response of a very long suspension bridge under gusty wind. Proceedings of 9th International Conference on Wind Eng., New-Delhi, India, 1995; 10061017. [85] Kovacs I, Svensson HS, Jordet E. Analytical aerodynamic investigation of cable-stayed Helgeland Bridge. J. Struct. Eng. ASCE 1992; 118(1): 147168. [86] Diana G, Cheli F, Zasso A, Collina A, Brownjohn J. Suspension bridge parameter identification in full scale test. J. Wind Eng. and Ind. Aerodyn. 1992; 41(1): 165176.

[87] Scanlan RH, Beliveau JG, Budlong KS. Indicial aerodynamic functions for bridge decks. J. Engrg. Mech. Div. ASCE, 1974; 100(4): 657672. [88] Scanlan RH. Role of indicial functions in buffeting analysis of bridges. J. Struct. Eng. 1984; 110(7): 14331446. [89] Yoshimura T, Nakamura Y On the Indicial Aerodynamic Moment Responses of Bridge Deck Sections. Proc. 5th International Conference on Wind Eng., Fort Collins, CO, 1979. [90] Wilde K, Fujino Y. Aerodynamic control of bridge deck flutter by active surfaces. J. Eng. Mech. 1998; 124(7): 718727. [91] Chen X, Matsumoto M, Kareem A. Time domain flutter and buffeting response analysis of bridges. J. Eng. Mech. 2000; 126(1): 716. [92] Scanlan RH, Jones NP. A form of aerodynamic admittance for use in bridge aeroelastic analysis. J. Fluids Struct. 1999; 13(7): 10171027. [93] Hatanaka A, Tanaka H. New estimation method of aerodynamic admittance function. J. Wind Eng. Ind. Aerodyn. 2002; 90(12): 20732086. [94] Costa C. Aerodynamic admittance functions and buffeting forces for bridges via indicial functions. J. Fluids Struct. 2007; 23(3): 413428. [95] Livesey FM, Larose GL. The Pont de Normandie during construction, aeroelastic modeling of behavior. J. Wind Eng. Ind. Aerodyn. 1996; 65(1-3): 203215. [96] Miyata T, Yamada H, Katsuchi H, Kitagawa M. Full-scale measurement of AkashiKaikyo Bridge during typhoon. J. Wind Eng. Ind. Aerodyn. 2002; 90(12): 15171527. [97] Irvine M. Cable Structures. Dover Publication, Mineola, New York, USA, 1981. [98] Yamaguchi, H. and Fujino, Y. Stayed cable dynamics and its vibration control, Proceeding of International Symposium on Advances in Bridge Aerodynamics, Rotterdam, Netherlands, 1998; 235253. [99] Caetano E. Cable vibrations in cable-stayed bridges. Struct. Eng. Docum. 2007, 9. IABSEAIPC-IVBH Zurich, Switzerland. [100] Kumarasena S, Jones NP, Irwin P, Taylor P. Wind-induced vibration of stay cables. Report FHWA-HRT-05-083, US Federal Highway Administration 2007. bridge/pubs/05083/index.cfmS. [101] Paiudoussis MP, Price SJ. The mechanisms underlying flow-induced instabilities of cylinder arrays in crossflow. J. Fluid Mech. 1988; 187(1): 4559. [102] Granger S, Paidoussis MP. An improvement to the quasi-steady model with application to cross-flow-induced vibration of tube arrays. J. Fluid Mech. 1996; 320: 163184. [103] Price SJ. Wake induced flutter of power transmission conductors. J. Sound Vib. 1975; 38: 125147. [104] Simpson A. Wake induced flutter of circular cylinders-Mechanical aspects (Mechanical support system role in determination of aeroelastic stability of leeward cylinder immersed in wake using undamped flutter theory). Aeronaut. Quart 1971; 22: 101118. [105] Zdravkovich MM Flow Around Circular Cylinders, Applications. Vol.1-2. Oxford: Oxford University Press, 2003.


