Professional Documents
Culture Documents
Abstract: In recent years, increasing attention has been paid to the collapse failures of long-span continuous rigid-frame bridges under strong
earthquake excitations. This paper presents the results of a study in which a 1:15 scaled two-span prestressed concrete continuous rigid-frame
bridge model with box-type piers was tested using the shake-table array test system to investigate the seismic response characteristics. Two
nonlinear finite-element (FE) models were constructed. The first was a single-girder model that was used to simulate the seismic response
under weak seismic waves. The second was an explicit dynamic FE model that was used to simulate the collapse and failure mechanisms of
the scaled bridge under strong earthquakes. Testing revealed that the response of the central pier of the prestressed concrete continuous rigid-
frame bridge was the largest under seismic excitation, and the damage first appeared at the lower end of the central pier in all cases. The numer-
ical simulations revealed that traveling wave effects have a beneficial effect on the displacement at the top of all piers. The explicit dynamic
model was able to predict the failure modes and collapse process of the scaled bridge model. The plastic hinges emerging at the ends of the
piers were considered the main failure modes, and the collapse process changed with different seismic wave excitations. Such tests and analy-
ses can provide useful reference for the seismic-strengthening and anticollapse design of prestressed concrete continuous rigid-frame bridges
with a long span. DOI: 10.1061/(ASCE)BE.1943-5592.0000912. © 2016 American Society of Civil Engineers.
Author keywords: Prestressed concrete continuous rigid-frame bridge; Collapse failure modes; Shaking-table array testing; Strong earth-
quake excitations; Explicit dynamic FE model; Traveling wave effects.
Introduction potential collapse failure may also occur in this type of bridge
under ultra-specification-level earthquake excitations (Han et al.
In the past 30 years, the prestressed concrete continuous rigid- 2009; Fan 2013).
frame bridge has become one of the main selections for use in In recent years, the shaking-table array system has become one
crossing valleys and gulfs in modern highways and high-speed of the most favorable techniques to perform earthquake simulation
railways. Usually, excellent seismic performance can be obtained tests of whole-bridge structures under multiple-support excitations
with this type of bridge because of the strong deformation (Huang et al. 2008). For example, Johnson et al. (2008), Chen et al.
capacity of the tall piers used in the bridge design. At present, the (2008), Noguez and Saiidi (2012), and Saiidi et al. (2013a, b) have
behavior of the continuous rigid-frame bridge under the action of separately conducted experimental research on two-span and four-
small and medium earthquakes can be well extracted from testing, span continuous concrete bridges using the three-table array at the
numerical simulation, and real earthquake damage, but the col- University of Nevada at Reno. Zong et al. (2014) selected a 1:100
lapse failures under very strong earthquake actions are not as scaled model of a three-tower cable-stayed bridge to conduct shak-
clear. For example, damage of the piers in prestressed concrete ing-table testing under uniform and nonuniform excitations. Li et
continuous rigid-frame bridges was discovered in 2008 following al. (2015) presented the shake-table test of a 1:10 scaled typical
the Wenchuan Earthquake in P.R. China, which revealed that the curved bridge with concrete box girders under spatial ground
motions, including traveling wave effect, local site effect, and
ground motion multidimensionality. Huang et al. (2014) used the
1
Professor, School of Civil Engineering, Southeast Univ., No. 2, Ganhaizi Bridge as prototype. A 1:8 scaled bridge model with two
Sipailou Rd., Nanjing 210096, P.R. China (corresponding author). E-mail: spans and three lattice high piers was designed for multiple-
zongzh@seu.edu.cn
2 shaking-table tests at Fuzhou University, China. Li et al. (2014)
Associate Professor, School of Civil Engineering, Fuzhou Univ., No.
