Po Laws SEISMIC DESIGN OF HIGHWAY BRIDGE FOUNDATIONS. :."°~" @ VOL. I] DESIGN PROCEDURES ANG,GUIDELINES |=“ “* US Department FHWA/RD-86/102 ‘of Fanspertation Federal Highway Finel Report Administration June 1986 Earthquake Protection of a Transportation Structures his acura i enable te US public rag te National Tachi! efecmatin Service, Sringi, Vin 22164Technical Report Documentotion Poge [repent ier 2 Caveinment Recension Nor 3 Recipients G FHWA/RD-86/102 Tied 3 Repent Dare SEISMIC DESIGN OF HIGHWAY BRIDGE FOUNDATIONS June 1986 Volume II. Design Procedures and Guidelines ‘. Pectorming Organtsatton Code Pevtooming Veet mle er eeepc eeepc eee Ignatius (Po) Lam and Geoffrey R. Martin ¥. Peiloming Organ jon Wane and Ad V8. Wor Unit No: (FRAT ‘The Earth Technology Corporation 35a1062 3777 Long Beach Boulevard TH. Contato Gren Hay Long Beach, California 90807 DIPH61-83-c-00138 TS. Type of Report and Pavtod C Final Report September 1983 - June 1986] To Office of Engineering and Highway Operations Federal Highway Adninistration 7 6300 Georgetown Pike TA Spensorng haency Code McLean. Virginia 22101-7296 [TE svpptamentay FHWA contract manager (COTR): John O'Fallon (HNR-10), 1985-1986 James Cooper (HNR-10), 1984-1985 fring Agency Nome nd Ad Ta Abenecr aris report provides specific procedures for the setemic design of bridge foundations and abutments based on hand-calculation methods using design ehaces and Hiamene puethods. This report suppleuents the Seismic Design Guidelines For Highway Bridges published as a design specification by the Auerican Areociation of State Highway and Transportation Officials (AASHTO) in 1983. Design’ procedures are presented for footings, piles, drilled shafts, and abutments. Comments on site investigation procedures and in situ and laboratory seat saB 2*¢ also provided in relation to determination of site soil parameters or analyses. | Additional comments are provided on earthquake-induced liguefaction and slope stability as they affect bridge foundation design. The selected design Procedures vere verified by comparing analytical predictions with experimencal data This volume is the second in a series. The others in the series are: FHWA No. Vol. No. Short Title RD-86/101 I Executive Summary RD-86/103 ur Example Problems and Sensitivity studies Wey Wore Bridges, Seismic design, Earthquakes, | No restrictions. This document is Foundations, Abutments, Piles, available to the public through the Footings, Drilled shafts National Technical Information Service, Springfield, Virginia 22161. 19S Cland. Ta eed Te Waal Faas | Pico Unclassified Unelassi fied 181 Form DOT F 1700.7 (#72)TABLE OF CONTENTS VOLUME I: EXECUTIVE SUMMARY Section BACKGROUND» 2. eee ee eee ee OBJECTIVES AND SCOPE OF WORK. ~~... SENSITIVITY STUDIES-.-...-..... SOIL PROPERTIES... - see ee ee eee DYNAMIC GROUND STABILITY... 2... 0 FOOTING FOUNDATIONS. -- 2.1... ee PILES AND PILE GROUPS... +. ee. eee DRILLED SHAFTS... eee ee ee ABUIMENTS- 2 ee eee eee ee ee REPORT OUTLINE. + eee ee ee eee REFERENCES 6 + eee ee pe eee ee ee VOLUME II: DESIGN PROCEDURES AND GUIDELINES 1.0 2.0 Section INTRODUCTION» 2 ee ee ee Lil Background 2. eee eee ee ee Leo Hi Pnepooetsdaiteretisraes test eee cere 1.3 Project Organization .. +... 1-4 Report Organizations see. eee REVIEW OF SEISMIC DESIGN GUIDELINES AND RESPONSE ANALYSIS PROCEDURES... 1. 2.1 FHWA/AASHTO Seismic Design Guidelines 2.2 Foundation Modeling for Bridge Respons: 2.4 Modeling Foundation Damping... « « it e Analysis 2.3 Simplification of the Foundation Stiffness system Page war aaonne 10 cy 12 Page3.0 4.0 5.0 6-0 TABLE OF CONTENTS (continued) Section SOIL PROPERTY EVALUATION. 2-2 ee ee eee eee Sele eamarad ete tetiectecteatte diet crtc ered a tiateeteaetettectactete 3.2 Field Program and General Site Soil Classification ~ 3.3 Soil Parameters Derived From Standard Penetration Tests 3-4 Parameters Derived From Coné Penetrometer Tests . . - 3.5 Laboratory Testing Program -- +--+ ++ e eee ee 3.6 Modulus of Elasticity... +--+ ee eee eee DYNAMIC GROUND STABILITY EVALUATION «+++ ess ee ee aad 42 4.3 4h 45 Introduction + see eee ee ee ee Liquefaction Potential Evaluation - +... +s +s Liquefaction of Clayey Soils ss... eee ee ee Slope Stability Evaluation. . 6s. ee eee eee Dynamic Settlement Evaluation... +++ eee eee ANALYSIS OF FOOTINGS 6 6 eee ee ee ee eee 51+ Ganamad tier teeeiiaeetelittestsiecederetiste etme edits te 5.2 General Form of Stiffness Matrix... +... eee 5.3 General Stiffness Equation of a Rigid Footing... « 5.4 Stiffness Coefficients of a Circular Surface Footing 5.5 Rectangular Footings... +e eee eee eee ee 5.6 Enbedment Effects 2 ee eee ee ee eee ee 5.7 Stability of Footings Under Earthquake Load... 5.8 General Guidelines and Commentary» ++ +--+ ++ Amacesisiod priest eesot tices a oer tense nits Gat Gamacad tte iist et stated da tstisiststecetistetstetsiiete 6.2 Development of p-y Curves for Sand - ++ ee eee 6.3 Development of pry Curves for Clay - +++ eee + uh Page Ww Ww 18 19 aL 21 22 27 27 27 ae & 40 40 40 42 42 44 44 a7 49 52 52 55 587.0 8.0 9-0 TABLE OF CONTENTS (continued) ANALYSIS OF DRILLED SHAFTS. - +. + - Te 762 763 Teh General eee eee eee ee Lateral Load-deformation Behavior Axial Load-deformation Behavior . General Guidelines and Commentary ANALYSIS OF ABUTMENTS . 1... +e Bel 8-2 8.3 General ss ee eee te eee Seat Type Abutments.-. 1+ + Integral Abutments .-. +--+ EXAMPLE PROBLEM OF SEISMIC DESIGN OF A 9.1 9.2 9.3 904 9.5, 9.6 General se see ee ee ee ee Problem Description. +. + + + + Liquefaction Potential... 2. - Foundation Stiffmess - ++ ++ + ‘TYPICAL Section 6.4 Linear Subgrade Modulus... . 6.5 Analysis Methods for Lateral Loading on Piles « 6.6 Development of Load Transfer Characteristics for Axial Loading... + 1s 6.7 analysis Methods for Axial Loading on Piles . . 6.8 Pile Head Stiffness Matrix of a Single Pile . . 6.9 Development of Pile Group Stiffnesses . . . . 6-10 Individual Pile Head Forces and Moments - 6.11 Pile Group Effects -. +++. 6.12 General Guidelines and Conmentary BRIDGE Dynamic Response Analysis of the Overall Bridge Conclusion 2+ +e eee eee iv Foundation 61 65 72 76 81 83 86 87 89 LoL 101 103 1 un 122 122 123 13h ian Lan 141 14 144 147 157‘TABLE OF CONTENTS (continued) section 0.0 SUMMARY AND RECOMMENDATIONS © s+ +s ttt ol Summarys seer te ee tte 10.2 Recommendations +--+ ests ttt 11-0 REFERENCES. 2 ee ee eerste VOLUME III: APPENDIXES APPENDIX A. EXAMPLE PROBLEM FOR LIQUEFACTION CALCULATION. APPENDIX B SENSITIVITY STUDY AND EXAMPLE PROBLEMS FOR FOOTINGS Bel Dynamic Stiffness of a Footing - +++ +++ B.2 Embedment Sensitivity Study. + + +++ B.3 Example Problems for Footings. + + + + APPENDIX C EXAMPLE OF PILE ANALYSES. = + 2 s+ r 8 tt Gel Generales eee eee ¢.2. Example Problem for Computer Solution. ¢.3 Abstract of Program “BNCOL". - + ++ + G.4 Example of Hand-Calculation Procedure. c.5 Example of Pile Group Stiffness Calculation. ‘APPENDIX D SENSITIVETY STUDY OF PILES AND PILE GROUPS. APPENDIX E COMPUTER PROGRAM PILECAPs + + + + +s 5 0 * gel Introduction. seer eect a2 Program Descriptions + +--+ ett 159 159 160 162 18 19 19 20 gest 52 67 67 67TABLE OF CONTENTS (continued) E.3 Solution Process... eee ee eet ete ee 69 BA OUtpUE . ee ee 69 E.5 PILECAP Input Guide.» 2 ee eee ee ee n £.6 PILECAP Source Code Listing... - +++ eer tees 73 E.7 Example of PILECAP Input Data... 1-2-2 +++ eee 80 B.8 Example of PILECAP Output Data... +--+ sees eee aL APPENDIX F EXAMPLE PROBLEM FOR DRILLED SHAFT. --- - ee eee eee 96 Fel Generale se eee ee ee ee 96 F.2 Problem Description, . +e e+e eer eee eee 97 F.3 Top-of-Shaft Stiffness Matrix Détermination. . .. - + + 7 P.4 Dynamic Response Analysis.» ++ ee ee eee ee ee 104 P.5 Load Distribution Along Drilled shaft... ..- +--+ 105 APPENDIX G EXAMPLE PROBLEM FOR ABUIMENT © eee eee ee eee eee | 107 G.l Problem Description 2. eee eee eee eee ee 107 G.2 Stiffness of Abutment Wall... eee ee ee es 107 G.3. Stiffness of Footing... -ee eee ee ee 108 G.4 Combined Stiffness Matrix at Bottom of Abutment Wall se eee ee ee ee ee ee 109 APPENDIX H SENSITIVITY STUDY OF BIGHT BRIDGES ... +--+ +++ eee 1BLIST OF TABLES Table Le 2. 3. 10. ne a. 13. ue 15. 16. W. Description of bridges used in sensitivity study... « Linear elastic pile solution for sand, Ey Linear elastic pile solution for clay, Ey where kg/KjT=O.l- +--+ ee eee Linear elastic pile solution for clay, Ey where Kg/K]T = 0.2 es ee ee rene Linear elastic pile solution for clay, EB where Kg/kjT = O05 ee eee ee eee Linear elastic pile solution for clay, By where Ko/KjT = 1-0. +0 - ee ee Linear elastic pile solution for clay, Bs where ko/kyT = 2-0. - 0+ e+ eee Linear elastic pile solution for clay, Eg where kg/KjP = 5-00 - eee eee ee Linear solution for a rigid pler or pile, where ko/kjL = Ole eee eee eee Linear solution for a rigid pier or pile, where ko/kyL = 0.2 - 02+ ee eee Linear solution for a rigid pier or pile, where kg/kqL = 05 se ee eee eee Linear solution for a rigid pier or pile, where ko/kqb = 10+. eee ee eee Linear solution for a rigid pier or pile, where kg/kqL = 2.0.00 eee eee Linear solution for a rigid pier or pile, where ko/kjL= 50+ eee ee ee ee Linear solution for a rigid pier or pile, where Kg "Oe e eee ee eee Linear solution for a rigid pier or pile, sigue Lig’ bie ioe 3 ako + ke ae By = ko + Eg = ko + Ey = "ko + By ko + By = ko + Ex = ko + Ex = kg + tae ke ke az ke kz ke ke where kp 0. eee cee eee ee tte ee te Bridge model input parameters and multimode vibration solutions for upper and lower bound soil conditions... +--+. +++ vit Page 14 94 95 96 7 98 99 100 114 1s 116 ay 118 ug 120 a2. 152LIST OF FIGURES Figure Page Le Work plans oe es ee ee et ee ee eee 4 2. Highway bridge foundations»... +. ee ee ee aL 3. Comparison of vibration frequency of fixed base model with foundation springs model for 8 bridges»... -- +2 +05 15. 4, Soil behavior type classification chart. . +... 22. ee ee 23 5. Expanded soil behavior type classification chart with equivalent overburden normalized friction angle and relative density trends. se eee ee eee et ee ee 6. Typical variation of shear modulus with shear strain amplitude . 25 7. Chart for evaluation of Liquefaction potential for different magnitude earthquakes see ee ee te ee ee ee OD 8. Relationship between Cy and effective overburden pressure... . 31 9. Principles of analytical approach (total stress) to Liquefaction potential evaluations. eee ee ee ee ee eR 10. Effective stress approach to ‘liquefaction evaluation showing effect of permeability. ©. +e ee eee ee ee ee 88 Al. Maximum distance to significant liquefaction as a function of earthquake magnitudes 6 eee ee ee tee ee ee te 12s, Schematic illustration of sliding block approach for displacement caleulationge » see tee tee ee tee eens |B 13. Effect of relative density on settlement in 10 cycles... 1. 39 14. Form of footing stiffness matrix... . ee. eee eee ML 15. Definition of geometry parameters and coordinate systems for a doobhagt cred eteders ree rctatte tne iis ernie rsne test tretttertas 16. Procedure for calculating equivalent radius of a rectangular itocetnig Pete eeeriet oeeettc terertertsreertstieacerieh eeerierraeeeee ete eetIENSy 17. Shape factor for rectangular footings... ee ee ee eee 46 18. Embedment factor of footings... es. eee ee ee ee eB 19. Consideration of footing stability for earthquake Loading fend Eonar stereo erot mre rate tee outa ai ente erst saiect sree ttetr50. 20. Three-dimensional soil~pile interaction (after Bryant and Matlock cq 97? yetistiet-sieiteteiecodistioeiitteacaiisteepineraaistt ais eeeentt5: 21. Cj), Cg, and C3 coefficients for p-y curve calculation for sand (after Bogard and Matlock, 1980) ...+-.+....- 37 22. Modulus of subgrade reaction, kj, for sand (after O'Ned11 and Murchison, 1983)) + sete ee et ee ee ee ee ee 5D 23. Hyperbolic p-y curve shape for sand and clay... .. 2... 60 24. Comparison of linear and nonlinear pile solutions for sand... 63 25. Comparison of linear and nonlinear pile solutions for clay... 66 26. Variation of equivalent linear subgrade modulus for sand and 27. Linear pile solution in sand (see table 2)...-..-+...- 71 28. Linear pile solution in clay (see tables 3 through 8)... ... © 73 29. Representation of axially loaded pile behavior and soil reaction 74 30. Procedure for development of axial load-displacement characteristics of a pile. se ee eee eee ee eB Bl. Effects of superposition of cyclic load on static lod... 1. 79 32. Form of a pile-head stiffness matrix. s+... eee ee ee 82 vittLIST OF FIGURES (continued) Figure 33. Calculation of pile group stiffness matrix. +... ++. ee 34. Problem description of a drilled shaft... ...-+...+- 35. Idealized subsurface model test for a clay profile... ..- 36. Idealized subsurface model test for a sand profile. ..... 37. Back£itting analysis of pier test for a clay site... ... 38. Backfitting analysis of pier test for a sand site... ..~- 39. Comparison of skin friction with ultimate lateral resistance CAPT OVC sree eee tee tee teee 40. Solutions for a rigid pier or piles +. eee eee eee ee 41. Representative seat-type (freestanding) abutments... .- - 42. Yielding abutment walls: active wedge force diagram... 43. Effect of seismic coefficients and friction angle on seismic active pressure coefficient... eee ee ees 44. Relative displacement, acceleration, and velocity time histories of yielding wall ss eee eee ete eee 45. Representative integral abutments»... se eee ee ee 46. Procedure to incorporate wall stiffness in abutment stiffness imide eros eter steetece tetistee uctietee etioacteteees tiers 47. Wall pressures from forced inward movenent.. +. +--+ ee 48. Iterative procedure for design of integral abutments. . . . « 49. Typical bridge and corresponding model. --- +--+ ese 50. Soil profile, shear strength, and stiffness parameters, and Liquefaction potential. vee eee eee eet eee 51. Single pile stiffness solutions for upper and lower bound soil conditions. ». eee eee eee eee 52. pile properties and lateral soil stiffness. -- +... ++ 53. Pier foundation stiffness matrix... ++ +e ee ee ee 54. Abutment foundation stiffness matrix. +... e 2 eee 55. Single mode solution of equivalent static seisnic design load 56. Solution for governing loads and displacements of the bent foundation from both single mode and multimode anaylses. « 57. First four mode shapes from multimode analysis for upper and lower bound soil conditions. -- +++ e+e ee eee ix Page 84 102 104 105 107 108 109 12 124 126 128 130 132 134 136" 142 143 146 148 149 150 153 156 158LIST OF ABBREVIATIONS AND SYMBOLS Symbol, A a Empirical factor for py calculation for sand. Acceleration Coefficient... +e ee ees Cross-sectional pile area (L2)... 22... Half footing width (L). 6 6 ee ee ee eee Width of abutment wall (L).- +--+ ee ee Overburden normalization constant for CPT correlation Elastic seismic response coefficient... . . Coefficient for p-y curve calculation for sand. Coeffictent for p-y curve calculation for sand- Coefficient for p-y curve calculation for sand. Soil Cohesion (F/L2). 6+ +e ee eee eae Footing thickness (L) +. e ee ee eee Pile/shaft diameter (L).- + eee ee eee Bottou-of-footing depth (L) ss se eee ee Relative densitys se e+e ee eee eee Displacement of abutment wall (L)... +. - Young's Modulus of Blasticity (F/L2)..... Active force on yielding abutment wall (F/L). Passive force on yielding abutment wall (F/L) Soil Modulus (force per unit depth per unit deflection) Eecentricity (Les eee ee ee eee eee Horizontal force (F). + ee eee ee ee eee Vertical force (F). ee eee eee ee ee Unit skin friction (force per unit pile surface area) Reference Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section 6 244,89 3,6 6,7 3,4 3,5,8 6,7 6,7 67Symbol £ Frequency (hertz). + ses eee eee G Shear modulus (F/L2) 2. 2 ee ee ee Guax Low-strain dynamic shear uodulue (F/L2)...- +. + + Fa Acceleration due to gravity (32.2 ft/sec? or 9.81 m/s?) H Depth below ground surface... + - H —- Height of abutment wall (L)- ...- + - h Thickness of ground cover above footing (L). « I Area moment of inertia (L4). 6... ee eee IC Bridge 1mportance classification... .- ~~ 4 Back£411 slope angles ee ee ee ee J Empirical constant for py calculation for clay K Spring stiffness value (F/L) -- «+ - Kag Seismic active pressure coefficient. . KPE Seismic passive pressure coefficient . Kg Translational spring stiffness (#/L) « K, Rotational spring stiffness (FL). . - Kg Coefficient of active earth pressure « Ky Coefficient of at-rest earth pressure. Kp Coefficient of passive earth pressure. Ky, Horizontal translational stiffness in x-direction (F/L) Kg Wordzontal translational stiffness in Y-direction (F/L) K33 Vertical translational stiffness (F/L) - +--+ +++ Kyg Rocking rotational stiffness about X axis (FL)... + Ks55 Rocking rotational stiffness about Y axis (FL)... - LIST OF ABBREVIATIONS AND SYMBOLS (continued) xi Reference Section Section Section Section Section Section Section Section Section Section Section Section Section Section thon Section Section Section Section Section Section 249 34,5LIST OF ABBREVIATIONS AND SYMBOLS (continued) of pile/shaft length cF/L) Symbol Kg Torsion rotational stiffness (FL) x, Horizontal acceleration coefficient. ky Vertical acceleration coefficient. kg Soil stiffness at ground surface (¥/L2). kj Modulus of subgrade reaction (/L3). . Meet Memenk (RL) eee eee Me Pile/shaft head moment (FL). « + N SPT bloweount (blows/ft). +. +1. N Yield acceleration ing... ~~ N, Modified SPT blowcount value (blows/ft)- PHetePoree: Ce) ede teded steers eisai Pe(x) Equivalent static seismic load per unit bridge representing primary mode of vibration (F/L) .. ++ + Pg Assumed uniform load per unit bridge length used to calculate period (F/L) s+ +e tet eet rete Pt Lateral pile/shaft head force (F).- + + +e eee p Lateral soil resistance of pile/shaft (force per unit pile length) sss ee eee Py‘ ULtdmate lateral soil resistance Q — Axlal pile/shaft load (F). + « q Unit tip resistance (F/L2) . . « de Cone resistance (F). e+e eee Qe. Normalized cone resistance (F) - R Radius of circular footing (L) - R Equivalent radius of rectangular xi footing Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Reference 56 8 8 6,7 6,7 67 67 67 6,7 6,7 6,7LIST OF ABBREVIATIONS AND SYMBOLS (continued) Symbol R Ta wx) Ye Response modification factor». +/+ eee eee Reduction factor for earthquake~induced shear stress calculation. ss eee ee ee ee eee Site coefficient... ee eee eee Undrained shear strength of soil (F/L2). . Sedemic performance category. +--+ + ++ Characteristic Pile length = (EI/k,)1/5 (1). Natural period of vibration (T). +. +++ + Axial soil resistance due to pile/shaft skin CL)o eee tet eee eee Translation of abutment wall (L)-- +++ Shear (PF). 