Scientific Paper

Structural Engineering International 3/2013

[106] Matsumoto M, Yokoyama K, Miyata T, Fujino Y, Yamaguchi H. Wind-induced Cable Vibration of Cable-Stayed Bridges in Japan. Proceedings of Canada-Japan Workshop on Bridge Aerodynamics, Ottawa, Canada, 1989. [107] Yoshimura T. Aerodynamic stability of four medium span bridges in kyushu district. J. Wind Eng. Ind. Aerodyn. 1992; 4142: 12031214. [108] Furuya M, Miyazaki M. Wind induced vibration of parallel hangers in Akashi Kaikyo Bridge and its aerodynamic remedy. Proceedings of 2nd Cable Dynamics Seminar, Norway, 1998. [109] Hikami Y, Shiraishi N. Rain-and windinduced vibration. J. Wind Eng. Ind. Aerodyn 1988; 29: 409418. [110] Yoshimura T, Savage M, Tanaka H, Wind-induced vibrations of bridge stay-cables. Proceedings of 1st International Seminar on Cable Dynamics, 1995, Liege, 437444. [111] Geurts C, Staalduinen P. Estimation of the effects of rainwind induced vibration in the design stage of inclined stay cables. Wind Engineering into the 21st Century, Larsen A, Larose GL, Livesey FM (eds). Balkema: Rotterdam, 1999; 885892. [112] Larsen A, Lafrenie`re A. Application of a limit cycle oscillator model to bridge cable galloping. Proc.6th International Seminar on Cable Dynamics (ISCD), Charleston, South Carolina, US, 2005. [113] Zuo D, Jones NP. Interpretation of field observations of wind-and rain-wind-induced stay cable vibrations. J. Wind Eng. Ind. Aerodyn. 2010; 98(2): 7387. [114] Yamaguchi H. Analytical study on growth mechanism of rain vibration of cables. J. Wind Eng. Ind. Aerodyn. 1990; 33: 7380. [115] Geurts C, Vrouwenvelder T, Staalduinen P, Reusink J. Numerical modelling of rain-windinduced vibration: Erasmus Bridge, Rotterdam. Struct. Eng. Int. 1998; 2: 129135. [116] Matsumoto M, Shiraishi H, Shirato H. Rain-wind induced vibration of cables of cablestayed bridges. J. Wind Eng. Ind. Aerodyn. 1992; 44: 20112022. [117] Saito T, Matsumoto M, Kitazawa M. Rain-wind excitation of cables on cable-stayed Higashi-Kobe Bridge and cable vibration control. Proceedings of International Conference on Cable-stayed and Suspension Bridges (AFPC), 2, 1995; 507510. [118] Macdonald JH, Larose GL. Two-degreeof-freedom inclined cable gallopingpart 1: general formulation and solution for perfectly tuned system. J. Wind Eng. Ind. Aerodyn. 2008; 96(3): 291307. [119] Macdonald JH, Larose GL. Two-degree-offreedom inclined cable gallopingpart 2: analysis and prevention for arbitrary frequency ratio. J. Wind Eng. Ind. Aerodyn. 2008; 96(3): 308326. [120] Matsumoto M, Yagi T, Hatsuda H, Shima T, Tanaka M, Naito H. Dry galloping characteristics and its mechanism of inclined/yawed cables. J. Wind Eng. Ind. Aerodyn. 2010; 98(6): 317327. [121] Matsumoto M, Yagi T, Liu Q, Oishi T, Adachi Y. Effects of axial flow and Karman vortex interference on dry-state galloping of inclined

stay-cables. Proceedings of 6th International Symposium Cable Dynamics, 2005; 247254. [122] Matsumoto M, Shiraishi N, Kitazawa M, Knisely C, Shirato H, Kim Y, Tsujii M, Aerodynamic behaviour of inclined circular cylinderscable aerodynamics. J. Wind Eng. Ind. Aerodyn. 1990; 33: 6372. [123] Cheng S, Larose GL, Savage MG, Tanaka H, Irwin PA. Experimental study on the windinduced vibration of a dry inclined cablePart I: phenomena. J. Wind Eng. Ind. Aerodyn. 2008; 96(12): 22312253. [124] Cheng S, Irwin PA, Tanaka H. Experimental study on the wind-induced vibration of a dry inclined cablepart ii: proposed mechanisms. J. Wind Eng. Ind. Aerodyn. 2008; 96(12): 22542272. [125] Larose GL, Zan SJ, The aerodynamic forces on stay cables of cable-stayed bridges in the critical Reynolds number range. Proceedings 4th International Symposium Cable Dynamics, Montreal, Canada, 2001; 7784. [126] Macdonald JHG, Larose GL. A unified approach to aerodynamic damping and drag/lift instabilities, and its application to dry inclined cable galloping. J. Fluid Struct. 2006; 22: 229252. [127] Cremer JM, Counasse C, de Ville de Goyet V, Lothaire A, Dumortier A. The stays, their dynamic behaviour, their equipments: Bridges at Ben-Ahin, Wandre and upon Alzette. Proceedings of 1st International Symposium on Cable Dynamic, Liege, 1995; 489496. [128] Virlogeux M. Cable vibrations in cablestayed bridges. In Bridge Aerodynamics. Larsen A, Esdahl S (eds), Balkema, 1998; 213233. [129] Fujino Y, Pacheco B, Nakamura S, Pennung W. Synchronization of human walking observed during lateral vibration of a congested pedestrian bridge. Earthq. Eng. Struct. Dyn. 1993; 22: 741758. [130] Fujino Y, Pennung W, Pacheco B. An experimental and analytical study of autoparametric resonance in a 3DOF model of a cablestayed-beam. Nonlinear Dyn. 1993; 4: 111138. [131] Lilien JL, Costa AP. Vibration amplitudes caused by parametric excitation of cable stayed structures. J. Sound Vib. 1994; 174(1): 6990. [132] Da Costa AP, Martins JAC, Branco F, Lilien JL. Oscillations of bridge stay cables induced by periodic motions of deck and/or towers. J. Eng. Mech. ASCE 1996; 122: 613622. [133] Krenk S, Hogsberg JR. Damping of cables by a transverse force. J. Eng. Mech. ASCE 2005; 131: 340348. [134] Pacheco B, Fujino Y, Sulekh A. Estimation curve for modal damping in stay cables with viscous damper. J. Struct. Eng. ASCE 1993; 119(6): 19611979. [135] Fujino Y, Hoang N. Design formulas for damping of a stay cable with a damper. J. Struct. Eng. ASCE 2008; 134(2): 269278. [136] Hoang N, Fujino Y. Analytical study on bending effects in a stay cable with a damper. J. Eng. Mech. ASCE 2007; 133(11): 12411246. [137] Nakamura A, Kasuga A, Arai H. The effects of mechanical dampers on stay cables with high-damping rubber. Constr. Build. Mater. 1998; 12: 115123.