2, Xueyuan Rd., Fuzhou 350108, P.R. China. E-mail: xiatian@fzu.edu.cn tested a 1:40 scaled model of the Taizhou Changjiang Highway
3
Engineer, Henan Provincial Communications Planning & Design Bridge incorporating different seismic structural systems on the
Institute Co. Ltd, No. 70, Midlonghair Rd., Zhengzhou 450052, P.R. shake tables at Tongji University, China. In addition, the influences
China. E-mail: liuhaihong_02@163.com of variations in those viscous dampers and elastic cables on the
4
Doctoral Student, School of Civil Engineering, Southeast Univ., No. response of the model were also investigated. These kinds of seis-
2, Sipailou Rd., Nanjing 210096, P.R. China. E-mail: liyale@seu.edu.cn mic testing enhance the understanding of seismic behavior of
5
Engineer, Fujian Academy of Building Research, No. 52, Jintang Rd., whole-bridge structures. However, there are few experimental stud-
Fuzhou 350014, P.R. China. E-mail: hxy3730334@163.com
ies focused on the seismic behavior of the whole prestressed con-
Note. This manuscript was submitted on July 22, 2015; approved on
January 12, 2016; published online on March 9, 2016. Discussion period crete continuous rigid-frame bridge under multiple excitations.
open until August 9, 2016; separate discussions must be submitted for Meanwhile, collapse analyses of bridge structures have gradu-
individual papers. This paper is part of the Journal of Bridge ally gained more attentions. For example, Miyachi et al. (2012)
Engineering, © ASCE, ISSN 1084-0702. carried out progressive-collapse analysis on a three-span continuous
span simply-supported bridge and a three-span truss-arched bridge model was tested using the three-table array at Fuzhou
bridge. Deng et al. (2015) reviewed the main causes and mecha- University, China, and simulated until collapse failure using an
nisms of bridge collapse under different factors, including flood, explicit dynamic model with the aid of the LS-DYNA software pro-
earthquake, landslide, hurricane, typhoon, vehicle overloading, gram. The objective of the study was to investigate the seismic
and terrorist attack. The collapse analyses just described response until the progressive-collapse failure of a prestressed con-
focused mainly on the concise but comprehensive analysis of crete continuous rigid-frame bridge under multiple-support and
engineering examples under static loading; they seldom multidimensional excitations, and provide technical support for the
involved the analysis of collapse failure under seismic loading. anticollapse design of this type of bridge.
Zhou et al. (2014) presented a numerical simulation of the seis-
mic response until to collapse failure for a scaled multispan
cable-stayed bridge model under uniform excitations and non-
uniform excitations. Progressive-collapse analysis of pre-
stressed concrete continuous rigid-frame bridges under strong-
earthquake loading have seldom been investigated.
The current design standards (AASHTO 2007; Ministry of
Transport of the People's Republic of China 2008) cannot meet the
Similarity
Properties Physical quantities Dimension ratio
Fig. 2. Cross section of main girders (units are centimeters)
Geometry Length L 1/15
properties Linear displacement D 1/15
Angular displacement B 1
Material Modulus of elasticity E 1
properties Density P 1
Poisson ratio M 1
Strain « 1
Stress Eɛ 1
Equivalent quality density r 1.67
Dynamic Frequency v ¼ l1 ð r E1 Þ1=2 10
properties Damping ratio j 1/225
Horizontal acceleration E=ð r lÞ 9
Structural stiffness El 1/15
Structure weight r l3 1/2025 Fig. 3. Reinforcement setup of pier (units are centimeters)
Fig. 4. (Color) Construction and installation of the bridge model (images by Haihong Liu): (a) template of bridge model; (b) hoisting preparation; (c)
installation of the bridge model on the shaking table
Fig. 5. Accelerometer setup and numbering [Note: H1#, L1#, and V1# mean No. 1 accelerator in horizontal (x), longitudinal (y), and vertical (z); the
rest are similar]
Fig. 6. (Color) Details of the data-acquisition system (images by Haihong Liu): (a) accelerometers; (b) strain gauges; (c) laser displacement sensor;
(d) data-acquisition system
model should be simple in terms of its fabrication and construction. The bridge model dimensions are shown in Figs. 1–3. The total
(2) The similarity ratios of the model were determined according to length of the main girder was 12 m consisting of two spans, and the
the size and capacity of shaking tables. (3) The bridge model should height of the concrete box girder was 0.45 m at the pier top and 0.35
be as large as possible to eliminate the effects of irregularity among m at the midspan, and the bridge deck was 0.84 m wide, as shown in
its configuration and variations in the properties of the model mate- Fig. 2. The main girder had a 0.4-m linear gradient segment. The
rials. Based on these requirements, the model bridge was geometri- height of the pier was 3.2 m without considering the concrete foun-
cally scaled to near 1:15 according to the basic size and construction dation height of 0.3 m. A box-type cross section was used in the
of an actual bridge structure, and the capacity of the three shaking piers, the external size of the pier was 0.32 0.44 m, and the inter-
tables. The stiffness of the flexible foundations was not considered nal dimensions were 0.18 0.26 m, as illustrated in Fig. 3.
