Efficient Longitudinal Seismic Fragility Assessmen

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/228346801
Efﬁcient Longitudinal Seismic Fragility Assessment of a Multispan Continuous

Steel Bridge on Liqueﬁable Soils
Article in Journal of Bridge Engineering · January 2010

DOI: 10.1061/(ASCE)BE.1943-5592.0000131
CITATIONS READS
37 434
7 authors, including:
Leonardo Dueñas-Osorio Jamie Ellen Padgett

Rice University Rice University
121 PUBLICATIONS 2,115 CITATIONS 193 PUBLICATIONS 2,581 CITATIONS
SEE PROFILE SEE PROFILE
Reginald Desroches
Rice University
234 PUBLICATIONS 6,023 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Urban water sustainability View project
1st International Workshop on Risk and Resilience of Industrial installations Against Natural Threats and Mitigation Strategies View project
All content following this page was uploaded by Jamie Ellen Padgett on 09 October 2015.
The user has requested enhancement of the downloaded file.

Efficient Longitudinal Seismic Fragility Assessment of a
Multispan Continuous Steel Bridge on Liquefiable Soils
Bayram Aygün, S.M.ASCE1; Leonardo Dueñas-Osorio, A.M.ASCE2; Jamie E. Padgett, A.M.ASCE3; and
Reginald DesRoches, M.ASCE4
Abstract: The increased failure potential of aging U.S. highway bridges and their susceptibility to damage during extreme events
necessitates the development of efficient reliability assessment tools to prioritize maintenance and rehabilitation interventions. Reliability
communication tools become even more important when considering complex phenomena such as soil liquefaction under seismic hazards.
Currently, two approaches are widely used for bridge reliability estimation under soil failure conditions via fragility curves: liquefaction
multipliers and full-scale two- or three-dimensional bridge-soil-foundation models. This paper offers a computationally economical yet
adequate approach that links nonlinear finite-element models of a three-dimensional bridge system with a two-dimensional soil domain
and a one-dimensional set of p-y springs into a coupled bridge-soil-foundation 共CBSF兲 system. A multispan continuous steel girder bridge
typical of the central and eastern United States along with heterogeneous liquefiable soil profiles is used within a statistical sampling
scheme to illustrate the effects of soil failure and uncertainty propagation on the fragility of CBSF system components. In general, the
fragility of rocker bearings, piles, embankment soil, and the probability of unseating increases with liquefaction, while that of commonly
monitored components, such as columns, depends on the type of soil overlying the liquefiable sands. This component response depen-
dence on soil failure supports the use of reliability assessment frameworks that are efficient for regional applications by relying on
simplified but accepted geotechnical methods to capture complex soil liquefaction effects.
DOI: 10.1061/共ASCE兲BE.1943-5592.0000131
CE Database subject headings: Bridges, steel; Soil liquefaction; Soil-structure interactions; Structural reliability; Uncertainty
principles; Simulation; Seismic effects.
Author keywords: Bridges; Liquefaction; Soil-structure interaction; Fragility; Reliability; Uncertainty; Simulation.
Introduction defined performance levels as a function of earthquake hazard

intensity and liquefaction potential. These fragility curves, ob-
The United States has approximately 601,000 bridges 关Federal tained from sampling nonlinear finite-element 共FE兲 models of
Highway Administration 共FHWA兲 2008兴, with a significant three-dimensional 共3D兲 bridge, two-dimensional 共2D兲 soil, and
portion of them requiring retrofit for protection against natural one-dimensional 共1D兲 lateral coupling p-y springs, enable effi-
hazards such as earthquakes. However, the costs to tackle these cient reliability assessment of bridge components including 1D
improvements range in the billions of U.S. dollars. Hence, it site amplification and liquefaction effects. These calculations
is critical to prioritize pre-earthquake risk mitigation and post- offer sufficient accuracy within reasonable computational de-
earthquake response strategies. Fragility curves can play an im- mands, while also opening the possibility of studying entire in-
portant role in infrastructure risk assessment and investment ventories of bridges and associated bridge networks.
decision making. This study develops new fragility relationships Heterogeneity in foundation soils results in highly variable soil
that capture the conditional probability of coupled bridge-soil- strength and stiffness parameters, which contribute significantly
foundation 共CBSF兲 system components to reach or exceed pre- to the uncertainty of soil-bridge system component performance
predictions 共Brandenberg et al. 2008兲. This highlights the need for
1 bridge reliability assessment models that explicitly account for
Graduate Research Assistant, Dept. of Civil and Environmental En-
gineering, Rice Univ., 6100 Main St., MS-318, Houston, TX 77005. foundations and underlying soil strata, especially in regions with
2
Assistant Professor, Dept. of Civil and Environmental Engineering, low frequency of seismic events. For this reason, the proposed
Rice Univ., 6100 Main St., MS-318, Houston, TX 77005 共corresponding 3D/1D framework is illustrated with a test bed CBSF system
author兲. E-mail: [email protected] considering soils from the coastal area of South Carolina, United
3
Assistant Professor, Dept. of Civil and Environmental Engineering, States. This area is susceptible to liquefaction-induced ground
Rice Univ., 6100 Main St., MS-318, Houston, TX 77005. failure, and the consequences of bridge network damage from
4
Professor and Associate Chair, School of Civil and Environmental seismic events could be high 共USGS 2002兲. Emphasis is placed
Engineering, Georgia Institute of Technology, 790 Atlantic Dr., Atlanta, on the relative effects of liquefaction on CBSF system compo-
GA 30332-0355.
nents, since bridge configuration, soil strata, and ground motions
Note. This manuscript was submitted on May 8, 2009; approved on
April 30, 2010; published online on May 7, 2010. Discussion period open
are typically not all available for the same South Carolina area,
until June 1, 2011; separate discussions must be submitted for individual although collectively are representative of the more general
papers. This paper is part of the Journal of Bridge Engineering, Vol. 16, central/eastern United States 共CEUS兲.
No. 1, January 1, 2011. ©ASCE, ISSN 1084-0702/2011/1-93–107/ Several researchers have studied seismic vulnerability of
$25.00. bridges using analytical and computational methods in the past.
JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 93
Downloaded 15 Dec 2010 to 168.7.217.223. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelibrary.org
However, the steady increase in computing power has changed Recently, Kwon and Elnashai 共2009兲 also developed fragility
the nature of soil-structure interaction 共SSI兲 analyses of bridges, curves with a sophisticated and typical CEUS bridge model in-
from practical and efficient 共Vlassis and Spyrakos 2001; My- cluding SSI but excluding potential liquefaction effects.
lonakis and Gazetas 2000; Ciampoli and Pinto 1995兲 to more To reach a compromise between acceptable accuracy and com-
sophisticated and computationally complex 共Elgamal et al. 2008; putational efficiency that affords the explicit modeling of lique-
Zhang et al. 2008b兲. Additional examples of prior research on faction effects in a probabilistic framework, this study creates a
bridge vulnerability analysis may be divided into three categories: nonlinear FE soil-foundation-bridge model that explicitly ac-
共1兲 analysis of models with emphasis on soil-foundation elements counts for SSI and liquefaction potential via 2D soil elements and
along with simplified structural components 共Brandenberg et al. 1D p-y springs, coupled with critical bridge and foundation com-
2007; Boulanger et al. 1999兲; 共2兲 analysis of models with empha- ponents in three dimensions. The components of the structural
sis on structural details but simplified soil-foundation elements or model comprise nonlinear steel bearings 共fixed and rocker兲, abut-
methods of analyses 共Zhang et al. 2008a; Nielson and DesRoches ments, deck impact elements, and fiber model representations of
2007a; Choi et al. 2004; Saadeghvaziri et al. 2000兲; and 共3兲 analy- concrete bents, columns, and buried piles. This CBSF model has
sis of computationally expensive models with detailed bridge- a run time that is almost 35 times more efficient than recently
soil-foundation elements and methods of analysis 共Kwon and published bridge-soil-foundation systems that include sophisti-
Elnashai 2009兲. cated soil elements suitable for nonlinear time-history analyses.
Boulanger et al. 共1999兲 studied seismic soil-pile-structure in- The efficiency is observed mainly because of the 2D soil element
teraction via dynamic p-y spring analyses, where they captured and 1D p-y spring approach, which still captures some complex
lateral pile-soil interaction forces at a given depth with 1D springs soil behavior and enables large-scale reliability studies.
that solely depend on the pile displacement at that depth and Instead of predefining a homogeneous or a repetitive identical
modeled the superstructure as a lumped mass. In contrast, Nielson soil profile for different bridge pile groups, this study also con-
共2005兲 performed a comprehensive fragility analysis using non- ducts a statistical analysis of bridge foundations and underlying
linear time-history FE analyses of several sophisticated bridge soils for South Carolina to capture the relative effects of spatial
models with multiple damage mechanisms typical of CEUS heterogeneity on CEUS bridges. Spatial heterogeneity refers to
bridges, but the bridge foundations and soil material were mod- the type of soil columns, which vary across bridge bents and
eled as surface springs. Similarly, Mackie and Stojadinovic abutments, and include different thicknesses of liquefiable layers
共2006兲 conducted fragility analyses using nonlinear time-history topped by different crust types of soft or stiff clays. This soil
FE tools on multispan California highway bridges where bridge heterogeneity and its potential effects on the CBSF system com-
foundations were either fixed or modeled as springs. ponent response due to distinct load transfer patterns highlight the
Zhang et al. 共2004兲 investigated the behavior of a detailed 2D limitations of using “liquefaction multipliers” to obtain structural
nine-span composite bridge structure model underlain by liquefi- response from nonliquefiable soil models, and the need for afford-
able soil in a probabilistic framework. Although they adequately able models that can handle reliability assessment studies. Spe-
modeled substructure and superstructure elements 共e.g., bearings, cifically, the proposed model accounts for uncertainty in structural
joints, piers, and piles兲, the computational time for one time- and soil parameters through fragility relationships where multiple
history analysis seems impractical for future applications on sen- levels of bridge functionality are considered. The application of
sitivity analyses or regional bridge network risk assessments, this method to single bridges and bridge networks will enable
mainly due to the thorough soil domain model. Other studies have regional reliability and risk assessments at an affordable compu-
also reported sophisticated, nonlinear, computational models of tational time, and perhaps aid with emerging needs for interde-
the soil-foundation-structure system for the Middle Channel pendence analyses that focus on coupling among utility systems,
Humboldt Bay Bridge to demonstrate the effects of soil liquefac- but have not incorporated critical bridge networks yet 共Dueñas-
tion on bridge components, but at a high computational demand Osorio et al. 2007; Adachi and Ellingwood 2008兲.
共Zhang et al. 2008b; Elgamal et al. 2008兲. This paper is divided into six sections. The next section intro-
To improve modeling efficiency and still capture complex soil duces the proposed 3D/1D CBSF system framework, and explains
response phenomena, Dueñas-Osorio and DesRoches 共2006兲 in- the modeling details of its components. The adequacy of the
vestigated multispan simply supported bridges typical of CEUS CBSF model is also verified through benchmark comparisons.
incorporating not continuum soil models but p-y soil springs Afterward, the paper provides insights into the dynamic behavior
where the soil stratigraphy was deterministically treated, although of CBSF models, before the concepts of fragility analysis applied
the detailed structural model was probabilistic. In addition, Shin to bridge systems are reviewed. Then the paper presents a
et al. 共2007兲 attempted to approximate the probabilistic bridge component-level probabilistic reliability assessment and analysis
response in liquefiable soils using liquefaction multipliers, which of results. The last section concludes the paper and provides ideas
is practical but may introduce inaccuracies. Zhang et al. 共2008a兲 for future research.
efficiently evaluated the fragility of different bridge classes typi-
cal of California by considering soil liquefaction effects through
an equivalent static analysis procedure. Also, Sextos and Taskari Modeling of CBSF Systems
共2008兲 explored four different modeling approaches with different
complexity, including the computation of dynamic impedances at To construct a CBSF system FE model with realistic soil profile
the piles and abutments, and an elaborate computational multi- data and structural details, the authors use empirical information
platform scheme to simulate soil-structure-foundation interaction, representative of South Carolina to characterize the soil and foun-
although only in a deterministic fashion. Kwon et al. 共2008兲 de- dation system, which maybe subject to seismic hazards and liq-
veloped liquefaction-sensitive fragility curves for the Meloland uefaction potential typical of low-frequency/high-consequence
Road Overcrossing bridge in California, considering only concen- events. This soil and foundation system along with a common
trated damage at the bottom of the central pier and neglecting class of CEUS and South Carolina bridges provides an idealized
other potential damage mechanisms associated with liquefaction. CBSF system for analyzing the relative effects of soil liquefaction
94 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011
CL
Left abutment First bent Second bent Right abutment
2.44 m 1.22 m
2.44 m 3.66 m
3.0 m
4.88 m
5.5 m
18.3 m 4.88 m 18.3 m
14.6 m
13.4 m
Stiff Clay Medium Sand Loose Sand Soft Clay Water Table
Fig. 1. Representative soil profile for South Carolina. The Cooper marl strata, labeled as stiff clay, serves as the bearing soil layer for the majority
of deep foundations in the region. Note that the drawing is not to scale to observe soil strata details, and the bridge contour is a simplified version
of the actual bridge in Fig. 2.
on bridge components. Fifty blueprints of existing bridge soil and are especially vulnerable due to potential high demands on their
foundation systems in South Carolina counties, including columns, bearings and abutments due to the larger inertial deck
Dorchester, Berkeley, Charleston, and Orangeburg, are analyzed loads 共Nielson 2005兲. This bridge type accounts for a representa-
to characterize regional soil profiles and foundation types. Five tive 13.2% of the CEUS bridges and 8.27% of the South Carolina
idealized soil profiles and three foundation types emerge as rep- bridge inventory.
resentative of the available data set 共Aygun 2009兲. The chosen
soil profile for this study is shown in Fig. 1, which is typical of
Bridge Modeling
low lands and exhibits the simplest stratigraphy. Regarding foun-
dations, 12.2 m long concrete piles with 0.46 m ⫻ 0.46 m square Fig. 2 shows the MSCS girder bridge typical of the CEUS/South
section at the abutments and bents constitute one of the most Carolina regions chosen for this study. The steel girders are con-
common substructure geometries and dimensions, which are tinuous over the interior bents and the bridge spans have all the
adopted in this study. same length of 30.3 m. The abutments are seat type with cast-in-
Following the findings from Nielson 共2005兲 and Nilsson place concrete piles, and the abutment geometry is modeled in
共2008兲 on bridge inventory analyses, multispan continuous steel this study with rigid FE components 共Fig. 3兲. The bridge and pile
共MSCS兲 bridges are selected for this study. Continuous bridges foundation model is created in OpenSees 共McKenna and Fenves
17.8 m
30.3 m 30.3 m 30.3 m
Rocker 1.1 m Girders @ 1.83 m o.c