2+ + eee eee ee ee Peak ground velocity (L/T) - +++ friction Displacement (L)- + +++ e+e eee eee Static displacement profile due to applied load, Po (L) Weight of active soil wedge behind yielding abutment wall (F) eee ee ee Dead weight of bridge superstructure and tributary substructure per unit lenghth (F/L). +--+. +++ Lateral pile/shaft deflection (L). +++ ee eee ee Critical lateral pile/shaft deflection (at half py) (L) Depth below ground surface (L) --- +--+ Axial pile/shaft displacement (L). - 6. ++ Critical axial pile/shaft displacement (1) Footing shape correcion factor... . + ~ Coefficient used to calculate natural period wit of bridge Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Section Reference 9 6,7 67 67 5,6,7,8 6,7 67LIST OF ABBREVIATIONS AND SYMBOLS (continued) Symbol @ Footing embedment depth correction factor... . + + + 8 Slope of soil face. +++ ee B Coefficient used to calculate natural period Y Total Unit weight of soil (F/L3)...... Y Coefficient used to calculate natural period 6 Displacement (L) + ++ + ee ees Effective unit weight of soil (F/L3) 2... 6 pile compliance (L)e ++ ee eee ee eee of bridge of bridge 6 angle of friction between pile~soil interface. 6 Angle of friction between soil and abutment wall & Strain amplitude at half the peak deviator stress of unconfined compression test . v Poisson's Ration - ee ee ee ee ee 4 Angle of internal friction... eee eee 6” angle of internal friction from drained triaxial Average confining stress (F/L2).-- +++ ++ & Effective average confining stress (F/L2). - «+ G Vertical stress (F/L?) +--+ +e eee et eee q° Effective vertical stress (F/L2)- +--+ +++ % Lateral confining stress (#/12). 2.2 2. ee t Shear stress (F/L2). +e ee eee eee eee tt reedueney (rely cise ee ate eects eae Note: ~ L = Length F = force T= tlae xiv Reference Section Section Section Section Section Section Section Section Section Section 5 8 9 5565758 6,7 91-0 INTRODUCTION 1.1 Background In 1981, the Applied Technology Council (ATC), under the sponsorship of FHWA, developed recommendations for the seismic design of highway bridges. These recommendations were published as an FHWA report, Seismic Design Guidelines for Highway Bridges, in 1981.(1) With the report's recent approval by the American Association of State Highway and Transportation Officials (AASHTO) as a guide specification, many bridge designers will, for the first time, face the task of applying a new methodology for seismic analysis and design of bridges.(2) To carry out a meaningful seismic response analysis for even the more common types of bridges, the guidelines require an evaluation of the dynamic soil-structure interaction effects of the foundation system together with the evaluation of abutment behavior and site stability during earthquake Toading. While the FHWA/AASHTO guidelines make specific recommendations with respect to the structural analysis and design of bridges during earthquake loading, they are less specific with respect to foundation analysis and design.(1+2) This is in part due to the complexities which are associated with the different foundation systems encountered in bridge structures in combination with the wide variety of soil types encountered in practice. Whereas general analysis approaches and associated references are provided for foundation design in the FHWA/AASHTO guidelines, no one document exists where a bridge designer can find specific guidance on the various facets of seismic foundation design. Because of the clearly established significance of foundation behavior not only for the earthquake stability of a bridge structure but also for the overall dynamic response of the structure, a document specifically addressing analysis approaches to selsnic design of bridge foundations and providing simplified guidelines for routine design is clearly needed. 1-2 Purpose The ‘objectives of the research studies presented in this document are:© To advance the state of practice in earthquake resistance design of bridge foundations. 0 To provide simplified analysis approaches and foundation design guidelines to bridge designers. ‘The new design guidelines presented in this document are intended to supple- ment and complement the existing guidelines given in chapter 6 of the FHWA/AASHTO, Seismic Design Guidelines for Highway Bridges, and should be used in conjunction with the latter document. ‘The goal of the research studies presented in this report 1s to address questions related to the influence of foundation and abutment response on the overall seiemic design problem at several levels of complexity. This approach and related sensitivity studies provide the basis for establishing the simplified design guidelines. The so called “practicing bridge engineer” will, for the most part, be a team composed of the structurally oriented bridge designera supported by the geotechnical engineers. In this respect emphasis is given to so called “regular bridges” representative of conven— tional highway bridges with conventional foundation and abutaent systems. Representative foundation systems considered in the study comprise footings, piles, pile groups, pile bents, and drilled shafts. In the development of simplified foundation design guidelines, emphasis is placed on (1) methods for assessing the stiffness characteristics of foundation systems under moment and horizontal loading, and (2) methods for assessing the distribution of load throughout the foundation system arising from earthquake loading of the bridge structure. Bridge abutments are vital parts of the highway bridge foundation system, and design of the abutment system is often a critical part of the overall seismic design assessment. Because of the complexity of abutment systems, emphasis is primarily placed on (1) the behavior of integral abutment systems where the stiffness and deformation behavior of these abutments may significantly influence the overall bridge response behavior, and (2) a review of pressure distributions behind abutment walls, and the displacement approach to design of free standing walls.With an emphasis on use by the practicing bridge engineer, the docunent is particularly cogatzant of the nature of soil data generally obtained from routine foundation investigations of bridge sites. However, attention is also given to inprovenents in the state of practice of subsurface soil exploration, particularly that of in sftu testing, and its application to evaluation of the earthquake behavior of foundation soils. 1.3 Project Organization Work Plan. The overall work plan is schematically summarized in figure 1. A review and evaluation of existing literature on earthquake response and design of bridge foundation systems was first conducted. Based on the results of the review, various technical issues were identified. Also, a range of foundation parameters (sizes, configurations, etc.) were established for subsequent sen sitivity studies. Computer programs which were used to conduct the necessary analyses in the project were then identified. The information from all the above stepa were channelled into the planning process and also provided the background information in the sensitivity studies. The sensitivity studies constituted the most significant part of this project. In the course of these studies, the appropriateness of various foundation analysis procedures was first examined by comparing analytical predictions with experimental (e-g-, load test) data. These studies provided the basis for selecting the ultimately adopted foundation analysis procedures. Dynamic response analyses using the adopted analysis procedures were then conducted using several existing highway bridges. The sensitivity of various levels of refinement in modeling and assumptions on input parameters were examined in this second series of analyses. Results from this series of sensitivity study then Eormed the basis for the development of the simplified analysis models and design procedures. Various design charts and design guidelines were then developed and documented. Recommendations for areas of future research were also identified.REVIEW AND EVALUATION OF EXISTING LITERATURE ON EARTHQUAKE RESPONSE AND DESIGN OF BRIDGE FOUNDATION SYSTEMS, FINALIZE TECHNICAL ISSUES AND DATA RANGE FOR PARAMETER STUDIES ‘¢ IDENTIFY COMPLETE PROGRAMS FOR ANALYSIS © IDENTIFY TENTATIVE SIMPLIFIED APPROACHES. ‘TO FOUNDATION DESIGN ‘© PLAN SENSITIVITY STUDIES SENSITIVITY STUDIES: + SOIL PROPERTIES + FOOTINGS + STRUCTURAL MODELS + SITE LIQUIFACTION + SINGLE PILES» BRIDGE/FOUNDATION = PILE GROUPS: SYSTEM RESPONSE ‘DRILLED SHAFT DEVELOPMENT AND VERIFICATION OF SIMPLIFIED MODELS PREPARATION OF DESIGN GUIDELINE DOCUMENT, DESIGN CHARTS AND. RESEARCH COMMENTARY Figure 1. Work planWorking Team. The working team for the project reported in this document consisted of researchers, consultants, and d igners from three organizations: o The Earth Technology Corporation (Geoffrey Martin, Po Lam, Lucas Hekma). © University of Nevada at Reno (Bruce Douglas, Gary Norris). o Engineering Computer Corporation (Roy Imbsen, Wen David Liu). ‘A review committee was also established to monitor the progress of the project and to provide recommendations to the working team. The review committee con~ sisted o: © John O'Fallon who serves as the technical administrator from the FHWA (replacing James Cooper, who initiated the project for FHWA). © James Gates (Caltrans). © John Glark (Andersen-Bjornstad-Kane Jacobs Inc.)- Me. Po Lam served as the project manager and the principal investigator in the project. Dre. Geoffrey Martin and Bruce Douglas served as the project direc~ tors. Professors Hudson Matlock (Earth Technology Corporation) and Joseph penzien (University of California, Berkeley), were consultants to the project. ‘The background research on pile foundations by Hudson Matlock form the basis of the procedures for pile analysis recommended in this report. ‘The performance of the project was carried out as a team effort. The Earth ‘Technology personnel participated in all facets of the project. Professor Douglas provided input on the basic background information on the actual per- formance of various bridges and foundations types under dynamic loading con ditions (cyclic or earthquake). Roy Imbsen furnished the SEISAB progran which were used in the dynamic response sensitivity studies of several bridges. Dr. Gary Norris provided comments and input regarding design procedures and the general behavior of pile foundations. Dr. Wen David Liu developed the design chatts for footings.1.4 Report Organization Volume IT, Main Text. Results of research and the conclusions drawn during the course of this project are divided into two volumes. The first volume (this subject documentation) contains a condise description of procedures, guideli- nee and design charts reconmended for the seismic design of the various types of highway bridge foundations. The relationship between this document and the FUNA/AASHTO Seismic Design Guidelines for Highway Bridges is discussed in section 2. The basic background information on selection of soil properties ‘and soil parameters for analyses 1s summarized in section 3. A review of the evaluation procedures for stability of bridge sites under earthquake con- ditions is provided in section 4. Procedures for analysis and design of footing foundations are presented in section 5; piles in section 6; large diameter drilled shafts in section 7 and abutments in section 8. In each of the sections 5 through 8, detailed design procedures to evaluate foundation stiffness and to examine meuber stresses are first outlined: Then, miscella~ neous guidelines and commentaries on vartous design facets are furnished. In section 9 an example problem for a typical highway bridge is used to illustrate how the foundation analysis procedures are used to meet the requirements given by the FEWA/AASHTO design specifications for seismic design of highway bridges. Finally, conclusions and recommendations from the overall study are presented in section 10. Volume IIT, Appendixes. A second volume consisting of eight appendixes is included as part of this overall report. These appendixes provide documen- tation of various detailed studies conducted in the course of the project. Example problens illustrating the various procedures described in the main text are also included in the appendixes.2.0 REVIEW OF SEISMIG DESIGN GUIDELINES AND RESPONSE ANALYSIS PROCEDURES 2.1 FHWA/AASHTO Seismic Design Guidelines ‘The FHWA/AASHTO guidelines were prepared with a view of providing bridge designers with seismic design guidelines which reflected both the resulte of recent research and the benefits of observed performance of bridges during past earthquakes.(152) The guidelines and the associated commentary provide broad general guidelines for foundation and abutment design requirements. This docu- ment is intended to supplement the FHWA/AASHTO guidelines by providing simplified explicit approaches to the seismic design of bridge foundations and abutments. ‘The FHWA/AASHTO design guidelines (chapter 3) define four seismic performance categories (A, B, C, D) on the basis of an acceleration coefficient (a) and importance classification. Increasing degrees of complexity and sophistication of seismic analysis and design for foundations 19 specified for categories A through D. For category A bridges (a <.09), no specific seismic design require ments are needed. For category B bridges (.09 < a <.19) elastic seismic forces acting on the superstructure are determined using a simplified single mode spectral method. This determination requires the stiffness of the foundation systen to be modeled and combined with the superstructure stiffness to allow the calculation of the natural period of the bridge. A combination of the elastic seismic forces in two orthogonal directions is used for design. A strong emphasis in this docu- ment is given to simplified methods for evaluation of the stiffness of bridge foundation systems to be used in the natural period computations. In developing the stiffness characteristics of foundations, the FHWA/AASHTO guidelines recom mend due consideration of the nonlinear behavior of soils in the determination of the equivalent elastic modulus values or soil reaction for analysis. The specific recommendations given in this document take due cognizance of these recommendations. Setomic design forces for the eupporting substructure at the base of bridge columns are determined by applying an appropriate response modification factorto the elastic forces (see chapters 3 and 4 of refs. 1 and 2). Design forces for the foundations are in general taken as twice the seismic forces in the columns to ensure structural failure of components does not occur. Once design forces for the foundation system are determined, methods for distributing these forces throughout the foundation system are required. Methods for such distri- bution form an important element of the design methodology given in this report. For seismic performance categories C and D where acceleration coefficients are higher, a multimode spectral method is recommended for determining elastic for- ces for irregular bridges which have abrupt changes in stiffness over the span or between columns. Design forces for foundations may either be those deter~ mined in a similar manner to category B bridges or the forces at the base of columns corresponding to column plastic hinges. These forces are consistent with the design philosophy of minimizing damage beneath the ground surface that would not be readily detectable and reflect the maximum forces that can possibly be transmitted to the foundations by plastic hinging of the columas. For saturated sand and soft clay foundation soils, the FHWA/AASHTO guidelines recommend consideration of the potential for soil strength loss during earth- quake cyclic loading, particularly for saturated sand and soft clay soils sub- jected to high Levels of acceleration. Such considerations are important in assessing the ultimate capacity of foundations, the potential for degradation of foundation stiffness during earthquake loading, and the consequences of poten- tial Liquefaction of saturated sands. In this document, more specific recommen- Gations on how to take such factors into consideration are provided along with recommendations for methods of foundation investigation to obtain the appropriate soil parameters ueeded for seismic design of foundation systems. Abutments also play an important role in the overall earthquake performance of bridge structures. The FHWA/AASHTO guidelines provide recommendations for eva~ luating displacement of free-standing abutments and the importance of the stiff nese of monolithic abutments in influencing superstructure response. In this document, additional comments are provided on free-standing abutments displace- ment and the nature of pressure distributions behind abutment walls. Particular emphasis is placed on the development of a simplified methodology for deter~ mining stiffness of monolithic abutment wall to be used in calculating total bridge stiffness characteristics for natural period computations.2.2 Foundation Modeling For Bridge Response Analysis The need to account for foundation effects in assessing the overall earth- quake response of highway bridges has been repeatedly demonstrated in recent research. (3+4,5) Ideally, earthquake response of bridges should be evaluated using a single direct analysis which models the total discretized system, consisting of the structure, foundation, and the soil mass. However, such an approach is beyond “the state of practice” at the present time due to the complexity involved in modeling the foundation-soil system. Instead, so{l structure interaction analy- ses are usually conducted using the substructure method which utilizes a two-step approach.(6s7) In the first step, the response of the foundation system without the superstructure is analyzed. The force-displacement rela~ tionships of the degrees of freedom of the nodes at the interface between the superstructure and foundation are determined. These so-called dynamic-stiffness coefficients can physically be interpreted as a system of generalized springs and dashpots. The substructure method usually utilizes a linear representation of the foundation system. In the second step, the bridge structure supported on this spring-dashpot systea is analyzed to obtain the response characteristics of the overall bridge-foundation system for a specific earthquake. Because the stiffness modeling of the foundation system can play a dominant role in influencing overall response, the thrust of this report is placed on: 0 Development of simplified procedures for assessing the stiffness characteristics of foundation systems under moment and horizontal loading, to be used for elastic response analysis of the overall bridge system. © Development of wethods for evaluating the distribution of load throughout the foundation system arising from earthquake loading of the bridge structure. 