[138] Main JA, Jones NP. Free-vibration of taut-cable with attached damper I : linear viscous damper. J. Eng. Mech. ASCE 2002; 128, 10621071. [139] Main JA, Jones NP. Free-vibration of taut-cable with attached damper II: non-linear damper. J. Eng. Mech. ASCE 2002; 128, 10721081. [140] Chen ZQ, Wang XY, Ko JM, Ni YQ, Spencer BF, Young G, Hu JH. MR damping system for mitigating wind-rain induced vibration on dongting lake cable-stayed bridge. Wind Struct. 2004; 7(5): 293304. [141] Hui LI, Liu M, LI J, Guan X, Ou J. Vibration control of stay cables of the Shandong Binzhou Yellow River Highway Bridge using MR fluid dampers. J. Bridge Eng. 2007; 401: 409. [142] Gimsing NJ, Georgakis Supported Bridges, 3rd edn. Hoboken, NJ, USA, 2011. CT. Cable John-Wiley,

[143] Abdel-Ghaffar AM. Cable-stayed bridges under seismic action. In Cable-Stayed Bridges Recent Developments and their Futures, Ito M, Fujino Y, Miyata T, Narita N. (ed.). Elsevier, 1991; 171192. [144] Harichandran RS, Hawwari A, Sweidan B. Response of long-span bridges to spatially varying ground motion. J. Struct. Eng. ASCE 1996; 122(5): 476484. [145] Nazmy AS, Abdel-Ghaffar AM. Effects of ground motion spatial variability on the response of cable-stayed bridges. Earthq. Eng. Struct. Dyn. 1992; 21(1): 120. [146] Der Kiureghian A, Neuenhofer A. Response spectrum method for multi-support seismic excitations. Earthq. Eng. Struct. Dyn. 1992; 21(8): 713740. [147] Dyke SJ, Turan G, Caicedo JM, Bergman LA, Hague S. Benchmark control problem for seismic response of cable-stayed bridges, 2000. [148] Ito M. Cable-supported steel bridges: design problems and solutions. J. Const. Steel Res. 1996; 39(1): 6984. [149] Maeda K, Otsuka A, Takano H. The design and construction of the Yokohama Bay Bridge, Cable-Stayed Bridges Recent Developments and their Futures, Ito M, Fujino Y, Miyata T, Narita N. (ed.). Elsevier, 1991; 377395. [150] Combault J. The Rion-Antirion bridge when a dream becomes reality. Proceedings of IABSE Workshop, Shanghai, 2009, 4762. [151] Smyth AW, Jin-Song P, Masri SF. System identification of the Vincent Thomas Suspension bridge using earthquake records. Earthq. Eng. Struct. Dyn. 2003; 32(3): 339367. [152] Ganev T, Yamazaki F, Ishizaki H, Kitazawa M. Response analysis of the HigashiKobe Bridge and surrounding soil in the 1995 Hyogoken-Nambu Earthquake. Earthq. Eng. Struct. Dyn. 1998; 27: 557576. [153] Siringoringo DM, Fujino Y and Namikawa K, Seismic Responses Analyses of the Yokohama-Bay Cable-Stayed Bridge in the 2011 Great East Japan (Tohoku) Earthquake. J. Bridge Eng. ASCE, (In press) DOI: 10.1061/ (ASCE)BE.1943-5592.0000508 [154] Siringoringo DM, Fujino Y. Observed dynamic performance of the Yokohama-Bay