Fixed
Bearing 15.01 m
Bearings
Deck
520 mm
8 vertical piles
Rocker Fixed 0 .46 m

A
Bearing Bearing
0 .76 m
A 12 - #29 bars
1.6 m 10 67 mm
15 - #29 bars
4.6 m
5.0 m 5.0 m 0 .76 m

1219 mm
914 mm
0 .9 m 4 - #16 bars 0 .46 m

B B
2.4 m 0 .46 m 0 .76 m 0 .76 m 0 .46 m

1.1 m
C C #16 bars @ 305 mm o.c #13 bars @ 305 mm o.c

Footing
Bent beam Column C- C
A- A B- B
Fig. 2. Representative MSCS bridge for CEUS and South Carolina 共modified from Nielson and DesRoches 2007a,b兲
0.51 m 0.51 m independent multiyield” material and are also embedded into
FSPM’s. These FSPM elements impose incompressibility condi-
Embankment soil
tions to allow the generation of pore pressures. In this work,
1.52 m
1
spring in: sandy and clayey materials are implemented in a centerline 2D
active pasive
action G action mesh, which is key for the efficiency and sufficient accuracy of
Pounding the proposed 3D/1D CBSF model.
2 element
0.64 m 3 2
The soil mesh consists of 2D soil columns whose dimensions
Bearing
0.91 m G 3 spring are 23.16 m in depth and 25.01 m in width. The individual Op-
4 1
4 Rigid element enSees “quad” elements are 25.01 m wide, 0.61 m deep, and are
G Centroid of the positioned directly adjacent to the nonlinear beam-column ele-
ments representative of bridge piles 共Fig. 4兲. Although the aspect
abutment section
0.91 m 4 ratio of these soil elements exceeds standards for FE modeling,

