PACIFIC EARTHQUAKE ENGINEERING RESEARCH CENTER

Performance-Based Earthquake Engineering Design Evaluation Procedure for Bridge Foundations Undergoing Liquefaction-Induced Lateral Ground Displacement

Christian A. Ledezma and Jonathan D. Bray University of California, Berkeley

PEER 2008/05 AUGUST 2008

Performance-Based Earthquake Engineering Design Evaluation Procedure for Bridge Foundations Undergoing Liquefaction-Induced Lateral Ground Displacement

Christian A. Ledezma and Jonathan D. Bray, Ph.D., P.E. Department of Civil and Environmental Engineering University of California, Berkeley

PEER Report 2008/05 Pacific Earthquake Engineering Research Center College of Engineering University of California, Berkeley August 2008

ABSTRACT
Liquefaction-induced lateral ground displacement has caused significant damage to pile foundations during past earthquakes. Ground displacements due to liquefaction can impose large forces on the overlying structure and large bending moments in the laterally displaced piles. Pile foundations, however, can be designed to withstand the displacement and forces induced by lateral ground displacement. Piles may actually “pin” the upper layer of soil that would normally spread atop the liquefied layer below it into the stronger soils below the liquefiable soil layer. This phenomenon is known as the “pile-pinning” effect. Piles have been designed as “pins” across liquefiable layers in a number of projects, and this design methodology was standardized in the U.S. bridge design guidance document MCEER/ATC-49-1. A number of simplifying assumptions were made in developing this design procedure, and several of these assumptions warrant re-evaluation. In this report, some of the key assumptions involved in evaluating the pile-pinning effect are critiqued, and a simplified probabilistic design framework is proposed for evaluating the effects of liquefaction-induced displacement on pile foundations of bridge structures. Primary sources of uncertainty are incorporated in the proposed procedure so that it is compatible with the Pacific Earthquake Engineering Research (PEER) Center performancebased earthquake engineering (PBEE) framework. A detailed description of the problem and of the current approach to evaluating liquefaction-induced bridge damage is first presented. The PEER-PBEE methodology is described, with emphasis on the components of the method that are more pertinent to the problem under study. Several preliminary evaluations are performed before the proposed simplified procedure is applied. These preliminary assessments include the estimation of the seismic hazard at the site, a liquefaction triggering assessment, and an evaluation of the liquefaction-induced flow failure potential at the site. The details of the application of the PEER-PBEE methodology to the case of bridges founded on piles affected by liquefaction-induced lateral ground displacement are then provided. A procedure to estimate residual lateral displacements at the abutments for a given intensity of the ground motion is described, followed by a discussion of the structural models that can be used to evaluate how these displacements influence the response of the bridge superstructure. With these assessments of bridge response, which are couched in probabilistic terms with key sources of uncertainty characterized, well-defined damage states of the bridge system and its iii

Some of the most useful insights developed through the application of the proposed method are the identification and quantification of the parameters most affecting the uncertainty in the assessment of the seismic performance of a bridge. the point estimate method. and the simplified coupled model. They are the first-order reliability method. which was defined by the spectral acceleration at the degraded period of the potential sliding mass at the abutments. The results from this procedure compare well with the observations from these cases. Lastly. A simplified “user’s guide” that summarizes the proposed procedure follows this abstract. and they are also consistent with the results from an advanced finite element analyses performed by the University of Washington research team. iv . Showa bridge (1964 Niigata earthquake). Three alternative models are available in the estimation of repair cost ratios and downtimes for the different hazard levels included in the analysis. and one centrifuge model test performed at UC Davis. The proposed simplified procedure is validated through its application to three important “case” histories: Landing Road bridge (1987 Edgecumbe earthquake). is the most important source of uncertainty. Lastly.components can be estimated. The proposed procedure is applied to a realistic bridge example that was developed by PEER and Caltrans engineers to illustrate the use of the method and the insights that can be garnered from its application to a bridge evaluation at a site with potentially liquefiable soil. procedures for the final estimates of repair costs and downtimes for the most likely damage conditions are presented. The ground motion hazard. a procedure to incorporate ground motion time histories is described. and findings from the study and recommendations for future work are summarized. It is followed by the uncertainty in estimating seismic displacement at a specific ground motion hazard level and the uncertainty in estimating the residual undrained shear strength of the liquefied soil.

Any opinions. Partial funding of this project through PEER by Caltrans was also provided. findings. and Božidar Stojadinović of UC Berkeley were also invaluable. Steve Sahs. Discussions on the application of the PEER-PBEE methodology with Professor Armen Der Kiureghian of UC Berkeley and with Professor Kevin Mackie of the University of Central Florida were invaluable. Michael Keever.ACKNOWLEDGMENTS This work was supported by the Earthquake Engineering Research Centers Program of the National Science Foundation under NSF Award Number EEC-9701568 through the Pacific Earthquake Engineering Research Center (PEER) under subaward Project ID: 2422006. Caltrans engineers Tom Shantz. Scott Ashford of Oregon State University. Additionally. and conclusion or recommendations expressed in this material are those of the authors and do not necessarily reflect those of the National Science Foundation. Ahmed Elgamal of UCSD. Geoff Martin of USC. v . discussions with other fellow PEER researchers Professors Ross Boulanger of UC Davis. Pedro Arduino and Steve Kramer of the University of Washington. and Mark Yashinsky were of great help in reviewing this work and in developing estimates of the expected damage and consequences at specified levels of residual seismic displacement.

...... G3 Passive Forces versus Structural Capacity. vii LIST OF FIGURES ................................................................................... G8 PRIMARY ASSESSMENT......................................................................................................................................................................................................................... G4 Relationship between Yield Coefficient and Equivalent Undrained Shear Strength.................................2 6..................................................................................................................1...............................................G2 5........................................................................................................................................... iii ACKNOWLEDGMENTS ................................................................................1 1....G9 EXCEL SPREADSHEET ..........................................................................................................................................................................................................................G2 OUTPUT PARAMETERS................................. xiii LIST OF TABLES .................... G7 Passive Reaction at the Abutment ........................................................................................................................................................................................2 Current Approach to Evaluating Liquefaction-Induced Bridge Damage ..............12 vii ...............................................CONTENTS ABSTRACT................................................................................................ G8 Base Area.....................................................12 2..............................................................G9 EVALUATION OF DOWNTIMES AND REPAIR COSTS ......................11 De-aggregation of Hazard ......................................G1 ASSUMPTIONS AND LIMITATIONS .....9 2............................................................................................................................................2 Liquefaction Evaluation ...............................................1 Seismic Hazard....................................................................................................................................................................................................................................................................1 2........................................1 Problem Studied ..................G1 1 2 3 4 5 OVERVIEW .....................................................................................................................1 5.......3 6.............................................................................1 1...2 5.....1.....G1 INPUT PARAMETERS.9 2......................................................................................................................................G5 INCORPORATION OF UNCERTAINTY .......2 Uniform Hazard Spectra .......................................................................................................3 PEER-PBEE Methodology ................................................................................................................. xvii USER’S GUIDE TO AN EVALUATION PROCEDURE FOR ASSESSING BRIDGE DAMAGE DUE TO LIQUEFACTION-INDUCED LATERAL GROUND DISPLACEMENT....6 2 PRELIMINARY ASSESSMENT............................1 6.............................................................................. G2 Flow Failure Evaluation .............................................................3 1..G11 1 INTRODUCTION .......................................................................................................................v TABLE OF CONTENTS ...............................3 6 6..................................................................................................................................... G7 Shear Force in Piles ..........................................G2 PRELIMINARY ASSESSMENT...................................................4 7 8 9 Liquefaction Evaluation .................................................................................

...39 4....49 5......................................2 5..............47 Simplified Approach ......37 PROBABILISTIC EVALUATION OF BRIDGE DAMAGE.........2 Sources and Characterization of Uncertainty ..16 Other Correction Factors...................1.........................21 3 PROBABILISTIC EVALUATION OF SEISMIC LATERAL DISPLACEMENT .................................. Given Intensity Measure ...............................................................................................................1 Damage to Abutments...........................................................51 6............2..........................................................................2 Downtimes ..........21 3...................................................1.......................2 Simplified Approach ..............................................1..................................................................................1...................................................................4 2...........................35 3........................2..........43 Component-Based Approach ........1 5............................4 Passive Soil Forces versus Ultimate Structural Resistance...2...........1......................................................5 Overburden Corrected SPT Value ......1 5............2 4 Decoupled Models ..........................3 Damage to Piles .........................................................1 Repair Costs ......3 2...........................................................................27 3.........................1 Estimation of Residual Lateral Displacement..................................3 Flow Failure ...............15 Magnitude Scaling Factor ................................................................1........................................39 4........51 viii ................................................3 Procedures to Calculate Probability of Exceeding Engineering Parameter Threshold...................................................................................................................................35 Coupled Model.............40 4..............1....1 3........41 5 PROBABILISTIC EVALUATION OF CONSEQUENCES..................43 5.....................49 Component-Based Approach ..............................1.22 3......2...................................2 Derived Engineering Demand Parameters ........2......................................................................................................2.....1...........2............................................................................22 3.......................................1 Deterministic Estimation of Residual Lateral Displacement at Abutments................2......15 Influence of Fines Content ....................................1 2...................................................1 General ...............................2..........................................2 2.....2 Damage to Columns.........................2.................................................................17 2.......................32 3.........................................43 5...16 Correction for Low and High Overburden Stresses.............1 Lateral Displacement as Engineering Demand Parameter.........49 6 BRIDGE EVALUATION EXAMPLE ....................................................18 2.....21 3.............

............................................6.....................................87 8 VALIDATION OF SIMPLIFIED APPROACH ............................3..........3.............1.......................................................................................................67 6.............................3 7 Properties ..2 8........54 Liquefaction Assessment .............................1 6.....................60 Initial Fundamental Period of Potential Sliding Mass ............................89 8...10 Passive Reaction ..............3 8................................................................3...........................................................................59 6.........................66 6...........1 6....3.................3 Calculations................................................................................93 Soil Profile .....................................1 Intermediate Piers .......................................2 Assumptions........3.......................................................................4 1987 Edgecumbe Earthquake...................................................................................................................................................................................6 Results Using Simplified Coupled Model ..........................11 Probability of Liquefaction ..................................................................................................................................63 Size of Potential Failure Surface.......2 Comments ....................................2 Abutment .....54 6.........................60 6.................6.....................................................2 6........3....................................89 8................8 6......................................................1..............1 Results .............94 Preliminary Calculations........................................3 6.......94 ix ......3..........................................6 6.....................................................75 Results ...........................1.......74 Input ..............................................................3..............................................................................................9 Vertical Stresses ........................1 Landing Road Bridge during 1987 Edgecumbe Earthquake..........................5 6..........................57 Flow Failure Assessment ...............................4........................3.......3....5 Results Using Point Estimate Method (PEM)........61 Yield Coefficient as a Function of Equivalent Undrained Shear Strength of Critical Layer ............................................................6...........................76 SENSITIVITY ANALYSIS ..................2 6.......1 8..........................................................................................................................................83 7......................................57 Passive Soil Forces versus Ultimate Structural Resistance...........53 6.......................3..........4 Seismic Hazard.....67 6...........................6.......4 Results Using First-Order Reliability Method (FORM) ..............................................72 6.........7 6....66 6.................................................................1...........4.................3..74 6..................................................................................................................3......65 Ground Motion Intensity..81 7...................89 Landing Road Bridge .59 6...............................

.1........112 Results Using Simplified Approach................1 Passive Reaction at Abutment ....4 8..............112 8................................................110 Soil Properties .105 Results Using Simplified Approach..............................2 Showa Bridge during 1964 Niigata...........................................................................110 OTHER ISSUES.99 8............................6 Results Using Simplified Approach...........................................................4................5 Stability Analysis................................127 APPENDIX A: REPRESENTATIVE MATLAB ROUTINES .........1........................................................................................2 Principal Findings ................................123 10....................2 Incorporation of Ground Motion Time Histories........95 8........................................2 8...........4........................96 8.............................................................1 8...................................................97 Results .......106 Introduction .............................2 8................................................ Japan.......................................................3.................................5 9 1964 Niigata.....................................................3...........2.........101 8........ Japan................101 Showa Bridge ...........95 8.......................................4 8.........................................97 8................................................................................ Earthquake .............2 Shear Force in Piles .......................1 Comparison with Results from Advanced Finite Element Analysis.............................................3 Future Research.6...........................................................................117 9..................3..................96 8...1...........123 10..95 8.........................110 Foundation Properties ..........................1 8.................2..........................................1.........................121 10 CONCLUSIONS...............................................................................5...1 Damage Estimation..............................................................3..........125 REFERENCES......................................5 8............1.........................................2 Sensitivity Analysis ................3 Centrifuge Tests ...............3 8.......8........................2.3 Residual Undrained Shear Strength of Liquefied Soil ........................133 x ......................2...............................1........6................................................3 8.................1...................111 Input Motions .............................................123 10....................4.........................99 8..........100 8................................................................................................................................................................................1 Summary ...4...1.......................3.......................................................103 Soil Profile ........1......................1 Initial Fundamental Period of Potential Sliding Mass........................4 Equivalent Undrained Shear Strength from Passive Reaction and Piles Shear Force ...117 9............... Earthquake..........1.........................................................4......................

.....141 APPENDIX C: ULTIMATE CAPACITY CALCULATION SHEET—PYCAPSI ...............................................................................APPENDIX B: PILES DESIGN ...............161 xi .......

...............G8 Figure 5..................... 5....3 Fig...................................................................................................28 Dispersion in estimates of Sur / σ vc ' ratios from different methods..................................................................................2 Fig......1 Fig................................................................................... 5....................................................... 3.............................................3 Fig.................LIST OF FIGURES Figure 1........................6 Fig..........5 Fig...........2 Fig................................................1 “Pile-pinning” effect for case of pile locked into both soil above and below liquefiable soil layer ...1 Fig.. 4...2 Fig..........................G11 Fig...47 xiii ..................26 Mean estimates of Sur / σ vc ' ratios from different methods ..................7 Fig.......................... 1985) ....... 3... 3....................1 Fig. 3.....14 Fig........................................................35 Structural model to estimate pile-cap rotation as function of lateral deformation ........19 Passive soil forces versus ultimate structural resistance at abutments .................... 2001...............8 Fig..............................2 SPT clean-sand base curve for magnitude 7......................................... 5........46 Repair cost ratio cumulative distribution given damage state considering low (continuous line) and high (dashed lines) uncertainty............. 3........29 Structural model to estimate lateral displacement of intermediate piers’ pile caps from displacement of abutment pile caps .... modified from Seed et al..G6 Figure 4.................................. 3................41 Piles damage states as function of pile cap lateral displacement............................................................................................................................................................3 Passive soil forces versus ultimate structural resistance at piers (modified from Berrill et al............................................................................................ 3....... 2..........G1 Figure 2......... 4............................................................ 2001)... 4..................................................................2 Fig................................................ 2............G5 Figure 3......4 Fig... 1........19 Estimation of initial fundamental period of potential sliding mass (Ts) .........................3 Fig...........................23 Passive reaction versus deflection at abutments (after Caltrans 2006)...................................................................................................................................................... 3...................................................................37 Schematic representation of simplified coupled model.............1 Fig.....42 Bridge damage as function of residual lateral displacement ......................................5 earthquakes with data from liquefaction case histories (after Youd et al...............37 Abutment damage states as function of residual relative lateral deformation in longitudinal direction.. 2...........................36 Simplified 2D structural model ................................40 Column repair decisions as function of residual tangential drift ............................................................................44 Repair cost ratio distribution given damage state considering low (gray bars) and high (black bars) uncertainty .

............... 6........1 Fig.... 6......... k y = 0........14 Fig.... 6................ 6....21 Fig...... 6.....57 Post-liquefaction stability analysis of left abutment ( FS = 1......................53 Right abutment—soil profile and dimensions (not to scale) .......................................... k y = 0......3 Fig.............. 7......17 Fig..8 Fig.........61 Representative soil column of potential sliding mass at right abutment ...................58 Post-liquefaction stability analysis of right abutment ( FS = 1...................................53 Uniform hazard spectra.....................64 Fig.......... 6.....................................................18 Fig........................... 6........73 Deformed shape of bridge from OpenSees model...4 Fig........... 6............11 Fig.... 6.............................11 ) ...........65 Passive reaction versus displacement .....................................56 Factors of safety against liquefaction . 6................. 6........... 5% damping..9 Fig....................... Upper and bottom figures correspond to results for left and right abutments...84 xiv ................................................. 6.... 6.... 6.....................10 Fig.................................................79 Relative importance of random variables for different hazard and displacement levels..................... 6........79 Contribution to expected cost of repair actions for different hazard levels......................................................... 6............21 ........................... ......66 Probability of exceeding lateral displacement threshold (d) for different hazard levels......................055 ) ................................................................................76 Repair cost ratios using simplified coupled model...70 Repair cost ratios using FORM—high uncertainty .............. black symbols denote cases with liquefaction and white symbols denote cases without liquefaction.................................................................52 Piles at interior bents—dimensions and properties .....16 Effect of residual undrained shear strength on base length of potential sliding mass ..............2 Fig...................................................68 Fig........71 Damage states using PEM .....13 Yield coefficient ( k y ) versus equivalent total residual undrained shear strength ( Su ) relationship for cases without liquefaction ....Fig............19 Fig.................... respectively.............................37 .......................52 Piles at abutments—dimensions and properties .................................... 6.............................................. 6............................... 6... 6...7 Fig.............6 Fig...............................64 Fig..... In top figures.......52 Left abutment—soil profile and dimensions (not to scale) ........................................5 Fig...................62 Yield coefficient ( k y ) versus equivalent total residual undrained shear strength ( Sur ) relationship for cases with liquefaction....59 Representative soil column of potential sliding mass at left abutment ........ 6...23 Fig........... 6.12 Example bridge profile ..1 Damage states using FORM ....69 Repair cost ratios using FORM—low uncertainty ...15 Fig....... 6............................20 Fig.22 Fig........ at site.....

2 Normalized relative sensitivity ( δˆ ) of probability p f = P( D > d ) to variations in mean of random variables for different hazard and displacement levels. 7.........................7 Fig..... 8...........90 Acceleration time series and 5%-damping response spectrum of Matahina dam station (Source: PEER NGA database)...... 8......... Large (L). Landing Road bridge.......................91 Landing Road bridge—simplified SPT and shear wave velocity profiles .... (A) Roof.. and Showa bridge .102 Strong-motion earthquake records during Niigata earthquake...100 Relative importance of......9 Fig......10 Probability of bridge being in damage states of Negligible (N)... Continuous line is for Sur........94 Potential failure mass at north abutment... 8....96 Equivalent residual undrained shear strength versus yield coefficient...........93 Schematic elevation of Landing Road bridge (after Berrill et al.... 7. ......13 Fig..... dotted line for α.......4 Fig...... Mw=6......... 1964..................................................... 8...98 Probability of exceeding threshold residual lateral displacement (d) at north abutment of Landing Road bridge.................. Moderate (M)............. and Matahina dam...................104 Schematic diagram of collapse of Showa bridge (after Takata et al........92 Distance and site condition correction factor as function of period ..... Small (S)......... and dash-dotted line for ε .......85 Fig............... 8.......... and P( D > d ) 's sensitivity with respect to.................. 2001) .....................6 Fig................ J................3 Fig. 8.. Upper and bottom figures correspond to results for left and right abutments........... 8..............105 xv .........................2 Fig............... 8..... Source: earthquake Engineering Online Archive.......edu/elibrary/) ..............14 Epicenter of 1964 Niigata earthquake..... 8... 8.. considering contribution of passive reaction and piles shear force ......................... as function of expected level of residual lateral displacement... three random variables........................1 Fig.....5 Fig.........101 Fig.................................12 Fig......... 8... http://nisee...... 8.. (B) Basement of 4-story apartment building (after Iwasaki 1974) ............. respectively... NISEE..Fig....................99 Fig........................6 Edgecumbe earthquake in New Zealand ........... and Collapse (C) ...berkeley.............. given intensity of 1987....... 1965) .. respectively ...3 ˆ ) of probability p f = P( D > d ) to variations in Normalized relative sensitivity ( η standard deviation of random variables for different hazard and displacement levels....103 Collapse of Showa bridge (Photo by Penzien....8 Epicenter of 1987 Edgecumbe earthquake... Upper and bottom figures correspond to results for left and right abutments. 8......86 Fig... 8..............................................11 Fig.......................... ...

...........120 Comparison of results—residual long.... 2003).............3 Fig................108 Schematic diagram showing predicted loading based on JRA code (after Bhattacharya 2003).... Dimensions in mm (after Brandenberg et al...................117 Main components of analytical model (after Shin 2007) ................... modal..........122 xvi ......... 9.....109 SJB03 model layout—accelerometer instrumentation..................22 Fig...7 Soil profile and estimated liquefied layer (after Hamada et al.............. 8.........2 Fig.... (b) Newmark rigid model.....117 Comparison of results—residual pile cap displacement at abutments .1 Fig......... relative deck-end/abutment displacement ........... 2005).......... 9......120 (a) Problem analyzed.......... 8.... 8................................................106 Post-earthquake recovery and deformation of pile from Showa bridge................................................. 9.............15 Fig............ 9......115 Bridge example finite element mesh (after Shin 2007) ...............................................17 Fig...................... (d) nonlinear..........................111 Representative time series from SJB03 for medium Santa Cruz motion (after Brandenberg et al...................................16 Fig. lumped mass.. 9.............. 9.20 Fig...........19 Fig..............18 Fig.119 Comparison of results—residual tangential drift ratio at columns................ 8...118 Comparison of results—residual pile cap displacement at piers .........5 Fig......................... 8.23 Fig....4 Fig...................6 Fig....................... coupled sliding model (after Rathje and Bray 2000) ............ 9.113 Soil displacement versus pile cap displacement for SJB03 test ................... 1986)..... coupled sliding model.... 8.....................Fig.........114 Simplified foundation structural model . Note absence of an upper non-liquefiable crust (after Fukuoka 1966) ...................................................107 Idealized pressure distribution for pile foundations affected by liquefactioninduced lateral ground displacement (after JRA 1996) .... 8..........109 Deformed shape and bending moment diagram (units are kN and m) .... (c) linear elastic..... 8...................... 8.21 Fig................

.....8 Table 6.................39 Column repair decisions as function of residual tangential drift .......94 xvii ............13 Residual tangential column drifts .75 Table 6.......................................4 Table 6..............72 Probability of D > d given IM (PEM results) ......................................................................... G9 Damage State as function of seismic displacement .......................................73 Expected downtimes using PEM .................................................................2 Table 6..................... G10 Table 4....6 Table 6............................................. 2006) .............7 Table 6.........................56 Probability of D > d given IM (FORM results).............................................71 Expected downtimes using FORM....................................12 Input displacements .............................................................................................................................................44 Repair cost ratio distribution given damage state considering low and high uncertainty (values in parenthesis correspond to high uncertainty) .................... G10 Expected Downtimes as function of damage level .............................70 Expected repair cost ratios using FORM—high uncertainty........3 Table 5.......................................................10 Equivalent pile length under each pier ....................................3 Table 6.................2 Table 4......................................................68 Expected repair cost ratios using FORM—low uncertainty.................9 Abutment damage states as function of residual relative lateral deformation in longitudinal direction.................2 Table 5......1 Table 8.........................................1 Table 6.............................................................................72 Expected repair cost ratios using PEM—low uncertainty.........74 Table 6................................................................................................................................1 Characteristics of selected random variables.......1 Table 5.................................11 Stiffness of soil springs.......LIST OF TABLES Table 1 Table 2 Table 3 Table 4 Characteristics of selected random variables .......................................49 Uniform hazard spectra............................77 Table 7...74 Table 6............................................................ at site..............40 Piles damage states as function of pile cap lateral displacement....1 Table 4........................................3 Table 6........................................42 Bridge damage as function of residual lateral displacement ..........82 Soil properties (after Kashighandi and Brandenberg.............73 Expected repair cost ratios using PEM—high uncertainty.45 Expected downtimes given bridge damage level............................................ 5% damping....5 Table 6......................................................... G10 Repair cost ratio distribution given damage state ..........................76 Table 6................

The post-liquefaction static factor of safety at the abutments is assumed to be larger than 1 (i. In this simplified approach. The bridge performance is expressed in terms of repair costs and downtimes. This procedure is appropriate as a screening tool to identify bridges where the effects of liquefactioninduced lateral displacement are important. Figure 1 2. OVERVIEW This guide presents a method to estimate the effects of liquefaction-induced lateral ground displacement on bridges founded on piles.. It is also assumed that residual lateral displacement at the abutments is an appropriate index of bridge performance. ASSUMPTIONS AND LIMITATIONS It is assumed that the pile cap and the upper non-liquefiable crust will move together. the amount of residual lateral displacements at the bridge abutments is assumed to define the performance of the bridge (see Fig. no flow failure). It is assumed that the bridge superstructure restrains the horizontal movement of the abutment through the compression of the deck. and inertial effects from the structural response of the bridge are neglected.USER’S GUIDE TO AN EVALUATION PROCEDURE FOR ASSESSING BRIDGE DAMAGE DUE TO LIQUEFACTION-INDUCED LATERAL GROUND DISPLACEMENT 1. It is assumed that the piles are fixed against G-1 . 1).e.

4. i.. and SPT (or CPT) profiles.1 in the main report). and its de-aggregation with respect to the controlling event’s magnitude and distance. and some equations will need to be modified (see Section 3.. and their mechanical properties (i. Young’s modulus. This simplified method is applicable to cases with groups of vertical piles. 3. thickness of the different layers. (2003): G-2 . and plastic moment). INPUT PARAMETERS Information required to use this simplified method are the uniform hazard spectra at the site.1. such as number and configuration of piles. 5.1. In situations where all the piles are inclined.e. OUTPUT PARAMETERS The results of this evaluation are the median and dispersion of residual lateral displacement at the abutments and the repair costs and downtimes associated with the estimated displacement.rotation at some distance above and below the liquefiable material. and serves as a reasonable approximation for the case of groups of vertical piles with only a few battered piles.1 PRELIMINARY ASSESSMENT Liquefaction Evaluation The probability of liquefaction can be evaluated using the expression developed by Seed et al. i. the results of this method are only approximate and probably overly conservative. and foundation properties.e. For the case of a thin nonliquefiable upper crust and pinned pile connection it could be assumed that the piles are pinned at the pile/pile-cap connection.. the soil profile. the 5%-damped acceleration response spectrum.e. 5. moment of inertia.

The residual undrained shear strengths ( Sur ) are assigned to the potentially liquefiable layers identified in the previous step.. N1. CSR. The value of Sur can be estimated using the expression: ⎛ 0.5 ⎟ × 1 + S ur ≈ exp ⎜ ⎜ 128 ⎝ 8 ⎠ ⎜ ⎝ ( ) 2 ⎞ ⎟ . 0 ≤ N 1. M w is moment magnitude. and a conventional pseudostatic slope stability analysis is performed.3N 1.2 Flow Failure Evaluation The second step in the preliminary assessment is to verify that the static post-liquefaction factor of safety at the abutments is larger than 1. σ v '.004 FC ) − 13.53 ⋅ ln ( M w ) − 3.60−CS ⎛ N 1. g is the acceleration of gravity.60 is the standard penetration test (SPT) blow-count normalized to an overburden pressure of approximately 100 kPa (1 ton/ft2) and a hammer energy ratio or hammer efficiency of 60%.g. This expression is based on the combination of five of the most commonly used procedures to estimate the residual G-3 . and rd is the nonlinear shear mass participation factor (see Eq.60−CS ⎞ ⎜ − 3.97 ⎟⎟ ⎝ ⎠⎟ ⎜ PL ( N1. 12% fines is expressed as FC = 12 ). the NORMSDIST function in Excel).e. CSR = 0.5 in the main body of the report). FC is percent fines content (by dry weight) expressed as an integer (i.32 ⋅ ln(CSR ) − ⎞ ⎞ ⎜ ⎜ ⎟⎟ ⎜ ⎜ 29.⎛ ⎛ N1. CSR is the cyclic stress ratio. 5. M w . 2.60 ⋅ (1 + 0.70 ⋅ ln (σ v ') ⎟⎟ ⎜ ⎜ +0.60 . FC ) = Φ − ⎜ ⎟ 2.70 ⎜ ⎟ ⎜ ⎟ ⎜ ⎟ ⎝ ⎠ (1) where PL is the probability that liquefaction occurs.65 ( amax / g )(σ v / σ 'v ) rd . where amax is the peak horizontal ground surface acceleration. and Φ is the standard normal cumulative distribution function (e. σ v is total vertical stress.. σ 'v is effective vertical stress.05 ⋅ FC + 44.60−CS ≤ 20 ⎟ ⎟ ⎠ (2) where the top bar denotes the mean value of the respective variable.