2.3 Simplification of the Foundation Stiffness System At the beginning of this project, a survey was conducted to compile structural and foundation data for the commonly used highway bridges. A sensitivity study was then performed to evaluate the significance of various factors on overall bridge response. This in turn formed the basis for simplification of the morecomplicated design approaches to allow the development of simplified guidelines and procedures for practical applications. Representative foundation systems for highway bridges are shown in figure 2. ‘The range of foundation sizes most commonly found in typical highway bridges are listed below: Spread Footings Size: 8 ft to 30 ft (2.5 to 9 m) wide Thickness: 2 to 6 ft (0-6 to 1.8 m) thick Embeduent: Up to 8 ft (2.4m) ground cover Piles and Pile Groups Pile types: Reinforced concrete, timber and steel H or pipe piles Diameter: 8 to 24 inch (20 to 60 em) 0.D. Length: 15 to 100 feet (4.5 to 30 m) Pile Group Configuration: 1 by 2 to 6 by 6 square or rectangular pile group Spacing: 3 to 8 diameters center to center Batter angle: Vertical or battered piles battered at 1 horizontal to 4 vertical Drilled shafts Diameter: 4 to 8 ft (1 to 2.5 m) Length: 20 to 100 fr (6 to 30 m) Pier Type: Reinforced concrete Abutments Abutment types: Integral or seat type Height: 5 to 40 feet (1-5 to 12 a) Width: 30 to 150 feet (9 to 45 m) Support: Abutment wall commonly supported by spread footings or piles. 10INTEGRAL ON FOOTING SEAT TYPE ON PILES ABUTMENTS ae SPREAD FOOTING PILE FOOTING SINGLE COLUMN ON DRILLED SHAFT PILE BENT PIERS. Figure 2. Highway bridge foundations. 7‘The sensitivity study was conducted in the course of the project to evaluate the sensitivity of overall dynamic response on foundation stiffnesses using eight typical highway bridges (see table 1). Further details regarding the sen- sitivity study can be found in appendix H. In the study, the natural frequen- cles of the highway bridges were determined. In one set of analyses, the foundation stiffnesses were estimated and incorporated in the analyses. In the other set, a fixed base assumption was adopted. A summary comparison between the two sets of solutions is presented in figure 3. ‘The ratio of the frequen— ches for a fixed base model to the foundation stiffness model was plotted against the length of the highway bridge. It can be seen that the dynamic response characteristics of the overall bridge are significantly affected by the foundation system, especially for shorter bridges. Significant errors (over 50 percent) could occur for bridges below 300 feet in length if foundation effects are ignored in estimating the overall earthquake forces. To correctly account for the inertia forces in the foundation system under dyna~ mic loading conditions, the foundation stiffness coefficients would, rigorously speaking, be dependent on the frequency of loading. The frequency-dependent stiffness and damping coefficients for certain foundation systens (e.g., spread footings and single piles) have been developed and presented in the literature for certain soil conditions (e.g-, semi-infinite half space). On the basis of the sensitivity study, 1t was concluded that for the conditions encountered in the typical highway bridges, the inertia forces in the foundation system can be ignored in estimating the foundation stiffness. The above concept will be applied throughout the rest of the report in the development of simplified pro- cedures for each foundation type: footings in section 5, piles in section 6, drilled shafts in section 7, and abutments in section 8. Various forms of representation of foundation stiffnesses have been used by bridge engineers, including (1) translational and rotational spring elements for all types of foundation systems, and (2) extended column (cantilever beams) for Piles. A full 6 by 6 stiffness matrix to represent the foundation stiffness ts recommended for the highway bridge application and is used in the current study (section 9 and appendix G). The full 6 by 6 matrix is the aost general form of representation of foundation stiffnesses. It can be used for all foundation types including abutments, spread footings, pile footings (both vertical and 12pattered piles), pile bents, and drilled shafts. The full 6 by 6 stiffness matrix can also rigorously accomodate both the boundary spring element and the extended column concepts. 2.4 Modeling Foundation Damping. Proper implementation of foundation damping is a highly complex problem. Two primary sources of foundation damping needs to be considered: (1) radiation damping associated with the {nertia effects of the foundation-soil mass system, and (2) material damping. Incorporation of foundation damping in an overall bridge response generally involves a two-step approach. In the first step, the foundation damping characteristics for a range of loading frequencies and displacement amplitudes are evaluated. Due attention aust be given to the fundamental differences in the two sources of foundation damping. ‘The radiation danping factor ie dependent on the frequency of loading whereas the material damping factor is strain (displacement amplitude) dependent. A time-history solution approach is needed to properly account for both the above sources of damping in the dynamic response of the overall bridge model. If a modal superposition procedure is used, it is necessary to select a viscous damping coefficient such that when this viscous damper is used in conjunction with equivalent foundation stiffness, the proper amount of energy can be dissi~ pated through the viscous damper. Selection of the viscous damping coefficient requires a preliminary estimate of both the displacement amplitude and the approximate frequency of loading (which ie related to the natural frequencies of the overall bridge-foundation system) such that the proper magnitudes of radiation and material damping can be represented. After the development of the above foundation viscous coefficient, its contribu~ tion to the modal damping of the overall bridge response can be evaluated in the second step. Again, due to the imitation of the modal superposition technique, the contribution of the viscous damper to the overall response of the bridge can only be accounted for in an approximate manner. Generally, the eigen-vectors associated with the mass and stiffness matrices of the overall bridge model can be used to transform the viscous damper to a generalized damping matrix in the generalized coordinate system. The diagonal terms of the generalized damping 13——_— Ss SSeS settd uo adéy ae=g — a3eys peTT FAP vo umnqoo oT8uTs € 296 uoTaesedeg soiauu09 asva Teuueyg wokuey ronbnog setrd uo adé3 aeag — Suza005 atyd uo adka—TTeM OW z Ist eA0 prow youeY TTeyAAN aea7a bxeTO-e3ues seqFd uo adéa awag — BuyI005 eTTE uo adsa-TTEM “ory L 06s eno Kemyzed UesEoR s8upz003 oT1d serFd uo adé2 ae9g ——egeaedes vo guaq uuntoo 902uy, L 499 peoyzeag qeex3g PATIL s8up2003 at rd Buyq003 otyd uo qea8eaur aezedes wo qeq umnTOs~oMT z soz Bupssorsz0A9 yeaq3g x0qUTEY Buyz00z eTFd uo Tes8eqz —«Supa003 aTFd vo umnqoo oTsuTg z 800 Supssossseag peoy pueTOTAR seTtd uo edéa sees Sut300g eT Td uo uMnToo ayfuTS s oor ‘eSusyor0IUT e029 eso¥ 2uy3003 peoads uo Tex8oaut t 901 Sufssoasz0pun peoy yeTyIeSI0q aeTa saeds 30 G3) DEON SEPIA xeqmy 38097 —_ sApnag Aa;ayaysueg Uy pasq seSpyag Jo uotadzasseq oT STAVE “usa6p1aq g 40} jopow s6utids uonepunoy Yair japous aseq paxij Jo Aouenbay uoriesqa yo uostedwiog “¢ aunbiy d 5 2g | 3 & g & 2 £ g 3 8 i 3 i z a j 2 8 ears || rt (a) 002 099 008 oor oe 002 oot 9 H19N37 3oq1ua—<—t 2 1 L ° at e 8 8 $ . 7 ¥ 5 ° 8 v ° Bato. Y ° | z v ° s Vv 5 v ’ a a e ° 2 z ° ° z =— oa SE TORVeEIN —TOENAS 30 3008 v ‘SONINdS NOLLVONNOS aswe GaXis 5 15matrix is then used as the contribution of the foundation damping to the overall damping factor at the various modes. The off-diagonal terms cannot be accounted for in a modal superposition technique and are usually neglected. If the foun- dation damping is considered in the above approximate but rational manner, the demping factors for the various modes of vibration can be developed. It should be noted that the viscous damping coefficient for the foundation is not an intrinsic foundation parameter. The viscous damping model is merely a mathematical device to approximate the radiation and material damping phenome na. The viscous damping coefficient needs to be varied for a specific con~ dition: (1) frequency of loading, (2) displacement amplitude, and (3) foundation stiffness. Another noteworthy consideration is that the damping behavior for a given foun dation system will have different contributions to the modal damping of an overall bridge whea the particular foundation system is used for different superstructures.(8) For example, a given abutment system may contribute to a mich higher damping factor for a short bridge versus a longer bridge. Due to the above mentioned complexities and variabilities, development of a proper foundation damping factor can only be addressed on a case by case basis. ‘The FHWA/AASHTO seismic design guidelines (section 3.5) implicitly reflect a 5 percent system damping ratio for the recommended elastic design spectra. For category D bridges where dynamic elastic response analyses are recommended to determine the elastic design forces, a 5 percent system damping ratio is also recommended for most bridges. Refinements to these system damping ratios to account for damping from individual foundation elements is not warranted at this stage of the “state of the art", and must await further research. 163.0 SOIL PROPERTY EVALUATION 3.1 General One of the key elements in proper bridge design is the selection and design of foundation elements. In order to select and design the foundation system for a given bridge, the characteristics and behavior of the site soils need to be defined prior to the final design. Two different categories of site investiga- tion can be included in an exploration program. The first is a conventional drilling program followed by laboratory testing to establish the required infor- mation. The second utilizes in situ testing where soil characteristics, strength, and/or stiffness of the site soils are estimated by in situ testing devices such as the standard penetrometer test (SPT), cone penetrometer test (CPT), pressuremeter test, and geophysical tests. The relative advantages and disadvantages of both approaches have been widely discussed in the literature.(9) When executed properly, e{ther approach or combination of the two can provide satisfactory results. The primary objective of this section is to provide guidelines for the deter~ mination of necessary soll parameters to be utilized in conventional foundation stiffness and strength evaluations for seismic design. Since investigation and testing methods used for high quality conventional site investigations provide data needed for setenic design in most instances, no special field or dynamic laboratory testing programs over and above those commonly included in normal site investigation prograns will be discussed in detail. However, for seismic category D and/or for poor soil conditions, it is important to recognize the potential need for special field or laboratory tests for setemic design ot prototype in situ field testing programs. The following list summarizes the typical foundation and related coil parameters needed for the seismic analyses docunents in this report. WwFoundation Type/Problem ‘Typical Soil Parameters Needed Abutment Design Friction Angle, ¢ unit weight, Y Young's Modulus, Ey Footing Stiffness Shear Modulus, G Poisson's Ratio, U Piles and Drilled Shafts Friction angle, @ Shear strength, ¢ Unit weight, 7 Strain at 50 percent of the peak deviator stress of unconfined compression test, & Ground stability Liquefaction strength, ¥ 4," Unit weight, Permeability, k Coefficient of compressibility, my Relative density, Dy Normally the geotechnical engineer responsible for site investigations and laboratory testing would be expected to work closely with the structural engineer responsible for the design of the bridge to select appropriate values of the soll parameters Listed above. The following paragraphs provide only an overview of soil property evaluation. 3.2 Field Program and General Site Soil Classification In order to establish proper field programs for a given bridge site, one should review the geology of the site and, if possible, obtain information on nearby sites. Based on this information, one can determine the appropriate means of soil exploration and number of borings required for the site. For example, a gravelly soll site will not be sultable for using cone penetrometer test devi- ces. In most cases, a minimum of 2 to 3 borings or in situ soundings is recon mended for each bridge site. The depth of the boring can be influenced by the type of foundation selected in the preliminary design. In most cases, the depths of the boring should be greater than 2.5 diameters of the spread footing or 10 £t below the estimated tip of pile foundations. Where potential earth- quake rélated liquefaction or stability problems are suspected at a site, more and deeper boreholes may be required. Where borings are selected as the sole neans of soll exploration, good quality samples of cohesive soils should be 18obtained for further laboratory strength and consolidation tests. For cohe- sioniess sotls, the standard penetration test (SPT) blowcount values should be the mininum requirenent. ote that recent studies have shown that the azount of energy involved in the SPT is significantly influenced by the driller's practice and the type of equipment used. Therefore, special attention should be given to this isoue in order to establish any necessary correction factor to be used tater.(10) pase on the field boring logs and/or CPT soundings, the cross sec tional profile of the site should be established. This cross sectional profile can be used to establish needed Laboratory test programs, possible further site exploration, and to select probable ‘foundation types for the bridge. 3.3 Sotl Parameters Derived From Standard Penetration Tests ‘The SPT blowcount data 1s probably the most widely available data for soil sites. There are aumerous correlations established in the literature to esti~ mate soil property from SPT blowcount data. Some examples of these correlations ‘are given below. It should be noted that these correlations are general in nature and should be used with caution. If possible, special attention should be given to regionally established correlations. Relative Density and Angle of Internal Friction (4) for Cohesionless Soils(11,12) ‘Angle of Taterual Friction ¢ (Deg) Relative Resistance Density Peck et al.(11) meyerhof(12) ‘Type of Sotl N (blows/ft) Dr very loose sand 50 0.8 >4L 4s, 19Undrained Shearing Strength of Cohesive Soiis(13) Penetration Undvained Shear Resistance N Strength ¢ (blows/ft) (kips/£t2)* Consistenc: <2 <0.25 Very soft 24 0.25-0.50 Soft 48 0-50-1-00 Mediun 8-15 1.00-2.00 stiff 15-30 2.00-4.00 Very stiff >30 >4.00 Hard * 1 kip/ft? = 47.87 KN/m2 (KPa) ‘The above shear strength parameters can be used for evaluation of bearing capa~ chty of footings (section 5), and for pile and drilled shaft analyses (sections 6 and 7). For earthquake response analyses, dynamic modulus properties of the site soils are needed. The low strain dynamic shear modulus (Gpax) is a useful reference parameter, as discussed in section 3.6. Japanese researchers have conducted a detailed study to correlate the standard penetration test blowcount value (N blows per foot) to the Gmax of various soils.(14) The following simple equation provides the needed accuracy for most analysis procedures. Gaax (tonnes/m2) = 1200 n°-8 or Gag (Kip/£t2) = 245.4 NOB. ee ee ee eee eee a However, it should be noted that results of Japanese research did conclude several equations all in the general form of ANB for various soil types and site conditions. Therefore, if available data can justify the accuracy, one could select a slightly different equation for the determination of dynamic shear aodu- lus for the specific soil of concern. o1 The estimation of soil liquefaction strength from SPT blowcount values will be discussed in more detail in section 4. 203.4 Parameters Derived from Cone Penetrometer Tests In recent years, usage of the cone penetrometer test (CPT) has been increasing steadily. The ability to interpret the CPT sounding has also improved eignifi- cantly.(9515,16,17) Tt is now a commonly accepted alternative to use CPT instead of blowcount values for geétechnical soil profile description. Charts as shown in figure 4 in conjunction with the continuous cone-resistance versus depth curve can be used to establish soil layering and soil types at a site of concern. The accuracy of such charts has been verified in many comparison studies.(9) In addition to soil classification (see figure 4), the CPT data can be converted to equivalent SPT blowcount values using the simple equation N= 4c/A where Ic is the cone resistance and A is a constant ranging from 3.5 to 4.5 for silty soil and 4 to 5 for sandy soils. With this correlation, one can easily convert a CPT sounding data into equivalent blowcount values and use the relationships given in section 3.4 to estimate various soil stiffness and strength values. Other correlations between qe and N have also been established. (17) Furthermore, there are existing established procedures to estimate the undrained shearing strength (c) for cohesive soil and internal frictional angle (4 of cohesionless soil. A typical correlation between CPT data and ¢(17) is given in figure 5. In the past, CPT has been more popular in Europe than in the United States. However, the CPT has gained significant popularity and acceptance among geotech- nical engineers in the U.S. in the last 15 years. Tt {6 anticipated that the use of the CPT will increase and the correlation between soil properties and CPT data will be further refined. 3.5 Laboratory Testing Program For a conventional bridge foundation (in addition to the general purpose soil tests such as moisture and density, water content, grain size distribution, and index property tests), the following table may be of use to select the necessary laboratory tests. 21Design Problen/Soil Parameter Type of Test Needed Liquefaction strength relative density, cyclic triaxial or cyclic simple shear. dynamic compaction cyclic triaxial or cyclic simple shear (cohesionless soils) dynamic stiffness (Young's resonant column or cyclic triaxial modulus E or shear modulus ¢) static strength (c, ® static triaxial or direct shear If the available tn situ test data or the significance of the problem does not justify the inclusion of all the desired laboratory tests, a literature review to identify typical ranges of properties for a given soil may be appropriate. From this range, 1t may be possible to evaluate the likelihood of potential problems. Alternatively, the range of values may be used for sensitivity stu- dies. This approach is often appropriate for liquefaction strength and dynamic stiffness evaluation 3.6 Modulus of Elasticity ‘The computation of stiffness values for footings and abutments for use in dyna~ mic response analyses requires an average modulus of elasticity for the soil. In the case of footings (chapter 5), the value of the shear modulus G is required, while in the case of integral abutments a value of Youngs modulus is required. Selection of the appropriate value of modulus for use in stiffness calculations is complicated by the fact that the modulus is a function of strain amplitude and confining stress and varies widely with soil type.