in the bridge model. The stiffness of the abutment and the backfill
soil also was not included because of limitations of the experimental
Material Properties
conditions. According to the dynamic similarity regulations (Wang
2006), the similarity ratios of the model with respect to the proto- C35 strength grade concrete was adopted in the bridge model for
type bridge were as shown in Table 1. both the main girder and the piers. The piers were reinforced with
Fig. 7. Input acceleration time history used for the shaking-table tests: (a) time-history curve of EC wave; (b) time-history curve of JJ wave
9 1.00 x/y 1.73 1.57 1.74 1.59 1.64 0.97 1.02 1.02
Order number FE Model 1 frequency (Hz) FE Model 2 frequency (Hz) Testing frequency (Hz) Characteristic of mode
1 8.003 8.235 8.115 First longitudinal floating vibration
2 14.937 15.040 15.012 First lateral bending vibration
3 21.533 21.123 21.235 First vertical bending vibration
4 30.836 28.913 29.803 Second vertical bending vibration
5 35.200 32.343 34.226 Second lateral bending vibration
Note: FE Model 1 is the single-girder model, and FE Model 2 is the explicit dynamic FE model.
Analysis of Test Results Compared with the peak acceleration responses at the upper
ends of piers subjected to uniaxial earthquake excitation, the peak
Before starting the testing, the five-order frequencies and modal acceleration responses in action of biaxial seismic excitation were
shapes were extracted from measured acceleration signals with the slightly larger, as shown in Fig. 9, as in the example of El-Centro
stochastic subspace identification method (Ren and Zong 2004), as earthquake excitation with a PGA of 1.0 m/s2. It can be concluded
shown in Table 6. Because of the high piers, longitudinal vibration that the biaxial earthquake excitation increased the peak accelera-
of the model bridge occurred first, followed by transversal vibra- tion response of the bridge model. When the PGA of the input was
tion. The vertical bending vibration was listed in the third order. over 2.0 m/s2, the corresponding relationship was stepped into non-
This conforms to the basic dynamic characteristics of a continuous linear, especially in the x-direction, as shown in Fig. 10.
rigid-frame bridge with high piers and long spans.