The reaction of the liquefied material can be estimated using the recommendations by JRA (1996). Olson and Stark (2002). Expressions to estimate N1. direct CPT correlations for evaluating liquefaction triggering and residual undrained shear strength may be used. and Kramer (2007). the following CPT-SPT correlation proposed by Kulhawy and Mayne (1990) may be used to convert tip resistances into SPT blow-counts: ( qc Pa ) 0. If this condition is not satisfied. Additionally. Pa is atmospheric pressure (~100 kPa).60−CS proposed by Idriss and Boulanger (2007). 2).3 times the overburden pressure. If information from cone penetration tests (CPTs) is available.e. to assume that the lateral pressure distribution is 0. the recommendation is to design the foundations to withstand the passive pressures created by the soil flowing around the structure. 5. i.60−CS are shown in sections 2.2.2.1 and 2. N is the measured SPT value. and the two correlations of Sur with N1. and Sur / σ vc ' with N1. The passive reaction of the upper non-liquefiable crust against the pile cap can be estimated using the procedure developed by Mokwa (1999). The simplified procedure described in this guide assumes that movement of the foundation occurs in concert with the soil..60−CS .. The assessment requires a comparison between the estimated passive soil forces that can be exerted on the foundation system and the ultimate structural resistance that can be developed by the structure itself (see Fig.undrained shear strength of the liquefiable layer. that the structural resistance is smaller than the soil’s capacity. G-4 .e. and D50 is the particle size that 50% passes by weight.2 of the report.26 = 5. i. Sur which are: Seed and Harder (1990).3 Passive Forces versus Structural Capacity An assessment is made whether soil flows around the foundation or movement of the foundation occurs in concert with the soil.44 ( D50 ) N (3) where qc is tip resistance.

10. 2007).566 ln(k y ) ln( Sa) + 3.05 s . the first term of this equation. the term -1.Figure 2 6. the amount of seismically induced permanent slope displacement is estimated using the expression developed by Bray and Travasarou (2007): ln( D) = − 1. PRIMARY ASSESSMENT When the analysis in Section 5.66 .5Ts + 0.10 − 2. and ε is a normally-distributed random variable with zero mean and standard deviation σ = 0.22 when Ts < 0.333 ln(k y ) 2 ( ) 2 + 0. i. Sa represents the spectral acceleration at 1. Figure 3 shows how to estimate Ts for the case of a soil mass. and Vs is the average shear wave velocity of the sliding mass.83 ln(k y ) − 0. where H = the average height of the potential sliding mass. The initial fundamental period of the sliding mass (Ts) is estimated using the expression: Ts = 4H/Vs for the case of a relatively wide potential sliding mass that is either shaped like a trapezoid or segment of a circle where its response is largely 1D (Bray.244 ( ln( Sa) ) + 1..278( M w − 7) ± ε (4) where M w is moment magnitude.e.3 indicates that the soil forces are less than the total structural resistance.04 ln( Sa) + − 0. To eliminate a potential bias in the model for very low values of Ts . should be replaced with the term -0. Ts is the initial fundamental period of the potential sliding mass. underlain by a stiff G-5 .5Ts .

The value of the yield coefficient k y that goes into the Bray and Travasarou (2007) expression can be calculated as follows: N ⋅ VP ( D / 2) FP ( D / 2) ⎞ ⎛ k y = p ⎜ Sur + + ⎟ A A ⎝ ⎠ (5) where p represents the relationship between k y and Su . and A is the area of the horizontal portion of the failure surface. where the weighting factor would be the total weight of the respective slice.crust. Figure 3 In Equation 4.1. the estimated displacement is strongly dependent on the slope’s dynamic resistance. its yield coefficient. Since k y itself is dependent on the displacement D due to pile-pinning and other effects. i. FP ( D / 2) represents the passive reaction at the abutment. k y . See Section 3.1.e. VP ( D / 2) is the shear force on any abutment pile.. G-6 . one must use iteration to solve for D . N is the number of piles under the abutment.1 of the main report for more details. the location of the equivalent 1D soil column should be estimated as the weighted average of the location of the slices used in the slope stability analysis. In those cases. sliding on top of liquefied material.

e. and LPile is the equivalent length of the pile. Then. in that case Su = Smax ≈ S ur + N × 2M p / ( H × A) + FPmax / A . and FP maximum passive reaction at the abutment. A reasonable lower limit for Su is the mean residual undrained shear strength of the liquefiable layer.. and α top and α bottom are factors that indicate how many diameters above and below the critical layer the piles are fixed against rotation. G-7 . Su = Smin ≈ S ur .e. i. that the distance between points of fixity is equal to the thickness of the critical layer. the relationship between k y and Su is approximately linear. and that the passive reaction at the abutments has reached its maximum. i.2 Shear Force in Piles The shear force in any pile. M ( D) = 6 EPile I Pile D / L2 Pile < plastic moment of pile. M p the plastic max the total bending moment of each pile. 2 R is the diameter of the pile. Preliminary results for a series of sensitivity analysis indicate that for cases when there is a distinct weak layer where sliding is localized.. H the thickness of the liquefiable layer. D is the relative lateral displacement between the ends of the pile. Let S ur be the mean residual undrained shear strength of the liquefiable layer. between Smin and Smax . the shear force in any pile is calculated as VP ( D) = 2M ( D) H + ( α top + α bottom ) ⋅ 2 R (6) where M ( D) is the bending moment at the ends of the pile. I Pile is the moment of inertia of the pile.6. An upper limit for Su can be estimated considering that the piles have reached their plastic limit. H is the total thickness of the critical layer.1 Relationship between Yield Coefficient and Equivalent Undrained Shear Strength A relationship between k y and Su can be obtained by solving the pseudostatic slope stability problem at the abutment with different “reasonable” values of Su . 6. is estimated by assuming that the piles are fixed against rotation at some distance above and below the critical layer. where EPile is Young’s modulus. LPile = H + (α top + α bottom ) ⋅ 2 R . VP .

where l and t are its plan dimensions in the longitudinal and transverse directions.1. is estimated using the recommendations from Section 7. because the overall size of the potential sliding mass varies as the equivalent undrained shear strength of the critical layer changes. For details see Section 3. 6. a function of Su . the value of l (from the 2D pseudostatic seismic stability problem at the abutment) should be considered. Because the size of the potential sliding mass usually increases as the value Su increases. a conservative approach G-8 .4 Base Area The horizontal part of the base of the potential sliding surface can be assumed to be rectangular in shape. respectively..6. where for seat abutments (see Fig. 4): ⎛ h ⎞ K abut = Ki × w × ⎜ ⎟ ⎝ 5.0 ksf × h 5.e.5 Ae = h × w (area of the backwall) (7) Figure 4 where w and h are the width and height of the backwall (in feet). in general. of Caltrans’ Seismic Design Criteria (2006).3 Passive Reaction at the Abutment The passive reaction at the abutment. respectively. First. i. A = l × t .1 of the main report.1. FP .1. Two effects should be considered in the estimation of l and t .5 ⎠ kip / in Ki = 20 ft Pbw = Ae × 5.8.

40 Standard Deviation = 0. For given values of Sur . which have the characteristics given in Table 1. 7. EVALUATION OF DOWNTIMES AND REPAIR COSTS Based on the results of the previous section.that would lead to a smaller increase in resistance due to the pile-pinning effect would be for l to be for the case in which Su = Smax . which can be replaced in the Bray and Travasarou (2007) expression. as Boulanger et al. The result is a nonlinear. Table 1. 3 and 4 can be used to estimate the expected damage state of the bridge. G-9 . and Sa . α . Tables 2. M w . the tributary width t will be greater than the abutment crest width. Their recommendation of assuming that one half of the slide slope masses is restrained by the piles is adopted in this procedure. Second. implicit equation in D that can be solved iteratively. ε . (2006) have stated. and its associated downtime and repair costs. INCORPORATION OF UNCERTAINTY The uncertainty in the estimation of D is captured at a specified hazard level using the point estimate method (PEM) with the parameters Sur and ε . Ts . the expression for k y is a function of only the lateral displacement D .66 ε Normal 8. because there will be some influence of the abutment side slope masses in the abutment response. Characteristics of selected random variables Random Variable Sur Description Residual undrained shear strength of the liquefied material Error term in the estimation of the residual lateral displacement Probability Density Function Log-normal Uncertainty Measure Coefficient of Variation = 0.

15 0.55 0.9 | ds ) Pr( RCR = 1.7 | ds ) Pr( RCR = 0.2 | ds ) Pr( RCR = 0.20 0.8 | ds ) Pr( RCR = 0.5 to 1.05 0 G-10 .) 0 – 1” Negligible 1” – 4” Small 4” – 20” Moderate 20” – 80” Large > 80” Collapse Table 3.1| ds ) Pr( RCR = 0. Repair cost ratio distribution given damage state Pr( RCR = 0.3 | ds) Pr( RCR = 0.4 | ds ) Pr( RCR = 0.Table 2.15 0 0 0 0 Collapse 0 0 0 0 0 0 0.70 0.5 | ds ) Pr( RCR = 0. Damage state as function of seismic displacement Seismic Displacement Damage Level (in.0 | ds ) Pr( RCR = 0.5 days) 60 days (25 to 95 days) Table 4. Expected downtimes as function of damage level Damage Level Negligible Small Moderate Large Collapse Median downtime (and likely range) 0 0 0 1day (0.20 0.6 | ds ) Pr( RCR = 0.0 | ds ) Negligible 1 0 0 0 0 0 0 0 0 0 0 Damage State Small Moderate 1 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Large 0 0 0 0 0.

EXCEL SPREADSHEET An Excel spreadsheet is provided to perform the calculations indicated in steps 7 and 8. and considering that Sur and ε are the only random variables in the estimation of the seismic displacement D . A sample layout of the spreadsheet is shown in Figure 5.9. assuming a constant value for α and A . Figure 5 G-11 . using the point estimate method.

Hamada 1992). then other ground remediation techniques may not be required to achieve satisfactory performance. mitigation may be difficult and costly at some sites.1 1. and it may not be necessary if the structure’s foundation piles can be designed to withstand the displacement and forces induced by the liquefaction. Even if flow slides do not occur. Compressional forces in the structure. Japan earthquake (e. the Showa bridge failed catastrophically in part due to liquefaction-induced lateral spreading produced by the 1964 Niigata. When liquefaction-induced displacements threaten the performance of an engineered structure. ground improvement. which induce large bending moments. may all result due to liquefaction-induced displacement. soil remediation techniques may be considered to minimize the expected damage. Each of these additional loads may lead to severe damage or collapse of structures. if the foundation piles can be designed to “pin” the upper layer of soil. that would normally spread atop the liquefied layer below it. . These techniques usually involve ground modification. In cases where the post-liquefaction static factor of safety is less than one.g. However. earthquake-induced seismic displacements of the ground may progressively develop and damage piles and the overlying structure.1. and significant relative ground displacements across the supporting piles.. flow slides may occur and induce large lateral and vertical deformations causing extensive damage. or stabilizing measures that significantly reduce the liquefaction hazard or reduce the level of ground deformations to an acceptable level. Moreover. This technique has been coined the “pile-pinning” effect. For example. into the stronger soils below the liquefiable soil layer. additional shear forces in connections.1 Introduction PROBLEM STUDIED Past earthquakes have shown that pile foundations can be vulnerable to liquefaction-induced ground failures. which is illustrated in Figure 1.

Dynamic effective stress–large-strain finite element analyses (Finn et al. Case histories have shown that piles have developed plastic hinges without leading to structural collapse. As Martin (2004) indicates. 1997) indicated that the potentially large seismic displacements due to laterally spreading atop the liquefiable silt layer could be arrested by 2 . as well as near pile heads. 1991). in the Kobe earthquake. Also. The liquefiable silt layer was located between stronger soil strata above and below the potentially weak soil layer. The pile-pinning effect was first implemented in the United States as a result of the seismic safety evaluation of Sardis dam (Finn et al. Sardis dam was judged to be potentially unstable due to excessive seismic deformations resulting from liquefaction of a thin layer of silt in the foundation of the dam.. by including the pile-pinning effect) and allowing increased displacement corresponding to limited plastic hinge rotation (Martin 2004). Considerable cost savings could result by including structural resistance in displacement demand estimates (i. lateral analyses using p-y interface springs together with pile-deformations induced by estimated ground displacement profiles were consistent with observed pile performance.1 “Pile-pinning” effect for case of a pile locked into both soil above and below liquefiable soil layer.Fig. field investigations using borehole cameras and slope indicators showed that pile damage in lateral spread zones concentrated at the interface between liquefied and non-liquefied layers. 1.e.

or more advanced techniques.g. 1. Martin and Lam 2000.4 to 3.7 m apart) were installed in the mid-1990s to remediate Sardis dam. The assessment requires a comparison between the estimated passive soil forces that can be exerted on the foundation system and the ultimate structural resistance that can be developed by the structure itself. Assessment of the impacts to a bridge structure can then be made by considering the proximity of the failure block to the foundation system. This may be accomplished using the simplified Newmark charts. This would include the depths of soil likely to move and the plan extent of the likely soil failure block. Step 4: An assessment is made whether soil flows around the foundation or movement of the foundation will occur.. Work by Martin and others led to the development of a simplified design procedure for evaluating the effects of liquefaction-induced lateral spreads as proposed in the MCEER/ATC-49-1-recommended seismic design document of bridges (ATC/MCEER Joint Venture 2003). The pile-pinning effect is now recognized as a legitimate remediation option when bridge or wharf structures built on pile foundations are located in areas susceptible to liquefactioninduced lateral spreading (e.“pinning” the competent soils above the weak layer into the strong soils below the weak layer.” is delineated in the MCEER/ATC-49-1 document (ATC/MCEER Joint Venture 2003). Newmark time history analysis. and to determine the extent of such movements. Step 3: The maximum displacement of the soil is estimated. 2002). This assessment requires estimating the forces that can develop if soil is to actually flow around the foundation system and comparing 3 . The steps involved in the current MCEER/ATC (2003) design procedure are: Step 1: Step 2: The soil layers that are likely to liquefy are identified.2 CURRENT APPROACH TO EVALUATING LIQUEFACTION-INDUCED BRIDGE DAMAGE A simplified design procedure for evaluating the effects of liquefaction-induced lateral spreads on bridge performance. A total of nearly 2600 prestressed reinforced concrete piles (60 cm in diameter and spaced 2. A stability analysis is executed to determine the likelihood of soil movements. Martin et al. which includes the “pile-pinning effect.

because such damage in the foundation is not likely to pose a collapse hazard. The resulting induced movements of the foundations may produce substantial plastic hinging in the foundations.e. and in such cases the foundation is likely to move with the soil. the assessment indicates that movement of the foundation is likely. Step 6: If on the other hand. then the structure must be evaluated for adequacy at the maximum expected displacement. The induced forces are effectively the largest forces that the structure will experience.. the foundation is designed to withstand the passive pressures created by the soil flowing around the structure. The recommended acceptance criteria are the same as for the Seismic Analysis and Design Procedure E (SDAP E). it may be immediately obvious which condition. and the allowance of plasticity in the foundation is believed to be reasonable. In cases where a crust of non-liquefied material may exist at or near the ground surface.them with the likely resistance that the structure will provide. “large” is taken relative to the structural yield resistance. In many cases. there are realistically only two ways to restrict the foundation and substructure forces to acceptable values. The first method is to design the foundations to resist the forces that would accompany passive flow of the soil around the foundations. soil flow or foundation movement. the full structural resistance is likely to be less than the flow-induced passive forces.05 radians are used for an upper level event. Step 5: If flow of soil around the structure is likely. The implication of this assessment is that for relatively large ground movements. Step 7: If such deformations are not acceptable. the pushover method. is more likely. soil displacements are likely to induce similar magnitude movements of the foundation.1. and may induce relatively large reactions in the superstructure. even though plasticity may be below grade. and for this reason it is conservative to design a structure for such forces. In this context. Plastic hinge rotation limits of 0. The other method would be to limit the ground movement by providing either ground or structural remediation. i. It is the structural option that provides the most 4 . This is the mechanism illustrated in Figure 1.

the piles are elements that limit the acceptable displacements of the system. Step 8: The determination of the plastic mechanism that is likely to occur in the presence of spreading should be done in a reasonable manner. then. Thus. A more precise method of determining the plastic mechanism would be to use an approach that ensures compatibility of deformations between the soil and piles. and this makes use of the “pinning” or dowel action that pile or shaft foundations contribute as they cross the potential failure plane of the moving soil mass. and this shear can then be incorporated back into the stability analysis. The lateral shear that produces the plastic mechanism can be decreased to account for the P-Δ effect. Step 9: The system then must be assessed for a prescribed displacement field to represent the likely soil spreading deformation. The acceptance criteria are basically the structural deformation criteria for SDAP E. then using p-y curves from a program such as LPILE (Reese et al. an estimate of the likely shear resistance that the foundation will provide is estimated. great precision in the determination may not produce more accuracy. which uses the pushover method.rational first path. which accounts for plastic deformations in the piles themselves. one method is to use the upper bound method of plasticity and postulate potential mechanisms. In fact. Step 10: If substantial resistance is provided. then its effect on limiting the instabilitydriven movement of the soil block should be introduced into the analysis. This second requirement could be satisfied by using software that is capable of performing pushover analysis. simple estimates of the mechanism and its corresponding lateral resistance capability are often adequate. assess the mechanism that is likely to control using judgment. 2000) to produce boundary support elements that ensure compatibility. the reduction in shear required to produce a mechanism due to P-Δ should be considered. This step is typically not included in current assessments of potential foundation 5 . For instance. Due to the range of inherent uncertainties. From this analysis. Because the lateral soil force that produces a plastic mechanism in the foundation/substructure system is required.

such as the number of deaths and injuries. and downtime. 1. repair cost. The likely seismic performance and the consequences of that performance in terms of quantities that a client understands. for instance. then the design is complete.3 PEER-PBEE METHODOLOGY As opposed to judging that a design is satisfactory or unsatisfactory based on compliance with defined standards. it does not offer a systematic way of evaluating the performance of existing and proposed bridge designs at sites that could undergo liquefaction-induced lateral ground displacement. Once a realistic displacement is calculated. stone columns. Step 13: If the structure’s behavior is acceptable. The selection of structural or geotechnical remediation methods is based on the relative economy of the system being used. This implies that plastic rotations. The performance-based earthquake engineering (PBEE) framework developed by PEER provides a systematic procedure and it also allows the incorporation of relevant sources of uncertainty in the performance evaluation. and these may not need to connect to the foundation (passive piles). and potentially large ones. It is at this point that more permissive displacements than for substructure design can be relied upon. performance-based earthquake engineering (PBEE) provides a comprehensive methodology for assessing the likely performance of a design for an array of seismic loadings.movements. then the engineer must assess whether to try to produce adequacy either through additional piles or shafts. may be allowed to occur in the foundation under such conditions. Although the MCEER/ATC methodology provides a useful framework for analyzing this type of problem. Steps 11&12: The overall displacement is re-calculated with the revised resistance levels considered. although inclusion of such resistance may often improve the structure’s expected performance. is 6 . then the foundation and structural system can be assessed for this movement. Alternately. if not. ground improvement approaches may be considered. The process is repeated by returning to Step 8 and modifying the available resistance until the slope is stabilized.

The following sections present procedures to define those functions for the case of bridges with foundations that respond to liquefaction-induced lateral ground displacement. Initially.1. dm denotes a damage measure. Defining the seismic hazard in terms of the intensity of ground motions 2. the PEER-PBEE methodology is expressed through a framework equation that estimates the mean annual frequency of events where a specified decision variable exceeds a given threshold. Der Kiureghian (2005) provides a comprehensive review of the PEER-PBEE framework. In general terms. Estimating the damage produced for each calculated dynamic response 4. these steps are: 1. To evaluate Equation 1. the bridge/foundation system is evaluated to identify the governing mechanisms. although related. it is necessary to define the conditional complementary cumulative distribution functions G ( x | y ) . edp is an engineering demand/response parameter. which includes the pile-pinning effect. The PEER-PBEE equation is (Cornell and Krawinkler 2000): λ (dv ) = dm edp im ∫ ∫ ∫ G(dv | dm) dG(dm | edp) dG(edp | im) d λ (im) (1. Assessing the consequences for each damage state In mathematical terms. The Pacific Earthquake Engineering Research (PEER) Center PBEE methodology is based on the premise that the overall bridge performance assessment can be divided into a series of discrete steps that. G ( x | y ) = Pr( X > x | Y = y ) is the conditional complementary cumulative distribution function of the random variable X given Y = y . dG ( x | y ) is the differential of G ( x | y ) with respect to x . and λ ( x) is the mean frequency of { X > x} events per year. Evaluating the dynamic response of the system for each intensity level 3.1) in which im is the ground motion intensity measure. dv denotes a decision variable such as repair cost or downtime.estimated for a sliding scale of seismic hazard levels from frequent to rare earthquake events. can be analyzed independently. 7 .

The following is a short summary of the paper by Abrahamson (2000).” The calculation of the annual frequency of exceedance involves several probability distributions for each seismic source. magnitude. Probabilistic seismic hazard analysis follows the basic approach developed by Cornell (1968). a scenario earthquake is defined (i.2 Preliminary Assessment 2. the rupture dimension and location of the earthquakes. but is sometimes estimated using seismological simulations of the ground motion.e.. the distance to the site. i. The ground motion for the scenario earthquake is usually estimated using attenuation relations. and time of occurrence are explicitly considered. In a deterministic seismic hazard analysis. and in some cases rupture direction). and .. earthquake scenarios are evaluated separately.1 SEISMIC HAZARD Seismic hazard analyses involve the quantitative estimation of the ground shaking hazard at a particular site (e. and the fault type. This number of events per year is also called the “annual probability of exceedance. This estimation can be performed deterministically. and in some cases rupture directions. The basic methodology involves computing how often a suite of specified levels of ground motion will be exceeded at the site. For each source. location. which briefly describes both approaches. the earthquake magnitude.. number of standard deviations. style of faulting. in which uncertainties in earthquake size.e.” The inverse of this parameter is called the “return period. assuming a particular earthquake scenario. or probabilistically. see Kramer 1996 for more information). the tectonic regime. with emphasis on the probabilistic approach. The hazard analysis computes the annual number of events that produce a ground motion parameter that exceeds a specified level. at least five attributes need to be considered in developing a deterministic design earthquake: classification of active or inactive. To estimate the ground motion for the scenario earthquake requires selecting the appropriate site condition. with the addition that the variability in the ground motion (for a given magnitude and distance) is included. distance. For faults. ground motion model. the frequency of earthquakes of various magnitudes.g.

that describe the relative likelihood of different earthquake scenarios. r is the distance measure. The ground motion at the site is estimated from the attenuation relation. rupture width. ε ) is the probability that the ground motion exceeds the test level z for magnitude m . the magnitude.g. and number of standard deviations ε . x) fε (ε ) P( A > z | m. M maxi is the maximum magnitude for the i-th source. P ( A > z | m. r . Because in the context of attenuation relationships. and rupture location dip should be considered. r . The location of the earthquake depends on the geometry of the seismic source. distance. distance r . rupture location along strike.. fε (ε ) is the probability density function for a ground motion variability. m is magnitude.2) … f Exi ( x) f Eyi (m. and number of standard deviations fully define the intensity of the ground motion.1) where Ni ( M min ) is the rate of earthquakes with magnitude greater than M min from the i-th source. W ) f RAi (m. For point sources. known faults) the finite dimension and location of the rupture should be considered to compute the closest distance. For planar sources Equation 2. r . and P( A > z | m. Specifically. ε ) is either 0 or 1. The distance from the rupture to the site is computed from the earthquake location and rupture dimension. the randomness in the rupture length. ε )dmdrd ε (2. RA. respectively. y. ε )dWdRAdxdydmd ε 10 .the attenuation of the ground motion from the earthquake rupture to the site. For planar sources (e.1 becomes: M maxi ν i ( A > z ) = N i ( M min ) ∞ W = 0 RA= 0 Ex = 0 Ey = 0 m = M mini ε =ε min ∫ ∫ ∫ ∫ ∞ 1 1 ∫ ε max ∫ f mi (m) fWi (m. The occurrence rates of the earthquakes of various magnitudes are estimated by magnitude recurrence relations. RA)… (2. the rate at which the ground motion from the i-th source exceeds the test level z at the site is given by: ε =ε max ∞ m = M maxi ε =ε min r = 0 m = M mini ν i ( A > z ) = N i ( M min ) ∫ ∫ ∫ f mi (m) f ri (r ) fε (ε ) P ( A > z | m. f mi (m) and f ri (r ) are the probability density function for the magnitude and the distance. W ). ri ( x. ε is the number of standard deviations of the ground motions from the median ground motion.

. Since the hazard is computed independently for each period.W ) . For a selected return period. 2. the hazard is computed independently for each spectral period. rupture area.g. For multiple seismic sources. 11 . That is. respectively (e. x = 0 is one end of the fault and x = 1 is the other end of the fault).. The term “uniform hazard spectrum” is used because there is an equal probability of exceeding the ground motion at any period.2 may appear to be complicated. and f Eyi ( y ) are probability density functions for the rupture width. the total annual rate of events with ground motions that exceed z at the site is just the sum of the annual rate of events from the individual sources (assuming that the sources are independent): ν ( A > z) = N source i =1 ∑ ν ( A > z) i (2. A uniform hazard spectrum is developed by first computing the hazard at various spectral periods using response spectral attenuation relations.3) where N source is the total number of fault and areal sources. x and y give the location of the center of the rupture in terms of the fraction of the fault length and fault width. magnitude and distance combinations) with the full range of possible ground motion levels for each scenario while keeping track of which scenarios lead to ground motions that exceed the test value z .2 defines a complete set of possible earthquake scenarios (i. The hazard integral in Equation 2. Equation 2. location of the center of rupture along strike and location of the center of rupture down dip. These ground motions are then plotted at their respective spectral periods to form the uniform hazard spectrum.2.1.e. a uniform hazard spectrum does not represent the response spectrum of any single earthquake. In Equation 2.1 Uniform Hazard Spectra A common method for developing design spectra based on the probabilistic approach is through the uniform (or equal) hazard spectra. f Exi ( x) . respectively.where fWi (m. RA) . but most of the computations are related to finding the distribution of the closest distance. the ground motion for each spectral period is measured from the hazard curves. f RAi (m. in general.

The most common form of de-aggregation is a two-dimensional deaggregation in magnitude and distance bins.g. The de-aggregation by magnitude and distance bins allows the dominant scenario earthquake (magnitude and distance pairs) to be identified. This process is called deaggregation. Increased pore-water pressure is induced by the tendency of loose to moderately dense granular materials with poor drainage to compact when subjected to cyclic shear deformations. liquefaction may lead to ground oscillation or lateral spread as a consequence of either flow deformation or cyclic mobility. 2. The results of the de-aggregation will be different for different probability levels (e. expressed in terms of the induced cyclic stress ratio 12 . magnitudes. it is difficult to get an intuitive understanding of what is controlling the hazard from the hazard curve itself. such as beneath a foundation or sloping ground. the soil stratum softens. 100 year. Because all of the sources. allowing large cyclic deformations to occur. In loose materials.2 LIQUEFACTION EVALUATION Liquefaction is defined as the transformation of a granular material from a solid to a liquefied state as a consequence of increased pore-water pressure and reduced effective stress (Marcuson 1978).versus 1000 year return periods) and for different spectral periods. Estimation of two variables is required for evaluation of liquefaction resistance of soils: (1) the seismic demand on a soil layer. that may lead to large shear deformations or even flow failure under moderate to high shear stresses. the softening is also accompanied by a loss of shear strength.1. at a given ground motion level. and distances are mixed together.2. A common practice is to break down the hazard back into its contributions from different magnitude and distance pairs to provide insight into what events are the most important for the hazard. With de-aggregation.2 De-aggregation of Hazard The hazard curve gives the combined effect of all magnitudes and distances on the probability of exceeding a specified ground motion level. Beneath gently sloping to flat ground. to a residual level known as the residual undrained shear strength ( Sur ). the fractional contribution of different subsets of the events to the total hazard is computed. A deaggregation by seismic source allows the dominant seismic source to be identified. Loose soils may also compact after liquefaction with reconsolidation leading to ground settlement. As liquefaction occurs..