(18, 19) The low strain shear modulus Gyox may be obtained for site soils through the use of SPT values (as previously described) or through geophysical tests. These values correspond to shear strain amplitudes of less than 0001 percent. Earthquake-induced shearing strains related to footing or abutment response during earthquakes will typically range from 0.01 to 0.5 percent, depending on the maximum acceleration levels and the type of soil. Based on laboratory test data where shearing strains of this order have been applied using resonant column or cyclic triaxial equipment, values of G will be reduced to the order of 0.9 to 0-2 Guay depending on the strain level (figure 6). 2210° HEAVILY OVERCONSOLIDATED SANDY OR CEMENTED SOILS | crave to GRAVELY ‘SAND TOSAND g ee siury sano L ai rasan sy 8 g z = & ] a sanoy str 70 8 auaver uy @ 2 é TeLavey SILT 70 curv evar 3 g N 5 leer cxay 70 Sag oreanie z coat S 2 wicuey stusiTive Sous peaTs—> ° 10 : 1 1 t 1 . 0 1 2 3 4 5 6 FRICTION RATIO (percent) NoTEs: 1, FRICTION RATIO = (SLEEVE FRICTION / TIP RESISTANCE) x 100 2. acl = TIP RESISTANCE x (1 ~ logyq OVERBURDEN PRESSURE) Figure 4, Soil behavior type classification chart.NORMALIZED CONE RESISTANCE, qc1 (tsf) ‘CEMENTED OR HEAVILY SAND TO. ‘OVERCONSOLIOATED SILTY SAND SILTY SAND TO. SANDY SILT SANDY SILT TO ‘CLAYEY SILT CLAYEY SILT TOSILTY CLAY TOCLAY LAY To ORGANIC. clay PeATS—> FRICTION RATIO (percent) Figure 5. Expanded soil behavior type classification chart with equivalent ‘overburden normalized friction angle and relative density trends. FrySHEAR MODULUS RATIO, G/Gmax (percent) a Hf @ UPPER BOUND we LOWER B0UND 0 poi 10-7 10-3 19°? 107! 1 SHEARING STRAIN (percent) Figure 6. Typical variation of shear modulus with shear strain amplitude. 25Where appropriate, the effect of strain amplitude on modulus values should be taken into account when assigning values for analyses. In any event, it is recommended that upper and lower bound values equal to say + 50 percent of the best estimate for the average modulus value be used in sensitivity studies in dynamic response analysis. A typical range of values of shear modulus as a function of soil type is shown in the table below: Type of Soil Shear Modulus G (psi) Soft Clays 400 - 2,000 Firm Clays 1,000 - 5,000 Silty Sands 4,000 - 20,000 Dense Sands and Gravels 10,000 - 50,000 Note: 1 psi = 0.0069 MPa Values will vary within the range depending on strain amplitude, confining stress, initial static shearing stress, gradation (sands), and shear strength (clays). Laboratory tests should be used to confirm a best estimate modulus where bridge response is thought to be very sensitive to the selected modulus values. The value of Young's modulus E may be obtained from the shear modulus from the relationship: E=-20+ 496 where U= Poisson's ratio For practical design purposes, a value of U= 0.35 may be used for cohesionless solls and v= 0.45 for cohesive soils. 264.0 DYNAMIC GROUND STABILITY EVALUATION 4.1 Introduction Liquefaction of saturated granular foundation soils has been a major source of bridge failures during historic earthquakes.(20) For example, during the 1964 Alaskan earthquake, 9 bridges suffered complete collapse and 26 suffered severe deformation or partial collapse. Investigations indicated that Liquefaction of foundation soils contributed to much of the damage, with loss of foundation suport leading to major displacement of abutments and piers. EZarthquake-induced instability of slopes at abutments may also lead to signifi- cant displacements and related abutment damage, as will earthquake-induced settlement of backfill soils or gross settlement of the soils at a bridge site. In this chapter, methods for evaluating the site liquefaction potential of level ground are firet reviewed. Liquefaction of foundation soils is a significant factor in the design of pile foundations, as discussed in chapter 6. Although a detailed discussion of the stability of slopes under earthquake Loading is beyond the scope of this report, a brief overview of design approaches is given. General comments on earthquake-induced settlement are also provided. 4.2 Liquefaction Potential Evaluation The potential for Liquefaction of a given site is conmonly expressed in terme of a factor of safety. This factor of safety is defined as the ratio between the available liquefaction strength and the earthquake-induced dynamic stress. It has been found that both the strength and the induced stress can be normalized with respect to the effective confining pressure. As described in the FHWA/AASHTO guidelines, there are basically two different approaches to evaluate the factor of safety against Liquefaction for level ground site.(152) ‘The first method (empirical method) uses field performance data to establish the Liquefaction strength of a given soil using SPT blovcount data. ‘The average earthquake-induced shearing stress 1e estiaated fron the expected peak ground acceleration at the site using a aimple equation. The second method (analytical method) utilizes laboratory cyclic triaxial or simple a7shear tests to establish the liquefaction strength of a given soil. The earth~ quake-induced stress can be estimated from the above mentioned simplified equation or alternatively can be calculated using @ one-dimensional site response analysis program. Both of these approaches are commonly used. A comprehensive review of the field performance or empirical approach is given in a state-of-the-art paper presented by Seed, et al.(10) studies on the Liquefaction problem as applied to highway bridges have been presented by Martin, (20) and by Ferrito and Forrest.(21) 4 brief outline of the steps involved in each method is given below: o Empirical method: Barthquake-induced average shear stress Ty = 0-65 Anax Berg 2. ee ee ee £ where 4aax ~ "ax. acceleration at ground surface Y= total unit weight of soils H = depth to soll of concern rq = a deformation related reduction factor varying from 1.0 at the surface to 0.9 at 30 ft (9 m) depth and extends onward at a value of 0.9 to depth below 30 ft. Uaually the normalized tress ratio ty/q' ™ ‘tby/( 1H) is used in all calculations. 2, Liquefaction strength of the soils Q) A detailed compilation of case histories with respect to whether sites have or have not liquefied under previous earthquakes, resulted in a plot as shown in figure 7 where the liquefaction strength (design shear stregs ratio) is expressed in terms of corrected blowcount values-(10) These curves are updated versions of those given in the FHWA/AASHTO guidelines.(1,2) The corrected blowcount for sand ts obtained from the following equation. gga aotearoa «3 where Cy is a constant to normalize the effects of overburden pressure, and N is the SPT blowcount value. For silty sands and silts with Ds < 0.15 mm, blowcounts should be corrected using the equation: 28Jae ton per sq ft 10% 2 z 8 3s e « 3 § 2 2 é s & = é g E & $ Limited Strain Potential for Effective Overburdeni 40 Modified Penetration Resistance, N, ~ blows/ft tL 0 600) 800 1000 1200 Average Shear Wave Velocity in Top 50 ft -(Ips) NoTES: (1) 1 ton/ft2= 95.74KN/m? (KPa) (After Seed et al., 1983), Ua eee aoe 1 03m Figure 7. Chart of evaluation of liquefaction potential for different magnitude earthquakesMeum ceeT eS ee treristeriiis ct eu seat eeeeatia a The relationship between Gy and effective overburden pressure q~ is shown in figure 8. The reason that Cy is equal to 1.0 at an overburden pressure of 1 tsf (95.74 KN/n2) is that the majority of data collected have an effective overburden pressure of around 1 tsf (95.74 KN/m2), For any soil at a depth of 10 ft (3 m) or less, the field blowcount value (N) must be aultiplied by 0.75 before obtaining the N, value. Analytical method ‘The analytical approach for evaluating liquefaction potential is based on a comparison between field liquefaction strengths established from cyclic laboratory tests on undisturbed samples, and earthquake-induced shearing streases- In this approach it must be recognized that the development of a field Liquefaction strength curve from laboratory test results, requires data adjustment to account for factors such as correct cyclic stress simula~ tion, sample disturbance, aging effects, field cyclic stress history, and the magnitude of in situ lateral stresses. These adjustments require a con~ siderable degree of engineering judgment and may be guided by in situ test data. Also in many cases it is impossible to obtain undisturbed sand sample: I£ a total stress approach is used, the liquefaction potential is evaluated from comparisons of estimated earthquake~induced shear stresses with adjusted field liquefaction strength curves, as shown in figure 9. The earthquake-induced shear stress levels may be established from a simplified procedure (equation 2), or from more sophisticated assessments using one dinensional “equivalent linear” dynamic response programs such as SHAKE.(22) Average stress levels are established using the equivalent number of cycles concept (approximately 10 cycles for M7 and 30 cycles for MB.5 earthquakes). More recently, nonlinear programs have been introduced for response calculations. 4n improved representation of the progressive development of liquefaction is provided by the use of an effective stress approach where pore water pressure increases are coupled to nonlinear dyna~ mic response solutions, and the influence of potential pore water pressure dissipation during an earthquake is taken into account.(23,24) This approach provides data on the time history of pore water pressure increases during an earthquake, as shown in figure 10. It is of interest to note that a rough indication of the potential for Liquefaction may be obtained by making use of empirical correlations established between earthquake magnitude and the epicentral distance to the most distant field manifestations of liquefaction. Such a relationship has been described by Youd and Perkins (figure 11) and has been used as a basis for preparation of liquefaction~induced ground failure susceptibility maps.(25)Cn Q__02 OG OS aOsecul canal anaes oO r t T T 0.5}- 1 1.0) — i | \ s 3 et { B25p ; & F = 30) 8 L 6 35h 4 ; £4 fl a asr 4 : 5.0 1 1 1 1 ( l NOTE: 1 taf = 96.74 KN? (a \ . Figure 8, Relationship between Cy and effective overburden pressure. |nena | SISAWNY 3SNOdS3E INYNAD wows) S3Y¥NDALYY3 18 $3949 NOs 034073A30 SS3YLS DNTOAD ANF TWAINDI 2 uonsesanbyj 02 (ssens (e101) yoeoudse (2: ool ue yo saidioulig “6 aunfiry NOWOVS3NDET 3SN¥D OL $3792 40 Y3ENNN ol (siS31 8¥1 oo Woes) $32049 |S ViYO S31 AUOLVHORY? 2 NORE NOILO¥S |= MOHJ O3A1B30 3ANND HLON3ULS 8 30017 9NISM¥ HO1L9¥S3ND17 O1314 O2LYKILSI as SS3B1S JIVIAI 7 a 4 i juste ere ee seer SS3U1S B¥3KS INI9AI MBOSINO LN3TWAINDS Vo wo 8 roy ae ro so 32DEPTH (FEET) PORE PRESSURE (PSF) PORE PRESSURE (PSF) oo 1000 2000 oo a 2000 a in SSS | \ cf \\ t 1 / \iN 50} ee sop a i a A\ ge See zit zi le isoh = | | 4 10h =| l| 4 | | | 200 | L, 200 | Ll (a) PERMEABILITY (k) = 0 (b) k= 0,03 FT/SEC {=1 CM/SEC } NOTE: 1 pat = 47.9 N/m Figure 10. Effective stress approach to liquefaction evaluation showing effect of permeability9 T T 1964 ALASKA © 1906 SAN FRANCISCO © 1 1 1 1 1 8 = 1964 NIIGATA a SEF 1916 GUATEMALA 34 = ¥ 6 NN souvo USED BY Youd, 1978 = & ra MEAN EPICENTRAL OISTANCE CURVE 5 FROM KURISAYSHT ANG TATSUOKA, 1975 1 KM 0.62 MILES 4 * - 1 10 100 1,000 OISTANCE FROM SOURCE TO FARTHEST SIGNIFICANT LIQUEFACTION EFFECT, R, IN KILOMETERS Figure 11. Maximum distance to significant liquefaction as a function of earthquake magnitude. 34The above two methods (empirical or analytical) are equally acceptable when carried out with the proper procedures. The selection of the approach to be used for a particular project is usually made according to available data and tools. 4.3 Liquefaction of Clayey soils At the early days of liquefaction research, attention was given only to cohe- sionless soils such as sand or nonplastic silts. However, after the Tangshan earthquake in China, data collected suggested that certain low plasticity clayey soils may also suffer severe strength lose during and immediately after an earthquake. At the present time, available information suggests that clayey soils having the following characteristics may be vulnerable to severe strength loss as a result of an earthquake(10); © Less than 15 percent finer than 0.005 mm. © Liquid limit less than 35 percent. o Water content greater than 0.9 times the liquid limit. If site clayey soils fall into this critical range, cyclic laboratory tests should be utilized to assess the strength loss potential of the soils. 4s4 Slope Stability Evaluation The effects of earthquake loading on nonsaturated sandy or clay soil slopes or insensitive saturated clay slopes can be determined using conventional pseudostatic stability analyses. In a pseudostatic analysis, the effects of the earthquake are represented as an equivalent static horizontal force acting on the slope. The factor of safety for a given seismic coefficient can be esti- mated by using traditional slope stability calculation methods. A factor of safety greater than one indicates that the slope is stable for the given lateral force level, and a condition of limiting equilibrium or yield will not be exceeded. A factor safety of less than one indicates that the slope will yield and slope deformation can be expected. However, the factor safety does not pro- vide any quantitive measure of the amount of deformation to be expected. Tf the maximum ground acceleration is used for analyses, slope designs will be highly 35conservative as slope yield is not being permitted. Historically seismic coef- ficients of about one-third the maximum value are often selected for pseudosta~ tic analyses. ‘The approach implicitly recognizes that some slope yield, during periods when peak accelerations exceed the yield acceleration, is acceptable. ‘An approximate procedure for estimating earthquake~induced slope displacement for the above case, using the concept of a frictional sliding block, is now an accepted practice.(27,28,29) his approach solves for the accumulative move- ments of the sliding mass by integrating increments of movement which occur when the earthquake ground accelerations exceed the yield acceleration, as show schematically in figure 12. The cumulative permanent displacement will depend poth on the maximum magnitude and duration of the earthquake. The ratio of yield acceleration to maximum accelerations of 0.5 will result in block displa~ cements of the order of a few inches for a magnitude 6 1/2 earthquake and several feet for a magnitude 8 earthquake. For saturated silts and sands and sensitive saturated clays where significant pore pressure increases may be induced by earthquake loading, and @ potential exists for a significant undrained strength loss, earthquake stability analyses are more complex.(27) In the case of dense saturated sands, significant undrained shear strength can be mobilized even when residual pore pressure increases induced by an earthquake have reached high values. However for loose sands, the residual undrained strength which can be mobilized after high pore pressure build-up (associated with Liquefaction on level ground) may be very low, and in fact, less than the static undrained shear strength. Such con~ ditions may trigger flow slides or large ground deformations where seams of such material exist. Where such conditions are identified during site investigations of abutments area, specialized slope stability analyses are required. 4.5 Dynaaic Settlement Evaluation Dynamic compaction or volume reduction of cohesionless soll deposits is com nonly encountered during earthquakes. Compaction will also occur in saturated cohesionless deposits following dissipation of excess pore water pressure Induced by the earthquake. In the case of bridge structures, dynamic compaction 4s commonly encountered in cohesionless backfill soils behind abutments.Laboratory tests results from cyclic simple shear tests indicate that the amount of dynamic compaction is primarily a function of the cyclic shear strain ampli- tude, the number of cycles and the relative density.(30) Representative values for reductions in volume or compaction as a function of relative density and shear strain amplitude for dry sand (after 10 load cycles) are shown in figure 13.(30) Multidirectional shaking table test results indicate the compaction resulting from multidirectional loading may be about twice the value obtained from one dimensional shaking. In general, the compaction or settlement resulting from earthquake loading of cohesionless soils is unlikely to be much greater than 0.5 percent of the layer thickness for most well compacted soil backfills. In many cases, the effects of potential settlement may be minimized by the use of approach or settlement slabs behind abutments. 37DIRECTION OF PERMANENT DISPLACEMENT ACCELERATION YIELD ACCELERATION OF SOIL BLOCK (I.E, ACCELERATION REQUIRED TO INDUCE MOVEMENT) VELOCITY OF SOIL BLOCK TIME DISPLACEMENT OF SOIL BLOCK TOWARD BLUFF TIME Figure 12. Schematic illustration of sliding block approach for displacement calculations.VERTICAL STRAIN DUE TO COMPACTION IN TEN CYCLES (PERCENT) 2.0 = 45 percent 0.5 ° 0.01 = 60 percent = 80 percent 0.1 1.0 CYCLIC SHEAR STRAIN (PERCENT) Figure 13, Effect of relative density on settlement in 10 cycles. 10.0 ES5.0 ANALYSIS OF FOOTINGS 5-1 General this section presents reconmended procedures to evaluate footing stiffnesses which should be incorporated in the bridge model to evaluate the overall response of the bridge-foundation system. Additional design guidelines are Included to assess the Integrity of the footing-soil syeten under earthquake loading conditions. Footings are commonly used as foundations for piers and abutment walls. They are also used as pilecaps for pile group systems. ‘The procedures described in this section can be used to evaluate the stiffness characteristics of the footings for both conditions. The relative stiffness of a pilecap (footing) versus that from the pile members for a typical pile footing is discussed in section 6- ‘he current state of practice in soil-structure interaction analyses for footings involves solving for the response of a rigid footing foundation on a semi-infinite elastic half space-(31) For most typical highway bridges dyna~ mic effects can be ignored (see appendix B) in evaluating the stiffness characteristics of footings. Descriptions on the procedure to solve for the static stiffness of footings are provided below. 5.2 General Form of Stiffness Matrix ‘The general form of the stiffness matrix for a rigid footing is summarized in figure 14. the 6 by 6 stiffness matrix can be incorporated in aost computer programs such as SEISAB or STRUDL to account for foundation effects in the overall response of highway bridges.