they can be used as long as their adoption does not distort seismic
demand levels on the bridge/foundation structural components
0.16 m
1.83 m and are compatible to the loading conditions. For instance, the
0.61 m depth of the quad elements is controlled by the minimum
Fig. 3. Bridge abutment model in OpenSees: seat-type abutment with shear velocity in the soil profile 共Vs = 100 m / s兲 and the maximum
contact elements and nonlinear springs representing gaps, bearings frequency level of the seismic excitation signals 共f = 15 Hz兲 to
and embankment soil resistance ensure that the shear waves propagate accurately 共Zhang et al.
2008b兲. The width of the soil elements is controlled by the need
to generate inertial forces that when propagated through the pile
2001兲 making use of fiber sections with distributed plasticity for and superstructure cause nonlinear levels of structural demand
the columns and piles, and linear elastic beam-column elements forces and deformations that are expected on similar structures
for the decks. The median design strength of concrete and the under seismic excitation. For instance, for a ground motion with
yielding strength of reinforcing steel are 33.5 and 466 MPa, re- maximum peak ground acceleration 共PGA兲 of 0.65g and low-
spectively. The steel girders are supported on high-type steel bear- frequency content that amplifies force transmissibility, the choice
ings which transfer the loads to the columns using a single row of of 25.01-m width under nonliquefiable conditions triggers nonlin-
fixed bearings over each of the bents and a row of rocker bearings ear CBSF moments and curvatures in the columns on the order of
at each of the abutments. The square footing dimensions are 2.4 1 , 600– 2 , 000· kN· m, and 0.01–0.05 1/m, respectively, and forces
m on the side with a thickness of 1.2 m. Nonlinear bearings, and deformations in expansion bearings on the order of 30–50
abutment soil, and pounding elements are modeled as in Nielson kN, and 100–150 mm, respectively. Comparable levels of struc-
共2005兲. The skeleton model for the bridge superstructure and sub- tural demands are observed on components of similar bridges that
structure is shown in Fig. 4. are modeled with linear surface springs to capture soil and foun-
dation systems under seismic loading 共Choi 2002; Nielson 2005;
Padgett 2007兲. When considering liquefaction potential, these lev-
Soil Modeling els of structural demands are also observed at the onset of lique-
OpenSees is also used to model liquefiable soil materials and their faction for soil element widths of approximately 25.01
nonlinear behavior. Liquefiable sandy soils are modeled with a m—smaller widths yield demands that are too small. Adequate
“pressure dependent multiyield” material created by Yang et al. match of component structural demands is also observed for
共2003兲, which is embedded into a “fluid solid porous material” ground motions with low PGA and moderate to high transmissi-
共FSPM兲 to simulate undrained and saturated soil conditions. Non- bility; although in these cases the liquefiable and nonliquefiable
liquefiable saturated clayey soils are modeled with the “pressure CBSF systems show slightly higher moments, curvatures, forces,
and deformations than the surface springs model. This is because
all bridges respond almost linearly, but the structure with soil and
1st bent columns
2nd bent columns liquefaction potential becomes more flexible.
Bridge deck The established pure shear soil model dimensions simulate the
free-field soil response, where the width of the soil elements af-
fects inertial forces, but it does not affect the soil effective vertical
Right abutment
stresses and cyclic shear stresses that relate to pore pressures
Left abutment because these are computed as a 1D site response. Pore pressures
Right abutment
soil column
influence the pile-soil interaction near the piles and their effect is
captured by nonlinear dynamic 1D p-y springs. This essentially
2nd bent soil column 1D site response approach prevents the additional computational
cost of refined soil meshes. Additionally, the out-of-plane thick-
Transverse rigid
beams for 2D soil/ 1st bent soil column ness of the soil elements is taken for convenience as 15 m, equal
3D bridge coupling to the width of the bridge, although the thickness does not affect
the 2D soil elements that attach to p-y springs, which are con-
Left abutment
ceived to remain in a state of plain strain. Other studies that use
soil column actual 2D meshes to account for soil-structure interaction 共Zhang
et al. 2008b兲, or 1D site response analyses with p-y springs 共Bow-
Fig. 4. FE model for the CBSF system: coupled 3D bridge and 2D ers 2007兲, also use the bridge width as representative soil element
soil model with 1D p-y springs 共not visible between piles and soil thickness. The chosen set of soil dimensions also yield a close
elements兲 for 1D site response and liquefaction simulation match between the two dominant longitudinal modal periods of
the developed CBSF model 共T3 = 0.46 s and T5 = 0.40 s兲, and the Near-f ield element Far-f ield element
dominant longitudinal mode of the MSCS bridge with surface Drag
springs calibrated for CEUS by Nielson 共2005兲共T1 = 0.44 s兲. All Plastic Elastic
other input beyond geometry, mainly soil mechanical parameters, Pile
is adopted following the recommendations of Yang et al. 共2004,
2008兲 for typical soil conditions. Closure
1.0
Soil-Pile Interaction Modeling
P/Pult
The coupled SSI approach to model liquefaction effects in this
work relies on the dynamic p-y method, which links the 2D soil 0.0
mesh response to 1D p-y nonlinear springs to capture pile-soil Drag Plastic
complex interactions synchronously. This approach offers im- spring spring
provements over the equivalent foundation springs and dampers -1.0
-0.1 0.0 0.1 -0.1 0.0 0.1
method, where the input ground excitation applied at the base of yg yp
foundation springs is generally a free-field motion unable to cap- 1.0
ture the extent of SSI effects. The fundamental assumption of the
dynamic p-y method is that the pile-soil interaction force at a
given depth only depends on the pile displacement at that depth
P/Pult
0.0
共Wang et al. 1998兲. Dynamic p-y methods model the far-field soil
response with a multilayered soil shear column, and the near field, Closure Combined
where strong soil-pile interaction occurs, with nonlinear p-y spring spring
-1.0
springs and gap elements. These simplified p-y geotechnical tech- -0.1 0.0 0.1 -0.1 0.0 0.1
niques are ideal for adoption by structural and infrastructure en- yg y
gineers, since they capture essential features of soil behavior, are
conceptually simple, and reduce otherwise excessive computa- Fig. 5. Components of nonlinear p-y elements along with their indi-
tional costs for system reliability studies 共Reese and Van Impe vidual and combined behavior 共modified from Boulanger et al. 1999兲
2001兲.
The development and performance of dynamic nonlinear p-y
elements has been evaluated against the results of dynamic cen-
trifuge tests with reasonable agreement between calculated and
recorded responses 共Boulanger et al. 1999兲. Specifically, the non- units of force per unit length, is adaptive to the temporal pore
linear p-y behavior is modeled with elastic, plastic, and gap com- pressure change in the soil elements connected to it, thereby en-
ponents in series 共Fig. 5兲. The elastic and plastic springs abling the simulation of soil liquefaction. This temporal effect is
determine the backbone of the p-y curve. Radiation damping, synthesized by a degradation multiplier of soil resistance 共1
which can be significant if the foundation material is soft relative − ru兲, where ru denotes the excess pore pressure ratio that is
to the stiffness of the structure and which represents the loss of dependent on the effective soil stresses before and during cyclic
energy due to radiating stress waves that the pile transmits to the loading. Note that the scaled ultimate soil capacity cannot be less
soil, is modeled by a damper located in parallel with the elastic than the specified residual capacity of the material 共i.e., pres兲
component. The radiation damping coefficient for the springs, which is to be preserved even in the case of full liquefaction.
which is a function of pile diameter, soil density, and wave ve-
OpenSees adopts a value of residual soil capacity pres = 0.10. The
locities, is taken to be 1% for sands 共Shin et al. 2006, 2007兲.
normal condition strength parameters to determine the backbone
Radiation damping in soft cohesive soils is also assumed at the
1% level. The gap component consists of a nonlinear closure force-deformation curve of the interface springs are based on the
spring in parallel with a nonlinear drag spring. The gapping re- American Petroleum Institute 共API兲共1993兲 sand and Matlock
sistance represents a drag force on the sides of the pile as it moves 共1970兲/O’Neill 共1983兲 clay relationships.
within a gap in the soil. A value of 0.30 to model typical gapping To transfer the soil-pile interaction forces from the 2D soil and
resistance for a pile is adopted as in Boulanger et al. 共1999兲, 1D p-y springs model to the 3D bridge-foundation model, rigid
although they also found that the soil-pile interaction was not transverse beams are included from the center line of the bridge to
very dependent on this parameter. For cohesionless soils, the gap the line of piles in the transverse direction. The central p-y spring
element is not intended to capture physical gaps, but to match curves are defined for single piles equivalent to pile groups in the
observed stress-strain behavior of saturated sands that influence longitudinal direction at the bents and abutments. Using the group
the shape of the p-y backbone curves. efficiency factors presented by Mokwa 共1999兲, pult values are
The PyLiq1 material 共for softening soils兲 and PySimple1 ma-
scaled down so that the overlapping stress effect due to close
terial 共for nonsoftening soils兲 define the constitutive behavior of
spacing of piles is accounted for in the model. Emphasis is placed
p-y springs used to model the SSI in OpenSees. These Py mate-
rials approximate the resistance of the soil surrounding piles by on the longitudinal CBSF system behavior in this study, but a
integrating the passive and active pressures induced by pile de- similar strategy can be used to analyze its transverse response by
flection. The main input parameters to define the PyLiq1 material aligning the 2D soil and 1D p-y springs transversely and using
to capture liquefaction effects are pult, the ultimate resistance rigid links longitudinally. Note that most of the soil and structural
force that the soil can exert under lateral loading, and y 50, the CBSF model input parameters have been ascertained and verified
displacement when 50% of pult is mobilized. The PyLiq1 resis- in past studies 关Pacific Earthquake Engineering Research Center
tance of the 0.61-m spaced springs in this work, and expressed in 共PEER兲 2006; Zhang 2006; Nielson and DesRoches 2007b兴.
Dynamic Response of a CBSF Model under liquefaction potential all of the saturated sand layers liquefy as a
Scenario Earthquake Hazard consequence of the cyclic loading caused by the earthquake. Spe-
cifically, the 1st bent soil column exhibits full liquefaction—pore
Andrus and colleagues developed a synthetic accelerogram of ap- pressure ratio ru = 1.0—at the loose sand layer depth of 3.6 m
proximately 40 s with a peak acceleration of 0.3g for rock outcrop below ground surface, while softening of p-y springs occurs at
specific to Charleston, S.C. 共Andrus et al. 2006兲. They used the 3.84 s. The lateral displacement time history of the soil deposits
point-source stochastic model presented by Boore 共2000兲 assum- show that bent soil columns make large displacements up to 60
ing similar ground motion parameters as Chapman et al. 共2006兲. mm, whereas abutment soil columns displace by approximately
Before initiating any reliability assessment, this synthetic ground 15 mm owing to the limited thickness of the liquefiable sand. A
motion record is used to gain deterministic insight on the effects thorough analysis of the structural response reveals several key
of soil liquefaction on the response of CBSF system components. liquefaction effects on bridge components. For instance, liquefac-
Note that for a full probabilistic seismic assessment of bridge tion provides a means of natural base isolation decreasing the
components, multiple ground motions are required to capture haz- curvature demands on the columns by more than 50% compared
ard uncertainty, so the writers rely on available synthetic ground to the nonliquefiable foundation soil case. This decrease can be
motions for other sites within the CEUS region. attributed to a significant reduction 共up to 27%兲 of the PGA at
As a preliminary step in the dynamic response investigation, ground surface level, and to the rapid build up of pore pressures
eigenvalue analysis of the CBSF model is performed, yielding which make the structure very flexible at the liquefiable soil layer
five dominant modes within the first 15 natural vibration modes. interface, thus reducing demands at the superstructure level.
The first dominant mode, corresponding to the third natural vibra- However, this behavior can change as a function of soil strata,
tion mode in the longitudinal direction, has a mass participation ground motion, and CBSF system acceleration transmissibility.
factor 共MPF兲 of 23.62%, where the bridge vibrates in phase with Also, a decrease in ground surface peak accelerations due to liq-
the soil and its period is T3 = 0.46 s. The second dominant mode, uefaction is accompanied in this case by large deformation de-
corresponding to the fifth natural vibration mode in the longitu- mands on the piles at the liquefying sand layers and the overlying
dinal direction, has an MPF of 49.58%, but with the bridge oscil- clay layers. Bent pile caps undergo significantly large lateral dis-
lating out of phase with the soil and its period is T5 = 0.40 s. placements near the liquefiable soil layers 共150% increase com-
These interchanging longitudinal mode shapes are also in accor- pared to the nonliquefiable foundation soil case兲, whereas
dance with a recent study from Zhang et al. 共2008b兲 and when abutment pile caps do not undergo significant displacements
added with the rest of the first five natural vibration modes they partly because of the limited thickness of liquefiable abutment
account for approximately 76% of the longitudinal MPF. The soil layers and the fact that lateral spreading is not captured in the
third, fourth, and fifth dominant modes 共sixth, eighth, and 14th model. Pile cap displacements are used to quantify pile damage
natural vibration modes兲 are vertical with the deck spans vibrating because of three reasons. First, quantitative pile damage limit
out of phase, transverse, and vertical with the deck spans vibrat- states can be easily computed in terms of pile cap response 共Le-
ing in phase, respectively. The cumulative MPF in the longitudi- dezma 2007兲; in fact, this paper establishes pile limit states via
nal, transverse, and vertical directions among the first 15 modes is pushover analysis as a function of the soil column in which they
78, 65, and 46%, respectively. are embedded. Second, although maximum curvature of pile
Defining reasonable bridge-soil-foundation system damping is would be an apparent alternative to maximum pile cap displace-
another challenge that needs resolution before conducting dy- ment, the latter gives a better account of the overall pile and
namic SSI analyses. When full 2D plain strain soil mesh models foundation response whereas the former is a highly “localized
are used, they are known to increase the amount of radiation measure of demand” 共Bradley et al. 2009兲. And third, it is easier
damping unrealistically 共Luco and Hadjian 1975兲 and other to collect postevent data on pile cap displacements in practice to
damping sources should be adjusted. However, in the present assess pile foundation integrity and performance.
model, whose 2D soil elements are used in conjunction with p-y Liquefaction also increases the demand on the expansion bear-
springs for 1D wave propagation, no energy absorbing boundaries ings causing significantly large deformations under soil liquefac-
exist and all radiation damping is easily captured by the dashpot tion 共approximately 90 mm corresponding to a 30% increase
of the p-y spring assembly. In addition to this radiation damping, compared to the nonliquefiable foundation soil case兲, while fixed
other damping sources are also explicitly accounted for in this bearings at the bents undergo relatively small longitudinal defor-
study, such as soil and structural material damping through non- mations in both the liquefaction and nonliquefaction cases. Also,
linear constitutive models. Under these damping conditions, the 7.62-cm gap between the abutment and deck closes when
system-level Rayleigh damping should be low. Hence, only 2% liquefaction is considered, which could cause pounding damage at
stiffness proportional damping is adopted for the overall system the abutment back wall.
in order to approximate unaccounted energy dissipation mecha- The results from this deterministic dynamic analysis confirm
nisms. expected trends in the behavior of soil columns, foundation piles,
and bridge components. Hence, the CBSF system is used to
probabilistically investigate component response under multiple
Comparative Response with and without ground motions, soil failure effects, and parameter uncertainty.
Liquefaction Effects
Dynamic analyses of the CSFB system with soil liquefaction ef-
fects are conducted using the synthetic ground motion record Methodology for Bridge Fragility Analysis
from Charleston, S.C., for component response assessment. A set
of CBSF system models representative of nonliquefiable sites are Fragility curves capture the conditional probability of a structure
also developed for relative response comparison, in which the soil to reach or exceed predefined damage states given a hazard inten-
columns and p-y springs are defined as stiff clays. sity measure 共IM兲共e.g., PGA for earthquake hazards兲. Fragility
The analysis of results indicates that in the model with soil curves are developed here for bridges underlain by liquefiable
Table 1. Probability Distributions and Associated Parameters for CBSF System Modeling
Distribution parameters
Mean of Standard deviation Coefficient of Mean Distribution
Random variables Distribution ln X 共␭兲 of ln X 共␰兲 variation 共COVX兲 of X bounds
Concrete strength 共ksi兲 Normala — — 0.13 4.86 —
Steel strength 共ksi兲 Lognormala 6.13 0.08
Coefficient of friction 共COF兲 for expansion bearing Lognormala ⫺3.22 0.50 — — —
共longitudinal direction兲
COF for expansion bearing 共transverse direction兲 Lognormala ⫺2.30 0.50 — — —
COF for fixed bearing 共longitudinal direction兲 Lognormala ⫺1.56 0.50 — — —
COF for fixed bearing 共transverse direction兲 Lognormala ⫺0.99 0.50 — — —
Initial stiffness of fixed bearing Uniforma — — — — 0.50–1.50
Rayleigh damping 共stiffness proportional兲 Normala — — 0.28 0.02 —
Gap at the left abutment 共cm兲 Normala — — 0.32 7.62 —
Gap at the right abutment 共cm兲 Normala — — 0.32 7.62 —
Passive abutment resistance 共kips/in兲 Uniforma — — — — 11.50–28.80
Ground motion 共random sampling from 48 accelerograms兲 Uniforma — — — — 1–48
Soil shear modulus Lognormalb — — 0.30 d
—
Undrained shear strength Lognormalc — — 0.22 d
—
Friction angle Normalc — — 0.12 d
—
Contraction parameter 共defines the rate of pore pressure Normalb — — 0.20 d
—
buildup兲
Ultimate capacity of p-y spring 共sand兲 Normalb — — 0.10 d
—
Ultimate capacity of p-y spring 共clay兲 Normalb — — 0.30 d
—
a
Adapted from Nielson 共2005兲 and Nielson and DesRoches 共2007b兲.
b
Adapted from Pacific Earthquake Engineering Research Center 共PEER兲共2006兲.
c
Adapted from Zhang 共2006兲.
d
Variable with depth and/or soil type.
soils by using nonlinear structural response data from a time- analysis; Sc = median capacity for the predefined performance lev-
history simulation approach. First, a CBSF model is developed els; ␤d兩IM = conditional logarithmic standard deviation of the de-
where the structural and geotechnical parameters, such as con- mand; ␤c = logarithmic standard deviation of the limit state; and
crete and steel strengths, bearing friction coefficients, damping ⌽关 • 兴 = standard normal cumulative distribution function.
ratio, gap between deck and abutment, ultimate strength of the Multiple CBSF system failure modes are considered in this
p-y springs, and cohesion and friction angles of the foundation study by monitoring bearings, abutment soil, columns, piles, and
soil are all treated as random variables with specific probability unseating at the abutments. Except for the piles, the quantitative
distributions 关Nielson 2005; Nielson and DesRoches 2007b; Pa- and qualitative limit states for most of the bridge and foundation
cific Earthquake Engineering Research Center 共PEER兲 2006; components 共along with their variability兲 are adopted from Niel-
Zhang 2006兴共Table 1兲. Then, 48 CBSF models, which are as- son and DesRoches 共2007a,b兲 and Ledezma 共2007兲. Limit states
sembled following a Latin hypercube sampling 共LHS兲 technique, for bridge components combine a qualitative description of their
are subjected to an equal number of available synthetic ground level of damage and associated traffic closure times with a quan-
motions. In order to assess response variability, several repetitions titative metric of their physical state. For example, fixed steel
of the LHS approach are conducted yielding a total simulation bearings limit states are expressed in terms of longitudinal defor-
number of 200. After each transient analysis, the maximum re- mations, where 6 mm represent slight damage or the appearance
sponse of major bridge components is recorded and a probabilis- of cracks in the concrete pier, 20 mm represent moderate damage
tic seismic demand model 共PSDM兲 is constructed using all or prying of the bearings and severe deformation in anchor bolts,
maxima per component. This PSDM establishes a relationship 40 mm represent extensive damage or fracture of the bolts that
between the ground motion IM 共i.e., PGA from ground motions allows bearing toppling or sliding, and 186 mm represent com-
before site amplification兲 and maximum component response plete damage or sliding beyond a typical seat width. The writers
共i.e., demand兲, usually in the form of a power law 共Cornell et al. of this manuscript also established pile limit states from push-
2002兲. over analysis and the newly available OpenSeesPL tool for the
Assuming that both demand 共D兲 and capacity 共C兲 models fol- analysis of lateral pile-ground interactions. First yield, second
low a lognormal probability distribution, where C is a quantitative yield, 75% ultimate, and ultimate curvatures are used to capture
threshold of component response beyond which a level of physi- slight, moderate, extensive, and complete damage states. These
cal damage can be detected, the component fragilities can be cal- levels of curvature are related to lateral pile cap deformation
culated as follows 共Nielson 2005; Song and Ellingwood 1999兲: using OpenSees for specific soil profiles of the foundation system.
P关D ⬎ C兩IM兴 = ⌽ 冉 ln共IM兲 − ␭