Acceleration Response of Main Girders
Acceleration Response of Piers
The acceleration responses of the main girder were characterized by
In each testing case, the form of the acceleration time-history the following features: (1) Under bidirectional horizontal seismic
response curve was basically the same at the piers, whereas the action, the longitudinal acceleration response showed a W-shape
peak acceleration responses at the upper ends of the piers were symmetry. The values at the top of the piers were greater than those
larger than those at the lower end. The testing results are shown in at the middle spans [Fig. 11(a)]. (2) Under bidirectional horizontal
Table 4 and Table 5. The output peak accelerations were larger than seismic action, the lateral acceleration response presented the
the PGAs of input seismic motions, which means that the PGAs of inverted V shape symmetrically; the value at the top of the central
input earthquake motions were amplified at the upper ends of all piers was the maximum, the values at the middle spans were second,
piers. The largest peak acceleration responses in the central pier and the values at the tops of the two side piers were the minimum
were 182% of peak acceleration of input El-Centro seismic excita- [Fig.11(b)]. (3) The vertical vibration of the main girder was more
tion and 174% of peak acceleration of input Jinjiang seismic excita- obvious under horizontal earthquake excitations. The values at the
tion in the x-direction, respectively, and 194% of peak acceleration middle spans were larger than those at the tops of the three piers, as
of input El-Centro seismic excitation and 189% of peak acceleration shown in Figs. 11(c) and (d), for example, when the 2.0-m/s2 PGA
of input seismic Jinjiang excitation in the y-direction, respectively. of the EC wave excited in both the x- and y-directions, the vertical
Fig. 9. Comparison of acceleration response under uniaxial and biaxial input excitation: (a) longitudinal accelerations; (b) transverse accelerations
Fig. 10. Relationship between output peak acceleration and PGA of input EC wave
peak acceleration happened in the middle of the span between Pier represent the longitudinal displacement, the transverse displace-
1 and Pier 2, and arrived at 0.84 m/s2. ment of the upper ends of the piers, and the transverse displacement
of the upper end of the central pier, respectively. To obtain the rela-
Pier Strain Response tive displacement of the main girder, the shaking-table displace-
ments were subtracted from the absolute displacement measured at
The basic feature of the strain response of the continuous rigid-frame the upper ends of the piers.
bride model can be obtained according to the strain time-history When the input peak accelerations were 0.5 and 1.0 m/s2, respec-
response. The peak strains at the piers derived from each case showed tively, the displacements in both the longitudinal and transverse
that both the concrete strain responses and the steel strain responses directions excited by uniaxial seismic waves were obviously
under bidirectional excitation were larger than the corresponding smaller than those excited by biaxial seismic waves. The EC wave
values under unidirectional excitation using the same seismic wave, was selected as an example and is shown in Fig. 15. Biaxial earth-
as shown in Fig. 12. The strains of the central pier were larger than quake excitation amplified the peak displacement responses of the
those of the side piers. bridge model.
The nonlinear correlation relationships among the output peak
strains and input peak accelerations were also extracted while the
PGA of the input seismic wave was increased from 0.05 to 0.2g, as Traveling Wave Effects Analysis
shown in Fig. 13. It can be seen that the model was in the elastic
state when the PGAs of input seismic excitation were smaller than Single-Girder Model
0.1g. When the PGAs were bigger than 0.1g, the structural response
displayed nonlinear features. The finite-element (FE) model of the scaled model was established
with the use of ANSYS (Shang et al. 2008). Spatial beam element
Beam188 was used to simulate the main girder with variable section
Displacement Response
and piers. The truss element (Link8) with only tension stiffness was
Fig. 14 presents the longitudinal and transverse displacement time- adopted to model the prestressed tendons. Supplement mass was
history curves at the upper ends of the piers under the input seismic simulated by Element Mass21. There were 395 nodes, 172
wave excitations with peak acceleration of 0.1g. D1, D2, and D3 Beam188 elements, 48 Link8 elements, and 113 Element Mass21
Fig. 11. (Color) Comparison of acceleration response of the main girder (0.1g): (a) longitudinal peak acceleration response; (b) lateral peak accelera-
tion response; (c) vertical peak acceleration response; (d) vertical acceleration time history at seven points
Fig. 13. Relationship between output strain and PGA of input EC wave in x- and y-directions: (a) reinforcement strain; (b) concrete strain
Fig. 14. (Color) Displacement time-history curve under 0.1-g acceleration peak (units are millimeters): (a) longitudinal displacement at top of Pier 1;
(b) transverse displacement at top of Pier 1; (c) transverse displacement at top of Pier 2
largely adopted to conduct seismic response analysis of large- seismic responses of three piers were approximately symmetric
span bridges under multiple-support excitation. In this study, the under horizontal earthquake excitations. The measured and calcu-
LMM was adopted to analyze the seismic response of a pre- lated displacement in test points D1, D2, and D3 were also compared
stressed concrete continuous rigid-frame bridge under strong and are shown in Fig. 16. It can be seen from the figure that the dis-
earthquake excitation. Compared with the RMM, the structural placement peak values from the shaking-table test agreed well with
response using LMM is the total response, which cannot be dis- those from the FE model.