Structural Engineering International 3/2013

Scientific Paper


Bridge from system identification using seismic records. Struct. Cont.Health Monit. 2006; 13: 226244. [155] Paultre P, Chaallal O, Proulx J. Bridge dynamics and dynamic amplification factorsa review of analytical and experimental findings. Can. J. Civ. Eng. 1992; 19: 260278. [156] Moses F, Schilling CG, Raju KS. Fatigue evaluation procedures for steel bridges. NHCRP Report 299, TRB, Washington, DC, 1987. [157] Fryba L. Vibration of Solids and Structures under Moving Loads. Noordhoff International Publishing: Gronigen, 1972. [158] Fryba L. Dynamics of Railway Bridges. Thomas Telford: London, 2001. [159] Yang YB,Yau JD, Wu YS. Vehicle-Bridge Interaction Dynamics with Applications to HighSpeed Railways. World Scientific Publishing: Singapore, 2004. [160] Diana G, Cheli F. Dynamic interaction of railway systems with large bridges. Vehicle Syst. Dyn. 1989; 18: 71106. [161] Zivanovic S, Pavic S, Reynolds P. Vibration serviceability of footbridges under humaninduced excitation: a literature review. J. Sound Vib. 2005; 279: 174. [162] Inglfsson ET, Georgakis CT, Jnsson J. Pedestrian-induced lateral vibrations of footbridges: a literature review. Eng. Struct. 2012; 45: 2152.

[163] Venuti F, Bruno L. Crowd-structure interaction in lively footbridges under synchronous lateral excitation: a literature review. Phys. Life Rev. 2009; 6: 176206. [164] Matsumoto Y, Nishioka T, Shiojiri H, Matsuzaki K. Dynamic design of footbridges. IABSE Proceedings, No. P-17/78 1978; 115. [165] Bachmann, H, Ammann, W. Vibrations in Structures induced by man and machines. IABSE Struct. Eng. Docum. 1987, 3e. IABSEAIPC-IVBH Zurich,Switzerland [166] Dallard P, Fitzpatrick T, Flint A, Low A, Smith RR, Willford M, Roche M. London millennium bridge: pedestrian-induced lateral vibration. J. Bridge Eng. ASCE 2001; 6: 412417. [167] Abe M, Fujino Y. A simulation on pedestrian-induced lateral vibration of a bridge using a two-leg human model (in Japanese). J. Struct. Eng. JSCE, 1992; 441(I-18): 199202. [168] Roberts TM. Lateral pedestrian excitation of footbridges. J. Bridge. Eng. ASCE 2005; 10(1): 107112. [169] Newland DE. Pedestrian excitation of bridges. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 2004; 218: 477492. [170] Strogatz SH, Abrams DM, McRobie A, Eckhardt B, Ott E. Crowd synchrony on the Millennium Bridge. Nature 2005; 438: 4344. [171] Eckhardt B, Ott E, Strogatz SH, Abrams DM, McRobie A. Modeling walker synchroni-

zation on the Millennium Bridge. Phys. Rev. E 2007; 75(2): 110. [172] Macdonald JHG. Lateral excitation of bridges by balancing pedestrians. Proc. R. Soc. A 2009; 465: 10551073. [173] Ingolfsson ET, Georgakis CT. A stochastic load model for pedestrian-induced lateral forces on footbridges. Eng. Struct. 2011; 33: 34543470. [174] Zivanovic S, Pavic S, Reynolds P. Probability-based prediction of multi-mode vibration response to walking excitation. Eng. Struct. 2007; 29(6): 942954. [175] Dallard P, Fitzpatrick T, Flint A, Le Bourva S, Low A, Smith RR, Willford M. The London Millennium Footbridge. Struct. Eng. 2001; 79(22): 1733. [176] Setra. Footbridge. Assessment of Vibrational Behavior of Footbridges under Pedestrian Loading, 2006. [177] Guidelines for the Design of Footbridges, FIB, 2005. [178] Human induced Vibrations of Steel Structures (HiVoSS), Design of Footbridges Guideline, 2010. [179] Ko JM, Ni YQ. Technology developments in structural health monitoring of large-scale bridges. Eng. Struct. 2005; 27(12): 17151725. [180] Spencer BFJ, Cho S, Sim SH. Wireless monitoring of civil infrastructure comes of age. Struct. Mag-Bridges. Oct. 2011; 1216.


Scientific Paper

Structural Engineering International 3/2013