(32) As shown in figure 14, two of the degrees of freedom (vertical displacement and torsional rotation) are uncoupled with the other degrees of freedom in the stiffness matrix. However, the two components of horizontal translation are coupled to the two degrees of freedom of rocking (tilting) rotation in the stiffness matrix. In general, the off-diagonal (cross-coupling) teras are neglected for two reasons: 40xi) ya) 213) 213) 5 By 8k yO ky 0 0 oO KB 0 oO 2 O Ke OO 0 Of Ol aKa 0140.0) “Ker OOK Figure 14. Form of footing stiffness matrix. a(1) the values of these off-diagonal terms are small especially for shallow footings, and (2) they are difficult to compute. For the typical highway bridge conditions, the off-diagonal terms in the stiffness matrix for a footing can be neglected. However, for some unusual deeply embedded footings, where the ratio of the depth of embedaent to the equivalent footing diameter is greater than five, the contribution of stiffness from the cross-coupling tera may need to be included. ‘The procedure to solve for the coefficients of the stiffness matrix will be presented below. The definition of the geometric parameters and the coor dinate system which will be used exclusively for the rest of this section is summarized in figure 15. 5.3 General Stiffness Equation of a Rigid Footing ‘The stiffness matrix K of an embedded footing can be expressed as the following general equation: ste recta aerate teres erie erate eace treo renters tes ece(S)) where Ko is the stiffness matrix of an equivalent circular surface footing (see section 5.4, and figure 14) a is the foundation shape correction factor (see (section 5.5), and # is the foundation embedment factor (gee section 5.6) 5-4 Stiffness Coefficients of a Circular Surface Footing As defined in eq. 5, the solution for a circular footing bonded to the sur~ face of an clastic half space provides the basic stiffness coefficients for the various components of displacement as summarized below: Vertical translation (K33 term in figure 14) AB... 6) IY 42g 1 a EXCAVATED CASEY (A) EMBEDMENT. pe mE YY 1 (8) sizes Ss 2 {c) COORDINATE SYSTEMS A Figure 18. Definition of geometry parameters and coordinate systems for a footing. )Morizontal translation (Ky and Kp tn figure 14). 2B... (7) a 3 Torsional rotation (Xgg in figure 14) ~ 16,GR" . . ca) 3 Rocking rotation (K4g and K55 in figure 14) ee omer t1(9) 30-9 where G and Vare the shear aodulus and Poisson's ratio for the elastic half space, respectively. R is the equivalent radius of the footing as per figure 16. A discussion on values of G and Yto be used for footing analysis is given in section 3.6 of this report. 5-5 Rectangular Footings Application of the general equation (eq. 5) for rectangular footings involves the following steps: (1) Calculation of the radius of an equivalent circular footing for the various modes of displacement. The equivalent radius can then be used in eqs. 6 through 9 to solve for the baseline stiffness coef- ficients Ky in Eq. 5. The procedure to calculate the radius of an equivalent circular footing is summarized in figure 16. (2) Development of the shape factor a to be used in eq. 5. The shape factor for various aspect ratios (B/L) for the various modes of displacement are presented in figure 17. 5-6 Embedment Effects Three idealized conditions are commonly encountered related to enbedment effects for footings. These three conditions are schematically illustrated in figure 15. In the first situation (case 1 in figure 15), the grade level is at or below the top of the footing. The depth of embedment D is then uerely the depth of the bottom of the footing below the grade level. In cases 2 and 3, the top of the footing is below the grade level. Therefore, an addi- tional geouetric paraneter h needs to be considered. In case 2, h is the thickness of the ground cover whereas, in case 3, h is the depth of excava~ thon. 44RECTANGULAR, FOOTING EQUIVALENT CIRCULAR FOOTING EQUIVALENT RADIUS: TRANSLATIONAL: = R= [AEE ROTATIONAL: = jean au? 7 “Pa elie: a 2 + 4.2) py ~ [sseuatesnt sulk AXIS ROCKING) sol¥AXIS ROCKING) (2-AXIS TORSION) Figure 16, Procedure for calculating equivalent radius of a rectangular footing, sSHAPE FACTOR, < 1.20, 1.154 TORSION (Z--AX! us Figure 17, Shape factor for rectangular footings.To accommodate all of the above situations concerning the effects of embed~ ment, an extended set of design charts is needed to provide the embedment fac~ tor 6 ineq. 5. The design charts should accommodate the various modes of stiffness components (translational and rotational displacements) and the various combinations of the three basic geometric parameters: R, D, and h. In view of the above complexity, a sensitivity study was conducted (see appendix B) and a simplified procedure was developed. Simplified Procedure. The results of the sensitivity study (appendix B) indi- cate that in typical situations encountered in highway bridges, the effects of the ground cover or excavation can be ignored. The situation depicted in case 1 in figure 15 can be used to approximate the conditions in cases 2 and 3. ‘The foundation embeduent factor 8 for various embedment ratios ( D/R ) for the different modes of displacement are presented in figure 18. In the case where the top of the footing is below the grade level, it ts recommended that the thickness of the ground cover or excavation ( h ) be ignored and the thickness of the footing (not the actual depth of embedment Dz) be used to calculate the embedment ratio ( D/R ) in determining the embedment factor 8. 5.7 Stability of Footings Under Earthquake Load The above sections describe the procedure to determine the foundation stiff nesses vhich could be used in dynamic response analyses of the overall bridge- foundation system to determine the earthquake induced load level. After the earthquake-induced load level has been established, the next step is to exa~ mine the integrity of the footing foundation against the earthquake loading conditions. The stabillty of the footing-soil mass system should be eva~ luated. This is usually accomplished by using limit equilibrium methods which have been well documented in the literature and are not reproduced in this report.(33) A brief discussion is provided below regarding application of the limit equilibrium method for earthquake loading conditions. ‘The classical bearing capacity equation to evaluate the stability of a footing avises from the consideration of either a untform vertical pressure or a ver~ tical point load applied to the center of a footing. Earthquake Loading con- dictions are quite difterent from the above simple vertical loading condition (see figure 19). In addition to the static vertical load, it is necessary to a790 9 TWNOILWLOY ONY TYNOISHOL 70 30 80 20 2.0 15 1.0 ee. Wey Li) SH OR 5 0. 3.0 (IVINOZINOH ONY T¥OILU3A) TVNOLLYISNYEL 9 yoLows anawaaawa oR Figure 18, Embedment factor of footings. cyconsider cyclic vertical, lateral, and overturning moment from the earthquake. Usually, the cyclic nature can be ignored and the stability of the footing is evaluated by simply superposing the peak value of the earthquake-induced live load on the constant static vertical load. As shown in figure 19, the above loading condition 4s usually converted to an equivalent point load applied at some eccentricity e and at an inclined angle. The classical bearing capa- city equation is usually modified to account for the effects of the eccentri- city and the angle of inclination of loading.(33,34) the modified equation can be used to solve for the magnitude of the maximum allowable load that could be sustained by the soil prior to bearing capacity failure. The allowable load can then be compared to the load from the earthquake to assess the stability or the factor of safety. ‘The earthquake-induced monent could potentially cause partial uplift of the footing. The uplift areas can be calculated in the course of the bearing capacity determination discussed above. The FIWA/AASHTO guidelines recommend that the footing should be designed such that less than half of the footing base would be uplifted. 5-8 General Guidelines and Commentary In addition to the above described procedure for foundation analysis, guideli- nes to assist the bridge engineers in designing footing foundations for earth~ quake conditions are given below: 1. Liquefaction. An assessment of the liquefaction potential (see section 4.2) at the site should always be conducted wherever the use of footings are planned for highway bridges. Spread footing foundations (not supported by piles) should always be avoided wherever there is a strong likelihood for site liquefaction. 2. Dynamic Settlement. In addition to assessing the stability of the footing, the potential for settlement should also be examined. ‘Two forms of earthquake~induced settlement may occur. The first relates to the settlement of the overall site which has been discussed in section 4.5. The second relates to the localized settlement around the footing which could potentially cause differential settlement and causes misalignment of the bridge. A simple procedure is described here to assess the magnitude of 49Figure 19. Consideration of footing stability for earthquake loading condition.settlement of the footing. This procedure arises from field obser vations and model testing of the performance of footings under earthquake loading conditions. In general, loose and medium sand has been found to be susceptible to earthquake-induced settlements. Therefore the first step in evaluating the magnitude of settlement is to identify the presence and the layer thickness of the loose and mediua sand within the zone of influence of the footing (extending to a depth equivalent to the dimension of the footing, e.g-, the diameter). The earthquake-induced settlenent can then be assumed to be less than 0.005 of the thickness of the loose or medium sand layers within the zoue of influence. The highway bridges should be designed to withstand the above magnitude of settlement of the Footing. Uplift Load. The uplift capacity of the footing foundation is usually very small. ‘The magnitude of cyclic vertical load should always be smaller than the static vertical load for a footing to prevent uplift of the overall footing foundation. Uplifting of a footing should only be allowed when the consequences of redistribu- tion of loads on the overall bridge and also its effect on connec~ tion details are carefully examined. The FHWA/AASHTO guidelines allow partial uplift of the footing foundation, but the uplifted area should be restricted to less than half of the footing area.6.0 ANALYSIS OF PILES 6-1 General Pile foundations provide a means of support for the structure in difficult subgrade or loading conditions including the following situations where: © the upper soil strata are weak or compressible. © the shallower soil layers are susceptible to Liquefaction. © footings cannot transmit inclined, horizontal, or especially uplift forces. © scour is likely to occur. 0 future excavation may be adjacent to the structure. © expansive or collapsible soils extend for a considerable depth. Rarthquake-induced loading of a pile foundation is dominated by the inertia force of the superstructure. As a result, loading of a single pile from earthquake excitation bears many resemblances to a pile subjected to moment, shear and axial load at the pile top from structural loading. As shown in figure 20, the basic problem of a pile foundation involves the need to distribute a set of superstructure Loads into the surrounding soil mass through the pile member. ‘The general case involves consideration of three components of translational forces (an axial and two lateral shear for- cas) and three components of rotational moments (a torsional moment about the pile axis and tvo rotational moments about two orthogonal horizontal axes) along the pile member. For convenience in design or analysis, the axial soil support characteristics are assumed to be independent of the lateral soil sup- port characteristics. This is justified because lateral soil reactions are usually concentrated along the top 5 or 10 pile diameters whereas almost all of the axial soil resistance is developed at greater depths. Therefore, the axial and lateral soil support behavior can be studied and analyzed separa- tely. Under normal Loading conditions, some portion of the soil mass (at least near the surface) will always be yielded. Furthermore, in normal con- 52soning juoddns 108 soauyjuon (9) 1018107 tony (226) PoReW pue luessg Jar) 7) vonsesanu 3s JeuONUOWIP Sey. “OZ eNBI-} squawaroidsip aid suoi20a aid (4) Jouoysuawip- aarp (0) L! ao) os you9yssoy. ‘ *yuawaroydsy 1019107 z ‘wawaao)dsig toe 63ditions, several layers of soil will be encountered along the length of the pile. Therefore, a realistic analysis approach should account for the nonli~ near and the layered nature of soil conditions. In view of the above complexity, current design practice usually models the soil support charac- teristics along the pile by discrete Linear or nonlinear springs. Design or analysis of a soil-pile system usually involves aodeling of @ beam-columa sup ported by a set of lateral springs and a set of axial springs which can be Linear or nonlinear. A set of nonlinear support curves characterizing the lateral soil reaction versus lateral pile deflection 4s usually referred as a set of p-y curves. The corresponding set of nonlinear curves for axial load transfer characteristics 1s referred as a set of t-z curves. Torsional resistance against rotation of individual piles can usually be ignored or assumed to be very swall for highway bridges. Method of Analysis. A variety of methods is available for carrying out analy- ses of pile foundations.(35,36) Because.of the inherently variable and nonlinear nature of soil, there is seldom any advantage in attempting to apply closed form mathematical solutions or in developing extensive design charts. Instead, each case should be considered as an indivdual problem of obtaining compatibility of behavior among the structure, pile, and soil. This process 4e most conveniently handled by established computer models. Several computer codes now exist to provide nonlinear pile-soil solutions in practical design problens.(37, 38,39) such computer codes can yield reliable solutions with minimal amounts of effort (computer cost and labor) and can be easily used by engineers, even those with Limited computer training. Therefore, computer solutions are recommended to highway bridge engineers for pile and pile group analyses. In Limited cases, where the soil profile consists of a homogeneous layer, hand solutions with the aid of existing design charts can readily be used. Solutions by hand may also enhance understanding of the problem of soil-pile interaction. Therefore, the method of hand solution will also be presented for some simple soil conditions. The construction of a full set of p-y curves for lateral pile loading analysis involves calculating several p-y curves (one curve at each selected depth) along the entire length of the pile. Interpolations are done to provide p-y characteristics at all nodal points between input p-y curves. The nodal 54points should be closely spaced but only four to six input p-y curves are usually sufficient. Nodal point spacing of about one-half pile diameter is recommended to provide good resolution of the distributed soil support. Placement of p-y curves would typically be at the top and bottom (or assumed bottom for a very long pile) and perhaps 1, 3, and 7 diameters from the top. Since the lateral response action of a pile is concentrated close to the soil surface, p-y curves should be prescribed at closer spacings near the top. The pile response is usually relatively insensitive to p-y curves prescribed at greater depths. The pile solution can usually be cut off at 30 to 40 diame- ters without affecting the lateral behavior of the upper part. Construction of a p-y curve at each depth involves (a) formulation of the ultimate resistance py (in force per unit of pile length), (b) development of the {nitial tangent stiffness E. , and (c) fitting a curve. shape to meet the above initial stiffness and ultimate resistance. The step-by-step proce- dure for a particular pile is given below. At the present time, procedures for other pile shapes (e-g-, square or H piles) have not been well established, but to a satisfactory degree of approximation the actual gross width of the pile may be substituted. 6-2 Development of p-y Curves for Sand Lateral soil-pile interaction analysis involves constructing a set of p-y cur~ ves. The most widely accepted basis of developing the p-y curves arises fron correlation of large scale pile load tests. For sand, the pile load test con ducted at Mustang Island and the correlation program by Reese( 38,40) serve as the basis for development of p-y curves for sand in the design of pile foun- dations for offshore platforms against wave loading. ‘This procedure is recom mended for highway bridge applications. The specific procedures recommended by Reese are rather complicated and tedious. ‘The approach has been simplified and refined by others.(41,42) The differences between the p-y curves from the various simplifications and refinements and the original Reese's criteria are negligible. Ultimate Resistance (py). For sand, the ultimate resistance (py) in force per unit length is developed based in concept on either (1) the failure of an 55assumed wedge around the upper portion of the pile or (2) a horizontal plane— strain plastic flow around the pile. The wedge failure usually governs the ultimate resistance (py) of soil at the upper portion of the pile. The second mechanism usually governs the ultimate resistance (pyy at a greater depth, usually below ten diameters for a typical pile used for highway bridges) but in reality loading to the ultimate resistance is never achieved at those depths. Solution of the ultinate resistance involves calculation of the two ultimate values (py, and pyg)+ The lower of the two values, reflecting the proper failure mechanisa, should be used to construct the p-y curves. Actually, the wedge and ultimate resistance concepts are modified significantly in the Reese method by constants A introduced to force a fit to the Mustang Island test data. The step-by-step procedure is given below: Step-1 Calculate pyy = (Cyz + CpD)Y2 settee ee eee + C10) vhere z 4s the depth below the soll surface, D is the average pile diameter between depth z and the surface. yz is the vertical effective stress at depth z, and Cy and C2 are parameters from figure 21. Step-2 Calculate pyg = CY str eee reste eee ee CD where C3 ie given in figure 21. Step-3 Use the smaller value of py or pyg for the theoretical ultimate resistance (py). Step-4 Obtain an empirical factor A which is developed from correlation of pile load test data, from the following equation: A= 3- 0.82 / D up to a limiting lower bound value of 0.9 Breese setepacleteeraierer ee ppastesraeeceeetitn@ V1 Step-5 The ultimate resistance (py) to be used for developing the p-y curve is given by multiplying A by the theoretical ultimate resistance value al mia eatcee terete retisteetotee aie ietee He Hated eaieteeeeeeECLS) 56VALUES OF COEFFICIENTS C, AND C, 5 100 30 a 80 (Oates ui 3 60 3 ra ra 50 W oO 2 40 bs 30 8 a | 20 = 10 0 oO 20 25 30 35 40 ANGLE OF INTERNAL FRICTION, }, DEG hee ae ie a : x oT 08s oF EAT OE vl (After Bogard and Matlock, 1980). Figure 21. C;,C2, and Cg coefficients for py curve calculation for sand. 37Initial Tangent Stiffness (Hy). The initial tangent stiffness of the p-y curve for sand is assumed to vary linearly with depth z as recommended by Terzaghi.