A plausible method for evaluating CRR is to retrieve and test undisturbed soil specimens in the laboratory. Those criteria are largely embodied in the CSR versus N1. N1. Seed and Idriss (1971) formulated the following equation for the calculation of CSR: CSR = ⎛ a ⎞⎛ σ ⎞ τ av = 0. and rd is the stress reduction coefficient. and specimens of granular soils retrieved with typical drilling and sampling techniques are too disturbed to yield meaningful results.5 1.(CSR) and (2) the capacity of the soil to resist liquefaction.000 − 0.001753 z1.04052 z + 0.60 is the SPT blow count normalized to an overburden pressure of approximately 100 kPa (1 ton/ft2) and a hammer energy ratio or hammer efficiency of 60%.5 + 0. In the remaining part of this section the evaluation of liquefaction resistance using the standard penetration test is presented.5 + 0.1 is a graph of calculated CSR and corresponding N1. Figure 2. CRR curves on this graph were conservatively positioned to separate regions with data 13 .4177 z 0.65 ⎜ max ⎟ ⎜ vo ⎟ rd σ 'vo ⎝ g ⎠ ⎝ σ 'vo ⎠ (2.05729 z − 0. The latter coefficient accounts for the flexibility of the soil profile. Criteria for evaluation of liquefaction resistance based on the SPT have become relatively robust over the years. Unfortunately. σ vo and σ 'vo are the total and effective vertical overburden stresses. g is the acceleration of gravity. Several field tests have gained common usage for evaluation of liquefaction resistance. including the standard penetration test (SPT) and the cone penetration test (CPT). expressed in terms of the cyclic stress ratio required to generate liquefaction (CRR).5 + 0.60 plot reproduced in Figure 2.60 from sites where liquefaction effects were or were not observed following past earthquakes with magnitudes of approximately 7.000 − 0. field tests have become the state of practice for routine liquefaction investigations. can sufficiently undisturbed specimens be obtained.001210 z 2 (2.4113 z 0.4) where amax is the peak horizontal acceleration at the ground surface generated by the earthquake. Only through specialized sampling techniques. 2001): rd = 1. To avoid the difficulties associated with sampling and laboratory testing.5. For routine engineering practice the following equation may be used to estimate the average values of rd (Youd et al. The cost of such procedures is generally prohibitive for all but the most critical projects.5) where z is the depth below ground surface in meters. in-situ stress states generally cannot be reestablished in the laboratory.1. such as ground freezing.006205 z1. respectively.

Percent Fines ≥ 35 Fig. Rauch approximated the cleansand base curve plotted in Figure 2.1 are valid only for magnitude 7.5 earthquakes with data from liquefaction case histories (after Youd et al.1 SPT clean-sand base curve for magnitude 7. The CRR curve for fines contents <5% is the basic penetration criterion for the simplified procedure and is referred to as the “SPT clean sand base curve. Curves were developed for granular soils with the fines contents of 5% or less. A.5 earthquakes. and 35% or more as shown on the plot. 15%. 2. 2001. F. modified from Seed et al. In personal communication to Youd et al.indicative of liquefaction from regions with data indicative of non-liquefaction. (2001). 1985).” The CRR curves in Figure 2.1 by the following equation: 14 .

0 β = 1. This factor is commonly calculated from the following equation (Liao and Whitman 1986): N1.2 Influence of Fines Content Youd et al.8) α =0 α = exp ⎡1.” Similar types of relationships are available to estimate CRR7. (2001) recommend Equations 2.60 −CS 1 50 + + 34 − N1. Idriss with the assistance of R. clean granular soils are often classified as being too dense to liquefy and are labeled as “non-liquefiable.60−CS : N1. M.2.8–2.60 (2.10) 15 .5 = N1.5 based on normalized CPT penetration resistances.6) This equation is valid for N1.2. 2.5 /1000 ⎤ ( )⎦ for 5%<FC < 35% for FC ≥ 35% (2.60−CS = α + β ⋅ N1. These equations were developed by I. an overburden stress correction factor is applied (Seed and Idriss 1982).60−CS 135 10 ⋅ N1.99 + FC1. where C N = Pa σ 'vo (2.CRR7.1 Overburden Corrected SPT Value Because SPT N-values increase with increasing effective overburden stress.7.10 as approximate corrections for the influence of fines content (FC) on CRR.60−CS < 30 .0 ⎣ β = 1. 2. Seed for correction of N1.9) α = 5.2 for FC ≤ 5% β = ⎡0.60 −CS + 45 ( ) 2 − 1 200 (2.76 − 190 / FC 2 ⎤ ⎣ for FC ≤ 5% ( )⎦ for 5%<FC < 35% for FC ≥ 35% (2.60−CS ≥ 30 . B. For N1.7) where CN normalizes the measured standard penetration resistance N m to an effective overburden pressure of approximately Pa = 100 kPa (1 atm). N1. CN should not exceed a value of 1.60 = CN N 60 .60 to an equivalent clean sand value.

The rate of increase.. Seed and Idriss 1982.6.2..5 is the cyclic resistance ratio for magnitude 7.12) 2. CRR7. To account for the nonlinearity between CRR and effective overburden pressure. Several relations have been proposed to estimate MSF (e. isotropically consolidated triaxial compression tests on sand specimens were used to measure CRR for high-stress conditions and to develop Kσ values. Seed (1983) introduced the correction factor Kσ to extrapolate the simplified procedure to soil layers with effective overburden pressures different than 100 kPa.4 Correction for Low and High Overburden Stresses Cyclically loaded laboratory test data indicate that liquefaction resistance increases with increasing confining stress. By taking the ratio of CRR for various confining pressures to the CRR determined for approximately 100 kPa (1 atm). is nonlinear. the equation for the factor of safety (FS) against liquefaction is written in terms of CRR.2. Seed and Idriss (1982) introduced correction factors termed “magnitude scaling factors (MSFs). Seed (1983) developed the original 16 .24 ⎛ 7.1 apply only to magnitude 7.2. i.” These factors are used to scale the CRR base curves upward or downward on CRR versus N1. In this report the relation recommended by Youd et al. To adjust the clean-sand curves to magnitudes smaller or larger than 7.3 Magnitude Scaling Factor The clean-sand base or CRR curves in Figure 2.5 is determined from Equation 2. CSR. (2001) is used.5 ⎞ FS = ⎜ ⎟ MSF ⎝ CSR ⎠ (2.60 plot.56 = ⎜ ⎟ Mw ⎝ Mw ⎠ 2. and Andrus and Stokoe 1997). Ambraseys 1988.e. Cyclically loaded.5 earthquakes.5.5 ⎞ MSF = 2. To illustrate the influence of magnitude scaling factors on calculated hazard.56 (2. 102. and MSF as follows: ⎛ CRR7. however.11) where CSR is the calculated cyclic stress ratio generated by the earthquake shaking and CRR7.5 earthquakes.g.

4 (2.3 FLOW FAILURE If after assigning undrained residual shear strengths ( Sur ) to the layers that are likely to liquefy and calculating the post-liquefaction static factor of safety. rod length. it is reasonable (and conservative) to assume that large ground deformations will be imposed on the bridge foundations. there are other important factors that should be evaluated. and correction for sloping ground ( Kα ). are measured in the same units. and for relative densities between 60% and 80%. borehole diameter. and sampling method. f = 0. and Pa . This case is termed the liquefaction flow case.g.. aging. including relative density.6 . Other investigators have added data and suggested modifications to better define Kσ for engineering practice.5 Other Correction Factors The correction factors presented in the previous sections are the most relevant for the application discussed in this report. in general. step 2 of the MCEER/ATC (2003) design procedure). stress history. Such large ground deformations will likely induce large damage to the bridge and may even lead to a partial or complete collapse of the bridge..Kσ correction curve. the post-liquefaction static factor of safety is found to be less than or equal to unity (i. Hynes and Olsen (1999) compiled and analyzed an enlarged data set to provide guidance and formulate equations for selecting Kσ . and overconsolidation ratio. ground deformations on the order of several meters are possible. effective overburden pressure. and f is an exponent that is a function of site conditions.8 − 0. if the bridge cannot be relocated.13) where σ 'vo . The equation they derived for calculating Kσ is: ⎛σ ' ⎞ Kσ = ⎜ vo ⎟ ⎝ Pa ⎠ ( f −1) ≤ 1.e.7 .7 − 0. 2. the “pile-pinning” effects procedure presented in this 17 . NCEER’s recommendation for the values for f are for relative densities between 40% and 60%. SPT corrections due to energy ratio. however. atmospheric pressure. f = 0. 2. In these situations.2. e. Case histories suggest that when the post-liquefaction factor of safety is less than one.

4 PASSIVE SOIL FORCES VERSUS ULTIMATE STRUCTURAL RESISTANCE The methodology that is presented in the next sections assumes that the pile cap and surrounding soil move together.3 ⋅ σ vc ⋅ b . It is important to note that according to Boulanger et al. 2. the reaction capacity of the soil must be compared to the maximum lateral capacity of the piles and columns (see Figs. otherwise. uncontrolled bridge deformations and damage are likely to occur. the soil will continue to displace or flow around the stable foundation.2–2.6 ⋅ σ vc '⋅ b . the movement of the foundation will likely occur in concert with the soil.3). The pressures created by the flowing soil can be estimated using the procedure developed by Mokwa (1999) for the passive soil resistance of the non-liquefied upper crust reacting against the pile cap. Otherwise. the limit pressure for liquefied sand may be roughly represented as p ≈ 0. and significant ground improvement and/or stabilizing measures are required to ensure that the post-liquefaction factor of safety is greater than one. which are difficult to assess with confidence. and the recommended pressure distributions given by JRA (1996) for the liquefied material. significant. 18 . (2003) for beam on nonlinear Winkler foundation models where the lateral ground displacement loads are modeled as limit pressures (with the p-y springs removed in the lateral displacement interval). the lateral pressure can be approximately represented as p ≈ 0.document cannot be applied.5 ⋅ σ vc . Since σ vc ' ≈ 0. and the foundation should be designed to resist the pressures created by the flowing soil. which is the approach proposed by JRA (1996). To verify this. then. 2. If the soil capacity is smaller than the structural capacity. for liquefied sand.

2.3 Passive soil forces versus ultimate structural resistance at abutments.Fig. 2. 2001). 19 . Fig.2 Passive soil forces versus ultimate structural resistance at piers (modified from Berrill et al.

In this chapter. .1 After an exhaustive examination of the problem at hand (i. Although the amount of vertical settlement at the abutments can also influence the performance of a bridge. bridge undergoing lateral ground displacement wherein the soil crust moves with the pile cap). its effect is usually much more localized and it has not been incorporated in the simplified approach proposed in this study.e. If liquefaction-induced vertical settlements need to be estimated. given a provided ground motion intensity measure. is presented. is presented. and the procedure developed by FMSM Engineers (2007) for settlements induced by deviatoric deformations.1).3 Probabilistic Evaluation of Seismic Lateral Displacement LATERAL DISPLACEMENT AS ENGINEERING DEMAND PARAMETER 3. im. which is required to evaluate Equation (1.. exceeds a certain threshold of seismic displacement. a procedure to estimate the function G (edp | im) . seismically induced permanent lateral displacement (D) of the ground undergoing liquefaction-induced lateral displacement was judged to be the most representative engineering demand/response parameter (edp). it is recommended to use the methods by Ishihara and Yoshimine (1990) and Tokimatsu and Seed (1984) for volumetric compression. Any resulting bridge damage could be best tied to the amount of lateral ground displacement for this problem. The probability that the residual lateral seismic displacement. d. which is described by the relationship Pr( D > d | im) . The probability that the liquefaction-induced lateral ground displacement exceeds a specified threshold for a given ground motion intensity measure is required to implement the PEER-PBEE methodology. which is denoted as D.

05 s . In those cases. k y . the expression Ts = 2.22 when Ts < 0.2). the amount of residual lateral displacement will be a function of the intensity of the ground motion and the dynamic resistance of the earth slopes that impact the bridge system. In the proposed procedure.1 Deterministic Estimation of Residual Lateral Displacement at Abutments In general. i. and let L be the event that liquefaction does not occur. 3. can be estimated as: Pr( D > d | im) = Pr( D > d | im. and Vs is the average shear wave velocity of the sliding mass. For the case of a triangular-shaped sliding mass that largely has a 2D response.1. D . L) Pr( L | im) + Pr( D > d | im. underlain by a stiff crust.10 − 2..3.04 ln( Sa ) + − 0. For these seismic stability calculations.83 ln(k y ) − 0. Ts is the initial fundamental period of the potential sliding mass.10.1) can be estimated.5Ts .1. This dynamic resistance is represented by the yield coefficient. and ε is a normally-distributed random variable with zero mean and standard deviation σ = 0.6H/Vs should be used. Sa represents the spectral acceleration at 1.e.1. the term -1. where H = the average height of the potential sliding mass. sliding on top of liquefied material.1 shows how to estimate Ts for the case of a soil mass. the expected level of residual longitudinal displacement at the abutments.66 . Pr( D > d | im) . the first term of Equation (3.5Ts + 0. the initial fundamental period of the sliding mass (Ts) can be estimated using the expression: Ts = 4H/Vs for the case of a relatively wide potential sliding mass that is either shaped like a trapezoid or segment of a circle where its response is largely 1D (Bray 2007).566 ln(k y ) ln( Sa) + 3. To eliminate a potential bias in the model for very low values of Ts .1 Estimation of Residual Lateral Displacement The total probability theorem is used to estimate the required conditional probability Pr( D > d | im) . Figure 3. L) Pr( L | im) (3. should be replaced with the term -0.278( M w − 7) ± ε (3.2) where M w is moment magnitude. the location of the 22 .333 ln(k y ) 2 ( ) 2 + 0.244 ( ln( Sa) ) + 1. The probability that the residual lateral deformation exceeds a certain threshold given an intensity measure.1) The remaining parts of this chapter show how the terms of Equation (3. is estimated using the relationship developed by Bray and Travasarou (2007): ln( D) = − 1. Let L be the event that liquefaction occurs.

FP ( D / 2) represents the passive reaction at the abutment. Fig.equivalent 1D soil column should be estimated as the weighted average of the location of the slices used in the slope stability analysis. (2006). The lateral resistance of the slope is principally derived from three sources: the residual undrained shear strength of the critical layer ( Sur ). and the passive reaction at the abutment. 3. D . These three forces can be combined and expressed in terms of an equivalent undrained shear strength ( Su ): Su = Sur + N ⋅ VP ( D / 2) FP ( D / 2) + A A (3. 23 . recognizing that in these cases the maximum residual displacement usually occurs at the end of the ground motion and not at the beginning. the resistant forces are calculated using half of the maximum residual lateral deformation. and (ii) as recommended by Boulanger et al. VP ( D / 2) is the shear force on any abutment pile. and A is the area of the horizontal portion of the failure surface. the forces induced by the pile-pinning effect.1 Estimation of initial fundamental period of potential sliding mass (Ts). where the weighting factor would be the total weight of the respective slice.3) where N is the number of piles under the abutment. The terms VP ( D / 2) and FP ( D / 2) indicate that: (i) the shear force in the piles and the passive reaction at the abutment are a function of the lateral displacement of the abutment.

and A can be estimated. that the distance between the points of fixity is equal to the thickness of the critical layer. where EPile is Young’s modulus. I Pile is the moment of inertia of the pile. M ( D) = 6 EPile I Pile D / L2 Pile < plastic moment of pile.e. the function p( Su ) is roughly linear. i.e.. H is the total thickness of the critical layer. The shear force in any pile.2. then: N ⋅VP ( D / 2) FP ( D / 2) ⎞ ⎛ k y = p ⎜ Sur + + ⎟ A A ⎝ ⎠ (3.4) If this expression is replaced in Equation 3. M p the plastic max the total bending moment of each pile. The remainder of this subsection shows how the required terms p .e. H the thickness of the liquefiable layer. VP . between Smin and Smax . i. An upper limit for Su can be estimated considering that the piles have reached their plastic limit. and FP maximum passive reaction at the abutment. D is the relative lateral displacement between the ends of the pile.. The function p( Su ) can be estimated by solving the pseudostatic seismic stability problem at the abutment with different “reasonable” values of Su . i. k y = p ( Su ) . 2 R is the 24 . A reasonable lower limit for Su is the mean residual undrained shear strength of the liquefiable layer.Let p be a function that relates the value of the yield coefficient ( k y ) and the total equivalent undrained shear strength of the critical layer ( Su ). and LPile is the equivalent length of the pile. which needs to be solved iteratively. the result is an implicit equation for D . Preliminary results for a series of sensitivity analysis indicate that for cases when there is a distinct weak layer where sliding is localized. LPile = H + (α top + α bottom ) ⋅ 2 R . FP .5) where M ( D) is the bending moment at the ends of the pile. Su = Smin ≈ S ur . can be estimated assuming that the piles are fixed against rotation at some distance above and below the critical layer. VP . This idea follows the approach proposed in the document MCEER/ATC-49-1 (2003) where the assumed plastic mechanism for the piles considers the formation of plastic hinges at some distance above and below the liquefiable material.. in that case Su = Smax ≈ S ur + N × 2M p / ( H × A ) + FPmax / A . Let S ur be the mean residual undrained shear strength of the liquefiable layer. and that the passive reaction at the abutments has reached its maximum. Then. the shear force in any pile can be calculated as: VP ( D) = 2M ( D) H + ( α top + α bottom ) ⋅ 2 R (3.

e. Strictly speaking. The passive reaction at the abutment.8. conservative) approximation for the less common case of pile groups consisting of only inclined piles. they are a valid approximation for pile groups with only a few battered piles. the previous recommendations for the estimation of VP ( D ) are only valid for groups of vertical piles.2): ⎛ h ⎞ K abut = Ki × w × ⎜ ⎟ ⎝ 5.diameter of the pile. and α top and α bottom are factors that indicate how many diameters above and below the critical layer the piles are fixed against rotation.1 of Caltrans’ Seismic Design Criteria (2006). For the case of a thin nonliquefiable crust and pinned pile connection. where for seat abutments (see Fig. and they also work as a lower bound (i.0 ksf × h 5. respectively.5 Ae = h × w (area of the backwall) (3. 25 . can be estimated using the recommendations from Section 7. and pinned at the pile/pile-cap connection. In this simplified approach it has been assumed that α top ≈ α bottom = α . the shear force in any pile can be calculated as: VP ( D) = M ( D) H + α bottom ⋅ 2 R + d * (3. However.5 ⎠ kip / in Ki = 20 ft Pbw = Ae × 5. 3. FP ( D) .6) where d * is the vertical distance between the top of the liquefiable layer and the bottom of the pile cap. In such case..7) where w and h are the width and height of the backwall (in feet). a reasonable assumption would be to consider that the piles are fixed at a α bottom ⋅ 2 R below the liquefiable material (as before).

.2 Passive reaction versus deflection at abutments (after Caltrans 2006). (2006) have stated. The horizontal part of the base of the potential sliding surface can be assumed to be rectangular in shape. A = l × t . it is considered reasonable to use the abutment’s passive resistance without reductions. 3. Their recommendation of assuming that one half of the slide slope masses is restrained by the piles is adopted in the proposed procedure. because the overall size of the potential sliding mass varies as the equivalent undrained shear strength of the critical layer changes. it is important to note that the MCEER/ATC-49-1 (2003) procedure recommends reducing the prescribed passive capacity by 50%. However. assuming that the abutment fill has slumped “somewhat” due to the movement of the soil block. a “reasonable” factor according to this document. the value of l (from the 2D pseudostatic seismic stability problem at the abutment) should be considered. it might be argued that a reduction factor should be applied to account for the fact that this force usually acts well above the potential failure plane and is less effective in resisting the rotation of the soil mass. Nevertheless. a conservative approach that would lead to a smaller increase in resistance due to the pile-pinning effect would be for l to be for the case in which Su = Smax . the tributary width t will be greater than the abutment crest width because there will be some influence of the abutment side slope masses in the abutment response. a function of Su . In the context of a simplified approach like the one proposed in this study. as Boulanger et al. where l and t are its plan dimensions in the longitudinal and transverse directions. respectively. Also. Two effects should be considered in the estimation of l and t. 26 . Second. Because the size of the potential sliding mass usually increases as the value of Su increases.FP(D) D Fig. i.e. First. in general. it is still questionable how much this force should be modified in a more general case.

The uncertainty in Sur can be estimated if the probability density function (PDF) of the normalized standard penetration test blow count ( N1.60 has a lognormal distribution and a coefficient of variation (c.60 )dn1.) of 0. Five of the most commonly used procedures to estimate the residual undrained shear strength of the liquefiable layer. The characterization of the uncertainty of ε . by the total probability theorem: +∞ f Sur ( sur ) = −∞ ∫f Sur ( sur | n1. The parameters α top and α bottom are assumed to have a uniform distribution between 2 and 5.66 . The probability density function of the Sur / σ vc ' ratio given a corrected standard penetration test (SPT) blow-count ( N1.8) where f Sur ( sur ) is the resultant PDF of Sur . the variable ε is a normally distributed random variable with zero mean and standard deviation σ = 0.3 shows the 27 . From the model by Bray and Travasarou (2007).1. In the proposed approach it is assumed that N1.60 ) and the PDF of Sur for a given SPT value are assumed to be known.1.60 (n1. ε .60 (n1.3. are combined to estimate its mean and coefficient of variation. Figure 3.45 is given as the typical range for the coefficient of variation of N1. α top and α bottom .o.o. α top .60 ) are the given probability density functions. from the limits of the liquefiable layer to the points of fixity of the piles. Sur .2 Sources and Characterization of Uncertainty Three important parameters are treated as random variables in the proposed soil-foundation model used to estimate the engineering demand/response parameter: the error term from the model by Bray and Travasarou (2007). Sur . expressed in terms of number of pile diameters. and f N1.60 ) and f Sur ( sur | n1. the distance. In that case.15 to 0.3.60 (3.60 ) f N1. and α bottom is relatively straightforward. with the restriction that α top ⋅ 2 R cannot be greater than the vertical distance between the top of the liquefiable layer and the bottom of the respective pile cap. The selected c.60 . is based on Table 3 of Duncan (2000).60−CS ) is also assumed to be lognormal.v. It is important to note that above the liquefiable layer the assumption of fixity of the piles against rotation might be unconservative in situations where the upper crust is too thin or weak. wherein 0. and the residual undrained shear strength of the liquefiable layers.v.

28 . Preference was given to methods that estimate the ratio Sur / σ vc ' as opposed to just Sur and to the most recent method of Idriss and Boulanger (2007).5 ⎟ 0 ≤ N1. The result is shown in Figure 3.60−CS . respectively.4. which are shown in Figure 3. Olson and Stark (2002). 3.60 −CS ⎤ ⎦ = exp ⎜ 8 ⎝ ⎠ (3.60−CS proposed by Idriss and Boulanger (2007).mean estimates from the methods proposed by Seed and Harder (1990). and the correlations of Sur with N1. The relationship that was fitted to the combined results expressed by the black X’s. and the plus and minus one standard deviations of the models proposed by Olson and Stark (2002) and Kramer (2007).9) The upper and lower limits of the Seed and Harder (1990) relationship. These models were linearly combined using relative weights of 2:3:3:2:5. were used to estimate the coefficient of variation of this random variable.3 is: ⎛ N1. shown as the thick black line in Figure 3.60−CS .60−CS ≤ 20 E⎡ ⎣( Sur / σ vc ' ) | N1. to estimate the mean of ( Sur / σ vc ') | N1. and Sur / σ vc ' with N1.60−CS ⎞ − 3.3 Mean estimates of Sur / σ vc ' ratios from different methods. Fig.3 with black X's. and Kramer (2007).

i.4.e.v.o. the second-order approximation for the first moment of ( Sur / σ vc ') | N1. 3. c.. If it is assumed that ( Sur / σ vc ') | N1.o. of Sur / σ vc ' given N1. ⎡ ⎣( Sur / σ vc ') | N1. and that cov N1.40 0 ≤ N1.40 (constant).10) The result of this assumption is shown with thick black lines in Figure 3. Based on these results.v. Note that every pair of curves is shifted with respect to the others because every pair is centered on different mean values.8 and 3.60−CS ⎤ ⎦ = 0.9. shown in Equations 3.60−CS has the mean and c.60−CS ≤ 20 (3.v.Fig. a reasonable assumption is to consider that the c.60−CS is equal to 0.3 (constant).60−CS ( Sur / σ vc ') : gives the following recommended mean and standard deviation for 29 .60 = 0.o.4 Dispersion in estimates of Sur / σ vc ' ratios from different methods.

70 ⋅ ln (σ v ' ) ⎟⎟ ⎜ +0.13) 2 3 ⎛ q ⎞ ⎛ ⎛ q ⎞ ⎛ q ⎞ ⎛ q ⎞⎞ = exp ⎜ −⎜ +⎜ − 4. The probability of occurrence of liquefaction given an intensity measure.5 ⎝ 61. CSR. and D50 is the particle size that 50% passes by weight.11) If information from cone penetration tests (CPTs) is provided.60 ⋅ (1 + 0. one could use expressions that directly estimate Sur from CTP results.70 ⎟ ⎟ ⎟ ⎠ 30 (3.97 ⎟⎟ ⎝ ⎠⎟ ⎟ 2.μS ur / σ vc ' σS ur / σ vc ' ⎛ 0. (2003): ⎛ ⎜ ⎜ ⎜ PL ( N1. Alternatively.004 FC ) − 13.42 ⎟ × ⎜1 + exp ⎜ − 9.5 MPa) (3. can be estimated making use of the expression developed by Seed et al.03 (for qc1 ≤ 6. M w .32 ⋅ ln(CSR) − ⎞ ⎞ ⎜ ⎟⎟ ⎜ 29. and φ ' is the effective stress friction angle.53 ⋅ ln ( M w ) − 3. FC ) = Φ ⎜ − ⎜ ⎜ ⎜ ⎜ ⎝ ⎛ N1.60−CS ⎞ ⎜ ≈ exp ⎜ − × + 3.13) and (3. N is the measured SPT value.0143 ( qc1 ) ± 0.05 ⋅ FC + 44.3μ N1.12) where qc is tip resistance.44 ( D50 ) N (3.60 .26 = 5.14) where σ v ' is the effective overburden stress.1 ⎠⎠ ⎝ ⎠ (3. σ v '. Pr( L | im) . Let L be the event that liquefaction occurs. the following CPT-SPT correlation proposed by Kulhawy and Mayne (1990) may be used: ( qc Pa ) 0. Pa is atmospheric pressure (~100 kPa). for this purpose: Sur σv ' Sur = 0. respectively. and L be the event that liquefaction does not occur. q ≡ qc1Ncs-Sr is the equivalent clean-sand CPT normalized corrected tip resistance.7 ⎠ ⎝ 106 ⎠ ⎟ ⎝ σv ' ⎝ 11. Equations (3.5 1 ⎟ ⎜ 8 ⎟ ⎜ 128 ⎝ ⎠ ⎜ ⎝ ≈ 0.03 + 0.82 ⎟ ⎟ ≤ tan φ ' ⎟ ⎟ ⎜ 24.14) were developed by Olson and Stark (2002) and Idriss and Boulanger (2007). qc1 is the normalized tip resistance (MPa). The last source of uncertainty in the proposed model is the occurrence of liquefaction.15) .60−CS ⎛ μ N1.4 μ Sur / σ vc ' ( ) 2 ⎞ ⎟ ⎟ ⎟ ⎠ (3.

1 = qc ⋅ Cq . CSR is the equivalent uniform cyclic stress ratio. FC is the percent fines content (by dry weight) expressed as an integer (i.848ln ( M w ) − 0. σ v ' c is the effective overburden stress (kPa).60 1. CSR = 0.60 ) .045 + qc .60 ) f N1.60 )dn1.60 (n1.60 . i. and Φ is the cumulative normal distribution.60 | im)dn1.110 ⋅ R f ) + c (1 + 0. where amax is the peak horizontal ground surface acceleration. the probability of liquefaction occurrence given an intensity measure can be estimated using the following expression (Moss et al.60 (3.60 . R f = ( f s qc ) ⋅100 is the friction ratio.g.1 ( 0. 12% fines is expressed as FC = 12 ). qc .60 is the normalized standard penetration test blow-count. Cq = ( Pa σ v ') is the tip normalization factor.110 ⋅ R f ) + ( 0. Equation (3.17) (n1.e. where qc is the raw tip resistance (MPa). Pa is the reference stress (~100 kPa).632 ⎞⎞ ⎟⎟ ⎟ ⎠⎟ ⎟ ⎟ ⎟ ⎟ ⎠ (3. The total probability theorem can be used to estimate the probability of liquefaction given an intensity measure through: +∞ Pr( L | im) = = −∞ +∞ ∫ Pr( L | im..60 (n1.15) provides an estimation of the probability of liquefaction given the values of N1. If the values of σ v ' and FC are assumed to be deterministic. CSR is the cyclic stress ratio.where PL is the probability that liquefaction occurs. 2005): ⎛ ⎜ ⎜ PL = Φ ⎜ − ⎜ ⎜ ⎜ ⎝ ⎛ qc . M w is moment magnitude.60 31 .00).1 is the normalized tip resistance (MPa). and rd is the nonlinear shear mass participation factor.60 ) be the PDF of the random variable N1. n ∫ Pr( L | im.35 and 1. σ 'v is the effective vertical stress.002 ln (σ v ' ) − 20. If CPT results are available. CSR . and Φ is the standard normal cumulative distribution function (e.11.. σ v is the total vertical stress. g is the acceleration of gravity. so this relation can provide an estimate of Pr( L | im. M w is the moment magnitude. the NORMSDIST function in Excel). N1.. n 0 1. and c is the normalization exponent (a value typically between 0.60 ) f N1.65 ( amax / g )(σ v / σ 'v ) rd .16) where qc .e. the sleeve to tip resistance expressed as percentage.177 ln ( CSR ) − 0. The values of CSR and M w are part of the earthquake ground motion intensity measures. Let f N1.923 1. n1. and M w .850 ⋅ R f ) + ⎜ ⎜ ⎝ −7.