␰
冊共1兲
This pile-ground analysis is needed because pile limit states are a
function of not only their design, but also the soil profiles in
which they are embedded. Overall, the four mentioned damage
where ␭ = 1 / b关ln共Sc兲 − ln共a兲兴 and ␰ = 1 / b冑␤d兩IM
2
+ ␤2c ; a and b states—slight, moderate, extensive, and complete—are defined
= regression coefficients from the probabilistic seismic demand for the majority of the CBSF system components as listed in
Table 2. The fragility methodology that relates capacity with de- which are not critical under any soil conditions, experience
mand is now applied to the test bed MSCS bridge. heightened fragility, such as a 43% increase at the moderate dam-
age state. The abutments in active action 共left and right abutments
display similar response兲 experience in turn an average decrease
Liquefaction Effects on Component Fragility Curves in median fragility of 12.5% and in passive action of 34% for the
moderate damage state. This decrease in abutment back soil fra-
Forty-eight synthetic accelerograms created by Wen and Wu gility under liquefaction is explained by the reduction of load
共2001兲 for three cities in CEUS 共Carbondale, Illinois, St. Louis, transfer from the deep foundation soil to the structure through a
and Memphis, Tennessee兲 are used in this study to generate liquefiable soil top layer, and the fact that lateral spreading is not
PSDMs using nonlinear time-history analyses. These simulations being modeled in this work.
account for soil liquefaction effects, and are the basis to construct The abutment and 1st bent piles have a similar behavior and
fragility curves for multiple bridge components. The PGA values also undergo a median fragility decrease that ranges from 37 to
of the synthetic ground motions range from 0.06g to 0.67g. Since 43% for the moderate damage state, and from 50 to 60% for
these ground motions are developed for bedrock, they are propa- abutment piles and 12% for 1st bent piles at the complete damage
gated through the 2D soil columns of the CBSF system directly state. The soft layers topping stiff clays do not impose large re-
triggering liquefaction in most cases. Note that this study’s em- sistance and associated displacement demands on the pile caps
phasis on the relative response of CBSF system components to under liquefiable conditions, which tend to remain stationary or
seismic hazards with and without liquefaction effects permits the exhibit limited displacement, thus decreasing their fragility. Note
use of distinct data sets 共e.g., CEUS bridge type from data analy- that if pile curvatures were measured instead, these same condi-
sis across 11 states 共Nielson 2005兲, soil profiles from South Caro- tions of small pile cap displacement can be accompanied by high
lina, and synthetic CEUS seismic hazard along the Mississippi curvatures at the pile section that interfaces sands with bottom
river兲, which are not necessarily collocated due to paucity of data clay layers. In contrast to abutment and 1st bent piles, the piles at
sources and events, but are still representative of the CEUS region the 2nd bent become more fragile in terms of pile cap displace-
for comparative response analyses. ment. For example, their median fragilities increase in moderate
The parameters establishing PSDM functional relationships and complete damage states by 7 and 19%, respectively, due to
between the median seismic component demand and PGA are displacements of the stiff clay crust that tops the liquefiable layer.
presented in Tables 3 and 4 for the cases with and without lique- These displacement demands are transferred by the piles from the
faction, respectively. It can be observed that the dispersion of bottom to top layers of stiff clays and onto pile caps and columns,
seismic demand increases while the coefficient of determination, bridging the liquefiable soil layer gap in between clayey material.
R2, decreases only slightly with liquefaction effects. The resulting Such behavior is typically observed when the ground motions
bridge component fragility curves for all limit states with and trigger a gradual build up of the pore pressures, as opposed to
without liquefaction effects are plotted in Figs. 6 and 7, respec- possible cases in which a rapid build up of pore pressures creates
tively. A summary of the medians and dispersions for the fragility a flexible liquefied soil layer that limits the force transfer or trans-
curves of all damage states is shown in Table 5. missibility to piles and columns 共and other structural components兲
Overall, the analysis of results under liquefiable soil conditions from subsequent excitation peaks. Also, the role of uncertainty
show that expansion bearings, abutment and its embankment soil propagation from ground motions to demand to capacity is im-
in active action, 2nd bent columns, and all piles are among the portant in pile fragilities. In fact, piles exhibit the largest fragility
most fragile bridge components, whereas fixed bearings, embank- dispersion among bridge components, making the shape of their
ment soil in passive action, and 1st bent columns are compara- fragility curves unusually concave.
tively less critical. Note that the relative vulnerability of these Liquefaction can also decrease the seismic demands on some
components differs by damage state, although the stated critical of the bridge columns. Such a decrease in pier moments and
component groups remain relatively similar, even under nonlique- curvature demands is similar to the effect of a base isolator in
fiable soils conditions. Since MSCS bridges experience signifi- accordance with previous studies 共Mylonakis and Gazetas 2000;
cant deck displacements owing to their large mass, continuity, and Adalier and Elgamal 2002; Kwon et al. 2008兲. Specifically, soil
the relatively low expansion bearing stiffness at the abutments, stiffness and strength degradation decreases the median fragility
they are prone to having dominant longitudinal vibration modes of the 1st bent columns by 75 and 78% for moderate and com-
where the bridge and the soil columns vibrate in and out of phase. plete damage states, respectively. However, liquefaction effects
This effect contributes significantly to the general criticality of increases 2nd bent column fragilities by 35 and 48% in moderate
abutments and expansion bearings across limit states and lique- and complete damage states, respectively. The difference between
faction conditions. The criticality of columns and piles is also 1st and 2nd bent columns is also generally attributed to the dif-
sensitive to the particular soil, structural conditions, and seismic ferent soil profiles associated with these bent pile-column assem-
excitation at their sites. blies. Note that the negative effects of strong crusts topping
Focusing on the actual average effects of liquefaction on com- liquefiable soil layers on pile and column systems has also been
ponent fragility relative to the cases without liquefaction, where experimentally confirmed by Ashford et al. 共2006兲 and Kash-
stiff clay dominates the soil profiles, the fragility of columns and ighandi et al. 共2008兲. Fig. 9 displays a sample deformation profile
piles changes the most, whereas the fragility of expansion bear- that highlights how the pile-column group attracts less demand
ings and abutments’ soil failure changes mildly. Fig. 8 presents when embedded in a liquefiable soil profile topped by a soft soil
select component fragilities at the complete limit states, where the layer 关Fig. 9共a兲兴 relative to pile-column groups embedded in liq-
deck unseating failure mode is also present, and reflects typical uefiable soil profiles topped by stiff soil layers 关Fig. 9共b兲兴.
trends across all damage states. In particular, expansion bearings Finally, the results also show that relative to the component
see an 8 and 13% increase in median fragility 共reduction of me- limit states considered in this work, expansion bearing failure,
dian PGA兲 due to longitudinal bridge displacements at the mod- pile foundation failure, abutment displacement in active action,
erate and complete damage states, respectively. Fixed bearings, and unseating are more likely to occur than column failure under
Table 2. Limit States for CBSF System Components 共Nielson 2005; Ledezma 2007兲
Monitored compo-
nent response Slight damage Moderate damage Extensive damage Complete damage
to detect failure
Bridge failure mode modes Med 共Sc兲 Disp 共␤c兲 Med 共Sc兲 Disp 共␤c兲 Med 共Sc兲 Disp 共␤c兲 Med 共Sc兲 Disp 共␤c兲
a
Column failure Concrete column 1.29 0.59 2.10 0.51 3.52 0.64 5.24 0.65
curvature ductility 共␮⌽兲
Bearing failure at the Fixed 6.00 0.25 20.00 0.25 40.00 0.47 186.6 0.65
bentsa bearing-longitudinal
deformation 共mm兲
Bearing failure at the Expansion 34.4 0.60 104.2 0.55 136.1 0.59 186.6 0.65
abutmentsa bearing-longitudinal
deformation 共mm兲
Abutment soil failure in Abutment soil 37.00 0.46 146.00 0.46 1,000c 0 1,000c 0
passive actiona deformation 共mm兲
Abutment soil failure in Abutment soil 9.80 0.70 37.9 0.90 77.2 0.85 1,000c 0
active actiona deformation 共mm兲
Pile foundation failure Pile cap displacement With liq. Without liq. With liq. Without liq. With liq. Without liq. With liq. Without liq. With liq. Without liq. With liq. Without liq. With liq. Without liq. With liq. Without liq.
at the left abutmentb 共cm兲
3.1 2.0 0.4 0.4 5.4 3.8 0.4 0.4 9.8 6.0 0.4 0.4 13.1 8.0 0.4 0.4
at the right abutmentb 共cm兲
3.0 2.0 0.4 0.4 5.2 3.8 0.4 0.4 9.7 6.0 0.4 0.4 13.0 8.0 0.4 0.4
at 1st bentb 共cm兲
6.0 2.8 0.4 0.4 8.5 4.2 0.4 0.4 11.8 8.6 0.4 0.4 15.8 11.5 0.4 0.4
at 2nd bentb 共cm兲
3.1 2.8 0.4 0.4 5.5 4.2 0.4 0.4 10.1 8.6 0.4 0.4 13.5 11.5 0.4 0.4
Unseating at the Displacement of the — — — — — — 21.60 0
abutment deck relative
to the expansion
bearing 共cm兲
a
Adapted from Nielson 共2005兲.
b
Calculated median capacities Sc using push-over analyses as a function of soil strata in this study, and adopted dispersions ␤c from Ledezma 共2007兲. The convention for cases with liquefaction is “With
liq.” and for cases without liquefaction is “Without liq.”
c
Padgett and DesRoches 共2007兲 determined that severe damage at the abutments does not necessarily mean long-term loss of bridge functionality.
Table 3. PSDMs for CBSF System Components in Log-Log Space with Liquefaction Effects
PSDM regression coefficients
关ln共response兲 = b ⫻ ln共PGA兲 + ln共a兲兴
Averaged
Response b ln共a兲 dispersion 共␤d兲 R2
ln共1st bent column curvature兲 1.46 0.64 1.07 0.62
ln共2nd bent column curvature兲 2.41 2.96 1.67 0.59
ln共expansion bearing deformation兲 1.21 5.37 0.79 0.57
ln共fixed bearing deformation兲 2.27 2.5 1.98 0.61
ln共abutment passive deformation兲 1.39 4.54 0.99 0.64
ln共abutment active deformation兲 1.65 5.32 1.14 0.63
ln共left abutment pile cap disp.兲 1.67 5.39 1.17 0.63
ln共right abutment pile cap disp.兲 1.71 5.50 1.22 0.63
ln共1st bent pile cap displacement兲 1.61 5.89 1.04 0.66
ln共2nd bent pile cap displacement兲 1.87 6.41 1.22 0.67
ln共unseating displacement兲 1.34 5.46 0.80 0.61
liquefaction conditions 共and nonliquefaction conditions except for putationally expensive for bridge network reliability applications,
the unseating case兲 for likely or low PGA levels. However, col- and liquefaction multiplier methods which simply scale median
umns are still the most widely used failure mode to describe fragility parameters to account for the effects of soil failure and
bridge damage mechanisms in academia and practice. For in- do not capture the complexity of SSI phenomena at any level. The
stance, given a PGA= 0.2g it is more likely to observe deck mis- 3D/1D approach is illustrated with a CBSF system built with
alignment and dowel fracture in expansion bearings than shear representative data that is not necessarily collocated, but reflective
cracks and spalling in the columns as failure modes representative of the CEUS region.
of a moderate damage state. The reduced dominance of column Overall, soil liquefaction increases the uncertainty, ␤d, in pre-
failure to adequately represent system-level behavior is also in dictive PSDMs by an average of 32% across bridge components,
agreement with recent studies 共Zhang et al. 2008b; Nielson and and consequently reduces the predictive power of PGA in com-
DesRoches 2007a; Zhang et al. 2004兲 and needs additional treat-
ponent response. Albeit an increase in prediction uncertainty, the
ment beyond the scope of this paper.
coefficient of determination, R2, between component demands de-
creases only by an average of 7%, indicating that the proposed
3D/1D CBSF model is able to provide insights about bridge com-
Conclusions and Recommendations for Future
Work ponent behavior under seismic hazards and soil failure without
sacrificing predictive adequacy. It is found that soil failure from
The introduction of a coupled 3D bridge-foundation system with liquefaction increases the fragility of expansion bearings, and the
2D heterogeneous soil strata FE models and 1D p-y springs pro- potential for deck-abutment pounding effects and deck unseating,
vides an efficient yet adequate framework to monitor multiple whereas it decreases the fragility of abutments without lateral
bridge failure mechanisms in the presence of 1D soil site ampli- spreading. Also, this study highlights that the fragility of piles and
fications and liquefaction effects. This 3D/1D modeling approach columns heavily depends on the soil stratum overlying the lique-
also enables detailed probabilistic analyses of CBSF systems, and fiable sandy layer, which can sway their response in positive and
offers the opportunity to incorporate individual bridge fragility negative ways. For instance, there is potential for natural base
modeling into future transportation network reliability studies. It isolation of columns and pile cap displacements 共which affect
provides a compromise between 3D-only models which are com- piles兲 when soft soils overlay sandy layers, and there is the pos-
Table 4. PSDMs for CBSF System Components in Log-Log Space without Liquefaction Effects
PSDM regression coefficients
关ln共response兲 = b ⫻ ln共PGA兲 + ln共a兲兴 Averaged
dispersion
Response b ln共a兲共␤d兲 R2
ln共1st bent column curvature兲 1.48 1.48 0.84 0.63
ln共2nd bent column curvature兲 1.48 1.48 0.83 0.63
ln共expansion bearing deformation兲 1.10 5.21 0.64 0.61
ln共fixed bearing deformation兲 1.70 1.68 1.36 0.53
ln共abutment passive deformation兲 1.50 4.95 0.83 0.71
ln共abutment active deformation兲 1.72 5.61 0.98 0.69
ln共left abutment pile cap disp.兲 1.75 5.72 0.96 0.71
ln共right abutment pile cap disp.兲 1.76 5.74 0.97 0.71
ln共1st bent pile cap displacement兲 1.79 5.86 0.94 0.74
ln共2nd bent pile cap displacement兲 1.79 5.86 0.94 0.74
ln共unseating displacement兲 1.16 5.23 0.61 0.65
1 1
P[Moderate damage|PGA]
P[Slight damage|PGA]
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
(a) PGA (g) (b) PGA (g)
1 1
P[Extensive damage|PGA]
P[Complete damage|PGA]
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
(c) PGA (g) (d) PGA (g)
Abutment disp. in passive action Fixed bearing