(43) The initial tangent stiffness of the p-y curve (Eg) in force per unit deflec~ tion per unit length of pile, which can be assumed to be independent of the pile diameter, varies linearly with depth and is related to a coefficient k,.(40) the recommendation of the coefficient for sands of various den- sities or friction angles is given in figure 22. Biri pm edaeieeetee ees tiatte a teeters Pads tearte tet eet et discs eee (49: Gurve Shape. A hyperbolic expression can be used to describe the relationship of the pry curve (see figure 23) which has an ultimate capacity (py) and an initial tangent stiffness (Eg). ee a ogee ee Pu «T+ y/¥e where Ye = Pu/Es An example is given in appendix ¢ for the construction of a full set of p-y curves in sand. 6-3 Development of p-y Curves for Clay The most widely used procedure to develop p-y curves for clay has been based on the research by Matlock.(44) The procedure recommended for pile design for highway bridges is again based primarily on the research by Matlock. Construction of a family of p-y curves for clay is very similar to that for sand. The step-by-step procedure is outlined below: Ultimate Resistance (py). The ultimate resistance (py) in force per unit length is developed by using the smaller value of py, or pyg which correspond to the shallow wedge or the deeper horizontal plastic flow failure mechanisms. Pat and.pyg depend merely on the undrained shear strength c. Shallower Failure Mechanism: Pul = 3¢D+ YD2t Jez. ee ee ee eee ee ee ee (16) 58FRICTION ANGLE, @ 28 2° 20 28 ae 45° y VERY veax. | Loose MEDIUM verse | MERY 3200 250 ——— nia Tacent mooutus {REESE 1974) 4 pee a vei J $ (TERZAGHI, 1955) § = 200 Z x é ¥o 2 5 3 2 % 150 = 3 a 6 3 10 3 3 y 2 oy Se 50 sae Vv we ee = 95 cae, ou = : See ° 20 40 60 80 100 RELATIVE DENSITY, Or (percent) NOTE: 1 Ib/in? = 271.4 KN/m? (After O'Neill and Murchison, 1983), Figure 22. Modulus of subgrade reaction, Ky for sand, See 5FORCE PER UNIT LENGTH, P FORCE PER UNIT LENGTH, P IVCes, EQUIVALENT LINEAR Pu ve= fe Kye {SEE SECTION 6.2 FOR Pu FOR SAND) : (A) SAND, €s, EQUIVALENT LINEAR Ye" 2.6 ED (SEE SECTION 6.3 FOR Pu FOR CLAY) PILE DEFLECTION, y (B)CLAY Figure 23. Hyperbolic p—y curve shape for sand and clay. 60Deep Seated Failure Mechanisns: ia He geese creat edt etd ei gee eee a) where D is the pile diameter Ya 4s the vertical effective stress at depth z, and J is a dimensionless empirical constant with values ranging from 0.5 for soft clay to 0.25 for stiffer clays. Initial Stiffness (Es). The initial stiffness of the p-y curve for clay is given below! eee cee eres eee eee eee eee eEe eS ceeteeEe sy Ye vhere Yes 25 €D where & is the strain amplitude which occurs at one-half the peak deviatoric stress on laboratory undrained compression tests on undisturbed soil samples. Usually § ranges from 0.005 to 0.025. In the absence of laboratory data, a value of 0.01 is suggested. Gurve Shape. A hyperbolic relationship {s again recommended to describe the relationship of the p-y curve (see figure 23) which has an ultimate capacity (Py) and an initial tangent stiffness (Eg). PB. Lye ee eee eee tcc as) Pu” T+ y/¥e 6.4 Linear Subgrade Modulus The method for construction of nonlinear p~y curves for both sand and clay has been presented in previous sections. Nonlinear pile solutions under lateral loading in layered soil deposits usually requires the aid of computer models. Im many ‘cases, due to the insensitivity of overall pile behavior to the variation of soil support characteristics, linear representation of the soil stiffness would yield pile solutions of reasonable accuracies. The following 61recommendation of Linear soil stiffness could lead to a reasonable approxima- tion of the nonlinear solution in limited ranges of pile deflection. Subgrade Reaction for Sand. The modulus of horizontal subgrade reaction for sand recommended by Terzaghi hae been widely used in practical applica~ tions.(43) The magnitude of the support spring on piles can be assumed to be independent of the pile diameter and varies linearly with depth. Raat cher eee eee eee eretcret ePelceett tia delty (20) where Eg is the stiffness of the support spring in force per unit length per unit deflection, ky is a coefficient which depends on the density or friction angle (see figure 22), and 2 is depth from grade level. As shown is figure 22, the ky value recommended by Terzaght 1s smaller than that recomended by Reese. The value recommended by Reese corresponds to the initial tangent. stiffness of the load-transfer characteristics. The value recommended by Terzaght corresponds to the secant stiffness of the load- transfer behavior at typical design load levels. To provide the practicing engineers with the feel for the range of validity of Terzaghi's recommendation of linear horizontal subgrade stiffness, a set of pile head load-deflection solutions are presented in figure 24. Both free and fixed pile head boundary conditions are examined. ‘The densities of the sand range fron loose to dense with corresponding friction angles ranging from 30 to 40 degrees. A comparison of the solutions using Terzaghi's Linear stiff- ness with the nonlinear p-y approach indicates that Terzaghi's recommendations leads to an equivalent secant stiffness of the pile at about 0.2 to 1.0 inch (0.5 to 2.5 em) of pile head deflection depending on the friction angle or the density of sand. Based on the sensitivity study (appendix 6), typical pile head deflections for most highway bridges could range from 0.02 to 0.5 inches (0-05 to 1.3 cm). = pret? kegs tres (Tt? , Koo = T4HED/T , KSe purer a oHlt Fea Mead Cod 62/ 30 3 60 i 3 ’ 2 3 3 $ 7 210i NOTES: 0 1 — — LINEAR SOIL (1) 1ft 03m lin) ReeeuoN (2) ip =4448N FIXED HEAD CONDITION (3) 1 tb fin? = 28.7 Nécm2- HYPERBOLIC 20 rea : ea 3 30 f- a ¥ e ‘ . z [| 2 20 5 2 S se 3 ’ g ZL 2 10 r L, 0 3 a5 7 1s 20 Bo) (8) FREE HEAD CONDITION Figure 24, Comparison of linear and nontinear pile solutions for sand 63Subgrade Reaction for Clay. Terzaghi also recommended a procedure to obtain the linear subgrade stiffness for clay. However, the magnitude of the support spring depends merely on the undrained shear strength of clay and experiments show that it does not adequately account for the failure pattern of soil at shallow depth. Therefore, the following recommendation is provided. The equivalent linear spring constant to account for the soll support on piles can be developed based on the nonlinear p-y procedures for clay (section 6.3). The following considerations lead to the resultant linear spring represen- tation. 0 At normal design load levels, the pile response is largely affected by the support characteristics at shallow depth. Therefore the deep seated horizontal plastic flow failure mechanism can be ignored and the equation to calculate the ultimate resistance for the shallower failure mechanism (eq. 16) can be used to develop further simplified linear procedures. © The secant stiffness at one-half the peak load level on the p-y curve can be used to develop the equivalent linear stiffness. This procedure has been widely used in geotechnical practice. Using the above line of reasoning, the equivalent linear secant stiffness for pile analysis is developed as follows: gleich ky eistitiietecteetetee titted teete ete stati tietee restate (a where Ko = 0-6 c/ & y= 2 C14 2D 2 is the depth below the grade level D is the pile dianeter € ts the undrained shear strength Tis the effective unit weight % is the strain amplitude at one-half of the peak deviatoric stress level (see eq. 18) 64A sensitivity study 1s conducted to compare the previously described procedure for linear and nonlinear pile solutions for clay. A set of pile head load- deflection solutions are presented in figure 25. Both free-and fixed-pile~ head boundary conditions are examined. The undrained shear strength examined ranges from 0.5 KSF to 5.0 KSF (24 - 240 kN/m2), It can be seen from the com parison that the linear subgrade stiffness approach yield reasonable solutions up to about 0.5 inch (1.3 em) of pile-head deflection. The above recommended procedures for clay have been shown to yield reasonable solutions for soft to medium clay. For stiff clay, 4t is important to recognize that it is not shear strength alone which is important. Figsures, cracks, and small irregular blocks in contact with each other, typically found in stiff dessicated clays, would result {n deformation and elip highly con- centrated along such surfaces which may substantially alter the p-y charac~ teristics. Some conservatism (e.g-, assuming a lower ultimate resistance, say, a half value) would be appropriate. 6.5 Analysis Methods for Lateral Loading on Piles Procedures for development of soil support characteristics for sand and clay have been presented in previous sections. This section describes the various methods which can be used to implement the above described support charac- teristics and the properties of a pile in an overall analysis of a single pile. There are two primary objectives of a soil-pile interaction analysis: 1. Obtain the load-deflection (stiffness) characteristics of a soil- pile system which can be used in an overall structure-foundation dynamic response analysis to obtain the earthquake load (overall and distributed load). 2. After obtaining the overall dynamic response and load distribution analysis to assess their load level on the pile, evaluate the magnitude and the distribution of bending moment and lateral shear force along the pile. 651090 80 3 & 60 7 2 40 = 20 ° NoTES: ° mo ve : LINEAR son. () ft =03m tin) REACTION (2) 1b = 4.448. Een K+ k (A) FIXED-HEAD CONDITION ot Kye (2) 1b /n2 = 28.7 Nlom? ee eoeaeatis (4) 1 kst = 47.9 nea poy cUAVE, 7 |_| = 20 4 & 2 2 3 g 20 é VEBee 19 = fe | 05 10 05 10 15 20 Sui} (8) FREE-HEAD CONDITION Figure 25, Comparison of linear and nonlinear pile solutions for clay.‘Two types of analysis approaches will be presented in this section which can be used to achieve the above objectives: 1. the computer method, with linear or nonlinear soil supports, 2. the hand-solution method, with equivalent linear soils supports. Computer Method. ‘The computer method represents the most practical and effi- cient approach for seismic analysis and design of the pile foundations. Several computer programs exist in the public domain which can yield reliable linear or nonlinear pile solutions in layered soil deposits and where the effects of the ground water table need to be considered. Most of the struc~ tural analyses, especially for aseismic design for highway bridges, are con- ducted with the aid of a computer. Computer analyses of the pile foundation are not expected to present major difficulties for the bridge engineers. After the preliminary phase of setting up a well-established and well-documented pile program, practical pile design analyses incorporating the linear or nonlinear so{l support characteristics (as described earlier) can be conducted on a routine basis. The following comments are provided regarding the various modeling aspects of such computer approach: 1. The pile member may be represented by finite elenents, difference equations, or by discretized sechanteal analogs.(38, 39,45) 411 the above approaches will produce essentially identical solutions. ‘The accuracy of the pile-soil interaction solution is primarily a function of the closeness of spacing nodal supports and not of the degree of sophistication of the pile elements.(%®) 2. Most of the widely used computer programs accept prescriptions of linear or nonlinear soil support curves along the length of the pile and will automatically perform interpolations to develop support curves for all nodal points throughout the entire length of the pile. Therefore, linear or nonlinear soil support springs are usually caculated and input at the boundaries of each soil layer. In addition, the pile solu- tions are more sensitive to the support springs over the top five diameters. Sometimes, even though the soil layer is homogeneous, the linear or nonlinear support springs at shallow depth may not vary linearly with depth (see the proce— dures of constructing support curves described earlier). Therefore, the linear or nonlinear support springs may need to be calculated at a more refined basis at the top 5 to 10 diameter of soil. 673. Im general, the pile behavior (lateral stiffness and monent distribution) is very significantly affected by the structural connections at the pile head (e.g-, fixed or pinned or inter mediate condition). Although, the fixed or pinned boundary conditions are often assumed for design purposes, the highway engineers should be cognizant of the fact that both the above are probably idealized conditions. It 1s difficult to design a structural detail to achieve the above idealized condition. In reality, the structural constraint at the pile head falls in between a fixed and a free-pile-head condition. Assuming a fixed-or free~head condition, though, may provide an upper and lower bound of the pile stiffness characteristics. However, they do not necessarily lead to bounding solutions of the magnitude of pile monent (47) The above discussion has presented the various modeling aspects of using the computer approach in conducting soil-pile interaction analyses. An example is provided in appendix C which presents some typical input parameters and output solutions. Linear Hand~Solution Method. As discussed earlier, the computer approach ts recommended to accommodate the uonlinear, layered eituation in pile design. It 1s often impractical to develop pile solutions by hand because numerous sets of design charts are needed for the layering and the variability of the soil conditions. In very limited situations where the sofl and groundwater table conditions are very simple (e.g-, homogeneous with the entire pile totally submerged or above the water table), hand-solution procedures can be used to develop reasonable pile solutions. The hand solution methods also can lead to imptoved understanding of the pile-soll interaction problem. The following discussion is provided to illustrate how the hand=solution approach can be used for the more idealized cases. The tables can also be used with an iterative procedure to provide approximate nonlinear pile solutions for piles with nonlinear soil supports. Several example problems are presented in appendix C to illustrate the various methods of usage of the nondimensional tables. Im the case where there are uniform soil deposits and where the water table is either above the pile head or located at a depth below the zone of influence (say below five diameters), the variation of subgrade stiffness with depth developed using the recommended linear subgrade Modulus approach, (section 6.4) 68will lead to situations summarized in figure 26 for a sand site and a clay site. Design charts developed by Matlock and Reese can be used to obtain the pile solution for any combination of moment or shear loading applied at the pile head.(48,49) In general, the pile solution will depend on the length of the pile member. However, for very long piles, the length L loses signifi- cance because the deflection may be very nearly zero for much of the Length of the pile. A characteristic length, (1) is usually used to evaluate whether the pile under consideration can be regarded as a long pile. The characteristic length T can be defined as follows: grime Vee eee eee eee es eee eee (22) where EI is the bending stiffness, and ky is the coefficient of variation of the subgrade modulus with depth as defined in figure 26. For typical highway bridge conditions where the ratio of length of the pile to the characteristic length (L/T) is greater than 4.0, the infinitely long pile solution would yield sufficient accuracy (within 10 percent of the actual finite length solution). Therefore, the infinitely long pile solution is adopted. For a linear problem, the principal of superposition may be applied. The effects of an imposed lateral load P; and imposed moment, Mr, may be con~ sidered separately as shown in figures 27 and 28. If Yq represents the deflection due to the lateral load, Pe, and if Yg is the deflection caused by the moment, Mr, the total deflection i IM erPAti Vm etetie calistie erie te tee cedseete reteset ie rutiets (23) Similarly, the solution for slope, moment, and shear and soil reaction (see figure 27) can be solved by superimposing the solution from lateral load Py and moment Mp. 69Es= Kye SEE FIGURE 22 FOR VALUES OF Ky ae ny er kgs Ki e058 eye 5 200+ 02+ Jez: 8 TEED © van (d : nose v02(h ° 9 UNFT WEIGH oF sol, COHESION (a) CLAY Figure 26. Variation of equivalent linear subgrade modulus for sand and clay,(z aie a9s) pues uy uo|rnjos ad seaUl7 “Zz asnB| 5 ue ‘a ‘a “a ‘a ‘a % y “y y *y fe fa(t/im) + 4v(/4) = dd vonsoes Hos fay) + *v('a) =A 030s aE “a(tn) + “v(Lk) = W quowow = MOTE Hd3G = *a(ia/tn) + ‘v(iazid) = s dog 73) = ‘guajw) + fwuayy = K vouseyedThe form of the solution for deflection Y, slope S, moment M, shear V and soil reaction P for the form of a subgrade modulus of Eg = kjz are shown in figure 27. ‘The A and B shape functions corresponding to the solution for lateral Load Py and monent My are tabulated in table 3. The form of the solution for deflection y, slope $, moment M, shear § and soil reaction for the form of subgrade modulus of Eg = kg + ky2 ts schematically presented in figure 28. A tabulation of the A and B coeffictents for various combinations of Pp and My are presented in tables 3 through 8. The above described nondimensional tables can be used to solve for the pile behavior. In addition to providing solutions for any combination of pile head forces and moments, the nondimensional tables can be used to provide solu- tions for a pile which has a rotational constraint on the pile head (including the fixed-pile-head condition). 6.6 Development of Load Transfer Characteristics for Axial Loading The aspect of lateral loading on piles is usually emphasized for earthquake consideration. However, the rotational (rocking mode) behavior of a pile group has a significant imfluence on the overall lateral load~deflection behe~ vior of the bridge-deck.(50) the rotational stiffness of a pile group is related primarily to the axial stiffness of each individual pile. Therefore the procedure to evaluate the load-deflection characteristics of an axially loaded pile will also be discussed. ‘The various components of an axial pile problen are illustrated in figure 29. The overall pile behavior depends on the axial pile stiffness (AE) and the shear-transfer characteristics (t-z curves) along the side of the pile and at the pile tip (tip q-z curve). The fundamental problem in an analysis of piles under axial loading relates to the uncertainties of the load transfer charac~ teristics at the side and at the pile tip. The following factors need to be considered in developing the load transfer characteristics: © The side-friction capacity along the length of the pile. o The ultimate resistance at the pile tip. 72*(g yBnoiys ¢ 21q20 40s) Aejo UI UoAA[Os ajtd seauI7 “gz eunb}4 is 5 we ‘a ‘a “a ‘a ‘g %y y ky feQa/'W) + 4v(L/4) * ‘woyo094 jos i sa(istn) + v('d) = apayg aov4uns ‘Gnriows 2 2 “a(n) + YW(L8) = OW juowoy — MOTSB HGS 2\ 2 13 *a(ia/w) + ‘v(3414) © edoig ‘s 7!) ‘gus4iw) + afi)» uoy2ourQ_CCT } | 5 z 2 z 6 2 8 ol [2 @ LO OE 2 Nie eee 3 8 g 8 — z —© The form of the load transfer-deflection curves associated with each of the above forms of soil resistance. The aspects of side-friction capacity and the ultimate end-bearing resistance have been the main topic of research and consideration for pile design. The ultimate capacity of a pile depends on numerous factors including (a) the soil conditions and pile type, (b) the geologic history, and (c) the pile installa~ tion methods. Numerous methods have been proposed to predict the capacity of a pile. They can be grouped into the following categories: © Empirical method based on correlation of shear strength of soil (e-g+, ¢ and #) and pile load test data.