14 is obtained by transforming the set of random variables x to standard normal space u .2). L) or Pr( D > d | im. where x represents a vector of random variables. D . and by linearizing the limit-state function in the standard normal space G (u ) at 32 .60 is independent of the intensity measure ( im ). the limit-state function of the problem is g ( x) = d − D ( x) . several methods are available to numerically calculate the required probabilities. and ε . Two of these methods are appropriate in the context of the proposed simplified design procedure: the first-order reliability method (FORM) and the point estimate method (PEM). an approximation to the probability integral in Equation 3. where “failure” is defined as the event { D > d } given an intensity measure and that liquefaction has (or has not) occurred. The vector of random variables in this case is x = ⎡ ⎣Sur α ε ⎤ ⎦ . FORM. and the random variables that have been selected and characterized are: Sur .19) where f ( x ) is the joint probability density function (PDF) of x . Given Intensity Measure At this point a pseudoprobabilistic model to estimate the residual lateral seismic displacement. the evaluation of Pr( D > d | im. and the probability of failure is defined as a three-fold integral: pf = Ω ∫ f ( x)d x (3.where it has been assumed that N1. and Ω ≡ { g ( x ) ≤ 0} . The probability that liquefaction does not occur given an intensity measure can be calculated as: Pr( L | im) = 1 − Pr( L | im) (3. because for most nontrivial selections of f ( x ) and Ω.1. and an associated set of random variables have been identified. In FORM. L) .18) 3. L) is thought of as the estimation of the probability of failure of a component. The evaluation of this integral is usually difficult. α . In reliability theory terms. In the first of these two methods. Once these two components of the problem are defined. and D ( x ) comes from T Equation (3. The pseudoprobabilistic model is a modification of the one developed by Bray and Travasarou (2007).1. L) or Pr( D > d | im.3 Procedures to Calculate Probability of Exceeding Engineering Parameter Threshold. no closed-form solution of the integral exists. Pr( D > d | im.

ce. The second method.e. 2006. PEM. and Y a deterministic scalar function of X . This method indicates that if X =⎣ ⎡ X1 X2 Xn ⎦ ⎤ T is a vector of n random variables.. Preliminary results indicate that FORM gives reasonably accurate results when compared to Monte Carlo Simulations or the results from second-order reliability methods (SORM)..berkeley. an assumption must be made about the PDF of D ( x ) .20) where ϕn (u ) is the multivariate standard normal probability density function. e.21) In essence.g. at least in the neighborhood of the optimal point. The first-order approximation of the failure probability is given by the probability content of the halfspace in the standard normal space.24) 33 . http://www. No information is required (nor obtained) about the probability density function of the different parameters. Y = g (X) . For a detailed explanation of PEM.an optimal point u * . lognormal. The program FERUM was used to evaluate the required probability Pr ( g ( x ) ≤ 0 ) using FORM (Der Kiureghian et al. Then: pf = g ( x )≤0 ∫ f ( x)d x = G (u ) ≤ 0 ∫ ϕn (u )du (3. and G (u ) ≈ G1 (u ) = ∇G (u*)(u − u*) = ∇G (u*) ( β − α u ) u* = arg min { u | G (u ) = 0} (3. was first proposed by Rosenblueth in 1975.22) where Φ is the standard normal cumulative density function. The fundamental assumption is that the limit-state function is continuous and differentiable. This means that to estimate the required probabilities. p f ≈ Φ (− β ) (3. which is completely defined by the parameter β . the reader is referred to Christian and Baecher (1999).23) where m is the moment order and: ⎧ ⎪+ if X i = μ X i + σ X i si = ⎨ ⎪ ⎩ − if X i = μ X i − σ X i (3. the low-order moments of fY ( y ) can be estimated as: E[Y m ] = ∑P s1s2 sn ( ys1s2 sn )m (3.edu/~haukaas/ FERUM/ferum. FORM replaces the failure domain G (u ) ≤ 0 by the halfspace β − α u ≤ 0 . i.html). With PEM the first moments of the function D ( x ) are computed in terms of the first few moments of the random variable x .

μ X 3 + σ X 3 ) + + σ X1 . μ X 2 + σ X 2 . μ X 2 − σ X 2 . μ X 2 − σ X 2 . μ X 2 − σ X 2 . μ X 2 − σ X 2 . μ X 2 − σ X 2 . then: D = PL DL + (1 − PL ) DNL It can be shown that in this case the expectation of D and D 2 are: E[ D ] = PL E[ DL ] + (1 − PL ) E[ DNL ] E[ D 2 ] = PL 2 E[ DL 2 ] + (1 − PL ) 2 E[ DNL 2 ] + 2 PL (1 − PL ) E[ DL ]E[ DNL ] (3.25) (3.27 and 3. 34 . μ X 2 + σ X 2 .28) Equation 3. For the bridge problem discussed in this report.27) ⎡Y 2 ( μ X 1 ⎢ 2 ⎢ 1 Y ( μ X1 E[Y 2 ] = ⎢ 8 ⎢Y 2 ( μ X 1 ⎢ 2 ⎢Y ( μ X1 ⎣ ⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ + σ X 1 . μ X 3 + σ X 3 ) + Y ( μ X1 − σ X1 . μ X 2 + σ X 2 . μ X 3 − σ X 3 ) + − σ X1 .18 reduces to: ⎡Y ( μ X1 ⎢ 1 ⎢Y ( μ X1 E[Y ] = ⎢ 8 ⎢Y ( μ X1 ⎢Y ( μ X1 ⎣ + σ X1 .30) from where μ D = E [ D ] and σ D = E[ D 2 ] − E 2 [ D] . μ X 2 − σ X 2 . μ X 3 + σ X 3 ) + + σ X 1 .Ps1s2 sn = 1 ⎡ ⎢1 + 2n ⎢ ⎣ sn ∑∑ i =1 n −1 ⎤ si s j ρij ⎥ ⎥ j =i +1 ⎦ n sn (3.28 as σ Y = E[Y 2 ] − E 2 [Y ] . μ X 3 − σ X 3 ) + + σ X 1 . μ X 3 − σ X 3 ) + + σ X1 .27 gives an estimation of the mean of Y . respectively. the estimated level of displacement D can be written as a linear combination of the displacements that would occur with and without liquefaction for a given intensity measure of the ground motion. μ X 3 + σ X 3 ) + Y 2 ( μ X1 − σ X 1 . μ X 3 + σ X 3 ) + Y ( μ X1 − σ X1 . μ X 2 + σ X 2 . μ X 3 − σ X 3 ) (3. μ X 3 − σ X 3 ) + Y 2 ( μ X1 − σ X1 . μ X 2 + σ X 2 .26) ys1s2 = Y (X s1s2 ) For the case of three uncorrelated random variables Equation 3. μ X 2 + σ X 2 . μ X 2 + σ X 2 .29) (3. Its standard deviation can be estimated from the results of Equations 3. and DL and DNL be the displacements with and without liquefaction. μ X 3 + σ X 3 ) + Y 2 ( μ X 1 + σ X1 . μ X 3 − σ X 3 ) + − σ X1 . μ X 3 − σ X 3 ) ⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦ (3. μ X 3 − σ X 3 ) + Y ( μ X1 − σ X 1 . μ X 3 + σ X 3 ) + Y ( μ X1 + σ X1 . μ X 2 − σ X 2 . μ X 2 + σ X 2 . μ X 2 − σ X 2 . μ X 3 + σ X 3 ) + Y 2 ( μ X1 − σ X1 . Let PL be the probability of the occurrence of liquefaction.

3.1 Decoupled Models In this section. and ultimately it will influence the amount of drift that the bridge columns will experience and hence. but it will also affect the lateral displacements and rotations of the other pile caps. The stiffness of these springs is based on y50 which is the displacement that corresponds to half of the ultimate passive resistance of the 35 .2 DERIVED ENGINEERING DEMAND PARAMETERS The amount of lateral displacement at the ends of the bridge will directly affect the local response of the abutments and their foundation. a fraction of that lateral movement will get transferred to the next pile caps through the compression of the non-liquefiable crust. The adopted concept is that as the embankments displace toward the center of the bridge. three simple structural models are proposed to estimate the lateral displacement of the intermediate piers’ pile caps.5).2. Then. 3. the pile-cap rotation. and the effect that these displacements and rotations will have on column drift. the abutments’ pile caps will move in concert with the upper non-liquefiable crust. The stiffness k of the springs used in this model can be estimated considering the pile cap passive earth pressure spring model (Mokwa 1999).5 Structural model to estimate lateral displacement of intermediate piers’ pile caps from displacement of abutment pile caps. overall bridge damage. 3. 3. A simple structural model is proposed to estimate the lateral displacement of the intermediate piers’ pile caps from the displacement of the abutments’ pile caps (see Fig. Fig.

25 + 2α ⎟ ⎜4+ Su b ⎝ ⎠ (3. each column of this model represents one row of piles.e.5 ⋅ (2 R). The pile cap is assumed to be infinitely rigid compared to the piles.cap ). L H Fig.6 Structural model to estimate pile-cap rotation as a function of lateral deformation. 3. so their area and moment of inertia need to be specified accordingly.32) where the factors of 3.cap is estimated using the φ = 0 sliding wedge method (Mokwa 1999): Pult. Figure 3. H eq ≈ H Liquefiable Layer + 3. the height of the columns needs to be consistent to what has been assumed previously.31) Pile-cap rotation can have an important effect on column drift. and d * is the vertical distance between the top of the liquefiable layer and the bottom of the respective pile cap. d *} (3. i.6 shows the simple structural model that is proposed to estimate pile-cap rotation as a function of its lateral deformation.5 correspond to the mean of α top or α bottom . the relationship between pilecap rotation ( θ ) and lateral displacement (D) is: 36 . finally. R is pile radius.cap = Su bH 2 ⎛ ⎞ H γH + 0.5 ⋅ (2 R) + Min{3..pile cap ( Pult. The value of Pult. If the response of the piles is assumed to be linear elastic.

. For other situations (e. it is proposed to use a structural model as shown in Figure 3. 37 . respectively.6. To evaluate the potential effects that foundation deformations may have on bridge column drifts. programs such as SAP2000.8 Schematic representation of simplified coupled model.33) where H and L are defined in Figure 3. Fig. The inputs for this model are the abutments’ lateral displacements applied at both ends of the bridge. and APile and I Pile are. or nonlinear behavior of piles). flexible pile cap.7. Fig. the area and moment of inertia of one pile. OpenSees.7 Simplified 2D structural model. and the lateral displacements and rotations of the intermediate pile caps. Given the potential levels of deformations involved. 3.θ= D clockwise ⎛ APile L ⎞ H + 16 ⎟ ⎜ ⎜ I Pile ⎟ ⎝ ⎠ 24 2 ⋅ (3.2. 3.2 Coupled Model Figure 3. the model used to represent the superstructure’s elements has to capture geometric and material nonlinearity. 3.8 shows a simple structural model that is proposed to estimate column drift given the amount of residual lateral displacement at the embankments of the bridge.g. and RISA can be used.

The stiffness of these springs can be estimated considering the pile cap passive earth pressure spring model (Mokwa 1999). and between the embankment and the first pile cap. because there are two springs acting in series between the pile caps. between pile caps. through the compression of the upper non-liquefiable layer.8 try to model the interaction that occurs.In this model. where Hliq is the thickness of the liquefiable material at the location of the respective pier. and R is the pile radius. The springs shown in Figure 3. the equivalent length (Leq) of the piles tries to represent the lateral stiffness they provide to the system.5(2R). 38 . Leq =Hliq+2×3. It is proposed to consider that the stiffness of the intermediate springs is half of the stiffness of the end springs.

1 shows. and backwall failure. As Table 4. columns. and engineering judgment.25 0. backwall spalling.5 in (0. and piles. Kevin Mackie.1 DAMAGE TO ABUTMENTS Four damage states (repair decisions) were identified for the abutments: cleaning.11 m) 5.1 is the graphical representation of the values given in Table 4.25 0. Figure 4. these states were considered to be a function of the residual relative lateral displacement between the abutment and deck structure.05 m) 4 in (0. Table 4. on discussions with other engineers.14 m) Dispersion Log-units 0.4 Probabilistic Evaluation of Bridge Damage Based on the assumptions made by Dr.30 0.10 m) 4.1. assembly. Damage State/ Repair Decision Cleaning Assembly Backwall spalling Backwall failure Median Residual Lateral Deformation 2 in (0.40 . the following models have been adopted for the estimation of damage (or repair decisions) to the abutments. 4.3 in (0.1 Abutment damage states as function of residual relative lateral deformation in longitudinal direction.

and allow failure.72% Dispersion Log-units 0.40 0.1 Abutment damage states as function of residual relative lateral deformation in longitudinal direction. thicken pier.2 DAMAGE TO COLUMNS Four repair decisions have been identified for the piers: do nothing.40 0.25% 2. Figure 4. re-center column.2 is the graphical representation of the values given in Table 4. Repair Decisions Do nothing Thicken pier Re-center column Failure Median Residual Tangential Drift 0. 4.2.50% 1.00% 6.30 0.35 40 .Fig. the repair decisions were considered to be a function of the residual tangential drift of the columns. As Table 4. Table 4.2 shows. 4.2 Column repair decisions as function of residual tangential drift.

If the plastic rotation in the pile is calculated to be greater than 0. the expected damage state for the pile would be estimated to be moderate. the repair decisions are a function of the pile cap lateral displacement and the thickness of the potentially liquefiable layer. Kevin Mackie for the assessment of pile damage.3 shows. the expected level of damage to the pile would be estimated to be negligible or small. large. and enlarge and add piles. 41 . analytical models or engineering judgment based on recommendations from a panel of experts are the only ways to estimate the amount of damage induced in a pile as a function of pile deformation or distortion. Two repair decisions have been identified for the piles: add pile. Using the analytical approach.2 Column repair decisions as function of residual tangential drift. Figure 4. Another alternative is to make use of the estimations made by Dr. or collapse.4 (log-units) has been assumed in this case. 4.3 is the graphical representation of the values given in Table 4.3. the most straightforward estimation would be to utilize the allowable limiting plastic rotation of 0. Consequently.05 radians.3 DAMAGE TO PILES Without sufficient case histories and experimental data. As Table 4.Fig. A constant dispersion of 0.05 radians. if the plastic rotation in the pile is calculated to be less than 0. 4.05 radians that is given in the MCEER/ATC-49-1 (2003) document.

3 Piles damage states as function of pile cap lateral displacement.Table 4.7 m) 31 in (0.4 m) Fig. 4.8 m) 31 in (0.7 m) 47 in (1.2 m) 55 in (1. 42 .3 m) 12 in (0.5 m) 28 in (0. Piles Left Abutment Column 1 Column 2 Column 3 Column 4 Right Abutment Median displacement required to “add piles” 12 in (0.8 m) Median displacement required to “enlarge and add piles” 20 in (0.3 m) 16 in (0.3 Piles damage states as function of pile cap lateral displacement.5 m) 20 in (0.4 m) 55 in (1.4 m) 28 in (0.

and collapse (see Table 5. which is modified to reflect the restraint provided by the bridge deck and pile-pinning effect (when it is appropriate to do so). Seismic Displacement (in. has the primary effect on the overall performance of the entire bridge system.1 5. Figure 5.) 0 – 1” 1” – 4” 4” – 20” 20” – 80” > 80” Damage Level Negligible Small Moderate Large Collapse .1 Bridge damage as a function of residual lateral displacement. small. it is assumed that the amount of seismically induced permanent lateral displacement (D) at the bridge abutments. Table 5.5 Probabilistic Evaluation of Consequences 5. moderate. large.1. Depending on the level of seismically induced residual lateral displacement calculated at the bridge abutments.1 shows the resultant fragility curves for these damage states.1 REPAIR COSTS Simplified Approach In this simplified approach. it is proposed to consider that the bridge can reach five potential levels of damage: negligible.1). given a seismic displacement.

it is assumed that there is a certain probability of reaching specific repair cost ratios. Moderate. Small.2). That is why it is recommended to apply this simplified approach considering at least two different levels of uncertainty. The resultant fragility curves for Repair Cost Ratios are shown in Figure 5.Fig. which for this case are categorized as low (gray bars/continuous line) and high (black bars/dashed lines). Each subplot corresponds to one specific repair cost ratio. For every damage state.3.1 (see Table 5. and the probability of exceeding a specific repair cost ratio for a given damage state is provided along the y-axis. where the five damage states (Negligible. 5. The potential repair costs of a bridge are a function of the level of damage the bridge has experienced.2 and Fig. 44 . 5. The level of uncertainty in the estimation of repair cost ratios as a function of damage state may have an important effect on the results of the proposed approach. Large. which have been discretized in bins of 0.1 Bridge damage as function of residual lateral displacement. which has been subdivided into eleven subplots for the sake of clarity. and Collapse) are listed across the x-axis.

2 Repair cost ratio distribution given damage state considering low and high uncertainty (values in parenthesis correspond to high uncertainty).25) 0 (0.20 (0.15 (0.20) 0.05) 0 0 Collapse 0 0 0 0 0 (0.1| ds ) Small 1 (1) 0 0 0 0 0 0 0 0 0 0 Moderate 0 1 (0.05) 0 (0.05) 1 (1) 0 0 0 0 0 0 0 0 0 0 Pr( RCR = 0.15) 0.8 | ds ) Pr( RCR = 0.9 | ds ) Pr( RCR = 1.05) 0.5 | ds ) Pr( RCR = 0.15 (0.4 | ds ) Pr( RCR = 0. Damage State Negligible Pr( RCR = 0.30) 0.0 | ds ) Pr( RCR = 0.25) 0.3 | ds) Pr( RCR = 0.1) 0 (0.70 (0.20 (0.2 | ds ) Pr( RCR = 0.95) 0 (0.0 | ds ) 45 .7 | ds ) Pr( RCR = 0.55 (0.20) 0.6 | ds ) Pr( RCR = 0.05) 0 0 0 0 0 0 0 0 Large 0 0 0 0 (0.1) 0 (0.05 (0.Table 5.25) 0.

5.Fig.2 Repair cost ratio distribution given damage state considering low (gray bars) and high (black bars) uncertainty. 46 .

2.3 Repair cost ratio cumulative distribution given damage state considering low (continuous line) and high (dashed lines) uncertainty. (2007) is selected to estimate the total repair cost ratio from the repair cost of different components of the bridge. and (3) a cost model relating damage states to the cost of available repair methods. 5. In this simplified approach. (2) a probabilistic damage model relating bridge demand to discrete damage states of key bridge components. and each repair quantity has a unit cost that is used to relate DVs to DMs.2 Component-Based Approach The method proposed by Mackie et al. Damage states are defined for 19 bridge components to relate DMs to EDPs.” This tool uses three probabilistic models to evaluate repair costs based on the Pacific Earthquake Engineering Research Center’s performance-based earthquake engineering framework: (1) a probabilistic seismic demand model relating earthquake intensity measure to bridge response parameters.1. Each damage state is associated with several repair quantities. 5. the results from the coupled model (see Section 3.Fig. Repair cost 47 .2) are used to estimate EDPs versus IM relationships. In this method the repair cost probability distribution is derived using first and second moments from a simple graphical tool called “Fourway.

the Fourway method still provides an exact value for the expected loss and a variance approximation. and information based on case histories of Caltrans bridges in previous earthquakes. In order to completely define the Fourway process. damage. (2007) is taken from different sources including Caltrans Construction Statistics. However. Performance groups are defined in terms of different EDPs to quantify damage to different bridge components. the Fourway method is based on the first and second (central) probabilistic moments and is a generalization of the closed-form solution to a single DV-DM-EDP-IM analysis. and loss models in order to obtain a relationship between intensity and loss. the expected loss and accompanying variance can be obtained in closed form. total repair costs (TC) are derived from a relationship between the total repair quantities (Q) and a unit cost (C) for the repair items. Three specific conditions must be valid: (1) the distribution of EDP is lognormal when conditioned on IM (and so forth for DM and DV when conditioned on EDP and DM. (2007) used repair quantity (Q) as the DV. The loss model by Mackie et al. The two probabilistic moments derived from the method can be used as lognormal or normal distribution parameters to describe both the probability of exceeding different repair costs at a given IM. even in the case of arbitrary conditional probability distributions and non-power-law median relationships. The repair quantities are combined between performance groups taking into account the correlation between performance groups. respectively). As indicated before.data in Mackie et al. The simplified Fourway solution requires certain conditions on each of the demand. the seismic demand model is formulated from simplified coupled analysis of the example bridge. Caltrans Bridge Design Aids. damage. and loss models must be formulated. When these three assumptions hold. According to Mackie et al. 48 . (2007). Each Fourway analysis predicts the repair quantity distribution (based on the moments of the distribution) associated with each performance group and repair method. as well as the probability of exceeding a single repair cost at varying levels of intensity. seismic demand. (2) the conditional dispersion of EDP given IM is constant across the range of intensities considered. A single Fourway analysis is performed for each performance group in the bridge and repair quantity. Finally. Uncertainty in the repair quantity and unit cost models were incorporated into the analysis. and (3) that the median relationships in each probabilistic model follow a power-law relationship.

1.5 to 1.2. the median downtime would be 60 days. the median downtime of the bridge would be 1 day.3 Expected downtimes given bridge damage level. (2007) (see Section 5.e. Damage Level Negligible Small Moderate Large Collapse Median downtime (and likely range) 0 0 0 1day (0.56 . c..v.2 Component-Based Approach This approach follows the same ideas described in Mackie et al. and that if collapse occurs. no downtime will occur. but where repair cost ratios have been replace by repair times.2. Table 5. it is assumed that if the damage state of the bridge is moderate or less. 49 .2 5. that if large damage occurs. i. = 42 days / 75 days ≈ 0.1 DOWNTIMES Simplified Approach In the simplified approach.5 days) 60 days (25 to 95 days) 5.5. The uncertainty around the median number of days has been assumed to be similar to the coefficient of variation of the HAZUS then: estimates for downtimes (add reference).o.2).

The soil below the left abutment consists of a medium stiff clay crust underlain by a thin. The pier columns are 4 ft in diameter with a 2% longitudinal steel reinforcing ratio.1 Bridge Evaluation Example GENERAL The methodology that has been presented in the previous sections will be applied to the general bridge example described in Figures 6.. The groundwater is located at the bottom of the surface clay layer. and it does not exist below the right embankment. triggered broad bridge damage. and yield strength of 60 ksi. The bridge consists of a five-span reinforced concrete structure with a post-tensioned reinforced concrete box girder deck section and monolithic piers. a layer of stiff clay. The deck is 6 ft deep. and a dense sand layer underlain by rock. and the four piers are 22 ft high. followed by a dense sand layer underlain by rock. The lower clay layer below the left abutment becomes thinner toward the center of the bridge.1–6. loose to medium dense sand. The bridge that has been selected for the analysis is a typical Caltrans highway bridge underlain by liquefiable soil susceptible to lateral ground displacement.5 to illustrate how the proposed methodology can be applied in engineering practice. The soil beneath the right abutment consists of the same clay crust underlain by a thicker layer of the loose sand. The properties of the loose and medium sand layers across the bridge were aimed to induce liquefaction under moderate ground shaking so that lateral ground displacement. The same type and number of piles was used at the abutments but distributed in only one row. thickness of 0. . especially in the vicinity of the right abutment. The embankments are 28 ft high and have 2:1 (H:V) slopes.5 in. Each bridge pier is supported by a group of 3×2 open-ended steel piles with a diameter of 2 ft. The three middles spans are 150 ft long and the two end spans are 120 ft long.6 6.

fy = 60ksi) L=70 feet Fig. 6.2 Piles at interior bents—dimensions and properties. fy = 60ksi) Fig. 4D Group of 6×1 circular open-ended steel piles (PP 24×0.1 Example bridge profile.Fig.5. 52 . 6. 3D 3D L=60 feet Group of 3×2 circular open-ended steel piles (PP 24×0.3 Piles at abutments—dimensions and properties.5. 6.

53 . 2.5 Right abutment—soil profile and dimensions (not to scale). 6.4 Left abutment—soil profile and dimensions (not to scale).= 0.v. Fig. There is only one critical sliding surface at each abutment. the SPT blow-count at the critical layer of the left abutment has a lognormal distribution with mean 15.3.o.2 ASSUMPTIONS 1. 6. and a standard deviation of 4. At the left abutment.5.60−CS = 15 . 6. Since it has been assumed that c. the slope stability analysis performed using SLOPE/W (GEOSLOPE/W 2004) shows that the base of the potential failure surface goes through the bottom of the layer with an SPT blow-count of N1.Fig.

assuming a c. the SPT blow-count at the equivalent critical layer of the right abutment has a lognormal distribution with mean 13.3 6. N1.o. (1997). The de-aggregation of the hazard at a period of 1 second shows that at all three hazard levels.= 0. The two layers have the same SPT blow-count.60−CS ⎛ μ N1.4 μ Sur / σ vc ' μS ur / σ vc ' ( ) 2 ⎞ ⎟ ⎟ ⎟ ⎠ (6.3μ N1. i.60−CS has a lognormal distribution with mean and standard deviation given by ⎛ 0. At the right abutment. Two liquefiable layers are involved in the potential failure surface. Again.1. the slope stability analysis shows that the base of the potential failure surface goes through the bottom of bottom layer with an SPT blow-count of N1.1 CALCULATIONS Seismic Hazard The 5%-damping uniform hazard spectra for a free-field soil condition from Somerville and Collins (2002) was used in this analysis.3. These spectra were derived from a probabilistic seismic hazard analysis using the ground motion model of Abrahamson and Silva (1997) for a site in Oakland.60−CS = 13 .9.60−CS ⎞ ⎜ 3.3. Based on the results of Section 3. The Hayward fault is a strike-slip fault that has the potential to generate earthquakes having magnitudes as large as 7. the 54 . 4.3. 10%. the hazard is dominated by earthquakes on the Hayward fault.1) σS ur / σ vc ' 6. According to Somerville and Collins (2002).e. and a standard deviation of 3. which is located about 7 km east of the site.2 assume that the undrained residual shear strength to effective vertical stress ratio ( Sur | σ vc ' ) given that a value of μ N1.1. Only the strike-normal component (parallel to the longitudinal direction of the bridge) is considered in this 2D analysis. The hazard levels considered by Somerville and Collins (2002) were those of events with probabilities of exceedance of 50%.. which were represented in the probabilistic ground motion hazard analysis using the empirical model of Somerville et al. California. and 2% in 50 years.60−CS = 13 .5 1 ≈ exp ⎜ − × + ⎟ ⎜ 8 ⎟ ⎜ 128 ⎝ ⎠ ⎜ ⎝ ≈ 0. it contains rupture directivity effects.v.

As the next sections of this report will show. where TR is the return period.95 = 1. Table 6. ≈ −0.458 ⇒ TR ≈ 29 years (6.5 / 6. The uniform hazard spectra for the smallest hazard level was estimated by making a linear extrapolation in the Sa ( soil ) / MSF versus ln (TR ) space. and 2% in 50 years.3776 + 01411 .largest contributions come from events with magnitudes 6. a peak ground acceleration of 0.6 show the 5%-damping uniform hazard spectra for the free-field soil condition and the four hazard levels considered in the analysis. ⋅ ln (TR ) 1. which would not likely cause liquefaction. for every period of vibration. and the Sa / MSF ratio for each period was estimated assuming TR = 29 years (82% in 50 years event). 6. Assuming that for this case M w ≈ 6.6) 2.2) Following the procedure given in Youd et al. 10%.1411 ⋅ ln (TR ) MSF (6.6 .3) A similar linear regression was performed for each period of vibration. For PGA (soil) the result of the linear regression was: PGA( soil ) ≈ −0. respectively. 55 .0 for the hazard levels corresponding to events with probabilities of exceedance of 50%. and 7. Then. the probability of liquefaction for the three hazard levels analyzed by Somerville and Collins (2002) is high (> 80%). then MSF ≈ (7.1 and Figure 6.458 and 014 . a fourth hazard level. (2001).3776 + 0.6.8. was added for this design example to illustrate the effects that liquefaction has on the likely seismic performance of the bridge.14g would not cause liquefaction at the site.