tinguished into pseudo-dynamic and dynamic components. The acceleration and displacement response time-history curves
at the upper ends of the three piers are shown in Fig. 17 and Fig. 18,
respectively. It can be seen that the response tendencies of the
FE Model Verification
FE analysis were consistent with those from the shaking-table
Based on the single-girder FE model, the time-history analysis of tests. From these comparisons, it can be concluded that the single-
the scaled bridge under single-directional excitations and bidirec- girder FE model can be used to predict the seismic response of the
tional excitations was conducted by the Newmark- b method. The two-span rigid-frame model bridge under small and medium
Fig. 16. Comparison of testing and calculating peak displacement: (a) under EC wave excitation (0.05g); (b) under EC wave excitation (0.1g);
(c) under JJ wave excitation; (d) under EC wave excitation (0.2g)
Fig. 17. (Color) Comparison of calculating and testing acceleration response in Case 9 (JJ = 0.1g, x þ y): (a) Pier 1; (b) Pier 2; (c) Pier 1
analysis of rigid-frame bridged. The test results of the internal force In the explicit FE model, concrete was simulated by solid ele-
of the piers also support this conclusion. ment Solid164, which has eight nodes, with each node having 9
degrees of freedom. The truss element Link160 with only tensile
stiffness was adopted in modeling reinforcement, and pre-
stressed tendons were simulated by element Link167 with only
Collapse Failure Analysis tensile stiffness. Supplement mass was modeled by Element
Mass166. The explicit FE model of the scaled bridge model was
Explicit Dynamic FE Model established using 20,993 nodes and 18,017 elements, including
14,296 Solid164 elements, 3,072 Link160 elements, 136
There are significant characteristics of discontinuity and highly Link167 elements, and 513 Mass166 element,. The boundary
nonlinear and large deformation for the process of structural col- conditions were also given considerable attention because of
lapse. With the traditional implicit FE method, it is difficult to simu- their great influence on calculating the results, and were selected
late the collapse and failure mechanisms of a prestressed concrete to simulate the real fixed situation in the shake-table test. One
continuous rigid-frame bridge under strong earthquake excitation. vertical and three rotational displacements were fixed, and 2 lat-
Therefore, LS-DYNA, a prominent explicit dynamic program, was eral degrees of freedom were released in the input direction of
adopted to simulate the collapse and failure process of the scaled earthquake excitation. The explicit FE model of the scaled
bridge model under strong earthquake excitation. bridge model is shown in Fig. 20.
Fig. 18. (Color) Comparison of calculating and testing displacement response in Case 8 (EC = 0.1g, x þ y): (a) longitudinal displacement at top of
Pier 1; (b) transverse displacement at top of Pier 1; (c) transverse displacement at top of Pier 2
Fig. 21. (Color) Damage situation at key times under JJ wave in x-direction: (a) t = 7.3 s; (b) t = 11 s; (c) t = 13.9 s; (d) t = 14.9 s; (e) t = 16.5 s; (f) t =
19.2 s
conditions are presented in Fig. 22. The times at which plastic 14.9 s, as shown in Fig. 21(d). Finally, the whole scaled bridge col-
hinges occurred are listed in Table 7. lapsed at 16.5 s, with a mass of concrete spall and steel bars ruptured
The entire scaled bridge began to collapse when the peak accel- as shown in Fig. 21(e). It was noted that the main failure mode of
eration of the JJ wave was 4.5g. All components of the scaled bridge the scaled bridge subjected to longitudinal seismic waves included
were within linear elastic conditions before 7.3 s. At 7.3 s, the con- the plastic damage at the upper and lower ends of the three piers and
crete at the lower end of Pier 1, the upper and lower ends of Pier 2, large deformation in the longitudinal direction [Fig. 21(f)]. No
and the lower end of Pier 3 stepped into the nonlinear stage, as obvious failure was observed in the main girder.