(45) © In situ testing methods which develop pile capacities from in situ measured data (e.g-, from blowcounts, CPT data, or pressuremeter tests) .(51) © Dynamic meagureaent from pile driving records and wave equation analyses.(5 o Engineering Formulas.(53) © Direct application of site-specific static pile load test data.(54) Although some of the above methods have been more widely used in practice (e.g, the empirical method in the first item above and the in situ testing method for CPT in sand), none of the above mentioned methods is accepted uni- versally. Incorporation of site-specific pile load test data has been per- ceived to be the most reliable method for pile capacity determination. In view of the above variation of viewpoints and the fact that the aspect of pile capacity is usually addressed in other design considerations (e.g., static design), no procedure will be recommended here for the axial pile capacity; instead, the author encourages the highway engineers to use thetr own local experience and site-specific geologic and geotechnical information to develop the ultimate capacity (for both side friction and end bearing of the pile) tor soil-pile interaction analyses. Load-Transfer Characteristics. In addition to the ultimate side friction and end-bearing capacity, some assumptions need to be made to develop the load transfer-displacement relationships (for both side friction and end bearing) to evaluate the overall pile behavior. 75The subject related to the form of the load transfer-displacenent relationship is again highly complicated. Again, at the present time, there has been no uniform agreement among the geotechnical practitioners on the above subject. The load transfer-displacement relationship described below is relatively simple to use and has been used by many designers. Side Friction: Pimte leet C2 vem) mately ieee steetetite ieee tees eattetce (24) where £ = unit friction mobilized along a pile segment at displacement, z, foax = maximum unit friction, and eae a teen eae ea fnax 18 fully mobilized. Az, value of 0.2 in (0.5 cm) is recommended for all soil types. End Bearing: a 259 cman qe pretetetaieiiereraieeetiais tergreters ste (25) where max = maximum tip resistance q* tip resistance mobilized at any value of z ze, and ¢ = critical displacement corresponding to dnax: A Ze value of 0.05 of the pile diameter is recommended. 6.7 Analysis Methods for Axial Loading on Piles At normal loading conditions (e.g., under the static dead weight), plastic slippage at the pile-soil interface will occur along a significant upper por- tion of the pile. ‘Therefore, a realistic analysis approach must be able to account for the nonlinear nature of the plastic slippage behavior. Computer Method. The computer approach can again provide the most con~ venient means of solving for the axial pile behavior.(37) Many of the well established computer programs allows for prescription of the t-z curves at various depths along the length of the pile (e.g-, at the boundaries of each 76soil layer) and will autonatically perfora interpolations to develop the sup- port curves at all the pile stations. The t-z curves for side friction can ugually be assumed to be symmetrical for both compressive and tensile loading (see figure 29). In general, the q-z curve at the pile eip will be nonsya~ metric (see figure 29). The tensile resistance at the pile tip can be assuned to be minimal. A series of analyses should be conducted to solve for the full range of the load-displacoment characteristics of the pile up to the load level where all the soil resistance has been mobilized. Working load behavior can then be examined in the light of the overall characteristics and reserve load-deflection characteristics. Hand~Solution Method In addition to the computer solution method, a simple hand solution method can also be used to develop the complete nonlinear load displacement characteristics of the pile which takes into account (a) the side shear transfer-displacement, (b) the end-bearing load-deformation, and (c) the elasticity of the pile member. The procedure is illustrated graphically in figure 30. The steps involved are outlined below: 1, Establish the unit side-friction profile and the end bearing capacities. 2. Construct the accumulative side-friction and the pile-tip resistance versus pile displacement curves, in accordance with the assumed distribution of side friction and tip resistance as well as the assumed soil-reaction-displacement curve shapes. 3. Construct a load-displacement curve by summing the load values of the accumulative side-friction and the tip-resistance curves at any displacement value resulting in a load-displacement curve that corresponds to the rigid pile case. 4. Add an additional component of displacement to the above rigid load-displacement curve to account for the pile compliance. The amount of additional displacement at each load level (Q) can be approximated by the following equation. 7‘and e Jo sonsuajoeseyo 1usws2e9|ds}p—peo} j21xe Jo yawidojerap 20} aznpeo01d “OE 2:n61y NX cb = 24 1 wBbOrO= wL w y9z0'0 = ut ‘ALON worsszuenog —|—raanian (eon Lani WoW 1 0 40 co ne fe ve co re te ese WONLDINS NIXS Ruane ie ane : eaue coumer nna 27 F aowwssisaea : sgepeeee i Autoveveanssnven09 oon i (won ANaRAastO WKY Pe 3 (ntene Vwongs-2071 107 maw teatat ALovarD ns WHY oo oe o * (a wouvwransa riS (ite Compliance) =e ee ee ee (28) where L is the pile length, A and E are the cross sectional area, and the Young's modulus of the pile, respectively. In most cases, except when the pile tip 1s embedded directly into bedrock, resulting in an extremely high tip resistance, the rigid pile load-displace~ ment curve (from step 3 above) provides a good approximation of pile-head stiffness. Most of the piles used in highway bridges are relatively short in terms of axial response behavior. ‘Therefore, the component of displacenent from the compliance of the pile (from step 4) is relatively suall. In the case where the pile tip is embedded directly into the bedrock the elastic shortening or elongation of the pile member will account for most of the compliance of the soil-pile system. Step (4) above, alone can be used to calculate the axial pile stiffness. g. # soedeHese oebin a oetee de ieceeds reset (27): Initial Static Loading. The procedure to solve for the axial load~ displacement characteristics for the couplete range of pile displacement has been given above. Under earthquake conditions, sone magnitude of cyclic axial load will be superimposed on a static bias load (e.g-, the static dead weight). Figure 31 illustrates the various factors which come into the pic~ ture due to the above static bias loading. As shown in figure 31, in normal design range, where the maximum load level (from superimposing the cyclic load on the static bias) does not exceed the pile capacities (for both the peak compressive or tensile load), the static dead weight can be neglected in solving for the secant stiffness of the pile. The magnitude of cyclic loading slong with the backbone load-displacement curve can be used to develop the secant stiffness of the pile at the various load levels. The above can be concluded by tracing the hysteretic load-displacement histories using the con- ventional Masing's hypothesis to extend the backbone curve to cyclic Loading environment. The above is generally valid when both the compressive and ten- 9AXIAL F LOAD + eyeLicLoaD + 2a w COMPRESSIVE CAPACITY STATIC LOAD wear STIFFNESS FROM C — D *» STIFFNESS FROM 0 — A AXIAL DISPLACEMENT UPLIFT CAPACITY Figure 31. Effects of superposition of cyclic load on static load. 30sile load levels have not been reached. The load-displacement behavior of the pile will be very complicated when the pile capacities (compressive or ten- sile) have been exceeded. In general, permanent displacement of the pile will oceur for the above condition and should be accounted for in highway bridge design. This aspect will be discussed further in appendix D. 6.8 Pile Head Stiffness Matrix of a Single Pile The above sections describe the procedure to evaluate the load-deformation characteristics of a single pile under lateral (shear and moment) and axial loading conditions. Utilization of the above in an overall response study requires developaent of the equivalent linear pile head stiffness matrix. In certain critical situations (e.g., for highly nonlinear behavior) and where the foundation has a significant effect on the overall response of the bridge, an iterative approach may be required to ensure that the assumed stiffness is appropriate for the resultant load or displacement level. For a given elastic representation of soil stiffness, the form of the pile-head stiffness matrix can be represented by a stiffness matrix as shown in figure 32. In general, the two components of horizontal translation (shear terms) are strongly coupled to the two degrees of freedom of rocking rotation (moment term) in the stiffness matrix. Due to the strong cross-coupling relationship between the shear and moment of the pile, a careful examination of the structural connec~ tivity (pile head rotational constraint) is needed to (a) evaluate the Lateral stiffness of the pile, and (b) the bending moment and stress distribution along the pile. For a situation where a single pile member is rigidly connected to (or con- tinuous with) a column, the full pile-head stiffness matrix (with the cross- coupling term) should be used in the overall bridge-foundation model. The full pile-head stiffness matrix can generally be deduced from the analysis procedure described above solving for the various combinations of translation and rotation corresponding to different combinations of shear and moment load. For a situation where a pile-head connection is pinned, the pile foundation can be represented by a set of physical springs (axial and lateral springs). aL(eC EuMOlnnOnt tine 1 | {oO | Ko fe} K2q fe) fo} Il fol Oo Ks 0 ° ° { [Heel ety leer detest” eek teetat esi kes] © © © Kee 0 - K nee O° ° ° ° 66 a (A) FULL PILE-HEAD STIFFNESS MATRIX FORCE VECTOR FOR 8, 1.0 Gen by ue es Kay’ oO °° ° Kaa" ° © 0 ka’ Kis Ks Kay's yy — RES Kes . Koa Kaz Kao! * Kaa ay Kaa’ = Koa [B) FREE-HEAD STIFFNESS AND ITS RELATIONSHIP TO THE FULL STIFFNESS Figure 32. Form of a pile—head stiffness matrix.The rotational springs can be assumed to be zero. The magnitude of the lateral spring stiffness can be directly obtained from a load-deflection solu- tion of a free-head pile. It can also be related to the stiffness coef- ficients of the full pile-head stiffness matrix as shown in figure 32. ‘The method to incorporate the above single pile stiffness representation to obtain the pile group stiffness and the procedure to evaluate the pile-head forces and moment of an individual pile in a pile group are given in section 6-9 and 6.10 respectively. 6-9 Development of Pile Croup Stiffnesses Piles are the most commonly used foundation system for highway bridges. Most of the foundation consists of a square pile group (from 2 by 2 up to 6 by 6 group) under a rectangular footing which also serves as a pile cap. Both ver- tical pile groups or battered pile group systems are commonly used as supports for the piers and abutments. The stiffness characteristics of a pile group depends on the following: o the stiffness characteristics of the individual piles. © the configuration (number of piles, configurations of the loading position, pile head coordinates, and batter angles). Development of the stiffness matrix of a pile group involves evaluating and integrating the reaction forces at each pile for a prescribed unit dis~ placement for each degree of freedom at the point of loading (structure sup- port point) at the pile cap. The steps involved in the solution of the full stiffness matrix of the pile group using a Linear representation of the stiff- ness atrix of a single pile {s schematically illustrated in figure 33 and briefly outlined below: 1, Calculate the displacements at each pile head corresponding to a unit displacement at each degree of freedom at the point of loading by using a rigid pile cap assumption. 2. Solve for the reaction forces at each pile head using the pile-head displacement in step (1) and the pile-head stiffness matrix. The stiffness matrix should correspond to the pile-head connection (see figure 32). 83RELATE 5; AND 8; TO UNIT dg AND Og, STEP 2 CALCULATE PILE HEAD FORCES: Fi = Geltad see Figure 32 FoR [Kk] STEP 3. CALCULATE STATICALLY EQUIVALENT PILE HEAD FORCES ACTING AT LOADING POINT STEP 4 Far SUMMATION OF ALL PILE HEAD FORCES: For Ey Fiz Moye EM, a8 NOTE: 5)" TRANSLATION OF i" PILE HEAD. 9, = ROTATION OF it PILE HEAD. 6g ~ TRANSLATION OF PILE GROUP 8g = ROTATION OF PILE GROUP Fj = FORCE AT ith PILE HEAD Mj=MOMENT AT ith PILE HEAD Fg = FORCE ON PILE GROUP LOADING POINT Mg = MOMENT ON PILE GROUP LOADING POINT n= NUMBER OF PILES IN GROUP Figure 33. Calculation of pile group stiffness matrix.3. For each set of pile-head forces from step 2, calculate the statically equivalent set of forces at the point of loading. 4+ Summation of the resultant forces in step 3 for all the piles will lead to a force vector corresponding to a specific column on the pile group stiffness matrix (see figure 32). Repeating the steps 1 through 4 for each degree of freedom of displacement of the pile cap will result in the full stiffness matrix. For a vertical pile group, solutions of the pile group stiffness matrix is relatively simple and can be carried out by hand solution methods. For a bat~ tered pile group, the calculation becomes very tedious in step 2 and a hand solution procedure becomes impractical and computer solutions are recommended. A Fortran program is presented in appendix E for such a purpose. It can be used to solve for the pile group stiffness as well as to distribute the total pile group load to each individual pile head. Some of the relevant comments regarding the pile group stiffness is provided below: 1. ¥or a rigid pile-head-cap connection, the full pile-head stiffness matrix in figure 32 including the off diagonal terms should rigorously be used in step 2 to solve for the pile group stiffness. However, neglecting the off-diagonal terms will usually yield reasonable solutions. The overall rotational stiffness of the pile group is usually dominated by the axial stiffuesses of the piles. Therefore enphasis should be placed on evaluating the axial and translational stiffness of an individual pile to develop the pile group stiffness. The above approximation (neglecting the off- diagonal and rotational terms) will usually lead to less than 10 percent error in the overall pile group atiffness in a smaller 2 by 2 group; and the error will be reduced for a bigger pile group. 2. For a pinned pile-head-cap connection, the full pile-head stiffness matrix should be modified in accordance to figure 32 before implementation of the pile head etiffness in step 2. The pile stiffness can rigorously be represented by a set of physical axial and translational springs. 3. The form of the stiffness matrix for a vertical pile group will be relatively simple and resembles the stiffness matrix of an individual pile (see figure 32). The offects of the off-diagonal terms will be less significant for a bigger pile group. 4. The form of stiffness matrix for a battered pile group will be very complicated and can potentially involve the full 21 85coefficients (the matrix will be symmetrical in any situation). The program provided in appendix E can readily provide the solution of the full pile group matrix. 5. The rocking rotational stiffness of the pile group (which is most affected by the axial pile stiffness) has been shown to have a significant effect on the overall bridge behavior. (50) 6-10 Individual Pile Head Forces and Moments In the previous sections emphasis has been placed on the procedures to develop the foundation stiffnesses for individual piles and pile groups for dynamic response analysis of the overall bridge-foundation eystem. Another facet of foundation design involves evaluation of the distribution of structural forces to the foundation system and the procedure to examine the integrity of the structural members of the foundation systea. In general, a structural Fesponse analysis would yield the solution of foundation forces at the base of each of the piers or columns. If a single pile is used for support of each column, the foundation forces from the above structural analysis can directly be used as the pile-head force and moment which can then be used along with the procedures outlined in section 6.5 and 6-7 to solve for the distribution of axial thrust, bending moment and shear along the length of the pile. The following comments are provided regarding the aspects of distribution of structural load on individual piles: © Distribution of structural load onto individual piles ta a pile group should be solved by reversing the steps described in section 6.10. The displacement of the pile group (at the loading point) can be used to solve for the displacements of each individual pile-head which can then be used along with the pile-head stiffness matrix to solve for the pile head loads. The above can easily be done by hand calcula- tion method for a vertical pile group. Tedious computations are involved for a battered pile group and a computer program (see appendix E) is enclosed to perform the necessary computation involved in distributing the total foundation forces to individual pile heads. © As described earlier, a reduced pile-head stiffness matrix should be used to solve for the axial and lateral forces for a pinned pile-head-cap connection. © As discussed earlier in section 6.9, the rotational stiff- ness of the pile group is insensitive to the individual pile 86head rotational stiffnesses due to the dominance of the axial pile stiffness. The overall overturning moment on the pile group will largely be resisted by the axial reaction of individual piles. Therefore, the individual pile-head monent will basically be unrelated to the overall over turning moment of the pile group and will largely be related to the shear introduced on the pile head and the connection details at the pile-head~cap. In view of the above, the full pile-head stiffness matrix (including the off-diagonal terms) must be used in the evaluation of the pile-head moment in step 2 for a rigid pile-head-cap connection. Solution of the pile head moment for a fixed-pile head boun- dary condition using the resultant shear forces on the pile head can also be used to obtain an approximate pile-head ‘moment « 6-11 Pile Group Effects ‘The discussions presented in sections 6.9 and 6.10 addresses the interaction between the pile and the pile cap. Another aspect of group effects relates to the interaction of the stress and strain field in the soil medium among the piles in the pile group. This phenomenon is commonly referred to as the “group effects” which could potentially alter the load transfer charac- teristics (p-y and t-z curves) for the pile. The aspect of group effect is highly complicated. Because of costs and physical size, there has been little experimental evidence upon which to base the pile group procedure. At the present stage of knowledge, the author feels that a general design procedure to account for the group effects would be unwarranted. A general comment is Provided below on the various aspects of group effects for both axial and lateral loading considerations. Discussions on the various sensitivities are provided wherever possible. Group Effects in Axial Loading. In designing a pile foundation consisting of closely spaced pile groups, the effects of spacing on the ultimate capacity and load-deflection characteristics need to be considered. Group effects on the capacity of a pile are conventionally depicted as a group efficiency fac~ tor, the ratio of the ultimate capacity of a group normalized by the number of piles to that of a single pile. The efficiency factor {s affected by (1) spacing and configuration of the group, (2) soll type, and (3) the pile installation procedure. The configuration will limit the load-transfer sur- face area along the shaft and at the base of the group. A number of 87researchers have suggested procedures to estimate the efficiency factor of the group based on the above concept (referred to as “the equivalent Caisson” concept) -(11,33) In general, the ultimate capacity will be underpredicted if it is based solely on the above reasoning as discussed below. As a result of pile insertion, especially by pile driving, the soil around the piles will be displaced and become more compacted or consolidated. The con solidation is more pronounced for a closely spaced group and will improve the group efficiency factor. ‘The increase is very pronounced in sand. An effi~ cdeney factor ae much af 2.0 has been measured.