528 0.847 0.70 1.085 56 10% 0.261 0.00 82% 0. at site.883 0.376 0.917 0.513 0.642 1.717 0.896 1.00 3. 5% damping.071 1.201 0.096 0.859 0.1 Uniform hazard spectra.316 1.142 0. 6.560 1.523 1.00 4.226 0.294 1.698 0.533 . Probability of Exceedance in 50 years Period (s) 0.094 0.275 2% 0.596 0.30 0.594 1.820 2.134 0.335 2.6 Uniform hazard spectra.50 0.337 0.058 0 0 0 50% 0.499 0. Table 6.312 2.683 0.20 0.00 2.03 0. at site. 5% damping.10 0.350 0.301 0.718 1.002 1.075 0.449 0.310 0.Fig.

for the 82% in 50 years event. Fig. On the other hand.. 6.3 Flow Failure Assessment The next step is to evaluate if flow failure is likely to occur at any of the abutments.6.2 Liquefaction Assessment The procedure recommended in Youd et al. 6.7 shows the estimated factors of safety against liquefaction for the two abutments.e. which indicates that chances of liquefaction for this event are low. Figure 6. (2001) was used to estimate the factor of safety against liquefaction of the different sublayers of the loose sand layers present at both ends of the bridge.3. The program SLOPE/W (GEO-SLOPE 2004) was used to perform the slope stability analyses. the factor of safety against liquefaction is rather high (~2 or higher). where the expected 57 . and the four hazard levels considered. This figure shows that for the 50%.3. to estimate the post-liquefaction static factor of safety at the abutments. and 2% in 50-years events.7 Factors of safety against liquefaction. 10%. i. liquefaction is (very) likely to occur at both ends of the bridge.

⎛ N1.8 and 6. indicating that in this case flow failure is not likely to occur. where it can be seen that the static factor of safety at both abutments is larger than 1.e.37 .60−CS ≤ 20 .9 show the results of the post-liquefaction stability analysis.4) Figures 6.8 Post-liquefaction stability analysis of left abutment ( FS = 1.5 ⎟ if 0 ≤ N1.11 ). i. Fig.residual undrained shear strength was assigned to the sublayers of loose sand where N1. 6. k y = 0..60 −CS ⎞ Sur ≈ exp ⎜ − 3. 58 .60 −CS ≤ 20 8 ⎝ ⎠ (6.

055 ).21 . Soil (right ) = 567 kips .4 Passive Soil Forces versus Ultimate Structural Resistance 6.Fig. and PLiqLayer = 328 kips for the left and right ends of the bridge. PPileCap = 213 kips .9 Post-liquefaction stability analysis of right abutment ( FS = 1.xls developed by Professors Mokwa and Duncan (which is provided in Appendix B) was used to estimate the passive capacity of the pile cap. 6. Then. Soil (left ) = 319 kips and Pult .4.3. 59 . The result of this estimation is PNonLiqLayer = 26 kips .1 Intermediate Piers The Excel spreadsheet pultSI2. The 1996 JRA-recommended pressures against piles in laterally flowing liquefied soils were used to estimate the additional forces exerted against the piles. 6. Pult . and PLiqLayer = 80 kips .3. respectively. k y = 0.

Pult . Structural (right ) = 893 kips . Then. which means that in case of lateral ground displacement.2 Abutment The Excel spreadsheet pultSI2.4.3. Structural = 2 × ( 6 × 1381) kips ⋅ feet H eq (6. Considering only the pile cap passive capacity we have that Pult . The structural capacity at the abutment is: Pult . Mokwa and Prof. Structural < Pult . Structural (right ) = 360 kips . Structural = (6. Pult . 6. Duncan was used to estimate the passive capacity of the pile cap.xls developed by Prof. respectively. σ v ' and σ v . Then.5) where H eq ≈ 25 feet and H eq ≈ 44 feet for the left and right ends of the bridge. evaluated at the center of the equivalent critical layer are: 60 . Structural (left ) = 1179 kips and Pult .The structural capacity at the intermediate piers is: 2 × 5674 kips ⋅ feet 2 × ( 6 × 1381) kips ⋅ feet + 22 feet H eq Pult . Passive = 5842 kips . Soil . Structural (left ) = 613 kips and Pult .6) where H eq = 27 feet and H eq = 46 feet for the left and right ends of the bridge. This means that at the intermediate piers the soil will continue to displace or flow around the stable foundation in case of lateral ground displacement. Pult .3. 6.5 Vertical Stresses The effective and total vertical stresses. the abutment structure will move in concert with the soil. respectively.

230 ksf .Left Abutment σ v ' = 135 × 14 + 110 × 8 + (120 − 62.428 ) × 2. and since γ = 135 pcf .e.7) Fig. The fill material has φ = 45 .842 = 0. 000 psf /120 pcf ≈ 770 ft/s (6. N60 ≈ 50 .116. then Gmax ≈ 325 ( 50 )0. 6. i. and since γ = 120 pcf .68 = 2. 6.2 feet/s 2 × 4.86 ≈ 17 and Gmax = 325 (17 ) 0. The loose sand has σ v 0 ' = 2842 psf and CN = 2. then N60 ≈ 15 / 0.73 / 2 = 2994 psf 6.6 Initial Fundamental Period of Potential Sliding Mass The initial fundamental period ( Ts ) of the potential sliding mass at the left abutment (see Fig.10) can be calculated as follows.68 (ksf) .5 / 2 = 2842 psf σ v = 135 × 14 + 110 × 8 + 120 × 2.10 Representative soil column of potential sliding mass at left abutment.73 / 2 = 2877 psf σ v = 135 × 14 + 110 × 8 + 120 × 3.86 .5 / 2 = 2920 psf Right Abutment σ v ' = 135 × 14 + 110 × 8 + (120 − 62.2 feet/s 2 × 2.428 ) × 3.68 = 4647 ksf . 000 psf /135 pcf ≈ 1.8) 61 . then: Vs ( fill ) = Gmax / ρ = 32. 050 ft/s (6..3.2 / 2. then Vs ( Loose Sand ) = Gmax / ρ = 32. 230. 647. From Imai and Tonouchi (1982) for cohesionless materials Gmax ≈ 325 ( N 60 )0.

There are two sublayers involved with thicknesses (from top to bottom) of 2. and N60 values equal to 15 blows/ft (value already corrected by CN ).45 ft.11 Representative soil column of potential sliding mass at right abutment.50 ft and 2. for cohesive materials with PI ≈ 20 − 25 . 000 psf /110 pcf ≈ 410 ft/s (6.5 = 680 ft/s 14 8 2.9) and H1 + H 2 + H 3 H1 H 2 H 3 + + Vs1 Vs 2 Vs 3 14 + 8 + 2. then Gmax ≈ 585 ksf . 6. the ratio Gmax / Su is about 600. then: Vs = Gmax / ρ = 32. and since γ = 110 pcf . Using the relation by Imai and Tonouchi (1982). and OCR ≈ 2 . the only difference is the loose sand material. 6.14 seconds (left abutment) Fig.11) can be calculated using the same procedure used for the left abutment. With respect to the left abutment. The initial fundamental period ( Ts ) of the potential sliding mass at the right abutment (see Fig. The clay material has c ≈ (1200 + 750 ) / 2 = 975 psf . the 62 .According to Weiler (1988). 050 410 770 4 ( H1 + H 2 + H 3 ) Vs Vs = Vs = (6.5 + + 1.10) ⇒ Ts = ≈ 0.2 feet/s 2 × 585.

45 = 680 ft/s 14 8 2. and their shear wave velocities are around 740 ft/s.95 ≈ 0.050 ksf.7 Yield Coefficient as a Function of Equivalent Undrained Shear Strength of Critical Layer Figure 6. Figure 6.maximum shear modulus Gmax of these layers is about 2. 63 . the potential failure surface goes through the bottom of the shallow clay layer.3.50 2.16 seconds (right abutment) 680 (6.12 shows the nearly linear relationship between k y and Su for both of the bridge abutments considered in this design example if liquefaction occurs.45 + + + 1050 410 740 740 4 × 26. Then: Vs = ∑H i i ∑V i Hi si Vs = 14 + 8 + 2.50 + 2.11) ⇒ Ts = 6.13 shows the relationship between k y and Su for this case (same for both abutments). If liquefaction does not occur.

Fig. 6.Fig.13 Yield coefficient ( k y ) versus equivalent total residual undrained shear strength ( Su ) relationship for cases without liquefaction. 64 .12 Yield coefficient ( k y ) versus equivalent total residual undrained shear strength ( Sur ) relationship for cases with liquefaction. 6.

65 . As Figure 6. did not increase significantly. The latter finding does not mean that the overall size of the potential sliding mass did not increase with Su . and for the right abutment l = 60 feet if Su ≤ 500 psf . Fig. and l = 75 feet if Su > 500 psf . However. l . there is a ∼30% increase in the size of the base area of the critical slip surface for the right abutment as Su increases.14 Effect of residual undrained shear strength on base length of potential sliding mass.3. For the left abutment.6. the length of the base of the critical slip surface was tracked as the equivalent undrained shear strength of the critical layer varied. for value of Su higher than 600 psf the value of l remained constant.14 shows. It actually did slightly. It indicates only that the longitudinal dimension of the base of the potential failure surface. Based on these results it is assumed that for the left abutment l = 55 feet (constant). there was no distinct change in the size of the l with changes in Su .8 Size of Potential Failure Surface To study the effect of Su on the size of the potential failure surface. 6.

3.35 g. assuming that one half of the slide slope masses must be restrained by the piles (Boulanger et al. and the total passive capacity is 6 ft × 43 ft × 5 ksf × 6 / 5. and 2.58 g. respectively for the left and right abutments without liquefaction ( Ts = 0.87 g.Finally.3. 1. 410 kips . 6. The assumed gap was 4 in. 0. 6.31 g.5Ts are 0. 1. 6. Fig. 10%.5 ≈ 940 kips/in . the resulting tributary width in this case is t = 70 feet .88 g. 2006).35 g.10 Passive Reaction According to Caltrans Design Criteria. the longitudinal stiffness of the abutment is 20 (kip/in)/ft × 43 ft × 6 / 5. 0. respectively for the right abutment with liquefaction ( Ts = 0. 50%. respectively (Somerville and Collins 2002). and 0. 0.15).87 g.59 g. and 2.33 g. and 0.14 g. and 2% in 50 years events are 0. and the spectral accelerations at 1. 0.33 g.64 g.14 s ). (see Fig.34 g. 1.13 s ).9 Ground Motion Intensity The peak ground accelerations for the 82%.90 g. respectively for the left abutment with liquefaction ( Ts = 0. and 2.15 Passive reaction versus displacement.58 g. 6.16 s ). 0.33 g. 66 .5 ≈ 1. 0.

and ~100%. 67 . ~100%.17 shows the estimated damages for the different hazard levels. 83%.2 is a summary of the results given in Figure 6.3. for the left abutment. Figure 6.4. Finally. Table 6.5 shows the expected bridge downtimes for the different hazard levels analyzed. and two levels of uncertainty in the estimation of repair cost ratios as a function of damage state.96/2842 = 0. show the expected repair cost ratios for different hazard levels. respectively.65 ( amax / g )(σ v / σ 'v ) rd . in this case the values of CSR are: Left Abutment ( z = 9. and ∼100%.65 × PGA × 2920 × 0. where the cases with and without liquefaction have been combined using the total probability theorem. and rd is the nonlinear shear mass participation factor.25 feet ): CSR = 0.3 and 6.18 and 6.2–6. σ v is total vertical stress. 92%. and 2% in 50 years events are 8%. σ 'v is the effective vertical stress.16. for the right abutment.11 Probability of Liquefaction Recalling that the in-situ cyclic shear stress ratio is defined as CSR = 0.65 × PGA × 2994 × 0. Figures 6.19. 99%.4 RESULTS USING FIRST-ORDER RELIABILITY METHOD (FORM) Figure 6. The bottom figures in each column show the final result. 50%.65 ⋅ PGA The probabilities of liquefaction for the 82%.86 feet ): CSR = 0. g is the acceleration of gravity.16 and Tables 6.6. Tables 6. 10%. where amax is the peak horizontal ground surface acceleration. 6. Table 6. the upper figures in each column represent the probability of displacement being greater than the threshold given in the x-axis for each liquefaction case. and 17%.64 ⋅ PGA Right Abutment ( z = 9.5 show the results of applying the proposed simplified approach using the first-order reliability method (FORM). where the lines with white symbols correspond to the case without liquefaction and the lines with black symbols represent the cases with liquefaction. respectively. In Figure 6.96/2877 = 0.16.

of Exceedance in 50 yrs. 82% 16th Percentile Median 84th Percentile Prob. Table 6. In top figures.5” < 0. black symbols denote cases with liquefaction and white symbols denote cases without liquefaction.5” 50% 3” 7” 13” 10% 9” 14” 28” 2% 12” 27” 59” 50% 1” 3” 7” 10% 4” 10” 15” 2% 9” 14” 28” < 0.5” < 0.5” < 0.5” 68 .2 Probability of D > d given IM (FORM results).Fig. 82% < 0. of Exceedance in 50 yrs. 6.16 Probability of exceeding a lateral displacement threshold (d) for different hazard levels. Left Abutment Right Abutment Prob.5” < 0.

17 Damage states using FORM. 69 . 6.Fig.

6. Table 6.11 0.Fig.07 10% 0.21 0.19 2% 0.05 10% 0.10 2% 0.3 Expected repair cost ratios using FORM—low uncertainty.19 ~0 ~0 σ 70 .37 0.01 50% 0.22 50% 0.08 0. Left Abutment Right Abutment Probability of Exceedance in 50 years 82% μ Probability of Exceedance in 50 years 82% ~0 0.18 Repair cost ratios using FORM—low uncertainty.03 0.21 0.

24 50% 0.22 0.11 2% 0.Fig. 6.21 ~0 ~0 σ 71 .11 0. Left Abutment Right Abutment Probability of Exceedance in 50 years 82% μ Probability of Exceedance in 50 years 82% ~0 0.01 50% 0.08 0.05 10% 0.21 2% 0. Table 6.22 0.4 Expected repair cost ratios using FORM—high uncertainty.19 Repair cost ratios using FORM—high uncertainty.39 0.04 0.08 10% 0.

5 hours ~0. Left Abutment Right Abutment Prob. of Exceedance in 50 years 82% 0.6–6.5 day 0 n/a Negligible 1.5 Expected downtimes using FORM.7 and 6.Table 6.5 hours σ 0.5 hours 10% 1 day 2% 6 days 50% n/a 10% ~1 hour 2% 1 day 0.4” 0.6 is a summary of the results in terms of residual lateral displacements at the abutments.5 days 6. Left Abutment Right Abutment Probability of Exceedance in 50 years 82% μ Probability of Exceedance in 50 years 82% 0 n/a 50% 0.8 show the expected repair cost ratios for different hazard levels.5 RESULTS USING POINT ESTIMATE METHOD (PEM) Tables 6. Table 6. Table 6.9” μ +σ 72 . Tables 6.9 show the results of applying the proposed simplified approach using the point estimate method (PEM).6 Probability of D > d given IM (PEM results). and two levels of uncertainty in the estimation of repair cost ratios as a function of damage state.3” 0.6” 0.7” 1” 50% 2” 4” 7” 10% 3” 15” 27” 2% 8” 33” 58” 50% 1” 2” 3” 10% 2” 7” 11” 2% 4” 15” 26” 0. of Exceedance in 50 years 82% μ −σ μ Prob.5 days 3. Table 6.9 shows the expected bridge downtimes using PEM for the different hazard levels analyzed.

of Exceedance in 50 years 82% ~0 ~0 50% 0.05 10% 0. of Exceedance in 50 years 82% μ Prob. Left Abutment Right Abutment Prob.04 0.19 2% 0.39 0.23 50% 0.37 0.07 2% 0.17 ~0 ~0 σ Table 6.08 0.8 Expected repair cost ratios using PEM—high uncertainty.20 Damage states using PEM.7 Expected repair cost ratios using PEM—low uncertainty.05 0.19 ~0 ~0 σ 73 .19 0. Left Abutment Right Abutment Prob.08 0. 6.19 0. of Exceedance in 50 years 82% μ Prob.01 0.08 2% 0.03 10% 0.21 50% 0. of Exceedance in 50 years 82% ~0 ~0 50% 0.18 0. Table 6.18 2% 0.03 10% 0.18 0.05 10% 0.01 0.Fig.

5(2R). Table 6.Table 6.5”.6 m) 46 ft (14. 6. In this model. Leq = Hliq +2×3.10 Equivalent pile length under each pier.0 m) Column .5 hours ~0. Under each pier there is a group of 3×2 circular open-ended steel piles (PP 24”×0.10 shows the estimated values of Leq under each pier. Kevin Mackie of the University of Central Florida.5 hours 2% 0. Figure 3. and it is a modified and simplified version of a fully nonlinear 3D model of the PEER bridge using OpenSees (Mackie and Stojadinović 2006). where Hliq is the thickness of the liquefiable material at the location of the respective pier.5 days 6.8 shows the main components of the bridge model.6. fy = 60 ksi).1 Properties The properties of the superstructure can be found in Mackie and Stojadinović (2006).7 m) 3 4 (right most pier) 74 41 ft (12. and R is the pile radius (R = 1 ft).5 days 0 n/a σ ~0. Equivalent length Leq 27 ft (8.5 days 3.2 m) 1 (left most pier) 2 32 ft (9. the equivalent length (Leq) of the piles tries to represent the lateral stiffness they provide to the system.6 RESULTS USING SIMPLIFIED COUPLED MODEL This analytical model was prepared by Dr. Left Abutment Right Abutment Probability of Exceedance in 50 years 82% μ Probability of Exceedance in 50 years 82% 0 n/a 50% 0 n/a 10% 2% 0.9 Expected downtimes using PEM. Table 6.5 days ~2 days 50% 0 n/a 10% 0. with a center-to-center spacing of 6 ft.5 days ~0.

01 m. Spring 1 (left) 2 3 4 5 (right) Stiffness 273 kips/in (47810 kN/m) 137 kips/in (23904 kN/m) 137 kips/in (23904 kN/m) 137 kips/in (23904 kN/m) 273 kips/in (47810 kN/m) The stiffness of the intermediate springs 2.8 try to model the interaction that occurs.12 shows the displacements that were imposed. Table 6. between pile caps. and between the embankment and the first pile cap.2 Input Horizontal deformations. The stiffness of these springs was estimated assuming that the ultimate passive capacity of the soil reacting against the pile cap is reached at a deformation of 0. 3. Table 6. which are the mean displacements from Table 6. 75 . The value of Pult was calculated using the φ=0 sliding wedge method developed by Mokwa (1999). were applied to the free ends of the end springs. 6.6.2. through the compression of the upper non-liquefiable layer. and to the two ends of the bridge. Table 6.11 Stiffness of soil springs.The springs shown in Figure 3. consistent with the different hazard levels under consideration.11 shows the stiffnesses that were used. and 4 is half of the stiffness of any of the end springs because that is the equivalent stiffness of the two springs that are acting in series between the pile caps.

Table 6. 76 .36 m 0.21 Deformed shape of bridge from OpenSees model.12 Input displacements. 6. Fig.21 shows the deformed shape of the bridge after the application of the deformation pattern indicated in the previous section.3 Results Figure 6.25 m 0. Return Period TR=29 years TR=72 years TR =475 years TR =2475 years Nodes Left end Right end Left end Right end Left end Right end Left end Right end Horizontal Deformation Negligible 0.18 m 0.69 m 6.08 m 0.13 presents a summary of the results obtained using this analysis.Table 6.6.36 m 0.

e.5% 0.4% 0.5% 2.Table 6.7% 5..5% TR=72 years 2 3 4 1 2 3 4 1 2 3 4 TR =475 years TR =2475 years From this analysis.4638ln( Sa) − 0.12) ln (δ 4 ) = 2.5Ts.3% 2.3246 ± c ln (δ 2 ) = 1. Left + 1.9695ln( Sa ) − 2. Return Period of Event Column 1 2 3 4 1 Column Drift TR=29 years Negligible 0.5 ⋅ (1. same as the input displacements).0545 ± c (6.23 s . Sa is the 5% spectral acceleration (in g's) at T = 0. 77 . Right ) = 0.5Ts .3166 ± c where δ i is the residual tangential drift of column i (in %).1% 0.2% 0. using linear regressions in the log-log space.5% 4. and c = 0. the following intensity measure (IM)–engineering demand parameter (EDP) relationships could be estimated: For column drift: ln (δ1 ) = 2.3% 0.13 Residual tangential column drifts.4% 0.75 (i.3504 ln( Sa) − 0.8695ln( Sa) − 1.9% 1.5535 ± c ln (δ 3 ) = 1.

2 for the case when liquefaction occurs.1108 ± c ln ( D3 ) = 1.45. The circles in Figure 6.1. 21%.0188 ± c ln ΔDLeft = 1.13) For the residual longitudinal relative deck-end/abutment displacements at the left and right ends of the bridge. for the residual pile cap displacements at the columns. while in the simplified coupled model the c. Di (i = 1.15) Figure 6.1353ln( Sa ) − 3. and 2475 years.14) And.0989 ± c ln DLeft = 1.v. DLeft and DRight (in inches): ( ) ln ( DRight ) = 1. which is based only on the amount of residual lateral displacement at the abutments.5281ln( Sa ) − 2. and 37%. and 49% for these three hazard levels. Using the simplified coupled model the expected repair cost ratios are 1%. their level of dispersion is different.22 show the expected repair cost ratios for return periods of 72. 2.2889 ln( Sa ) − 2.5006 ln( Sa) − 3. Although both methods estimate similar mean repair cost ratios.6073ln( Sa) + 1.1031 ± c ln ( D2 ) = 0. 78 .3563ln( Sa ) + 2.1165ln( Sa ) − 3.3876 ± c (6. ΔDLeft and ΔDRight (in meters): ( ) ln ( ΔDRight ) = 1.o. is ~0.0924 ± c (6. 475.3. Using the simplified approach.7418 ± c (6. 21%.9602 ln( Sa) − 4. respectively.22 shows the estimated total repair cost ratios using these relationships and the procedure described in Section 5. The coefficient of variation in the simplified approach is ~0.8. 4) (in meters): ln ( D1 ) = 2.For the residual pile head displacements at the left and right abutments. the expected repair cost ratios are 8%.8652 ± c ln ( D4 ) = 1.

79 . 6.23 Contribution to expected cost of repair actions for different hazard levels.22 Repair cost ratios using simplified coupled model.Fig. 6. Fig. Figure 6.23 shows the contribution of the different repair tasks to the total repair cost for different hazard levels.

In that case. the initial fundamental period of the potential sliding mass. sensitivity analyses may be performed that provide information about the relative importance of the random variables and the sensitivities of the first-order approximation of the failure probability with respect to parameters in the probability distributions. the spectral acceleration at 1. the passive reaction of the embankment against the backwall. Sa . and ε variables had already been presented in previous sections.5Ts . the residual undrained shear strength of the liquefied material.1) . The characteristics of the Sur . Ts . α . thickness of the liquefiable material. In this paper. The random variables that were considered are: Sur . and considering that the uncertainty in H is small. the normalized distance to the points of fixity of the piles.7 Sensitivity Analysis The Pacific Earthquake Engineering Research (PEER) Center performance-based earthquake engineering (PBEE) approach can provide important insights through de-aggregation of the results to identify those components that impact the results most significantly. the error term in the estimation of the residual lateral displacement from the model by Bray and Travasarou (2007). seven random variables were considered in the estimation of the probability of exceeding a specified level of displacement for a given intensity measure. Additionally. ε . The uncertainty in the parameter Ts was estimated recalling that Ts = c ⋅ H / Vs (where c is a constant). the first-order approximations for the mean and standard deviation of Ts are: cH μTs ≈ σ Ts μVs ⎞ ⎛ ⎟ σ Vs 2 = ⎜ − cH ⎜ μ 2 ⎟ ⎝ Vs ⎠ 2 ⎛ dT ≈ ⎜ s ⎜ dVs ⎝ μVs ⎞ cH σ 2 = σ Vs ⎟ ⎟ Vs μVs 2 ⎠ 2 (7. and Pa . α . H . The first-order reliability method provides a useful set of importance and sensitivity measures.

(2002). The uncertainty in the value of the spectral acceleration at 1.o. Thickness of liquefiable material Passive reaction against the abutment’s backwall 82 Probability Density Function Log-normal Uniform Normal Log-normal Log-normal Log-normal Log-normal Coefficient of Variation 0. it is assumed that the coefficient of variation of Ts is 0.6.Then.66 (Standard Deviation) 0.1 Characteristics of selected random variables.[Ts ] = (7. the coefficient of variation of Ts would be: σ Ts μ σ cH = σ ⋅ Vs = Vs = c. The coefficient of variation of the variables H and Pa was assumed to be 0.1 summarizes the estimated characteristics of the seven random variables used in this section. the Electric Power Research Institute undertook an investigation of appropriate methods for estimating earthquake ground motion in eastern North America.39 . Then. included extensive.v. The result was a coefficient of variation of about 0. The measured shear wave velocities were shown to be lognormally distributed with σ lnV = 0.o. high-quality field and laboratory testing of soils at more than 200 different sites. Random Variable Sur Description Residual undrained shear strength of the liquefied material Normalized distance to the points of fixity of the piles Error term in the estimation of the residual lateral displacement Initial fundamental period of the potential sliding mass Spectral acceleration at 1.5Ts ( Sa ) was estimated by averaging the error terms in the attenuation relationships by Chiou and Youngs (2006) and Campbell and Bozorgnia (2007) for periods between 0 and 1 second. Table 7. According to Jones et al.60 0.[Vs ] 2 Vs cH μVs μTs μVs c.v. Table 7.30 α ε Ts Sa H Pa .30 0. which involved numerous investigators.87 0.3. for the analysis that follows.4. This work.40 0.40 0.5Ts .2) This means that the coefficient of variation of Ts and Vs should be similar.

Figure 7. while for larger displacements the opposite is true.. which. This contribution is reflected in the unit vector γˆ (Der Kiureghian 1999). The most important random variable was Sa . Figure 7. The parameters ε and Sur were also important.. and it is also different for each threshold level of lateral displacement. and the residual undrained shear strength of the liquefiable material.1 RESULTS The order of importance of the random variables is identified by their relative contribution to the variance of the linearized limit-state function. Lastly. 83 . i. an important fraction of the uncertainty involved in this problem is due to the uncertainty in the estimation of the intensity measure. in our case. for all hazard and displacement levels. and Pa was rather small. Ts . which in this case is the spectral acceleration at the fundamental degraded period of vibration of the potential sliding mass. the overall uncertainty was also affected by the uncertainty in the estimation of the residual lateral displacement given an intensity measure. changes from hazard level to hazard level.1 shows the absolute value of the terms in γˆ for each level of displacement and hazard level. For small lateral displacements.1 shows that for all cases the importance of the parameters α .e. i. Sur is more important than ε .7. H .e.

for each level of displacement and hazard level. The sensitivity of the “failure” probability. ε . and the i-th term in the vector η is related to ∂ p f dσ i .e. and Sur . Vectors δ and η indicate the sensitivity of p f to variations in the mean and the standard deviation of the random variables.1 Relative importance of random variables for different hazard and displacement levels. i. 7. where σ i is the standard deviation of the i-th random variable. ˆ. to variations in the distribution parameters of the random variables can be quantified through the vectors δ and η (Der Kiureghian 1999). To facilitate the comparison. where μi is the mean of the i-th random variable. p f = P( D > d ) . the vectors δ and η were normalized to have a unit length. respectively. The probability p f = P( D > d ) is most sensitive to the parameters (mean or standard deviation) of the variables Sa .2 and 7. 84 .. the i-th term in the vector δ is related to ∂ p f d μi . Figures 7. Upper and bottom figures correspond to results for left and right abutments.3 show the absolute value of the terms in the normalized vectors δˆ and η respectively. respectively.Fig.

and least sensitive to the parameters of the variables α . Ts . respectively. but for large displacement. variations in the mean of ε and Sur had an important effect on P( D > d ) . or Pa .2 Normalized relative sensitivity ( δˆ ) of probability p f = P( D > d ) to variations in mean of random variables for different hazard and displacement levels. 7. H . 85 . For small to moderate displacements. P( D > d ) was more sensitive to variations in the mean of Sa than to the mean of the other parameters. Fig. Upper and bottom figures correspond to results for left and right abutments.

The effect of these random variables on the estimation of bridge repair cost ratios or downtimes was also assessed. The previous results show that the most relevant parameters in the estimation of the residual lateral displacement are Sa . i. 7.3 Normalized relative sensitivity ( η standard deviation of random variables for different hazard and displacement levels. and Sur . respectively. and their effect on the mean estimates of these decision variables (DVs) was recorded. The result was that all three variables were almost equally important. however.e. ε . 86 . Sa and ε had a slightly larger influence on the estimated decision variables for higher hazard levels. Upper and bottom figures correspond to results for left and right abutments.. each parameter had essentially the same effect on the estimated DVs.ˆ ) of probability p f = P( D > d ) to variations in Fig. Each variable was scaled up and down one standard deviation.

and the residual undrained shear strength of the liquefiable material ( Sur ). is relatively small for the problem studied. followed by the error term in the estimation of the residual lateral displacements induced by liquefaction ( ε ). 87 . The contribution of the other four variables to the overall uncertainty.7. and their effect on the calculated probability.2 COMMENTS The results of this analysis indicate that the most influential parameters in this problem are the intensity of the ground motion ( Sa ).