Abutment disp. in active action First bent pile
Expansion bearing Second bent pile
1st bent column Unseating
2nd bent column
Fig. 6. Component fragility curves with liquefaction effects: 共a兲 slight damage; 共b兲 moderate damage; 共c兲 extensive damage; and 共d兲 complete
damage
sibility of high curvature demands on structural components when Recommendations for Future Research
stiff soils top liquefiable sands.
In general, the increase in the median fragility value across There are a number of directions where future research in this
susceptible and critical bridge components 共e.g., expansion bear- soil-foundation-structural reliability field can be oriented. For in-
ings, and select columns and piles兲 ranges from 7 to 35% for stance, although pile cap displacements serve as reasonable dam-
moderate damage and 13 to 48% for complete damage limit states age indicators for deep foundations, quantitative pile damage
in the presence of soil liquefaction. The median decrease in fra- states depending on other mechanistic pile response metrics
gility among beneficiary components from liquefaction effects should be established through experimental research. Further-
共e.g., abutments, and select columns and piles兲 ranges from 12 to more, in order to capture the effects of lateral spreading, sloped
75% for moderate damage and from 34 to 78% for complete ground should be modeled in future studies. Since residual func-
damage. From these ranges of affected or improved component tionality of a bridge after seismic events is as important as re-
response from liquefaction, columns and piles provide the upper sidual capacity, vertical interface springs representing skin
levels of relative response change. This study also shows that friction 共t-z兲 and pile tip resistance 共q-z兲 can be incorporated into
expansion bearings, pile foundations, and deck unseating failures
3D/1D CBSF systems to account for liquefaction-induced vertical
under liquefaction are more likely to occur for probable PGA
settlement and enable comparisons with full 3D FE models. In
demands than susceptible column failures which are widely used
addition, improved approaches to establish a relationship between
as the sole failure mode to describe entire bridge damage states.
Finally, in addition to capturing complex behavior, such as ground motion IMs and structural demand in the presence of liq-
bridge column and pile fragility as a function of soil layer prop- uefaction should be explored to decrease the variability in
erties over the liquefiable soil, the 3D/1D approach also captures PSDMs, in particular for pile groups. Several studies also at-
and propagates uncertainty. Piles compound uncertainty in de- tribute uncertainty to the effects of using different IMs, such as
mand, capacity, and mathematical models, resulting in concave PGA, peak ground velocity, spectral acceleration, etc. 共Mackie
fragility functions that necessitate reduced local uncertainty to and Stojadinovic 2004; Bradley et al. 2009; Padgett et al. 2008兲.
improve system-level predictions in practical applications. However, determining the optimum IM is still a challenge espe-
1 1
P[Moderate damage|PGA]
P[Slight damage|PGA]
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
1 1
P[Extensive damage|PGA]
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Abutment disp. in passive action Fixed bearing