shown in Fig. 21(a). The damage developed under the strong earth-
quake, and the plastic hinges inevitably appeared at both ends of Failure Mode under Bidirectional EC Wave Excitation
Pier 2. The concrete began to spall at the lower ends of Pier 1 and
Pier 3 at 11 s, as shown in Fig. 21(b). Further destruction of the The collapse process of the rigid-frame scaled bridge was simulated
scaled bridge led to the plastic hinge occurring at the lower ends of under a bidirectional EC wave with a peak acceleration of 2.7g.
all three piers at 13.9 s. The damage conditions are shown in During the calculation process, the damage occurred at the lower
Fig. 21(c). Subsequently, the plastic hinges appeared at the upper end of Pier 2, followed by damage at the bottom of Pier 1 and Pier
end of Pier 1, and longitudinal displacement increased obviously at 3. This type of damage phenomenon was consistent with shaking-
Fig. 22. (Color) Final collapse situation under EC wave in x þ y direction (PGA = 2.7g, t = 26.4 s)
Table 7. Time Points of Plastic Hinges in Piers Table 8. Continuous Rigid-Frame Bridge with Variable-Height Piers
piers were considered as the main failure modes, and the collapse girder and lattice pier.” Eng. Struct., 75, 73–86.
Johnson, N., Ranf, R. T., Saiidi, M. S., Sanders, D., and Eberhard, M.
process changed with different seismic wave excitations. For
(2008). “Seismic testing of a two-span reinforced concrete bridge.” J.
example, when inputting the bidirectional EC waves, plastic Bridge Eng., 10.1061/(ASCE)1084-0702(2008)13:2(173), 173–182.
hinges emerged at the lower ends of all three piers, and then large Leger, P., Ide, I. M., and Paultre, P. (1990). “Multiple support seismic anal-
transverse deformation took place, which led to the transverse ysis of large structures.” Comp. Struct., 36(6), 1153–1158.
collapse of the scaled bridge. Li, J. Z., Yan, J. K., Peng, T. B., and Han, L. (2014). “Shake table studies of
6. The damage and plastic hinges always first appeared at the bot- seismic structural systems of a Taizhou Changjiang Highway Bridge
tom of the central pier of the scaled bridge, in accordance with model.” J. Bridge Eng., 10.1061/(ASCE)BE.1943-5592.0000650, 04014065.
the results from the shaking-table tests and numerical simula- Li, X., Zhang, D. Y., Yan, W. M., Chen, Y. J., and Xie, W. C. (2015).
tion. Parameter analysis revealed that the failure mode “Shake-table test for a typical curved bridge: Wave passage and local
site effects.” J. Bridge Eng., 10.1061/(ASCE)BE.1943-5592.0000643,
remained the same, but the peak collapse acceleration
04014061.
decreased obviously with an increase in the length and diameter LS-DYNA 971 [Computer software]. Livermore Software Technology
ratios of the central pier. Corporation, Livermore, CA.
Ministry of Transport of the People's Republic of China. (2008).
“Guidelines for seismic design of highway bridge.” JTG/T B02-01-
Acknowledgments 2008, China Communication Press, Beijing (in Chinese).
Miyachi, K., Nakamura, S., and Manda, A. (2012). “Progressive collapse
The authors gratefully acknowledge the financial support provided analysis of steel truss bridges and evaluation of ductility.” J. Constr.
by the Natural Science Foundation of China (Grant No. 51178101 Steel Res., 78, 192–200.
and No. 51378112). This work was also supported by the Open MSC Marc [Computer software]. MSC Software, Santa Ana, CA.
Fund from the National Engineering Laboratory for Technology Noguez, C. A. C., and Saiidi, M. S. (2012). “Shake-table studies of a four-
of Geological Disaster Prevention in Land Transportation, span bridge model with advanced materials.” J. Struct. Eng., 10.1061
/(ASCE)ST.1943-541X.0000457, 183–192.