Whakatane. the site at Matahina dam can be classified as a Site C using the Bray and Rodriguez-Marek SGS classification. The strike.1. as a Site C using the NEHRP classification. dip.. 16 km away from the rupture area (see Fig.33g ).8 8.1 Validation of Simplified Approach LANDING ROAD BRIDGE DURING 1987 EDGECUMBE EARTHQUAKE This section presents the application of the proposed simplified approach to the case of the Landing Road bridge. 39°. respectively.6 earthquake of March 2.e. occurred on a normal fault. 2001 indicate that PGA = 0. 1987.1).30 ≈ 425 m/s. approximately. earthquake. i. and -110°. According to the PEER NGA Database (NGA Database 2007). .29 g (Berrill et al. and that it dips toward the northwest. and rake of the fault rupture were 235°. which means that the strike of the fault is. which suffered moderate damage due to lateral ground displacement induced by the 1987 Edgecumbe. 8. weathered soft rock/shallow stiff soil. and since Vs. and the nearest strong motion accelerograph was at the Matahina dam site.2 shows the acceleration time series and the 5%-damping response spectrum of the Matahina dam station. Figure 8. The NGA database indicates that the peak ground acceleration at the Matahina dam site was PGA = 0. 8. in the SW/NE direction. New Zealand.1 1987 Edgecumbe Earthquake The M w = 6.

Fig. 8.1 Epicenter of 1987 Edgecumbe earthquake, Landing Road bridge, and Matahina dam.

90

Fig. 8.2 Acceleration time series and 5%-damping response spectrum of Matahina dam station (Source: PEER NGA database).

The Landing Road bridge site is located, approximately, 19 km away from the rupture area. Figure 8.3(a) shows the measured SPT values near the pier (Berrill et al. 2001), and a simplified SPT profile from those measurements. Considering that Gmax = 325 ( N 60 )0.68 (ksf) (Imai and Tonouchi 1982), that N ≈ N 60 , and that γ ≈ 17 kN/m3 , the shear wave velocity profile shown in Figure 8.3(b) was estimated. From the simplified profile a value of Vs,30 ≈ 230 m/s was estimated.

91

Fig. 8.3 Landing Road bridge–simplified SPT and shear wave velocity profiles.

Given that the Matahina dam site and the Landing Road bridge site are at a different distance from the rupture area (16 km versus 19 km, respectively), and that the site conditions are also different ( Vs ,30 ≈ 425 m/s versus Vs,30 ≈ 230 m/s, respectively) the response spectrum shown in Figure 8.2 should be adjusted for distance and site conditions. To estimate the required modifications to the spectrum, the NGA models by Chiou and Youngs (2006) and Campbell and Bozorgnia (2007) were considered. Most of the input parameters for these models were taken from the NGA database: M w = 6.6 (moment magnitude), Rrup = RJB = 16 or 19 km (distance to rupture area), Vs ,30 = 425 or 230 m/s , W = 13 km (rupture width), δ = 39 (rupture dip), ZTOR = 0 km (depth to top of rupture), and Z 2.5 ≈ 1.3 km (depth of 2.5 km/s shear-wave velocity horizon, from Hodder and Graham (1993)). The spectra were estimated for the Matahina dam and Landing Road bridge sites using the Excel spreadsheets provided in the NGA website, and the ratio between the two spectra was calculated at each period. Figure 8.4 shows the results of these calculations for a period range appropriate for this case.

92

the Landing Road bridge comprises 13 simply supported spans 18. for periods of vibration between 0 and ~0. cast-in-place concrete deck and two foot-paths. and the beams are bolted down to the piers. it was decided not to modify the spectra from the Matahina dam site.5 shows. the shear force in the piles was estimated assuming that the piles were 93 . 2001). Each span consists of five precast post-tensioned concrete I-beams. each supported by eight 406 mm (16 in) square pre-stressed concrete piles raked at 1:6 (Berrill et al. As Figure 8. and the order of magnitude of the uncertainties of the rest of the parameters present in this problem.Fig.2 Landing Road Bridge SI units will be used for this case history because data were provided in SI units. As Figure 8. Given the uncertainties involved in this estimation. The substructure comprises concrete slab piers running the full width of the superstructure.3 sec. In the preliminary analysis that follows. 8.4 Distance and site condition correction factor as function of period. 8. an estimation of the spectra at the Landing Road bridge site can be obtained by reducing the one at the Matahina dam site by ~5% to ~12%. The spans are not separated but are bolted together and to the abutments.4 shows.1. carrying a two-lane.3 m (60 ft) long.

3 Soil Profile The soil properties used by Kashighandi and Brandenberg (2006) have been adopted.5) is studied.2 m (~20 ft) Unit Weight. friction angle.5 kN/m3 (~80 pcf) 17 kN/m3 (~110 pcf) 18 kN/m3 (~115 pcf) 42° 35° 38° Cohesion.1 shows the location.2 m (~4 ft) 6. Preliminary elastic analysis show that for cases where the rotation and vertical displacement of the pile cap is negligible. Table 8. and cohesion of the three soils layers that were considered.4 Preliminary Calculations The stability of the north abutment (see Fig. 2001). 94 . this assumption introduces an error of less than 10% in the estimation of the shear force in the piles. 8. Soil Layer Crust Loose sand Dense sand Depth to Top of Layer 0 1. 8. Table 8. c 8 kPa (~165 psf) 0 0 8. Fig. unit weight.1. γ Friction Angle. 8. φ 12.1 Soil Properties (after Kashighandi and Brandenberg 2006).5 Schematic elevation of Landing Road bridge (after Berrill et al.1.vertical.

4.1.70. (2001).3/33)}=1.60-CS≈9 bpf.2 Shear Force in Piles There are eight piles at the abutment.5 × 0. and the piles are 0.2 m2 (width × height).4.406 m square piles. (2001) the sands were relatively clean.5 × 1.5=9 bpf.1.8. then N60≈6×(55/60)≈5.4=11.1) 8.7 FP (kN ) = Ae (m 2 ) × 239 kPa × FP = (8 × 1.4) × 239 × 1.7 FP = 2200 kN (~495 kips) (8. the passive reaction at the abutment is: h ( m) 1. According to Berrill et al. Assuming that N55≈N. then N1. then N1.4 1. and considering that: 95 . (2001) it is estimated that the backwall has dimensions 8×1.1 Passive Reaction at Abutment From the drawings in Berrill et al.3) From Berrill et al.5 σ v ' = 33 kPa (~690 psf) (8.3 Residual Undrained Shear Strength of Liquefied Soil Assuming that the ground water table is at the top of the liquefied material.4.70.2) 8.81) × 2. and each one has a yield bending moment of My=230 kN⋅m. and since CN=min{1.60≈1. Using the expression given in the manual design by Caltrans (2006).5. then 2 ( 8 × 230 ) VP ≈ 5 + 2(3. N≈6 bpf.2 + (17 − 9. The thickness of the liquefiable layer is about 5 meters.1.70×5.406) VP = 470 kN (~105 kips) (8. the effective vertical stress at the center of the liquefiable material is: σ v ' = 12.sqrt(101.

μ Sur / σ v '

⎛ 0.3 N1,60−CS ⎛ N1,60 −CS ⎞ = exp ⎜ − 3.5 ⎟ ⎜1 + ⎜ 8 128 ⎝ ⎠⎜ ⎝

(

)

2

⎞ ⎟ ⎟ ⎟ ⎠

(8.4)

then μSur/σv’≈0.0983, and μSur≈3.2 kPa (~ 65 psf).

8.1.4.4 Equivalent Undrained Shear Strength from Passive Reaction and Piles Shear Force
The stability analysis of the slope shows that the base length of the sliding mass is about 30 meters. Assuming that one half of the slide slope masses must be restrained by the piles, and a transverse slope of 2:1 (H:V), the base width of the sliding mass would be about 9.4 meters. Then, the base area of the sliding mass is about 280 m2 and the maximum additional strength provided by the passive reaction and the shear force in the piles would be ΔSur=(2200+470)/280 = 9.5 kPa. This means that if liquefaction occurs and the passive reaction and shear force in the piles are fully mobilized, the equivalent undrained shear strength of the liquefiable material would be about Sur,eq=12.7 kPa (~ 265 psf).

8.1.4.5 Stability Analysis
Considering the properties shown in Table 8.1, and assuming that for the liquefiable layer (loose sand) Sur,eq=12.7 kPa, the stability analysis of the slope indicates that ky=0.082. Figure 8.6 schematically shows the results of the stability analysis at the north abutment.

Fig. 8.6 Potential failure mass at north abutment, considering contribution of passive reaction and pile shear force.

96

8.1.5

Results Using Simplified Approach

8.1.5.1 Initial Fundamental Period of Potential Sliding Mass
Based on Figure 8.6, it is assumed that the representative soil column comprises the upper crust plus ~3/4 of the liquefiable layer. The upper crust (silty material) has φ=42º which means that N60≈50 bpf, and since Gmax=325(N60)0.68 (ksf), then Gmax(crust) = 4647 ksf = 222500 kPa. Since γ(crust)=12.5 kN/m3, then Vs(crust) = sqrt(Gmax /ρ) = sqrt(222500/(12.5/9.81)) = 420 m/s (~ 1400 ft/s). The loose sand has φ=35º which means that N60≈18 bpf, and since Gmax=325(N60)0.68 (ksf), then Gmax(loose sand) = 2320 ksf = 111100 kPa. Since γ(loose sand)=17 kN/m3, then Vs(loose sand) = sqrt(Gmax /ρ) = sqrt(111100/(17/9.81)) = 250 m/s (~ 820 ft/s). The average shear wave velocity within the representative soil columns is:
H1 + H 2 H1 H 2 + Vs1 Vs 2 1.2 + 5 × 0.75 1.2 5 × 0.75 + 420 250

Vs =

Vs =

(8.5)

V s = 275 m / s (~900 ft/s)

Then, the initial fundamental period of the potential sliding mass is:
4( H1 + H 2 )

Ts =

Vs Ts ≈ 0.07 s

(8.6)

The 5%-damping spectral acceleration at 1.5Ts=0.11 sec is about 0.46g (see Fig. 8.2). With these parameters an approximate result may be obtained. From Bray and Travasarou (2007):

ln( D) = − 1.10 − 2.83 ln(k y ) − 0.333 ln(k y )
2

(

)

2

+ 0.566 ln(k y ) ln( Sa) + 3.04 ln( Sa ) +

− 0.244 ( ln( Sa) ) + 1.5Ts + 0.278( M w − 7) ± ε

(8.7)

97

then:
ln( D) = 2.48 ± 0.66 D = 12 cm, (6 cm to 23 cm)

(8.8)

Since the estimated residual lateral displacement of the abutment is about 5 in., and in the context of a simplified estimation, this would mean that the bridge should have suffered small to moderate damage, which is consistent with the observations made after the earthquake. Note that if the sliding mass is considered to be rigid, then Sa(1.5Ts) = PGA = 0.30g and
ln( D) =
ln( D ) = 2.35 ± 0.66 D = 10 cm, (5 cm to 20 cm)

(8.9)

The last required component of the simplified approach is the relationship between the equivalent residual undrained shear strength versus the yield coefficient. The yield coefficient was calculated for a suite of Sur values, and the results are shown in Figure 8.7. As this figure shows, the relation between these two parameters is strongly linear and it can be expressed as:
k y ≈ 0.01362 × Sur ( kPa ) − 0.09406

(8.10)

Fig. 8.7 Equivalent residual undrained shear strength versus yield coefficient.

98

respectively—and that there were negligible probabilities of having either no damage or collapse. given the characteristics of the Edgecumbe earthquake. Mw=6.1 is used to estimate damage to the Landing Road bridge. the Landing Road bridge should have experienced moderate damage (probability of 74%). These results are consistent with the observations made to the bridge after the earthquake.8 Probability of exceeding a threshold residual lateral displacement (d) at north abutment of Landing Road bridge. 99 . 8. given intensity of 1987.8.6 Results Using the simplified approach. 8.6 Edgecumbe earthquake in New Zealand.9 shows that.1 Damage Estimation The simplified approach described in Section 5. Figure 8. New Zealand. Figure 8. This figure also shows that the probabilities of reaching small and large damage states were rather small—15% and 11%. the probability of exceeding different levels of displacements can be estimated.8 shows the result of this estimation for the north abutment of the Landing Road bridge given the intensity of the Edgecumbe.1.6. Fig.1.

The analyses performed by Kashighandi and Brandenberg (2006) predict a 18% probability of having less than moderate damage.1. a 40% chance of having more than moderate. damage. To make a direct comparison. followed by Sur (dashdotted and continuous lines.Fig. respectively).6. and Collapse (C). the most important random variable in this case is ε. a 37% chance of having more than extensive damage. but less than extensive. Moderate (M).9 Probability of bridge being in damage states of Negligible (N).2 Sensitivity Analysis As indicated in Chapter 7. As this figure shows. the vector δ was normalized to have a unit length. which is different for each threshold level of lateral displacement.10(a) shows the absolute value of the terms in γˆ as a function of the level of displacement. Vector δ indicates the sensitivity of P( D > d ) to variations in the means of the random variables. Figure 8. the order of importance of the random variables is defined by their contribution to the variance of the linearized limit-state function. 8. 100 . Small (S). 8. Large (L). but no collapse. This contribution is reflected in the unit vector γˆ . for pile cap lateral displacements in the order of 8 cm. and a 5% chance of having collapse.

earthquake The 1964 Niigata earthquake—caused by faulting along an almost vertical plane—had a moment magnitude of M w = 7. three random variables.2 SHOWA BRIDGE DURING 1964 NIIGATA. 8. and its hypocentral depth was about 20–30 km (see Fig. 101 .1 1964 Niigata. Fig.6 (IRIS website 2007). as function of expected level of residual lateral displacement. followed by Sur. 8.2. Continuous line is for Sur.10 Relative importance of. 8. dotted line for α. and dash-dotted line for ε. Japan.Figure 8. which collapsed due to lateral spreading induced by the 1964 Niigata earthquake in Japan. its epicenter was under the sea. As this figure shows. and P( D > d ) 's sensitivity with respect to.10(b) shows the absolute value of the terms in the normalized vector δˆ as a function of the level of displacement.11). JAPAN. EARTHQUAKE This section presents the application of the proposed simplified approach to the case of the Showa bridge. about 55 km north from Niigata city. 8. ε is again the most significant random variable.

11 Epicenter of 1964 Niigata earthquake. The earthquake caused a peak ground acceleration at the south bank of the Showa bridge of 159 gals (0.Fig. Figure 8.12 shows accelerograms at the roof and the base of a 4-story building located along the Shinano River. and Showa bridge. 102 . 8.16g) (Orense 2005).

8.14).64 + 13. and the piers were also of pile bents (nine single-row piles of the same diameter and length of 25 m.75). The total length was 303. The abutments were of pile bents (nine single-row piles with a diameter of 609 mm and a length of 22 m).2 horizontally. According to Iwasaki (1974) the left-bank abutment moved about 1 m toward the center of the 103 . The superstructures were of 12-span steel composite girders with simple supports. (B) Basement of 4-story apartment building (after Iwasaki 1974).12 Strong-motion earthquake records during Niigata earthquake.9 m (= 13.75 + 10 @ 27.2 Showa Bridge The construction of the Showa bridge over the Shinano river was completed a month before the Niigata earthquake hit Japan (Iwasaki 1974). Due to the Niigata earthquake the bridge sustained severe damage (see Figs.Fig. 8.2. see Fig.13) with collar braces and cap beams. 8. and the width was 24 m. 8.13–8. (A) Roof. The design seismic coefficient for the substructures was 0.

8. out of a total of twelve girders. NISEE. the right-bank abutment and the approach road sustained no significant damage. and the approach road subsided considerably. The sixth span fell down on both its ends due to the failure of the fifth and sixth piers. Five girders. http://nisee. however.river. which had supported the span. Fig. 104 . The first to fourth piers from the leftbank tilted toward the right-bank. The seventh to eleventh piers. 1964.13 Collapse of Showa bridge (Photo by Penzien. On the other hand. suffered only slight damage. J. fell down into the river bed (see Fig.berkeley. the third to the seventh from the left-bank. Source: Earthquake Engineering Online Archive.edu/elibrary/). 8. The fifth and sixth piers collapsed completely into the river bed..14).

comparatively loose near the left bank and comparatively dense near the right bank (Iwasaki 1974). As Figure 8.14 Schematic diagram of collapse of Showa bridge (after Takata et al. 1965).2. 105 . the liquefied layer at the left bank was more than 10 meters thick. while at the right bank it was very thin. 8.Fig. 8.3 Soil Profile The ground conditions at the site were of sandy soils.15 shows.

15 Soil profile and estimated liquefied layer (after Hamada et al.15 and 8.2.16 show.Fig. 1986).17. it is recommended to check the design of the piles assuming that the liquefied soil will push the piles with a pressure distribution like the one shown in Figure 8. 8. In situations like this. 106 . in the case of the Showa bridge there was no upper nonliquefiable crust. 8.4 Results Using Simplified Approach As Figures 8.

Fig. 8.16 Post-earthquake recovery and deformation of pile from Showa bridge. Note absence of upper non-liquefiable crust (after Fukuoka 1966).

107

Fig. 8.17 Idealized pressure distribution for pile foundations affected by liquefactioninduced lateral ground displacement (after JRA 1996).

A modified version of the model used by Bhattacharya (2003) (see Fig. 8.18), is considered for this analysis. The point of fixity at the bottom of the piles was assumed to be located 3.5 diameters below the bottom of the liquefied layer, instead of directly below it. The external and internal diameters of the steel pile were set to 609 mm and 591 mm, respectively, and the Young’s Modulus was assumed to be 210 GPa. Figure 8.19 shows the deformed shape of the pile and the respective bending moment diagram.

108

Fig. 8.18 Schematic diagram showing predicted loading based on JRA code (after Bhattacharya 2003).

(a)

(b)

Fig. 8.19 Deformed shape and bending moment diagram (units are kN and m).

109

3. “… it seems probable that the girders would fall down into the river bed if the relative displacement increases above 50 cm between the two adjoining piers.17 ).” Based on the analysis shown in Figure 8. which means that.3.17 m. 110 . As Figure 8. which supports a fixed shoe and a movable shoe. and 14. Properties in this section are specified in prototype units. the piles did not fail in bending. i. was 50 cm and that there are no devices to prevent the fall of the girders. and material properties of E = 68. the soil profile consisted of a non-liquefiable crust overlying loose sand overlying dense sand.3 8.e. However. Figure 8.2g.3 m long.20 shows a schematic model and partial instrumentation layout of their centrifuge tests.1 CENTRIFUGE TESTS Introduction Brandenberg et al.The maximum bending moment is below the plastic limit ( FS = 1620 /1380 = 1. Davis to study the behavior of pile foundations in liquefiable and laterally spreading ground. Test SJB03 was selected to validate the simplified approach proposed in this study.20 shows. Water was used as a pore fluid for all of the models. The sand layers beneath the crust consisted of uniformly graded Nevada Sand. according to Iwasaki (1974). according to this analysis. This is due to the fact that the width of the crest of each pier. of California. a thickness of 63 mm. This result is consistent with the results by Bhattacharya (2003). the lateral displacement at the top of the piles was estimated in ~60 cm. (2005) performed eight dynamic centrifuge tests on the 9 m radius centrifuge at the Center for Geotechnical Modeling at the Univ. large enough to cause the collapse of the bridge girders.2g to 57. The piles had an outside diameter of 1.2 m wide. The non-liquefiable crust consisted of reconstituted Bay mud. 9.19.9 GPa (Young’s modulus) and σ y = 216 MPa (yield stress).. 8. All of the layers sloped gently toward a river channel carved in the crust at one end of the model. 8.2 Foundation Properties The foundation consisted of a 2×3 pile group with a cap 2. The models were tested in a flexible shear beam container at centrifugal accelerations ranging from 36.2 m high.

loose sand. 2003).Monterey Sand A4a A4b A3 Clay A28 A10b A9 A7 A6 A5 A17 584 A11 A8 A10a A14 Loose Sand A13 A26 A27 A24 A2 A1 North MAH2 MAH1 Dense Sand A25 A12 A15.27 A16 1650 Fig.26. respectively.4 m of loose sand ( Dr ≈ 35% ) over dense sand ( Dr ≈ 75% ). and dense sand layers. Based on these properties the unit weights were estimated as 17.3. 8. Dimensions in mm (after Brandenberg et al.20 SJB03 model layout–accelerometer instrumentation. 19. 111 .7 m of clay ( Su ≈ 44 kPa) and 5. and 21 kN/m3 for the clay.4 m of Monterey sand underlain by 2.3 Soil Properties The soil profile consisted of 1. A16 South A15 A11 A1-A4a A4b 788 A24 A5-A9 & A10b A10a A12-14 A25 A28 A17. 8.

21 shows several time histories of raw and processed data illustrating the behavior of the soil and the pile group during the medium Santa Cruz motion. the results after this motion will be analyzed. Then: Pult.cap 44 × 9.25 + 2 × 0. Port Island (83 m depth. and the resultant force of the liquefied sand reacting against the piles would be Pult. The subsequent motions were a medium Santa Cruz ( amax. Channel 1.2 m.67 g ). north–south direction) during the Kobe earthquake. 1 2 Univ. CA. the total passive reaction of the soil against the foundation would be Pult.base = 0.5 Results Using Simplified Approach First.base = 0.cap = Pult.12) The passive pressures below the pile cap can be estimated using the pressure distributions proposed in JRA (1996).3. Figure 8.9 × 1.67 g ) (Brandenberg et al.cap = Su bH 2 (8.75 . 8.clay = 0. UCSC/Lick Lab.2 ⎛ 17 × 2.11) where Su = 44 kPa.3.base = 0.2 ⎝ ⎠ ≈ 2850 kN (8. H = 2.17 ≈ 1385 kN .2 ⎞ + 0. during the Loma Prieta earthquake. The first was a small Santa Cruz motion1 with a peak acceleration of 0.5(37 + 70) × 6 × 1. The passive force between the clay layer and the piles would be Pult.loose sand = 0.soil ≈ 4950 kN .17 ≈ 715 kN . Santa Cruz. 112 . and α = 0. and a large Kobe2 ( amax. b = 9.35 g ). γ = 17 kN/m3.13g. 2005). a large Santa Cruz ( amax. the passive soil resistance is compared against the ultimate structural capacity.5(21 + 52) × 6 × 5. In this section.75 ⎟ 4+ ⎜ 2 44 9.8. of California.25 + 2α ⎟ ⎜4+ Su b ⎝ ⎠ Pult.2 × 2.4 ×1. Then.4 Input Motions A series of four simulated earthquake events were applied to Model SJB03.2 2. The passive soil resistance against the pile cap can be estimated using the φ = 0 sliding wedge method (Mokwa 1999): ⎛ ⎞ H γH + 0.2 m.

113 .21 Representative time series from SJB03 for medium Santa Cruz motion (after Brandenberg et al. 2005). 8.Fig.

8. diameters away from the i. In situations like this.e. it is proposed to follow the recommendation given in the MCEER/ATC (2003) document. Assuming a distance between the ts of fixity of Leq = 11. i. and the residual bending moment was ~1250 kN·m.4) = 13100 kN > Pult.22 shows.22 Soil displacement versus pile cap displacement for SJB03 test.17 ) .e. Assuming that the points of fixity of the six piles are 3. 8. the residual pile cap displacement and residual bending moment in the piles (at the location where it was measured during the test) were estimated as ~0.4 m ..23). and M p = 216000 × 0. the soil should have flown around the foundation.soil . As Figure 8.0576 m3 . then Pult.str ≈ 6 × (2 × 12440 /11. it is important to note that the 114 .0576 = 12440 kN ⋅ m . When comparing these results.0444 / ( 32 × 1. the residual pile cap lateral displacement in the test was ~0. where 4 Z p = π Do − Di4 / ( 32 Do ) .17 4 − 1.On the other hand.03 m.21 shows.04 m and ~3300 kN·m. As Figure 8. ( ) then Z p = π 1. ( ) Z p = 0.5 liquefied layer.str ≈ 6 × (2 × M p / Leq ) . Since the soil capacity is smaller than the structural one. respectively.. Fig. this is consistent to what was observed during the test. Pult. and using a simple structural model (see Fig. to design the foundation to resist the passive pressures created by the soil flowing around the structure. the plastic moment of the pile section is M p = σ y Z p .

Finally. and cyclic degradation of the stress-strain behavior in the clay.MCEER/ATC (2003) document is for design. respectively. so its estimations are likely to be conservative. it is important to realize that inertial effects can play an important role in the response of these systems. however. the sequence of shaking events likely induced changes in the soil properties. respectively. However. Fig. the maximum lateral displacement and pile bending moment are ~150% larger than the residual ones. As Figure 8.02 m and ~1650 kN·m. 8.3 m) relative displacement between the pile cap and the upper crust would generally be considered sufficient to mobilize the passive resistance of the clay crust against the pile cap. The rather large (~0. ~0. the residual lateral displacement and pile bending moment would have been ~0. as Brandenberg et al.23 Simplified foundation structural model. the softer load-displacement response observed in the centrifuge tests compared with static load tests can be attributed in part to cyclic degradation of the clay stress. including densification of the sand layers due to post-shaking consolidation. (2005). (2005) have indicated.03 m and ~1250 kN·m. Also.23. which are reasonably similar to the results from the test. the extrapolation of this type of result to real cases is not straightforward. If half of the full passive reaction is used in the simple analysis shown in Figure 8. 115 .21 shows. According to Brandenberg et al. the peak crust load (estimated from the test) for the medium Santa Cruz motion was less than 50% of the peak load observed during the large shakes.

Figure 9. Fig.2 Main components of analytical model (after Shin 2007). .2 presents some of the main components of this model. putting emphasis in an accurate characterization of the structural and soil behavior. abutments.9 9. including modeling of bridge components such as piles. and backwall structure. pile caps. 9.1 Bridge example finite element mesh (after Shin 2007). and Figure 9.1 Other Issues COMPARISON WITH RESULTS FROM ADVANCED FINITE ELEMENT ANALYSIS Shin (2007) developed a detailed finite element model of the bridge example described in Chapter 6. Fig.1 shows the finite element mesh developed by Shin. 9.

The selected EDPs were: residual pile cap longitudinal displacement at the abutments and at the intermediate columns. 89. several engineering demand parameters (EDPs) versus intensity measures (IMs) were identified. 475. 72.3–9. the proposed simplified coupled model captures reasonably well the response of the bridge for the different hazard levels. and residual longitudinal relative deck-end/abutment displacement. Figures 9.3 Comparison of results—residual pile cap displacement at abutments. and 2475 years. As Figures 9. and four of the most relevant EDPs were chosen to make the comparison.Forty ground motions were applied to this model: ten motions for each hazard level corresponding to return periods of 29. respectively. 9.6 show how the EDP-IM plots from the results of the advanced FE analyses compare with the results of the simplified coupled model described in Section 3.2. Fig.3–9. 118 . and 149 cm/s for return periods of 72.2. From these analyses.6 show. residual tangential column drift. It was assumed that the PGV was 51. Peak Ground Velocity (PGV) was selected as IM. Scaled versions of the ground motions developed by Somerville and Collins (2000) were used for this purpose. 475. and 2475 years.

Fig. 9.4 Comparison of results—residual pile cap displacement at piers. 119 .

relative deck-end/abutment displacement. 9. 9.5 Comparison of results—residual tangential drift ratio at columns. Fig. 120 .Fig.6 Comparison of results—residual long.

and a larger set should be used to estimate the dispersion around that mean. However. given an intensity measure. To perform the analyses. A second alternative is to use a simplified analytical model that captures the key components of the problem to estimate the required seismic response. 121 . Alternatively.5Ts ) for each hazard level. Figure 9. Although this type of detailed analyses can be useful in illustrating important seismic response mechanisms.7 shows a range of simplified models that are possible.2 INCORPORATION OF GROUND MOTION TIME HISTORIES In the simplified approach described in Section 3. Bray and Travasarou (2007) used an iterative procedure with an equivalent-linear sliding block model that had been shown to capture the nonlinear response of a potential sliding mass to develop their recommended expression. At least seven ground motions should be used to estimate the mean residual lateral displacement. If a site-specific set of ground motions is provided. models like those shown in Figure 9. a more refined analysis may be desired. its implementation may not feasible in all but the most critical projects due to economic or time constraints.9. the uncertainty can be conservatively estimated using the error term ε given in the expression by Bray and Travasarou (2007). and to be consistent with the results by Bray and Travasarou (2007).7 can be used to estimate the mean response and to characterize the uncertainty around that mean. was estimated using the expression proposed by Bray and Travasarou (2007). the suite of ground motions should be scaled to match the respective 5%-damped spectral acceleration at the degraded period of vibrations of the potential sliding mass ( 1. the probability that the residual longitudinal displacement at the abutments exceeded a certain threshold. One alternative is to use a fully-coupled nonlinear finite element model like the one developed by Shin (2007) for the bridge example described in Section 6. For example. from a rigid Newmark sliding block to a nonlinear lumped mass coupled sliding model. in some cases.