Abutment disp. in active action First bent pile
Expansion bearing Second bent pile
1st bent column Unseating
2nd bent column
Fig. 7. Component fragility curves without liquefaction effects: 共a兲 slight damage; 共b兲 moderate damage; 共c兲 extensive damage; and 共d兲 complete
damage
cially for CBSF systems where variability in soil, structural ma- provide insightful and computationally inexpensive data regard-
terial, and liquefaction potential are jointly present. ing dynamic 3D bridge response. This information can guide
Lastly, similar fragility analyses to the ones presented in this reliability-based designs and retrofit programs where structural
study can be performed in the transverse direction using the same and soil material improvements can be prescribed to ensure target
3D/1D approach, or if liquefaction susceptible p-y springs in two reliability levels for bridges and bridge networks even in the pres-
orthogonal directions are developed, then coupled responses can ence of soil liquefaction.
Table 5. Median and Dispersion Parameters for CBSF System Component Fragility Curves
Slight Moderate Extensive Complete
With liq. Without liq With liq. Without liq. With liq. Without liq. With liq. Without liq.
Bridge comp. Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰ Med 共g兲 ␰
1st bent column 0.77 0.60 0.44 0.62 1.07 0.56 0.61 0.59 1.53 0.62 0.86 0.65 2.00 0.62 1.12 0.65
2nd bent column 0.32 0.56 0.44 0.63 0.40 0.54 0.61 0.59 0.49 0.57 0.86 0.65 0.58 0.57 1.12 0.65
Exp. bearing 0.24 0.73 0.23 0.74 0.55 0.70 0.60 0.71 0.69 0.72 0.76 0.73 0.89 0.76 1.02 0.78
Fixed bearing 0.73 0.52 1.07 0.59 1.24 0.52 2.17 0.59 1.69 0.55 3.27 0.64 3.32 0.58 8.10 0.69
Abutment passive 0.51 0.60 0.41 0.55 1.37 0.60 1.02 0.55 5.46 0.50 3.70 0.46 5.46 0.50 3.70 0.46
Abutment active 0.16 0.63 0.14 0.60 0.36 0.71 0.32 0.69 0.56 0.69 0.48 0.66 2.61 0.46 2.13 0.44
Left abut. piles 0.31 1.48 0.21 0.48 0.43 1.48 0.30 0.48 0.62 1.48 0.39 0.48 0.74 1.48 0.46 0.48
Right abut. piles 0.29 1.44 0.21 0.48 0.41 1.44 0.30 0.48 0.58 1.44 0.39 0.48 0.69 1.44 0.46 0.48
1st bent piles 0.33 1.52 0.24 0.47 0.41 1.52 0.30 0.47 0.50 1.52 0.46 0.47 0.60 1.52 0.54 0.47
2nd bent piles 0.20 1.33 0.24 0.47 0.28 1.33 0.30 0.47 0.38 1.33 0.46 0.47 0.44 1.33 0.54 0.47
Deck unseating — — — — — — — — — — — — 0.94 1.80 1.13 0.49
1 1
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Expansion bearing w/o liq. 1st bent column w/o liq.
Fixed bearing w/o liq. 2nd bent column w/o liq.
Expansion bearing w/ liq. 1st bent column w/ liq.
Fixed bearing w/ liq. 2nd bent column w/ liq.
1 1
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
First bent pile w/o liq. Unseating w/o liq.
Second bent pile w/o liq. Unseating w/ liq.
First bent pile w/ liq.
Second bent pile w/ liq.
Fig. 8. Comparative liquefaction effects on component fragility curves for complete damage state: 共a兲 bearings with and without liquefaction; 共b兲
columns with and without liquefaction; 共c兲 bent piles with and without liquefaction; and 共d兲 unseating
Displacement prof ile at maximum pile curvature w/ liquef action