Southwest Jiao Tong University, P.R. China (No. SWJTU-GGS-
Ren, W. X., and Zong, Z. H. (2004), “Output-only modal parameter identifi-
2014001). The viewpoints of this paper represent only the authors' cation of civil engineering structures.” Struct. Eng. Mech.. 17(3–4),
opinions and do not represent the views of the funding 429–444.
committees. Saiidi, M. S., Vosooghi, A., Choi, H., and Somerville, P. (2013a). “Shake ta-
ble studies and analysis of a two-span RC bridge model subjected to a
fault rupture.” J. Bridge Eng., A4014003.
References Saiidi, M. S., Vosooghi, A., and Nelson, R. B. (2013b). “Shake-table studies
of a four-span reinforced concrete bridge.” J. Struct. Eng., 10.1061
AASHTO. (2007). Guide specifications for LRFD seismic bridge design, /(ASCE)ST.1943-541X.0000790, 139(8), 1352–1361.
Washington, DC. Salem, H. M., and Helmy, H. M. (2014). “Numerical investigation of col-
ANSYS 12.0 [Computer software]. ANSYS, Canonsburg, PA. lapse of the Minnesota I-35W bridge.” Eng. Struct., 59, 635–645.
Bi, K. M., Ren, W. X., Cheng, F., and Hao, H. (2015). “Domino-type pro- Shang, X. J., Su, J. Y., and Wang, H. F. (2008), ANSYS/LSDYNA dynamic
gressive collapse analysis of a multi-span simply-supported bridge: A analysis method and engineering practice, China Water Conservancy
case study.” Eng. Struct., 90, 172–182. and Hydropower Press, Beijing (in Chinese).
Chen, Y., Feng, M. Q., and Soyoz, S. (2008). “Large-scale shake table test Wang, T. (2006). Structural testing of civil engineering, Wuhan University
verification of bridge condition assessment methods.” J. Struct. Eng., 10 of Technology Press, Wuhan, China (in Chinese).
.1061/(ASCE)0733-9445(2008)134:7(1235), 1235–1245. Wilson, E. L. (1998). Three dimensional static and dynamic analysis of
Clough, R. W., and Penzien, J. (1993). Dynamics of structures, 2nd Ed., structures: A physical approach with emphasis on earthquake engineer-
McGraw-Hill, New York. ing, Computer and Structures, Inc., Berkley, CA.
Deng, L., Wang, W., and Yu, Y. (2015). “State-of-the-art review on the Xu, Z., Lu, X. Z., Guan, H., Lu, X., and Ren, A. Z. (2013). “Progressive-collapse
causes and mechanisms of bridge collapse.” J. Perform. Constr. Facil., simulation and critical region identification of a stone arch bridge.” J.
10.1061/(ASCE)CF.1943-5509.0000731, 04015005. Perform. Constr. Facil., 10.1061/(ASCE)CF.1943-5509.0000329, 43–52.
DEWESoft 6.6 [Computer software]. DEWESoft, Trbovljo, Slovenia. Zhou, R., Zong, Z. H., Huang, X. Y., and Xia Z. H. (2014). “Seismic
Dymond, B. Z., Roberts-Wollmann, C. L., Wright, W. J., Cousins, T. E., response study on a multi-span cable-stayed bridge scale model under
and Bapat, A. V. (2014). “Pedestrian bridge collapse and failure analysis multi-support excitations. Part II: Numerical analysis.” J. Zhejiang Univ.
in Giles County, Virginia.” J. Perform. Constr. Facil., 10.1061 SCIENCE A, 15(6), 405–418.
/(ASCE)CF.1943-5509.0000474, 04014006. Zong, Z. H., Zhou, R., Huang, X. Y., and Xia, Z. H. (2014). “Seismic
Fan, Z. H. (2013). “Research on collapse of rigid-frame bridge with super response study on a multi-span cable-stayed bridge under multi-support
high-rise piers under earthquake.” Master’s thesis, Wuhan Univ. of excitations. Part I: Shaking table tests.” J. Zhejiang Univ. SCIENCE A,
Science and Technology, Wuhan, China (in Chinese). 15(5), 351–363.