(d) Nonlinear.7 (a) Problem analyzed. 9. (b) Newmark rigid model. modal. coupled sliding model. lumped mass. coupled sliding model (after Rathje and Bray 2000). (c) Linear elastic. 122 .Fig.

The “pile-pinning” effect is a critical element of the seismic performance of a bridge system that is founded on piles that pass through a potentially liquefiable soil layer with • . 10.2 • PRINCIPAL FINDINGS Liquefaction of the foundation soils dramatically affects the structural response and resulting performance of a bridge system. as well as to a few representative well-documented case histories and a centrifuge model experiment. The use of the proposed approach was illustrated through its application to a detailed bridge example. Some of the key factors that have been incorporated in this procedure are the “pile-pinning” effect. It is critical that ground conditions at a bridge site be adequately characterized and evaluated in terms of potential for liquefaction or severe strength loss.10 Conclusions 10. a simplified probabilistic procedure for assessing bridge damage due to liquefaction-induced lateral ground displacement has been developed.1 SUMMARY Based on the general concepts delineated in the MCEER/ATC-49-1 document. and characterizations of the uncertainty in parameters such as the residual undrained shear strength of liquefiable soils. The estimated results using the simplified procedure agreed reasonably well with the post-earthquake shaking observations in these cases. and the distance to the points of pile fixity in the firm soil layers above and below the liquefiable soil. the error term in the estimation of the abutments’ seismically induced deviatoric permanent displacement.

the most dominant factor was the spectral acceleration at the degraded period of vibration of the potentially sliding mass. Most importantly. Significant deformations and forces in the bridge can lead to severe damage or even the collapse of the bridge. i. It is important to include this effect in the estimation of the dynamic resistance of the bridge-foundation-soil system. The characterization of the ground motion hazard is an important step in the PEER-PBEE approach. Although its formulation may at first appear to be complex. For a selected ground motion hazard level. Performance-based earthquake engineering (PBEE) provides a practical methodological for evaluating the likely seismic performance of existing and new bridges. • Based on the cases analyzed in this study.firm soil layers both above and below the liquefiable soil layer. No other parameter was 124 • • • . dollars.e. and repair actions. The lack of robust relationships that connect the calculated response of engineered systems and the consequences of these responses in terms of death.. however. and downtimes remains one of the most significant limitations to wider use of this methodology. it takes advantage of recent advances in probabilistic seismic hazard assessment techniques. the residual longitudinal displacement of the abutments is judged to be an appropriate index of the overall seismic performance of pilefounded bridges affected by liquefaction-induced lateral ground displacement. and documents the application of engineering judgment and decisions during the design and evaluation process. tracks the uncertainty associated with key parameters of the analysis. the most important parameters in this problem were the error term in the estimation of the residual lateral displacement at the abutments and the residual undrained shear strength of the potentially liquefiable layers. One of the most important steps in the PEER-PBEE methodology is also one of the most challenging ones. to define a robust relationship between seismic response. Large residual displacements of the abutments are typically associated with large deformations of the foundation system. damage condition. which results in large deformations and forces being developed in the bridge superstructure. Overall. there are now standard analytical tools that facilitate its application to actual projects. Analytical models combined with engineering judgment based on recommendations from a panel of experts are currently the most reliable ways to estimate these types of relationships.

The results of these simple models can guide the decision of whether the use of more advanced analyses is required or not. Incorporation of the bridge’s transverse response in the seismic performance evaluation is required. Improved evaluation procedures that incorporate the effects that inertial loads have in cases where the bridge deck is present are required. More robust relationships are required between engineering demand parameters.more important than the ground motion parameter that was used to define the seismic hazard at the site. repair times. Ground motion selection is an important topic warranting additional attention for cases where the bridge evaluation requires more site-specific analysis. Given the importance of the characteristics of the ground motion. This would include development of guidelines of when and how these inertial loads should be incorporated in the analysis.3 • FUTURE RESEARCH Development of an enhanced. damage measures. 10. it is difficult to make reliable estimations in these terms. and decision variables. such as downtimes. analytical model suitable for dynamic analysis of a bridge-foundation-soil system is required. Hence. This model could be a combination of the nonlinear lumped mass coupled sliding model proposed by Rathje and Bray (2000) and the current simplified coupled model. Although it is possible to quantify the consequences of a seismic event in terms other than repair costs. the database available to calibrate these models is still lacking. Gathering sufficient real data is a critical next step. the development of guidelines for ground motion selection is of great importance. yet still simple. • The combination of engineering judgment and simple analytical models can lead to reasonable estimates of the seismic performance of bridge systems undergoing liquefaction-induced lateral ground displacement. • • • • 125 . or number of deaths.

2007. FY ’98..” Geotechnique. Eng. 306-318..” J. D. N. Christensen. S. “Simplified Procedure for Estimating Earthquake-Induced Deviatoric Slope Displacements.J. 2003. and Kutter... 131(11).. “MCEER/ATC-49-1. Vol. Boulanger. thesis. ASCE. B. L. 2003. R...L. Vol. ASCE. ATC/MCEER Joint Venture. Boulanger. D. Keenan. and Geoenviron.. 1997. Brandenberg. 133. Davis. No. “Empirical Response Spectral Attenuation Relations for Shallow Crustal Earthquakes.. 68.. Part I: Specifications.S.R. W. W. San Francisco. Brandenberg. S.” Proceedings of the 8th U. R. Chang. 2003... R.A.A. Bray. 2000. April 18-22. “Behavior of piles in laterally spreading ground during earthquakes – centrifuge data report for SJB03. L. of Geotech.. National Conference on Earthquake Engineering.” UCD/CGMDR-03/03.....” Ph. B... 2001. and Travasarou. Recommended LRFD Guidelines for the Seismic Design of Highway Bridges.. “Behavior of piles in laterally spreading ground during centrifuge tests. Melbourne.” Prepared under NCHRP Project 12-49. Okada. S. and Silva. Australia. 94-127. N. 2003. ATC/MCEER Joint Venture. Recommended LRFD Guidelines for the Seismic Design of High-way Bridges.J. 501-517.D.” J. Liquefaction Study Report. B. Gulerce..A. University of California. California. of Geotech. 2006. “NCHRP 12-49. Abrahamson.” Seismic Performance and Simulation of Pile Foundations in Liquefied and Laterally Spreading Ground.. Paper No. 127 . and Geoenviron.. U.W. H.S. Bilge. Kutter. 2003. 381-392. ASCE. “Case study of lateral spreading forces on a piled foundation. 4. Chang. Engrg.A. Department of Civil Engineering. Center for Geotechnical Modeling. “Evaluating pile-pinning effects on abutments over liquefied ground. Geotechnical Special Publication No. T. Brandenberg. J. 2005. “Probabilistic Models for the Assessment of Cyclic Soil Deformations. No.REFERENCES Abrahamson. U.. 1822.” Berrill. K. J. S. 51(6).D. “State of the practice of seismic hazard evaluation.O.P. Bhattacharya. 2006.. 1. 145. J. D. W. “Pile Instability During Earthquake Liquefaction. Cambridge University. R.” Seismological Research Letters. and Cetin.. W. J.. Boulanger.” GEOENG 2000 Conference Proceedings.T. and Kutter. S. and Pettinga. and Chang.B. April 2007. pp.

9th Inter.. S. Vienna. Chiou. February 26–March 1. “The Stabilization of the Upstream Slope of Sardis Dam Using Driven Prestressed Concrete Piles: Methods of Analysis.. M.. 2006.” Paper presented at Symposium Honoring Luis Esteva. September 2005. Soc. “Probabilistic Liquefied Shear Strength”.” Journal of Geotechnical and Geoenvironmental Engineering. Georgia.. pp. Gutierrez. G.edu/news/2000spring/ index. C. 779-786. http://peer. “Campbell-Bozorgnia NGA Ground Motion Relations for the Geometric Mean Horizontal Component of Peak and Spectral Ground Motion Parameters.D. Christian. Am.W. GeoCongress 2006.” Bull.H. Issue 4. “Progress and challenges in seismic performance assessment. B.. 1991. Duncan. Der Kiureghian.. Eddy. “Probability Concepts for Performance-Based Earthquake Engineering. 1999). S.” Journal of Geotechnical and Geoenvironmental Engineering. A.A. 1968. J.berkeley. 2006. W. Finn. W. 1999. “Seismic Design Criteria.. 125. and Stacy. CA. “Chiou and Youngs PEER-NGA Empirical Ground Motion Model for the Average Horizontal Component of Peak Acceleration and Pseudo-Spectral Acceleration for Spectral Periods of 0. Y. Cornell. 128 . 531-553. pp. pp.” Interim Report for USGS Review. Conf. and Youngs. Spring 2000.” PEER Center News.S..-J. “Point-Estimate Method as Numerical Quadrature. “Engineering seismic hazard analysis. Vol... USA. A. 2006. Finn.” Report PEER 2007/2. 2000.. and Baecher G.Caltrans. on Large Dams. Ledbetter. A. Proceedings of GeoCongress 2006. M..L. K. R..” Proc. and Wu. T.L. A. and Bozorgnia. J.D. Atlanta. 2000. 307-316 (April 2000). 2006. Mexico City. Seismol. No. pp.. H. 2007. “Dam on Liquefiable Foundation: Safety Assessment and Remediation. Y. Ledbetter. Geotechnical Engineering in the Information Technology Age. 59 (1968) (5). “Factors of Safety and Reliability in Geotechnical Engineering. Volume 126. 1583–1606.. 17th Inter. Pacific Earthquake Engineering Research (PEER). Richmond. M.. M. 2005. Center.H. and Lumbantoruan.T..html Cornell. Der Kiureghian. B. “Introduction to Structural Reliability: Class Notes for CE229-Structural Reliability” (Spring. and Krawinkler. C. June 14.01 to 10 Seconds.” Proc. 1997. 9..” Campbell. R.

“Influence of confining stress on liquefaction resistance. T.” Prentice Hall. J. 1-22.W.” Earthquake Spectra 9.Conf. A. pp. L..” Technical Rep..X. 2007. 141-150.. “Fragility Curves for Lateral-Spreading Induced Damage to the Landing Road Bridge. Hynes.. Hamada. National Center for Earthquake Engineering Research. Jones. 1996... 2007..L. Calgary. Seismic Design.” GEOSLOPE/W International. Center. “Correlation of N-value with S-wave velocity. 1974. China. S. 4th International Conf. “Large ground deformations and their effects on lifelines: 1964 Niigata earthquake. 1999. Buffalo. Public Works Research Institute. Int.. Case Studies of Liquefaction and Lifeline Performance During Past Earthquake.Z.). in Geomechanics.P. on Earthquake Geotechnical Engineering – Invited Lectures. 1. “Geotechnical Earthquake Engineering. “Earthquake Microzoning from Soil Properties. Iwasaki. 67-72. 2002. ed. Wuhan. “Response Analysis of Civil Engineering Structures Subjected to Earthquake Motions. Kashighandi. Imai. Japan. M.. A. 129 . W. Ltd.. 1996.Y. E. Kramer. Personal Communication. R. Upper Saddle River. Ministry of Construction. 653 pp. Part V. R. Vol. and Graham. and Tonouchi.and CPT-based Relationships for the Residual Shear Strength of Liquefied Soils. A. Richmond. JRA. and Brandenberg. Japanese Road Association. P. M. Kramer. 2004.” Earthquake Geotechnical Engineering. The Netherlands. 1992. T.” Report PEER 2002/16. and Olsen. NCEER-92-0001. CA.. M. S... The Netherlands. Canada. “SPT. Personal Communication. D. Rotterdam. I. M.. 1982.” Proc..A. Hodder... NJ. Of Soil Liquefaction. Idriss. and Arduino P. 145–152. 3-1–3-123. GEO-SLOPE/W International. S. Balkema. Specification for Highway Bridges. and Mech. L. Ltd. Alberta.” Draft Paper. Pacific EQ Engineering Research (PEER). 1993. 209.J. Kramer. on Computer Methods and Adv. and Boulanger. “Stability Modeling with SLOPE/W. “Estimation of Uncertainty in Geotechnical Properties for Performance-Based Earthquake Engineering. K.” Ground Vibration Section.” Proceedings of 2nd European Symposium on Penetration Testing. Pitilakis. Balkema/Rotterdam/Brookfield. N.. 2006. K. S. Yuan (Ed. S. Workshop on Phys. Springer..L.

15. Mackie. “Earthquake Resistant Design of Foundations: Retrofit of Existing Foundations. K.” Geotechnical Engineering for Transportation Projects (GSP No. and Tokimatsu. USA. San Francisco.. Mackie. NGA Database. K. Stewart. Anderson. Virginia Polytechnic Institute. Mishac K. M. Martin. “Quick (updated) summary of Bridges Types 1 & 11” (November 11.. 2006. 1978. 2005.D. Buffalo. and Stojadinovic. 1965.” J. Rathje. on Geotech.F. 139-160.. Mokwa. Geotech. Next Generation Attenuation (NGA) of Ground Motions Project. Geotechnique. L. March. thesis. S. GeoTrans 2004. B. ed. N. 2005) Mackie. 112(3).” Proc. “Definition of terms related to liquefaction. G.. R. and Lam. B. M.. and Stojadinovic. I. B. Engrg.. http://peer.” GeoFrontiers 2005. Mayes. Martin. 373–377.. M. Marcuson.R. 2005).. R. N. D. “The Seismic Design of Bridges-Geotechnical and Foundation Design Issues. 1. “Seismic vulnerability of typical multi-span California highway bridges. ASCE. R. 2004.. Melbourne. “Overburden correction factors for SPT in sand. S. 1025-1047. P. K. Kayen.. R. 2007. Los Angeles.” Proc... and Stojadinovic. R.” Ph.” In Proceedings of the 5th National Seismic Conference on Bridges and Highways.. Newmark..berkeley.” In Proceedings of ASCE Structures Congress. CA. Third National Seismic Conference and Workshop on Bridges and Highways.edu/nga 130 . Martin. Wong. Editors. G. “Recommended Design Approach for Liquefaction Induced Lateral Spreads. July 27–31. “PEER Methodology Digest” (February 18. Edward Kavazanjian. W. Int. 2000... “Comparison of post-earthquake highway bridge repair costs. E. 1999. S. 126). J.. M.. 2007. L. and Geol.. September 18-20. and Whitman. Seed. Vol. R. V. K. G. Conf.. MCEER-02-SP04. “Probabilistic Liquefaction Triggering Based on the Cone Penetration Test. v. p. Engrg. Yegian. 1986. and Power.. Mackie..Liao. K. E. pp. 2006. “Effects of Earthquakes on Dams and Embankments”. 2002.. J-M.P.. R. R. California.Y. “Investigation of the Resistance of Pile Caps to Lateral Loading.. 2004.” J Geotech Eng Div ASCE 104(9): 1197–1200.. B. G. E.L. Moss.

“Earthquake-resistant design of earth dams. 2000.. 41–64.M. Austin.. 2002. L. Symp. Wang. M. “Liquefied Strength Ratio from Liquefaction Flow Failure Case Histories. 97(9).. Moss.” Journal of Geotechnical and Geoenvironmental Engineering. 72. USA. J. “Nonlinear Coupled Seismic Sliding Analysis of Earth Structures. New York. C. Div.” Earthquake Engineering Research Institute Monograph. L. TX. “Recent Advances in Soil Liquefaction Engineering: A Unified and Consistent Framework. R. F. Cetin. I. 2005.. 25.A.B. Riemer.D. T.F. 225 – 240. Long Beach.” Proceedings of H.. 2003.. Sci. and Idriss. 1975. volume 2..D. No.. Nat. 1249–1273.. and Faris. Seismic Des.E.” Keynote Address. Queen Mary. and Idriss. 1983. 26th Annual Geotechnical Spring Seminar..P. 3812-3814.. E. A.” Proc. Geotech.” Canadian Geotechnical Journal 1999. A. ASCE. Kayen. April 30.. pp. ASCE. Inc. I. S. 10. and Bray J. Wu. M. Engrg. H. Reese. of Earth Dams and Caverns.” Ensoft..E. “Ground motions and soil liquefaction during earthquakes. E. J. Rathje. E.. S. Rathje. “Simplified procedure for evaluating soil liquefaction potential. B. Vol. J. Seed. W. “SPT-based analysis of cyclic pore pressure generation and undrained residual strength. Isenhower. Seed. R. 11.. 1002-1014. J. Vol.” Can. 39: 629–647. M. D. 1990.S.. Arrelaga. Seed. Calif.M. M. R. “An examination of Simplified Earthquake-Induced Displacement Procedures for Earth Structures.B..B.. Sancio. Bray.. Los Angeles Section of the GeoInstitute. K. “Assessment of Liquefaction Potential based on Peak Ground Motion Parameters.S. A.. Oakland.” Soil Dynamics and Earthquake Engineering. 126.M.0M. B. pages 351-376. Orense. Bolton Seed Memorial Symposium.Olson. “FEMA 351Recommended Seismic Evaluation and Upgrade Criteria for Existing Welded Steel MomentFrame Buildings” Seed.. SAC Joint Venture for the Federal Emergency Management Agency (FEMA). H.. J.O. and Stark. Kammerer. “Point Estimates for Probability Moments.. Seed..” J. and Bray. ASCE. 1999. 2000. M. and Harder. Geotech. B. J. R. American Society of Civil Engineers. T. No. M. pp. Acad. 1982. R. R. J.. Pestana. D. 36(1): 72-87. 131 . “LPILE Plus Version 4..” Proc. H. 1971. Rosenblueth.. and Hendrix. California. H.

and Collins. Youd. Richardo Dobry.” Seismological Research Letters 68. Joseph P. “A probabilistic methodology for assessing seismic slope displacements. Marcuson III. Seed.. Maurice S. 20. J.A. Geotechnical Special Publication No. John T.F. Ignacio Arango. Mary Ellen Hynes. William F. Somerville.G. P. 2002..” Ph. 2004 (CD-ROM).. (1988).. Robertson. 124(10). 132 . Gonzalo Castro. W. 2004. Christian. Graves. p331. Leslie F. Liao. Earthquake Engineering and Soil Dynamics II – Recent Advances in Ground-Motion Evaluation. Liam Finn. Weiler.. A. “Modification of empirical strong ground motion attenuation relations to include the amplitude and duration effects of rupture directivity. Mitchell. Yoshiharu Moriwaki. and Kenneth H. 1997.L. Stokoe II. H-S.. D. 199-222. Oakland. Ronald D.D.Shin. University of Washington. Peter K. W. Idriss. Andrus.W. ASCE. 2001. Raymond B. 2007. Sam S.” Somerville. N.” Journal of Geotechnical and Geoenvironmental Engineering.” Proceedings. August 1-6. T. “Numerical Modeling of a Bridge System and Its Applications for Performance-Based Earthquake Engineering. Harder Jr. R..A... Koester. “Ground Motion Time Histories for the I880 Bridge. Canada. Geoffrey R. and Abrahamson. P. M. 13th World Conference on Earthquake Engineering. thesis. 15 pp. Power. T. and Der Kiureghian.. Travasarou. James K. N. N. Martin. “Small strain shear modulus of clay”. “Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils. Vancouver. C. Smith. I. Bray.. Kenji Ishihara.

Sa .” In that routine the user needs to specify the set of displacement at which the required probability is going to be calculated.m. and also appropriate input values for the parameters Ts . α . μ Sur .m.Appendix A: Representative MATLAB Routines This section presents the MATLAB routines that were used to estimate the probability of exceeding a certain level of residual longitudinal displacement at the abutments due to liquefaction-induced lateral ground displacement. The main routine is called “master.” The latter function calculates the amount of lateral displacement for given values of Sur . The main routine then calls the toolbox FERUM with input file “inputfile_*. α . and M w . Ts . ε .m. 133 . σ Sur .” which in turns makes use of the function “GetD_*. For the evaluation of the sensitivity parameters it is also necessary to input the standard deviation of the random variables Sur . Sa . and ε along the diagonal of the diagonal matrix D . for a given intensity measure of the ground motion. and M w using the expression developed by Bray and Travasarou (2007).

6. for i=1:length(dlist) d=dlist(i).beta_sensi_thetaf(:. %sa = 2. % NonLiq case D(2.1)=sdSu.3).591.874.07) % psf -. 6.334.0. run inputfile_liqLEFT % run inputfile_liqRIGHT % run inputfile_noliq ferumFORM p(k)=formresults. 6.879.333. Mw = 6.LEFT (Jun29. RIGHT 6.0. sa = 1. D(3. % Liq case D(1. %sa = 0. 7.m % close all clear all clc dlist=[0.6.5 1].07) % psf -. % muSu = 975.Upper Clay -.NoLiq -. % inches % -.580.8.beta_sensi_thetaf(:.LEFT & RIGHT % New (9/7/07) % LEFT %sa = 0.1). k=1.LEFT & RIGHT % psf -.LEFT (Jun29. Mw = 6.869. Mw = 6.k)=formresults. Mw %sa = 2. %sa = 2.6.2)=0. % muSur = 512.8. muSur = 649. % sdSu = 244.07) % -.k)=D*formresults.6.6.% % Routine master. impt_delta(:.579.imptg.RIGHT (Jun29. Ts = 0. Mw %sa = 0.RIGHT (Jun29. Mw Mw Mw Mw = = = = & = = = = 6.1)=sdSur. % New (9/7/07) % NoLiq -. % Ts = 0.LEFT & RIGHT % psf -. %sa = 0.Upper Clay -.8.14.13. impt_gamma(:. Mw D=zeros(3.pf1.66.326.16.07) % psf -.6. end 134 . % RIGHT %sa = 0. %D(1. 6.LEFT (May17.3)=0.309.0. sdSur = 260. % sdSur = 205.RIGHT (May17. % New (9/7/07) 6.07) % -. 7. % Ts = 0.346. %sa = 1. Mw %sa = 1.LEFT %sa = 0.07) % psf -.348.k)=D*formresults. impt_eta(:. k=k+1.86603.2). Mw = 7.

0. % alpha (uniform) 0.00 sdSur 0. 0.parameter = 'no'.scis_max = 20000.marg(1. gfundata(i). towards the origin analysisopt.x(3).5 0. system.target_cov = 0.sim_point = 'dspt'. 1.il_max = 5. 0.05. analysisopt.type = 'expression'. gfundata(i).correlation = [1.0.marg(2. 'GetD_Left(x(1). system. search algorithm analysisopt. 0].00 0. randomfield = 0. otherwise: given value % 'DDM': direct differentiaclotion..:) = [ 1 muSur 3. 0.' .001. 'FFD': forward analysisopt.expression = [num2str(d(i)) '-' . 0]. (0 < s <= 1) is the step size. 0].sim_point = 'origin'. end system.05..grad_flag = 'FFD'. % Maximum number of global iterations allowed in the % Maximum number of line iterations allowed in the % Tolerance on how close design point is to limit% Tolerance on how accurately the gradient points % 0: step size by Armijo rule.' . finite difference %analysisopt. analysisopt.:) = [ 6 probdata. state surface analysisopt.001. 0. num2str(sa) ')']..86603 0.66 muSur 3.parameter = distribution_parameter(probdata. for i = 1:length(d) gfundata(i).0 0. probdata.:) = [ 2 probdata.marg).step_code = 0. % epsilon (normal) probdata.' . 135 .0 0.evaluator = 'basic'.0 1.system=[length(d)].marg(3.0 0. 0..cov_max = 0. 0.scis_min = 1000. femodel = 0.0]. num2str(Mw) '. 0.ig_max = 100.0.e1 = 0. 0.. search algorithm analysisopt. 0. analysisopt.e2 = 0. analysisopt.stdv_sim = 1.txt'. system.0 0.0 0. % Sur (lognormal) 0.m % output_filename = 'outputfile_liqLeft. gfundata(i)..5 0..x(2). num2str(Ts) '. analysisopt.num_sim = 10000..% % Routine inputfile_liqLEFT. probdata.

% Moment of inertia of one pile (in^4) H = 2.Ip).R).'My'.epsilon.04*log(sa) .50*Ts + 0. S = subs('Sur + (FPiles + FPassive)/A'). 136 .566*log(ky)*log(sa) + 3.0.LEFT % lbf % From Bray and Travasarou (2007) eqn = log(d) .'g_abut'.sa) % function [D] = GetD_Left(Sur.H).5*d.244*(log(sa))^2 + .H.function [D] = GetD_Left(Sur.'W'.p_abut).^2 + . FPassive = vpa(FPassive). assignin('base'.'FPassive'.R. FPiles = subs('GetFPilesSmooth(0. % cm dsol = dsol/2.g_abut). FPassive = subs('GetFPassiveSmooth(0.Es).k_abut).7.333*(log(ky)). % inches (10 cm) k_abut = 21. assignin('base'.Ip. % Radius of one pile (feet) % Parameters for FPassive (lbf) g_abut = 4.[1 1e6]).10 .54. % psf ky = GetKySmooth(S. S = vpa(S). % kips/in/ft p_abut = 32.'alpha'. assignin('base'.'H'.5Ts % Ouput D: Residual displacement of abutment in longitudinal % direction (inches) % Parameters for FPiles (lbf) Es = 29000000.LEFT % Base area of potential failure surface (ft^2) -. assignin('base'.sa) % % Input Sur: Post-liquefaction undrained shear strength (psf) % alpha: Number of diameters. assignin('base'.A). assignin('base'.83*log(ky) .LEFT N = 6.8.alpha).'FPiles'.Mw.. % inches end D = dsol.1)<0 dsol = fzero(inline(eqn).'N'. assignin('base'.epsilon. % Width of back wall (feet) % Other parameters A = 55*71.alpha. % char() is required for Matlab 7 if feval(inline(eqn). assignin('base'.Es.alpha. % Young's modulus of steel (lbf/in^2) Ip = 2549. % Number of Piles My = 1381.W). assignin('base'. to points of fixity % epsilon: Error term in Bray an Travasarou (2007) % Ts: Initial fundamental period of the potential % sliding mass % Mw: Moment magnitude % Sa: Spectral acceleration at 1.'Sur'. assignin('base'.g_abut.54.0.( -1. 0.FPiles).My).alpha.p_abut.2.5*d...Ts.'k_abut'.k_abut.278*(Mw-7) + epsilon ). % Calculations syms d real assignin('base'.Ts.'Ip'. above and below liquefiable % layer.'A'. % inches else dsol = 1/2. assignin('base'.N).. % Plastic moment of one pile (kips*ft) R = 1.'l').FPassive). % lbf FPiles = vpa(FPiles).'p_abut'. assignin('base'.Sur).'R'. assignin('base'.N.'Es'.Mw. % kips/ft W = 43. % ky(S) -.My)'). assignin('base'. 1. % Thickness of liquefiable layer (feet) -.W)'). eqn = char(vpa(eqn)).5.

L*ft2in.H.alpha.R.N.Ip.My) Input Parameteres d: H: alpha: R: N: Es: Ip: My: Ouput F: Displacement (inches) Thickness of liquefiable layer (feet) Number of diameters.Ip.^RR). = 10.R. F = Feq*Vy. kips2lbf = 1000.^(1/RR)).N.function [F] = GetFPilesSmooth(d. Feq = b*deq + (1-b)*deq. = 0. My*kips2lbf*ft2in.alpha. L L My Vy dy b RR = = = = = H + 2*(alpha*2*R). to points of fixity Radius of one pile (feet) Number of piles Young's modulus of steel (lbf/in^2) Moment of inertia of one pile (in^4) Plastic moment of one pile (kips*ft) Total shear force in piles system (lbf) ft2in = 12.My) % % % % % % % % % % % % % % function [F] = GetFPilesSmooth(d. % % % % % feet inches lbf*in lbf inches deq = d/dy.H. return 137 . My*L^2/(6*Es*Ip). above and below liquefiable layer.Es.Es./((1+deq. (2*My/L)*N.

% lbf deq = (d-g_abut-dy)/dy.g_abut.5*p_abut/k_abut.p_abut.W) % % % % % % % % % % % % RR dy Vy function [F] = GetFPassiveSmooth(d. % b=0 Feq = 1 + deq. = 0.W) Input Parameteres d: Displacement (inches) g_abut: Abutment gap (inches) -. % inches = p_abut*1000*W/2.^RR). return 138 .p_abut./((1+deq.^(1/RR)).g_abut.Caltrans units k_abut: Abutment linear stiffness (kips/in/foot) Caltrans units p_abut: Abutment total passive capacity (kips/foot) Caltrans units W: Width of back wall (feet) F: Passive reaction at the abutment (lbf) Ouput = 10. F = Feq*Vy.k_abut.k_abut.function [F] = GetFPassiveSmooth(d.

end 139 . % -Sys = 0./((1+Seq.^(1/RR)).0004.^(1/RR)). % b=0 kyeq = 1 + Seq.0004*Sy. % psf Seq = (S-Sys-Sy)/Sy. elseif side == 'r' Sy = 0.LEFT & RIGHT Sy = 0.1473/0. % psf Seq = (S-Sys-Sy)/Sy./((1+Seq./((1+Seq.5*50000.5*50000.0895/0. % -Sys = 0. if side == 'l' Sy = 0. ky = kyeq*kys.function [ky] = GetKySmooth(S.side) % % Input S: Total equivalente shear strength of liquefiable layer % (includes passive reaction. % psf Seq = (S-Sys-Sy)/Sy.^(1/RR)). % psf kys = 0.1132/0.^RR). % psf kys = 0.piles reaction) % (psf) % side: 'l' or 'r' (left or right) % % Ouput ky: Yield coefficient RR = 200.5*50000. ky = kyeq*kys.0004. % psf kys = 0.0004. % -Sys = 0. elseif side == 'nl' % No Liquefaction Case . ky = kyeq*kys.^RR). % b=0 kyeq = 1 + Seq.^RR).0004*Sy. % b=0 kyeq = 1 + Seq.side) % function [ky] = GetKySmooth(S.0004*Sy.