Displacement prof ile at maximum pile curvature w/o liquef action
Displacement prof ile at maximum column curvature w/ liquef action
Displacement prof ile at maximum column curvature w/o liquef action
0 0
Depth from surface in m
Depth from surface in m
-5 -5
-10 -10
-15 -15
-20 -20
-0.2 -0.1 0 0.1 0.2 -0.2 -0.1 0 0.1 0.2
(a) Lateral displacement in m (b) Lateral displacement in m
Pile cap
Bent cap
Fig. 9. Pile and column lateral displacement profiles at the bents for a PGA= 0.6673 and the instant of time where column base and pile
curvatures near the tip are maximum: 共a兲 at first bent; 共b兲 second bent
Acknowledgments typical bridges in moderate seismic zones.” Eng. Struct., 26, 187–
199.
Ciampoli, M., and Pinto, P. E. 共1995兲. “Effects of soil structure interac-
The writers gratefully acknowledge the support of this research
tion on inelastic seismic response of bridge piers.” J. Struct. Eng.,
by the National Science Foundation through Grant No. CMMI- 121共5兲, 806–814.
0728040 and the Department of Civil and Environmental Engi- Cornell, A. C., Jalayer, F., and Hamburger, R. O. 共2002兲. “Probabilistic
neering at Rice University. The writers also wish to thank the basis for 2000 SAC Federal Emergency Management Agency steel
following researchers for their valuable contributions during the moment frame guidelines.” J. Struct. Eng., 128共4兲, 526–532.
completion of this study: Bryant Nielson, Matt Bowers, Ahmed Dueñas-Osorio, L., Craig, J. I., and Goodno, B. J. 共2007兲. “Seismic re-
Elgamal, Hyung-Suk Shin, Ronald Andrus, Scott Brandenberg, sponse of critical interdependent networks.” Earthquake Eng. Struct.
Christian Ledezma, Güney Olgun, Varun Varun, Oh-Sung Kwon, Dyn., 36共2兲, 285–306.
Chuang-Sheng Yang, Kevin Mackie, and Terje Haukaas. The con- Dueñas-Osorio, L., and DesRoches, R. 共2006兲. “Effect of liquefaction on
tinuing help from Frank McKenna and Silvia Mazzoni in the the performance of multi-span simply supported bridges.” Proc., 1st
OpenSees Forum is also appreciated. Lastly, the writers wish to European Conf. on Earthquake Engineering and Seismology
express their appreciation to the South Carolina Department of (ECEES), European Association for Earthquake Engineering, Istan-
Transportation for sharing detailed bridge-foundation inventory bul, Turkey.
Elgamal, A., Yan, L., Yang, Z., and Conte, J. P. 共2008兲. “Three-
data.
dimensional seismic response of Humboldt Bay Bridge-foundation-
ground system.” J. Struct. Eng., 134共7兲, 1165–1176.
Federal Highway Administration 共FHWA兲. 共2008兲. National bridge in-
References ventory data, U.S. DOT, Washington, D.C.
Kashighandi, P., Brandenberg, S. J., Zhang, J., Huo, Y., and Zhao, M.
Adachi, T., and Ellingwood, B. 共2008兲. “Serviceability of earthquake- 共2008兲. “Fragility of old-vintage continuous California bridges to liq-
damaged water systems: Effects of electrical power availability and uefaction and lateral spreading.” Proc., 14th World Conf. on Earth-
power backup systems on system vulnerability.” Reliab. Eng. Syst. quake Engineering, International Association for Earthquake
Saf., 93共1兲, 78–88. Engineering, Japan.
Adalier, K., and Elgamal, A.-W. 共2002兲. “Seismic response of adjacent Kwon, O.-S., and Elnashai, A. S. 共2009兲. “Fragility analysis of a highway
dense and loose saturated sand columns.” Soil Dyn. Earthquake Eng., over-crossing bridge with consideration of soil-structure interactions.”
22共2兲, 115–127. Struct. Infrastruct. Eng., 6共1–2兲, 159–178.
American Petroleum Institute 共API兲. 共1993兲. Recommended practice for Kwon, O. S., Sextos, A., and Elnashai, A. 共2008兲. “Liquefaction-
planning, designing and constructing fixed offshore platforms— dependent fragility relationships of complex bridge-foundation-soil
Working stress design, 20th Ed., Washington, D.C. systems.” Proc., Int. Conf. on Earthquake Engineering and Disaster
Andrus, R. D., et al. 共2006兲. “Shear-wave velocity and seismic response Mitigation, Institut Teknologi Bandung, Indonesia.
of near-surface sediments in Charleston, South Carolina.” Bull. Seis- Ledezma, C. 共2007兲. “Performance-based earthquake engineering design
mol. Soc. Am., 96共5兲, 1897–1914. evaluation procedure for bridge foundations undergoing liquefaction-
Ashford, S. A., Juirnarongrit, T., Sugano, T., and Hamada, M. 共2006兲.
induced lateral spreading.” Ph.D. dissertation, Univ. of California,
“Soil-pile response to blast-induced lateral spreading. I: Field test.” J.
Berkeley, Berkeley, Calif.
Geotech. Geoenviron. Eng., 132共2兲, 152–162.
Luco, J. E., and Hadjian, A. H. 共1975兲. “Two-dimensional approxima-
Aygun, B. 共2009兲. “Efficient seismic fragility assessment of highway
tions to the three dimensional soil-structure interaction problem.”
bridges on liquefiable soils.” MS thesis, Rice Univ., Houston.
Nucl. Eng. Des., 31, 195–203.
Boore, D. M. 共2000兲. SMSIM Fortran program for simulating ground
Mackie, K., and Stojadinovic, B. 共2004兲. “Improving probabilistic seis-
motions from earthquakes, version 2.0—A revision of OFR 96-80A,
mic demand models through refined intensity measures.” Proc., 13th
OF 00-509, USGS, Denver.
World Conf. on Earthquake Engineering, International Association for
Boulanger, R. W., Curras, C. J., Kutter, B. L., Wilson, D. W., and Ab-
Earthquake Engineering, Japan.
ghari, A. 共1999兲. “Seismic soil-pile-structure interaction experiments
and analyses.” J. Geotech. Geoenviron. Eng., 125共9兲, 750–759. Mackie, K. R., and Stojadinovic, B. 共2006兲. “Seismic vulnerability of
Bowers, M. E. 共2007兲. “Seismic fragility curves for a typical highway typical multiple-span California highway bridges.” Proc., 5th Na-
bridge in Charleston, SC considering soil-structure interaction and tional Seismic Conf. on Bridges and Highways, Multidisciplinary
liquefaction effects.” MS thesis, Clemson Univ., Clemson, S.C. Center for Earthquake Engineering Research 共MCEER兲, Buffalo, New
Bradley, B. A., Cubrinovski, M., Dhakal, R. P., and MacRae, G. A. York.
共2009兲. “Intensity measures for the seismic response of pile founda- Matlock, H. 共1970兲. “Correlations for design of laterally loaded piles in
tions.” Soil Dyn. Earthquake Eng., 29共6兲, 1046–1058. soft clay.” Offshore Technology Conf., Paper No. OTC 1204, Society
Brandenberg, S. J., Boulanger, R. W., Kutter, B. L., and Chang, D. of Petroleum Engineers, Richardson, Tex.
共2007兲. “Liquefaction induced softening of load transfer between pile McKenna, F., and Fenves, G. L. 共2001兲. OpenSees command language
groups and laterally spreading crusts.” J. Geotech. Geoenviron. Eng., manual, Pacific Earthquake Engineering Research Center, Berkeley,
133共1兲, 91–103. Calif.
Brandenberg, S. J., Kashighandi, P., Zhang, J., Huo, Y., and Zhao, M. Mokwa, R. 共1999兲. “Investigation of the resistance of pile caps to lateral
共2008兲. “Sensitivity study of an older-vintage bridge subjected to lat- loading.” Ph.D. thesis, Virginia Polytechnic Institute and State Univ.,
eral spreading.” Proc., 4th Geotechnical Earthquake Engineering and Blacksburg, Va.
Soil Dynamics Conf., American Society of Civil Engineers, Reston, Mylonakis, G., and Gazetas, G. 共2000兲. “Seismic soil-structure interac-
Va. tion: Beneficial or detrimental?” J. Earthquake Eng., 4共3兲, 277–301.
Chapman, M. C., Martin, J. R., Olgun, C. G., and Beale, J. N. 共2006兲. Nielson, B., and DesRoches, R. 共2007a兲. “Analytical seismic fragility
“Site response models for Charleston, South Carolina and vicinity curves for typical highway bridge classes in the central and southeast-
developed from shallow geotechnical investigations.” Bull. Seismol. ern United States.” Earthquake Spectra, 23共3兲, 615–633.
Soc. Am., 96共2兲, 467–489. Nielson, B., and DesRoches, R. 共2007b兲. “Seismic fragility methodology
Choi, E. 共2002兲. “Seismic analysis and retrofit of mid-America bridges.” for highway bridges using a component level approach.” Earthquake
Ph.D. dissertation, Georgia Institute of Technology, Atlanta. Eng. Struct. Dyn., 36共6兲, 823–839.
Choi, E., DesRoches, R., and Nielson, B. 共2004兲. “Seismic fragility of Nielson, B. G. 共2005兲. “Analytical fragility curves for highway bridges in
moderate seismic zones.” Ph.D. dissertation, Georgia Institute of Song, J., and Ellingwood, B. R. 共1999兲. “Seismic reliability of special
Technology, Atlanta. moment steel frames with welded connections: II.” J. Struct. Eng.,
Nilsson, E. 共2008兲. “Seismic risk assessment of the transportation net- 125共4兲, 372.
work in Charleston, SC.” MS thesis, Georgia Institute of Technology, USGS. 共2002兲. “Earthquake hazard in the heart of the homeland.” Fact
Atlanta. Sheet FS-131-02, U.S. Dept. of Interior, Reston, Va.
O’Neill, M. W. 共1983兲. “Group action in offshore piles.” Proc. of the Vlassis, A. G., and Spyrakos, C. C. 共2001兲. “Seismically isolated bridge
Conference on Geotechnical Practice in Offshore Engineering, piers on shallow soil stratum with soil-structure interaction.” Comput.
American Society of Civil Engineers, Reston, Va. Struc., 79共32兲, 2847–2861.
Pacific Earthquake Engineering Research Center 共PEER兲. 共2006兲. “PEER Wang, S., Kutter, B. L., Chacko, J. M., Wilson, D. W., Boulanger, R. W.,
geotechnical group meeting.” 具http://geotechnical.ce.washington.edu/ and Abghari, A. 共1998兲. “Nonlinear seismic soil-pile-structure inter-
opensees/PEER/PEER-bridge/UW-Material/PEERBridgeMeeting action.” Earthquake Spectra, 14共2兲, 377–396.
081806b.pdf典共March 12, 2009兲.
Wen, Y. K., and Wu, C. L. 共2001兲. “Uniform hazard ground motions for
Padgett, J. E. 共2007兲. “Seismic vulnerability assessment of retrofitted
mid-America cities.” Earthquake Spectra, 17共2兲, 359–384.
bridges using probabilistic methods.” Ph.D. dissertation, Georgia In-
Yang, Z., Elgamal, A., and Parra, E. 共2003兲. “Computational model for
stitute of Technology, Atlanta.
cyclic mobility and associated shear deformation.” J. Geotech.
Padgett, J. E., and DesRoches, R. 共2007兲. “Bridge functionality relation-
ships for improved seismic risk assessment of transportation net- Geoenviron. Eng., 129共12兲, 1119–1127.
works.” Earthquake Spectra, 23共1兲, 115–130. Yang, Z., Lu, J., and Elgamal, A. 共2004兲. “A Web-based platform for live
Padgett, J. E., Nielson, B. G., and DesRoches, R. 共2008兲. “Selection of internet computation of seismic ground response.” Adv. Eng. Soft-
optimal intensity measures in probabilistic seismic demand models of ware, 35, 249–259.
highway bridge portfolios.” Earthquake Eng. Struct. Dyn., 37共5兲, Yang, Z., Lu, J., and Elgamal, A. 共2008兲. OpenSees soil models and
711–725. solid-fluid fully coupled elements user manual, Univ. of California,
Reese, L. C., and Van Impe, W. F. 共2001兲. Single piles and pile groups San Diego, San Diego, Calif., 具http://cyclic.ucsd.edu/opensees/典共Jan.
under lateral loading, Balkema, Rotterdam, The Netherlands. 12, 2009兲.
Saadeghvaziri, M. A., Yazdani-Motlagh, A. R., and Rashidi, S. 共2000兲. Zhang, J., Huo, Y., Brandenberg, S. J., and Kashighandi, P. 共2008a兲. “Ef-
“Effects of soil-structure interaction on longitudinal seismic response fects of structural characterizations on fragility functions of bridges
of MSSS bridges.” Soil. Dyn. Earthquake Eng., 20共1–4兲, 231–242. subject to seismic shaking and lateral spreading.” Earthquake Eng.
Sextos, A., and Taskari, O. 共2008兲. “Comparative assessment of advanced Eng. Vib., 7共4兲, 369–382.
computational tools for embankment-abutment-bridge superstructure Zhang, Y. 共2006兲. “Probabilistic structural seismic performance assess-
interaction.” Proc., 14th World Conf. on Earthquake Engineering, In- ment methodology and application to an actual bridge-foundation-
ternational Association for Earthquake Engineering, Japan. ground system.” Ph.D. dissertation, Univ. of California, San Diego,
Shin, H., Arduino, P., and Kramer, S. L. 共2007兲. “Performance-based San Diego.
evaluation of bridges on liquefiable soils.” Proc. of the Research Zhang, Y., Acero, G., Conte, J. P., Yang, Z., and Elgamal, A. 共2004兲.
Frontiers Session of the 2007 Structures Congress, American Society “Seismic reliability assessment of a bridge ground system.” Proc.,
of Civil Engineers, Reston, Va. 13th World Conf. on Earthquake Engineering, International Associa-
Shin, H., Mahadevan, I., Arduino, P., Kutter, B. L., and Kramer, S. L. tion for Earthquake Engineering, Japan.
共2006兲. “Experimental and numerical analysis of seismic soil-pile- Zhang, Y., Conte, J. P., Yang, Z., Elgamal, A., Bielak, J., and Acero, G.
structure interaction of a two-span bridge.” Proc., 8th U.S. National 共2008b兲. “Two-dimensional nonlinear earthquake response analysis of
Conf. on Earthquake Engineering, Earthquake Engineering Research a bridge-foundation-ground system.” Earthquake Spectra, 24共2兲,
Institute, Oakland, Calif. 343–386.
View publication stats Downloaded 15 Dec 2010 to 168.7.217.223. Redistribution subject to ASCE license or copyright. Visithttp://www.ascelibrary.org