Appendix B: Pile Design This section presents the design of the piles under the abutments and piers of the example described in Chapter 6. 141 .

142 .

143 .

144 .

145 .

146 .

147 .

148 .

149 .

150 .

151 .

152 .

153 .

154 .

155 .

156 .

157 .

158 .

159

Appendix C:

Ultimate Capacity Calculation Sheet—PYCAPSI

Below is a snapshot of the Excel spreadsheet developed by Professors Mokwa and Duncan to estimate the ultimate passive reaction against pile caps. The electronic file can be obtained by contacting Prof. Mokwa at: rmokwa@ce.montana.edu.

161

PEER REPORTS

PEER reports are available from the National Information Service for Earthquake Engineering (NISEE). To order PEER reports, th please contact the Pacific Earthquake Engineering Research Center, 1301 South 46 Street, Richmond, California 94804-4698. Tel.: (510) 665-3405; Fax: (510) 665-3420. PEER 2008/05 PEER 2008/04 Performance-Based Earthquake Engineering Design Evaluation Procedure for Bridge Foundations Undergoing Liquefaction-Induced Lateral Ground Displacement. Christian A. Ledezma and Jonathan D. Bray. August 2008. Benchmarking of Nonlinear Geotechnical Ground Response Analysis Procedures. Jonathan P. Stewart, Annie On-Lei Kwok, Yousseff M. A. Hashash, Neven Matasovic, Robert Pyke, Zhiliang Wang, and Zhaohui Yang. August 2008. Treatment of Uncertainties in Seismic-Risk Analysis of Transportation Systems. Evangelos Stergiou and Anne S. Kiremidjian. July 2008. Seismic Performance Objectives for Tall Buildings. William T. Holmes, Charles Kircher, William Petak, and Nabih Youssef. August 2008. An Assessment to Benchmark the Seismic Performance of a Code-Conforming Reinforced Concrete MomentFrame Building. Curt Haselton, Christine A. Goulet, Judith Mitrani-Reiser, James L. Beck, Gregory G. Deierlein, Keith A. Porter, Jonathan P. Stewart, and Ertugrul Taciroglu. August 2008. Bar Buckling in Reinforced Concrete Bridge Columns. Wayne A. Brown, Dawn E. Lehman, and John F. Stanton. February 2008. Computational Modeling of Progressive Collapse in Reinforced Concrete Frame Structures. Mohamed M. Talaat and Khalid M. Mosalam. May 2008. Integrated Probabilistic Performance-Based Evaluation of Benchmark Reinforced Concrete Bridges. Kevin R. Mackie, John-Michael Wong, and Božidar Stojadinović. January 2008. Assessing Seismic Collapse Safety of Modern Reinforced Concrete Moment-Frame Buildings. Curt B. Haselton and Gregory G. Deierlein. February 2008. Performance Modeling Strategies for Modern Reinforced Concrete Bridge Columns. Michael P. Berry and Marc O. Eberhard. April 2008. Development of Improved Procedures for Seismic Design of Buried and Partially Buried Structures. Linda Al Atik and Nicholas Sitar. June 2007. Uncertainty and Correlation in Seismic Risk Assessment of Transportation Systems. Renee G. Lee and Anne S. Kiremidjian. July 2007. Numerical Models for Analysis and Performance-Based Design of Shallow Foundations Subjected to Seismic Loading. Sivapalan Gajan, Tara C. Hutchinson, Bruce L. Kutter, Prishati Raychowdhury, José A. Ugalde, and Jonathan P. Stewart. May 2008. Beam-Column Element Model Calibrated for Predicting Flexural Response Leading to Global Collapse of RC Frame Buildings. Curt B. Haselton, Abbie B. Liel, Sarah Taylor Lange, and Gregory g. Deierlein. May 2008. Campbell-Bozorgnia NGA Ground Motion Relations for the Geometric Mean Horizontal Component of Peak and Spectral Ground Motion Parameters. Kenneth W. Campbell and Yousef Bozorgnia. May 2007. Boore-Atkinson NGA Ground Motion Relations for the Geometric Mean Horizontal Component of Peak and Spectral Ground Motion Parameters. David M. Boore and Gail M. Atkinson. May. May 2007. Societal Implications of Performance-Based Earthquake Engineering. Peter J. May. May 2007. Probabilistic Seismic Demand Analysis Using Advanced Ground Motion Intensity Measures, Attenuation Relationships, and Near-Fault Effects. Polsak Tothong and C. Allin Cornell. March 2007. Application of the PEER PBEE Methodology to the I-880 Viaduct. Sashi Kunnath. February 2007. Quantifying Economic Losses from Travel Forgone Following a Large Metropolitan Earthquake. James Moore, Sungbin Cho, Yue Yue Fan, and Stuart Werner. November 2006. Vector-Valued Ground Motion Intensity Measures for Probabilistic Seismic Demand Analysis. Jack W. Baker and C. Allin Cornell. October 2006. Analytical Modeling of Reinforced Concrete Walls for Predicting Flexural and Coupled–ShearFlexural Responses. Kutay Orakcal, Loenardo M. Massone, and John W. Wallace. October 2006.

PEER 2008/02 PEER 2008/01 PEER 2007/12

PEER 2007/11 PEER 2007/10 PEER 2007/09 PEER 2007/08 PEER 2007/07 PEER 2007/06 PEER 2007/05 PEER 2007/04

PEER 2007/03 PEER 2007/02 PEER 2007/01 PEER 2006/12 PEER 2006/11 PEER 2006/10 PEER 2006/09 PEER 2006/08 PEER 2006/07

and Nicolas Luco. Performance of Lifelines Subjected to Lateral Spreading. James Moore. CPT-Based Probabilistic Assessment of Seismic Soil Liquefaction Initiation. Allin Cornell.PEER 2006/06 PEER 2006/05 PEER 2006/04 PEER 2006/03 PEER 2006/02 PEER 2006/01 PEER 2005/16 PEER 2005/15 PEER 2005/14 PEER 2005/13 PEER 2005/12 PEER 2005/11 PEER 2005/10 PEER 2005/09 PEER 2005/08 PEER 2005/07 PEER 2005/06 PEER 2005//05 PEER 2005/04 PEER 2005/03 PEER 2005/02 PEER 2005/01 PEER 2004/09 PEER 2004/08 PEER 2004/07 PEER 2004/06 Nonlinear Analysis of a Soil-Drilled Pier System under Static and Dynamic Axial Loading. Porter. P. Ground Motions for Earthquake Simulator Qualification of Electrical Substation Equipment. Sackman. William Holmes. Eberhard. Seismic Qualification and Fragility Testing of Line Break 550-kV Disconnect Switches. Gee Liek Yeo and C. August 2006. Ibarra and Helmut Krawinkler. January 2006. Anne Kiremidjian. September 2005. A Taxonomy of Building Components for Performance-Based Earthquake Engineering. Gregory L. S.and Shelf-Mounted Equipment. September 2006. Workshop on Modeling of Nonlinear Cyclic Load-Deformation Behavior of Shallow Foundations. Kutter. July 2006. Stephen Tobriner. Yue Yue Fan. Jared M. Comerio. PEER Testbed Study on a Laboratory Building: Exercising Seismic Performance Assessment. Karl Telleen. and Don Clyde. Stanton. E. Allin Cornell. and Jerome L. Geoffrey R. Geoffrey Martin. Stewart. February 2005. Joe Maffei. First NEES/E-Defense Workshop on Collapse Simulation of Reinforced Concrete Building Structures. and Meredith Williams. Global Collapse of Frame Structures under Seismic Excitations. and A. Van Nuys Hotel Building Testbed Report: Exercising Seismic Performance Assessment. Gregory L. and Justin Phalen. Fenves. Performance Characterization of Bench. Scott A. Sivapalan Gajan. Yoojoong Choi. Probabilistic Seismic Evaluation of Reinforced Concrete Structural Components and Systems. . Tara Hutchinson. Takhirov. Fragility Basis for California Highway Overpass Bridge Seismic Decision Making. Nelson. April 2006. November 2006. and Bruce L. Danya Mohr. October 2005. and John F. Jonathan P. Helmut Krawinkler. Maziar Motahari. Mary C. June 2005. and Ariane Fehrenkamp. Mackie and Božidar Stojadinović. E. Comerio. R. September 2005. Dimitrios Konstantinidis and Nicos Makris. Ozgur Yazali. Performance-Based Regulation and Regulatory Regimes. April 2006. A Guide to Seismic Safety on the UC Berkeley Campus. Marc O. Tae Hyung Lee and Khalid M. January 2005. Charles Menun. editor. Pacific Earthquake Engineering Research Center Highway Demonstration Project. R. August 2005. November 2005. Gang Wang and Nicholas Sitar. April 2006. Peter J. June 2005. Ashford and Teerawut Juirnarongrit. May 2006. Paolo Bazzurro. Martin. Samit Ray Chaudhuri and Tara C. and Eric Fujisaki. Experimental and Analytical Studies on the Seismic Response of Freestanding and Anchored Laboratory Equipment. R. August 2006. Der Kiureghian. Advanced Seismic Assessment Guidelines. January 2005. Bruce L. Takhirov. Shakhzod M. Fenves. Zach Price. Kutter. Empirical Characterization of Site Conditions on Strong Ground Motion. April 2006. Luis F. Kevin R. and Yuki Nakayama. Chad Harden. Test Applications of Advanced Seismic Assessment Guidelines. editor. Numerical Modeling of the Nonlinear Cyclic Response of Shallow Foundations. Tyler Ranf. J. Eric Fujisaki. September 2004. Tara Hutchinson. Armen Der Kiureghian. C. September 2005. January 2005. and Robert W. May and Chris Koski. Seed. Mary C. Shakhzod M. Kayen. Moss. Christopher Stearns and André Filiatrault. Stochastic Characterization and Decision Bases under Time-Dependent Aftershock Risk in Performance-Based Earthquake Engineering. March 2006. Seismic Response and Reliability of Electrical Substation Equipment and Systems. Stewart. Bracing Berkeley. B. Damage Accumulation in Lightly Confined Reinforced Concrete Bridge Columns. Graves. Junho Song. Electrical Substation Equipment Interaction: Experimental Rigid Conductor Studies. Keith A. July 2005. Chad Harden. R. Hutchinson. Nesrin Basoz. Mosalam.

André Filiatrault. Mahmoud M. Shake Table Tests and Analytical Studies on the Gravity Load Collapse of Reinforced Concrete Frames. October 2004. Cox and Scott A. Kerri Anne Taeko Tokoro. Software Framework for Collaborative Development of Nonlinear Dynamic Analysis Program. Comerio. Performance of Circular Reinforced Concrete Bridge Columns under Bidirectional Earthquake Loading. Probabilistic Response Assessment for Building-Specific Loss Estimation. February 2004. A Technical Framework for Probability-Based Demand and Capacity Factor Design (DCFD) Seismic Formats. Allin Cornell. September 2003. Kenneth E. Analytical Investigations of New Methods for Reducing Residual Displacements of Reinforced Concrete Bridge Columns. Hachem. Eduardo Miranda and Hesameddin Aslani. April 2003. Timothy Graf and James O. Kevin Mackie and Božidar Stojadinović. Michael Berry and Marc Eberhard. and Vitelmo V. Anderson. Peter Fajfar and Helmut Krawinkler. March 2004. Catherine Pagni and Laura Lowes. Implementation Manual for the Seismic Protection of Laboratory Contents: Format and Case Studies. Jun Peng and Kincho H. August 2004. Bertero. Ashford. April 2004. John W. October 2003. and Scott A. Marco Casari and Simon J. Anderson and Vitelmo V.PEER 2004/05 PEER 2004/04 PEER 2004/03 PEER 2004/02 PEER 2004/01 PEER 2003/18 PEER 2003/17 PEER 2003/16 PEER 2003/15 PEER 2003/14 PEER 2003/13 PEER 2003/12 PEER 2003/11 PEER 2003/10 PEER 2003/09 PEER 2003/08 PEER 2003/07 PEER 2003/06 PEER 2003/05 PEER 2003/04 PEER 2003/03 PEER 2003/02 PEER 2003/01 PEER 2002/24 PEER 2002/23 PEER 2002/22 Performance-Based Seismic Design Concepts and Implementation: Proceedings of an International Workshop. Seismic Demands for Nondeteriorating Frame Structures and Their Dependence on Ground Motions. Sequencing Repairs after an Earthquake: An Economic Approach. Finite Element Reliability and Sensitivity Methods for Performance-Based Earthquake Engineering. Malley. Nilanjan Mitra. Wilkie. and Arash Altoontash. Law. Mahin. February 2003. Seismic Demands for Performance-Based Design of Bridges. Natalie Gibson. July 2004. William T. April 2002. Response Assessment for Building-Specific Loss Estimation. February 2004. Seismic Performance of an Instrumented Tilt-up Wall Building. Murat Melek. Experimental Assessment of Columns with Short Lap Splices Subjected to Cyclic Loads. Allin Cornell. Jack W. Uncertainty Specification and Propagation for Loss Estimation Using FOSM Methods. November 2003. Mahin. September 2004. Moehle. Fifth U. Evaluation and Application of Concrete Tilt-up Assessment Methodologies. Terje Haukaas and Armen Der Kiureghian. Greg L. Ashford. Kenneth John Elwood and Jack P. September 2003. Predicting Earthquake Damage in Older Reinforced Concrete Beam-Column Joints. September 2003. James C. and Jack P. February 2004. Eduardo Miranda and Shahram Taghavi. September 2003. Effects of Large Velocity Pulses on Reinforced Concrete Bridge Columns. Mahin. Seismic Performance of Masonry Buildings and Design Implications.S. August 2003. August 2003. Moehle. . Orozco and Scott A. Junichi Sakai and Stephen A. Ricardo Antonio Medina and Helmut Krawinkler. editors. May 2004. April 2002. Bertero.-Japan Workshop on Performance-Based Earthquake Engineering Methodology for Reinforced Concrete Building Structures. Laura N. April 2004. Characterization of Large Velocity Pulses for Laboratory Testing. Fatemeh Jalayer and C. James C. Rodgers and Stephen A. Baker and C. Janise E. October 2004. April 2002. Effects of Connection Hysteretic Degradation on the Seismic Behavior of Steel Moment-Resisting Frames. Stephen A. A Beam-Column Joint Model for Simulating the Earthquake Response of Reinforced Concrete Frames. Wallace. Performance Models for Flexural Damage in Reinforced Concrete Columns. Ashford. Holmes and Mary C. November 2003. and Joel Conte. Performance of Beam to Column Bridge Joints Subjected to a Large Velocity Pulse. Lowes.

Documentation and Analysis of Field Case Histories of Seismic Compression during the 1994 Northridge. September 2001. July 2002. Marc O. Assessment of Reinforced Concrete Building Exterior Joints with Substandard Details. September 2001. André Filiatrault. Allen L. and Harry W. August 2002. Bray. Investigation of Sensitivity of Building Loss Estimates to Major Uncertain Variables for the Van Nuys Testbed. and Jun Han Yoo. December 2002. October 2002. and Lawrence D. Moore II. December 2001. Brad T. September 2002. Centrifuge Modeling of Settlement and Lateral Spreading with Comparisons to Numerical Analyses. Hutchinson. and Khalid M. 4–5 October 2001. and Yan Xiao. California. and Chatpan Chintanapakdee. Reaveley. and I. Component Testing. Richardson. Zerbe and Anthony Falit-Baiamonte. Nicos Makris and Jian Zhang.W. Chai. November 2002. Economic-Engineered Integrated Models for Earthquakes: Socioeconomic Impacts. November 2001. Specifications. Anil K. Sivapalan Gajan and Bruce L. Asadollah Esmaeily-Gh. Estimation of Uncertainty in Geotechnical Properties for Performance-Based Earthquake Engineering.S.-Japan Workshop on Performance-Based Earthquake Engineering Methodology for Reinforced Concrete Building Structures. Structural Characterization and Seismic Response Analysis of a Highway Overcrossing Equipped with Elastomeric Bearings and Fluid Dampers: A Case Study. June 2002. Porter. Idriss. Richard O. Steven L. Patrick M. Nicos Makris. Beck. Rocking Response of Equipment Anchored to a Base Foundation. Boulanger. Y. May. Eberhard. Paolo Gardoni. Armen Der Kiureghian. and Pedro Arduino. Kutter. December 2002. September 2001.C. James M. May 2002. Kramer and Ahmed-W. January 2003. Elgamal. Takhirov. Kelly and Shakhzod M. Peter Gordon. Steven L. July 2002. Black. Jennifer Soderstrom. Statistics of SDF-System Estimate of Roof Displacement for Pushover Analysis of Buildings. Kramer. and Rustem V. Earthquake. Rakesh K. Stability Analysis and Characterization of Buckling-Restrained Unbonded Braces . and Ian Aiken. Goel. Cameron Black. and Michael P. Shaikhutdinov. James E. Peter J.PEER 2002/21 PEER 2002/20 PEER 2002/19 PEER 2002/18 PEER 2002/17 PEER 2002/16 PEER 2002/15 PEER 2002/14 PEER 2002/13 PEER 2002/12 PEER 2002/11 PEER 2002/10 PEER 2002/09 Fourth U. Stewart. John F. and Jonathan D. and Christopher Stearns. Justin Nadauld. Constantin Christopoulos. Pantelides. TM PEER 2002/08 PEER 2002/07 PEER 2002/06 PEER 2002/05 PEER 2002/04 PEER 2002/03 PEER 2002/02 PEER 2002/01 PEER 2001/16 PEER 2001/15 PEER 2001/14 PEER 2001/13 .6 Earthquake. Guidelines. Whang. Jon Hansen. T. Effects of Fault Dip and Slip Rake on Near-Source Ground Motions: Why Chi-Chi Was a Relatively Mild M7. Daniel H. The Use of Benefit-Cost Analysis for Evaluation of Performance-Based Earthquake Engineering Decisions. Roeder. Consortium of Organizations for Strong-Motion Observation Systems and the Pacific Earthquake Engineering Research Center Lifelines Program: Invited Workshop on Archiving and Web Dissemination of Geotechnical Data. R. Aagaard. and Seismic Performance Characterization of Nonstructural Building Components and Equipment. Modeling Soil Liquefaction Hazards for Performance-Based Earthquake Engineering. Seismic Performance of Pile-Wharf Connections. Damage to Bridges during the 2001 Nisqually Earthquake. September 2002. Stallmeyer. September 2002. Probabilistic Models and Fragility Estimates for Bridge Components and Systems. December 2001. Jonathan P. Comerio and John C. Chris P. and Thomas H. Analytical and Experimental Study of Fiber-Reinforced Strip Isolators. December 2001. December 2002. Nonstructural Loss Estimation: The UC Berkeley Case Study. Jones. Heaton. Chopra.H. August 2002. Charles W. Hall. Seismic Behavior of Bridge Columns Subjected to Various Loading Patterns. Tyler Ranf. Barriers to Adoption and Implementation of PBEE Innovations. R.M. December 2002. Mosalam. Mary C. Smith.-Japan Workshop on Performance-Based Earthquake Engineering Methodology for Reinforced Concrete Building Structures. The Third U. Keith A. December 2002. Inelastic Seismic Response of Extended Pile Shaft Supported Bridge Structures. James L. Nicos Makris and Cameron J. Berry. February 2001.S. Robert Graff.

Jonathan P. Andrew S. Structural Engineering Reconnaissance of the August 17. Rehabilitation of Nonductile RC Frame Building Using Encasement Plates and Energy-Dissipating Devices. Calderone. Moehle. November 1999. Jonathan D. September 2000. November 1999. Paul G. Andrew S. The Second U. Experimental and Computational Evaluation of Reinforced Concrete Bridge Beam-Column Connections for Seismic Performance. Anderson. Shyh-Jeng Chiou. Elwood. March 2001.PEER 2001/12 PEER 2001/11 PEER 2001/10 PEER 2001/09 PEER 2001/08 PEER 2001/07 PEER 2001/06 PEER 2001/05 PEER 2001/04 PEER 2001/03 PEER 2001/02 PEER 2001/01 PEER 2000/10 PEER 2000/09 Development of Geotechnical Capabilities in OpenSees. Anil K. Robert W. Kenneth J. March 2000. Stewart. Jonathan P. September 2001. Hose and Frieder Seible. Boris Jeremi . January 2001. and Shakhzod M. Jonathan P. Building Vulnerability Studies: Modeling and Evaluation of Tilt-up and Steel Reinforced Concrete Buildings. Framing Earthquake Retrofitting Decisions: The Case of Hillside Homes in Los Angeles. editors. Dawn E. 1999 Earthquake: Kocaeli (Izmit). Performance-Based Evaluation of Exterior Reinforced Concrete Building Joints for Seismic Excitation. May 2001. Moehle. Abrahamson. May. Seismic Evaluation and Retrofit of 230-kV Porcelain Transformer Bushings. The Rocking Spectrum and the Shortcomings of Design Guidelines. John W. Analytical and Experimental Study of Fiber-Reinforced Elastomeric Isolators. September 2001. Clay J. and Jerome L. Goel. Wallace. Stiffness Analysis of Fiber-Reinforced Elastomeric Isolators. PEER 2000/08 PEER 2000/07 PEER 2000/06 PEER 2000/05 PEER 2000/04 PEER 2000/03 PEER 2000/02 PEER 2000/01 PEER 1999/14 PEER 1999/13 PEER 1999/12 PEER 1999/11 . Jian Zhang and Nicos Makris. Whittaker. and Mehmet B. Popov and Shakhzod M. December 1999. Organizational and Societal Considerations for Performance-Based Earthquake Engineering. August 2001. ed. December 2001. and Alicia Kitsuse. Andrew S. Amplification Factors for Spectral Acceleration in Active Regions. Behavior of Reinforced Concrete Bridge Columns Having Varying Aspect Ratios and Varying Lengths of Confinement. Bertero. Kelly. Stewart. and Norman A. Mehrdad Sasani. Armen Der Kiureghian. Takhirov. and Andrew S.-Japan Workshop on the Effects of Near-Field Earthquake Shaking. Amir S. Amir S. Gregory L. Stanton. Takhirov. and Eric Fujisaki. Gilani. Whittaker. November 2001. Ground Motion Evaluation Procedures for Performance-Based Design. July 2000. Henry Ho. Thalia Agnanos. Whittaker. Amir S. J. Anthony J. November 2000. Chung-Che Chou and Chia-Ming Uang. January 2001. Nels Roselund. Seismic Evaluation and Analysis of 230-kV Disconnect Switches. Andrew H. Gregory L. Yoojoong Choi. July 2000. Liu. Performance Evaluation Database for Concrete Bridge Components and Systems under Simulated Seismic Loads. Takhirov. Wallace. Jack P. Whittaker. Nicos Makris and Dimitrios Konstantinidis. April 2001. James M. December 1999. Fenves. Lehman. Hsiang-Chuan Tsai and James M. Khalid Mosalam. July 2000. Somerville. Vitelmo V. Seismic Response Analysis of Highway Overcrossings Including Soil-Structure Interaction. A Modal Pushover Analysis Procedure to Estimate Seismic Demands for Buildings: Theory and Preliminary Evaluation. Baturay. July 1999. Pantelides. Graves. and John F. March 2000. December 1999. U. John J. Sackman. Kelly and Shakhzod M. Stewart. Reaveley. Taejin Kim. Bray. Andrew S. and Lawrence D. Fenves. Detlof von Winterfeldt. September 2001. An Evaluation of Seismic Energy Demand: An Attenuation Approach. Andrew Whittaker. Egor P. and Eric Fujisaki. Chun-Hao Chen. and Jack P. Gilani. Development of an Electrical Substation Equipment Performance Database for Evaluation of Equipment Fragilities.-Japan Workshop on Performance-Based Earthquake Engineering Methodology for Reinforced Concrete Building Structures. Further Studies on Seismic Interaction in Interconnected Electrical Substation Equipment. December 2000. April 1999. Turkey. Vitelmo V. Experimental Study of Large Seismic Steel Beam-to-Column Connections. Bertero. Whittaker. Kee-Jeung Hong.S. James C. Chris P. Gilani. and Khalid M. Chopra and Rakesh K. Chandra Clyde. Halil Sezen. Peter J. Cover-Plate and Flange-Plate Reinforced Steel Moment-Resisting Connections.S. Naito. Mosalam. Yael D.

-Japan Workshop on Performance-Based Earthquake Engineering Methodology for Reinforced Concrete Building Structures. Burby. Jonathan P. Policy Analysis. James C. Cofer. Bertero. Seismic Evaluation of 550 kV Porcelain Transformer Bushings. Chopra and Rakesh Goel. Empirical Evaluation of Inertial Soil-Structure Interaction Effects. Raymond B. Fenves. February 1999. Sackman. Raymond J. November 1998. Jerome L. December 2000. Peter J. Nicos Makris and Jian Zhang.PEER 1999/10 PEER 1999/09 PEER 1999/08 PEER 1999/07 PEER 1999/06 PEER 1999/05 PEER 1999/04 PEER 1999/03 PEER 1999/02 PEER 1999/01 PEER 1998/08 PEER 1998/07 PEER 1998/06 PEER 1998/05 PEER 1998/04 U. Pacific Earthquake Engineering Research Invitational Workshop Proceedings. Andrew S. Capacity-Demand-Diagram Methods for Estimating Seismic Deformation of Inelastic Structures: SDF Systems. Lehman and Jack P. Gregory L. November 1999. December 1998. Adrian Rodriguez-Marek. James C. Economics and Earthquake Engineering. Behavior and Failure Analysis of a Multiple-Frame Highway Bridge in the 1994 Northridge Earthquake. October 1999. May. November 1998. Vitelmo V. and Eric Fujisaki. Documentation of Strengths and Weaknesses of Current Computer Analysis Methods for Seismic Performance of Reinforced Concrete Members. May 1998. and Robert Wood. Jonathan D. Adoption and Enforcement of Earthquake Risk-Reduction Measures. Anderson and Xiaojing Duan. Nicos Makris and Shih-Po Chang. Seed. Repair/Upgrade Procedures for Welded Beam to Column Connections. Whittaker. and Gregory L. Fenves and Michael Ellery. Amir S. Amir S. October 1998. December 1999. Juan W. Anil K. T. Gilani. Charles Menun and Armen Der Kiureghian. Whittaker. and Raul Bertero. May 1998.S. October 1999. April 1999. Seismic Evaluation of 196 kV Porcelain Transformer Bushings. Fenves. September 1998. William F. February 1999. Armen Der Kiureghian. Bray. November 1999. Rocking Response and Overturning of Anchored Equipment under Seismic Excitations. Anderson. 1998: Defining the Links between Planning. Gregory L. Mary Comerio and Peter Gordon. Interaction in Interconnected Electrical Substation Equipment Subjected to Earthquake Ground Motions. Rocking Response and Overturning of Equipment under Horizontal Pulse-Type Motions. Task 3 Characterization of Site Response General Site Categories. and Kee-Jeung Hong. Envelopes for Seismic Response Vectors. Seismic Performance of Well-Confined Concrete Bridge Columns. Dawn E. Chavez. Performance Improvement of Long Period Building Structures Subjected to Severe Pulse-Type Ground Motions. Fenves. Jens Feeley. May 14–15. Effect of Damping Mechanisms on the Response of Seismic Isolated Structures. and Norman Abrahamson. Gregory L. PEER 1998/03 PEER 1998/02 PEER 1998/01 . July 1999. Gilani. and Andrew S. Nicos Makris and Yiannis Roussos. Moehle. Stewart.

Mosalam. Holmes.edu/publications/peer_reports.ONLINE REPORTS The following PEER reports are available by Internet only at http://peer.html PEER 2008/101 Seismic Performance Objectives for Tall Buildings. August 2008. PEER 2007/101 Generalized Hybrid Simulation Framework for Structural Systems Subjected to Seismic Loading. and Nabih Youssef. PEER 2007/100 Seismic Evaluation of Reinforced Concrete Buildings Including Effects of Masonry Infill Walls. July 2007. William Petak. July 2007. William T. Tarek Elkhoraibi and Khalid M.berkeley. Alidad Hashemi and Khalid M. Charles Kircher. Mosalam. .

Sign up to vote on this title
UsefulNot useful