Efficient Longitudinal Seismic Fragility Assessmen

Uploaded by

Copyright:

Available Formats

Efficient Longitudinal Seismic Fragility Assessmen

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Efficient Longitudinal Seismic Fragility Assessmen

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Efﬁcient Longitudinal Seismic Fragility Assessment of a Multispan Continuous

Article in Journal of Bridge Engineering · January 2010

Leonardo Dueñas-Osorio Jamie Ellen Padgett

SEE PROFILE SEE PROFILE

Urban water sustainability View project

The user has requested enhancement of the downloaded file.

Introduction defined performance levels as a function of earthquake hazard

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 93

94 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

Rocker 1.1 m Girders @ 1.83 m o.c

Rocker Fixed 0 .46 m

5.0 m 5.0 m 0 .76 m

0 .9 m 4 - #16 bars 0 .46 m

2.4 m 0 .46 m 0 .76 m 0 .76 m 0 .46 m

C C #16 bars @ 305 mm o.c #13 bars @ 305 mm o.c

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 95

0.91 m 4 ratio of these soil elements exceeds standards for FE modeling,

96 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 97

98 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

P关D ⬎ C兩IM兴 = ⌽ 冉 ln共IM兲 − ␭

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 99

100 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

102 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

Abutment disp. in passive action Fixed bearing

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 103

Abutment disp. in passive action Fixed bearing

104 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

Displacement prof ile at maximum pile curvature w/ liquef action

Depth from surface in m

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 105

106 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011

JOURNAL OF BRIDGE ENGINEERING © ASCE / JANUARY/FEBRUARY 2011 / 107

You might also like