(55) In clay, the increase is complicated by the rate of consolidation around the group, the sensitivity of the clay and the amount of disturbance created by driving. In addition to the capacity aspects, load transferred into the soil mass by the adjacent piles in the group will introduce additional elastic settlements. Procedures to estimate such effects are suggested by Poulos, and 0'Neil1.(56,57) However, these procedures involve estimating additional displ: cements caused by neighboring piles using Mindlin’s solution of a point load in elastic continuum. The elastic continuum assumption cannot reflect the reinforcing effects on the piles themselves and usually results in overestima~ tion of group effects in settlement prediction. In view of the above uncertainties, a detailed procedure to account for group effects for axial loading cannot be developed. Further research 1s needed in this area. In the absence of further evidence, the single pile procedure described above could be used without adjustments for group effects for center to center pile spacings bigger than 3 pile diameters. Group Effects in Lateral Loading. The group effects will generally result in 2 softening of the lateral pile stiffness. Similar to the axial pile group probles, the group effects have been estimated fron theoretical considerations in an elastic continuum.(56,58) The above procedure does not account for the fact that the more rigid pfle member is present in the soll medium and would act as a reinforcing element for the sof] mass. Furthermore, the elastic con~ tinuum cannot account for the compaction of the soll mass in the group caused by pile driving and che nonlinear soil behavior. Therefore, the elastic solu~ tion usually overpredicts the group effects. 88‘The compaction effects in strengthening the soil medium tend to be rather significant for sandy soils for closer pile spacings. Therefore, in the absence of a detailed study, no adjustment is recommended for the group effects in sand. In the case of clay soils, especially for soft clay, a significant reduction in the pile-head stiffness could potentially occur. For a big pile group (e.g-, 6 by 6) at close spacing (less than 3 pile diameters center-to-center spacing), the lateral pile head stiffness could be reduced by as auch as fifty percent. 6-12 General Guidelines and Commentary The general procedures for design and analysis of piles and pile groups have been presented. ‘he procedure is only intended to be a general guide and not to replace judgment on a case-by-case basis by bridge engineers. In addition to the general procedures described above, some specific design guidelines concerning various pile design aspects are given below: 1. Linear Versus Nonlinear Procedure. Both linear and nonli~ neat pile design procedures have been recommended. The linear procedure can be used as an initial trial. The resultant solution should be examined to ensure that the assumed equivalent linear stiffness is appropriate. Our sensitivity studies indicate that the linear pile solution would yield reasonable pile solutions up to about 0.5 inch (1.3 em) of pile head deflection. At higher deflection levels, a nonlinear or an iterative approach may be necessary. Due to the dominance of the structural compliance in an overall bridge-foundation model, the linear procedure in general would lead to an adequate solu~ tion of the dynamic load levels. However, a more detailed procedure, including considerations of (1) the anticipated lateral force level, (2) the structural connection details, and (3) the nonlinear soil behavior may be warranted to evaluate the magnitude and distribution of moment along the pile to evaluate the integrity of the pile itself. 2, Moment Distribution. As discussed earlier, an examination of the integrity of the pile requires determination of the magnitude and the distribution of the moment diagram along the length of the pile. The above depends largely on the anticipated shear force, the pile head rotational constraint and the soil resistance characteristics (p-y curves). In general, a fixed-pile~head assumption leads to 89ae a maximum bending moment at the pile head, vhereas, a free head pile assumption leads to a maximum moment at depth. More realistic solution can be made by specifying some intermediate rotational constraint. The approximate magni- tude of the rotational constraint can be estimated by exa~ mining the pile-head-cap details. In the course of the earthquake, the ratio- of the applied shear versus pile moment may vary, leading to variations in the distribution of moments with time. Bridge engineers should be cognizant of the above factor in assessing the integrity of the pile member by way of monent diagrams from pseudo static solu- tions. Sensitivity of Soil Characteristics. A sensitivity study ‘of the overall pile behavior has been conducted to examine various facets of soil-pile interaction behavior (appendix D). The sensitivity study encompasses the following effects: (1) the p-y curve shape, (2) the potential gapping effects at the pile head, (3) the shear strength of soil, and (4) liquefaction effects. Based on the sen~ sitivity study it can be concluded that the overall pile behavior is insensitive to minor variations of the p~y cur- ves (e.g-, initial stiffness and curve shape). Primary emphasis should be placed on assessing the overall soil strength and the potential of liquefaction of the site soil in soil-pile interaction analysis under lateral loading. Liquefaction. Pile foundations can be used in areas where Tiduefaction of the shallower soll may be a potential problem. The phenomenon associated with the liquefaction behavior is that the foundation stiffness may soften as time progresses or as the number of cycles of repeated shearing increases. Ideally, a time-history analysis simi-~ lar to the one by Matlock, et al., should be conducted to account for the degradation effects of soil with time.(45) However, for design purposes the above approach would be impractical. To design for the liquefaction phe~ nomena, a conservative approach to bound a range of foun- dation stiffnesees may be warranted. The lower bound pile stiffness can be developed by assuming that the soil sup- port in the liquefied zone has been lost. The dynamic response of the overall bridge-foundation system should be conducted using both the lower and the upper bound foun~ dation stiffnesses to obtain the more conservative load level. After the seiemic load for the pile foundation has been obtained, distribution of the structural load to the soil-pile system can be conducted by using the liquefied soil support representation which will lead to a more con~ servative solution for pile moment. Rotational Stiffness of a Pile Group. Based on our sen- sitivity study and the Findings by Douglas, et al., the 906 7 rotational stiffness of a pile group is the most signifi- cant foundation parameter in affecting the overall response of the bridge foundation system.(4) The rotational stiff— ness of the pile group is most sensitive to the axial stiffness and relatively insensitive to the rotational stiffness of the individual pile. Batter Effect in a Pile Group. The analysis procedure to Solve for the stiffness-aid to perform distribution of the foundation load of a pile group is rather tedious. A com- puter program has been included to design a battered or a vertical pile group (appendix E). Battering the piles in a pile group generally leads to a bigger lateral stiffness for the pile group. The stiffening effect is very dramatic in the case of a short end bearing pile (versus a friction pile). Pile Group Settlement. The nonlinear load-deformation Behavior of a pile group subjected to the conbined dead weight and aoaent loading on the pile group hao been solved by Lam and Martin.(47) ‘The major conclusion 1s that, under the combined deadweight and the moment loading on the pile group, a permanent settlenent may occur at high monents. The vertical settlement of the pile group is assoctated with the plunging faluce of the compression pile, where the combined axial loading from deadweight and moment mobi- Lizes the full pile capacity at the ost conpressed pile. Hence, noraal selenite design for a pile group should allow for axtal loading from seisaically induced moment (that 13 axial capacity not to be exceeded by dead plus selemte load) unless permanent displacement of the pile group can be tolerated. Uplift of a Pile in a Group. Under normal condition, the axial load from the effect of the deadweight would be bigger than the magnitude of the cyclic axial load level. In general, the pile group would remain stable even if some of the pile is under tension until the frictional capacity is fully mobilized. In general, it is more critical to prevent plunging failure of the most compressed pile to prevent settlement of the pile group. It is very rare that a pile group fails due to uplifting of some of the piles. Pile-Cap Stiffness. A sensitivity study to evaluate the Sec eet group-pile cap stiffness arising from the pile versus the footing (pile cap) has been conducted (see appendix D). On the basis of the sensitivity study, it can be concluded that the contri- bution from the pile cap ts relatively small as compared to the piles. The stiffness of the footing would further be reduced if the soil is compacted, creating an air gap at the base of the pile cap. aL10. a. Soil-Pile Interaction. The use of pile stiffness charac~ Eeristics to determine earthquake-induced pile bending moments based on a pseudostatic approach assumes that monents are induced only by lateral loads arising from inertial effects on the bridge structure. However, it must be remembered that the inertial loads are generated by Anteraction of the freefield earthquake ground motion with the piles, and that the free-field displacements then selves can influence bending moments. Near the pile heads, bending moments will be dominated by the lateral interac- tion loads generated by inertial effects on the bridge structure. At greater depth (e.g., greater than 10D) vhere soil stiffness progressively increases with respect to pile stiffness, the pile will be constrained to deform in a similar manner to that of the freefield, and pile bending moments become a function of the curvatures induced by freefield displacements. In general, the bending moment caused by the curvature in the freefield will not pose any danger to the pile. Margason suggests that freefield cur- vatures of up to 6 x 10-4 inl (2 x 1073 m1) could be induced by strong earthquakes. ‘The bending stress corresponding to the above curvature would be small and should pose no problem to the typical steel or prestressed concrete piles. In some cases where liquefaction occurs at some of the soil layers, or where ground movement may occur due to Instability of a slope, ete-, the bending stress induced on the pile should be examined. Evaluation of the bending stress on a pile has been conducted by specifying offsets (movements) of the p-y curve along the length of the pile.(59) Recomended Pile Types. The uncertainties of ground and bridge response characteristics lead to the desirability of providing tolerant pile foundation systems. Toughness under induced curvature and shears is required, and hence piles such as steel H-sections and concrete filled steel- cased piles are favored for highly seismic areas. Lightly reinforced concrete piles are brittle in nature, so ade— quate longitudinal reinforcing should be specified to reduce this hazard. The reinforcing steel should be extended into the footing to tie elements together and to assist in load transfer from the pile to the pile cap. A well designed (especially regarding detailing at the pile-cap connection) pile foundation generally performs very well against earthquake loading. Experience has shown that reinforced concrete piles tend to hinge or shatter immediately below the pile cap. Hence tie spacing is reduced in this area so that the concrete is better con fined. Driven precast piles should be constructed with considerable spiral confining steel to ensure good shear strength and tolerance of yield curvatures should these be 92imparted by the soil or structural response. Clearly, it is desirable to ensure that piles do not fail below ground level, and that flexural yielding in the columns is forced to occur above ground level. The additional pile design requirements imposed on piles for bridges classified as SPC © and D for which earthquake loading ie more severe, reflect a design philosophy aimed at minimizing below ground damage which is not easily inspected following a tajor earthquake. 93orot ors oro we o00%0 z00°0 Ty0%0 EIT 00270 goz"0 e920 Onto 650° =940"0 ~Lgt*0 ~192°0 ~h9EO “£050 -zEn0 6940 -156°0 9640 -10%*0 ~ENETO =662"0 ~Sytro 900° 000%0 6z0°0 120° ~LT0°0 -560°0 mETZIO ~0ser0 -95H40 -9Ln0 -LL4t0 95990 -1TH"0 ~0SE°0 ~21e*0 ~012°0 -922"0 -18T 0 LEO ~560°0 ~850°0 -820°0 -100°0 00% "a (uz 21n6y 995) 2! ye°3 ‘pues 104 uorIos aid onsej@ seoUrT Z AGEL o00°0 -920°0 ~6£0°0 -2n0°0 =9T0°0 65070 0020 20480 B6"0 76510 2890 SLL°0 75870 searo 41610 6670 0960 916° 186°0 76670 66670 oo0"T ooort "8 000°0 TTO*0 azor0 6n0°0 s90°0 £g0°0 -900°0 ~sstt0 -$h20 -9SEr0 -z97°0 -929°0 -T6L°O -BL8"0 -896"0 =190°T -9ST'T -esz' -1Se°T -0Sh'T -6nstt ~6H9'T ~6HLtT 8 0000 000°0 ~600°0 -8z0°0 -L50°O ~680°O ~s0T"0 -010"0 -0€0°0 620°O ztt70 €zz°0 9E°0 Bnet onsto 1940 zsL°0 €Lgto e00't enttt e62°T esatl ezott o00%0 9n0ro LIT*0 102*0 Lsz*0 92270 600 ~£8740 -su9"0 -609°0 ~19L°0 -588"0 ~296°0 -LL6*O ~£L6°0 -L¥6"0 -L68"0 -7zet0 ~81L 40 -985°0 -22b90 -L2z70 o00*0 000°0 €1uro -L2070 -901'0 -Ezz"0 -6nE0 ~H24ro -1LEt0 = 66270 ~E6TO -9g0'0 60tto 56770 z6E°0 68rd sagt 11990 E9Lt0 onero go6"0 966°0 88670 00" ooo -ceurd -2E0'0 000° 1800 szz*o zenro 9z9t0 96970 9mL20 2410 L9Lto Leto £690 6H9"0 565° Testo 6shr0 6LErO 16Z°0 B61 oot 000°o v o00ro -600°O -920"0 -050°0 -7L0°0 =sL0"0 -020°0 Tetto beta Teet0 nyse geL°0 z96°0 9gort 9iz°t eset sett enor gent zs6"t zt? €Lztz0000 -L00°0 o€0r0 160° wet o9z+0 €12°0 €LT?0 160°0 200°0 ~901°0 ~612*0 -S2E°0 -1LEt0 ~0Tnt0 ~6EHt0 -95"°0 ~86nt0 ~Z4H"0 -904*0 ~LbEt0 -192"0 -LbT+o ‘a “tg2 aumbyy 208) 179 = 14/04 asoyin 2b + O9 <3 ‘Aelo 405 vonnjos onseya se0UrT “€ oiaeL 000°0 6z0°0 7z0°0 -L00°O -LL0°0 ~68T*0 -926°0 ~E990 -OL4*0 -08%°0 -0L°0 -Leyto ~E8e°0 ~BHEtO -606°0 -992°0 -227°0 ~9LT40 -TET*O -680°0 -150°0 ~120°0 000%0 ‘8 000%0 ~€20°0 ~LEoro -€40°0 -920°0 T9080 69T"0 49Er0 sso Tss'0 9490 BEL*O ozer0 Lg8t0 0690 61670 €n6*0 £9640 8L6°0 68670 966°0 666°0 o00*T ‘8 000° 600° 4200 G90 €90°0 190° zr0°0 ~6tTt0 -102"0 -206*O “1290 -095°0 ~9TLt0 -008"0 -188"0 -L16°0 -0L0°T -99T'T ~e97'T -I9ET -09%'T -095°T -099°T ‘a 000%0 To0*0 -L00°0 =720°0 -150°0 -980°O -s0T*o -zg0°U -150°O -100°0 120° goto 96270 TLet0 9540 695° 1s9'0 £9L°0 788 °0 stort 9STIT LOE*T Bonet ‘a 000°0 9€0°0 6600 LLTO Sez*0 6Izt0 1L0°0 -92270 -LLE*O -vEG"0 -589°0 ~$TB*0 -606°0 ~9€6"0 -L46°0 -046°0 ~216*0 -198°0 ~98L°0 -989°0 -LSs*0 -10%*0 ~$tZ*0 v 000°0 Sto*o -8t0r0 ~190°0 -z61t0 -OTE"0 ~6BE"0 -sge°0 -562"0 -702"0 -z80'o 890°O THz"0 eecro azbro zzg°0 sT9*0 €0L°0 98L°0 6580 126°0 69690 ooort v oooro -0£0°O -060"O -900"0 €90%0 arto 19670 ogsto 9290 9190 9010 BOLO LL9*O 6990 119"0 £960 9050 Onno 99690 eazto 610 660°O 000° 00080 120°0 Leora Laon seoro -920°0 -E9T"O -s6er0 ~9TS*0 =r99"0 -€8L°0 -s26°0 -€90'T -OET'T ~e6TtT -1S2tt ~S0EtT -7geT -E6E'T -SeH'T ~654°T ~E9H"T -894"T 0000 -100°O -220°0 -€40°0 -390%0 ~TL0°6 -4z0°0 gorrs Berto aTEt0 9Gh"0 Lz9°0 9z8°0 9€6r0 egort LUT z0e't sent €LgtT HTL asert 00°2 Tse000% -110"0 zz20°0 5800 TLTtO e270 oazeo o0z*0 Tet"o z0°0 -79090 ~8LT'0 -262°0 one -16E70 ~0E%0 -854"0 -9L9"0 ~SL400 -8590 -0z4°0 -856°0 =022"0 ‘8 oooro gz0ro 9200 100"0 -£90°0 =OLT*O ~906°0 -Tento -S9n80 ~28H°0 -08%°0 -954°0 -604°0 -LLEtO -0nE°0 6620 -562"0 -807"0 -19T*0 =n1Tt0 -010°0 -TE0°0 0000 8 2:0 = ayy asoyn zh + 04 «83 ‘heya 40j uonnios od anseje seoury “y o1ae 200°0 -120°0 ~se0r =nH0°0 -0€0°0 9z0°0 srt recto Tzt0 915°0 E1920 LoL*o 6L°0 eee 69870 206°O 6z6*0 z6"0 TL6*0 $960 76690 666° ooort 0000 400°O 1z0°0 200 1900 790°0 "20°0 -260°O -L9T*0 -192°0 =HLErO -905°O -959°0 -LEL*O -ez80 -116"0 =€00°T -L60°T -e6ttT -162"T -06ETT ~689°T ~686*T o0uro zooro -s00°0 -0z0°0 -9%0°0 ~6L0°O -H0T 0 -160°0 -$90"0 -£20°0 040°0 Lztro enzo ETEt0 T6E"0 LLyo ELs*0 e19t0 26170 91670 Osort nett Brett ‘a ooo 6z0°0 seuro astro L240 €1Z0 9800 -z8T'0 ~EZE°0 =oLb0 ~929°0 -09L°0 -998"0 -€06°0 -926°0 -€€6°0 -£26°0 ~€68"0 -0n8"0 -S9L"0 -599°0 ~6ES*0 -996"0 v oovro st0r0 -210"0 -EL0°0 -a9T"0 -6LZ"0 -096°0 ~EnEt0 -262"0 E120 -E0T*O 960°0 66190 1827°0 6LEto zLyr0 5950 5590 2yLro z2zer0 7680 nG6°O o00't v 00070 -120°0 -620"U -600°0 64080 T9T*O zero 9050 o1sto 12970 9g9t0 1990 gegto nT9t0 19st BESO ganeo GzHt0 sser0 LLz"0 Té6tro 26070 000° oGoro 8t0ro €€0°0 h080 9€0°0 =s10°0 ~setto -Enero -05%°6 -015"0 ~869°0 ~6728°0 -096%0 -220'T -290'T -BET'T ~6ettt ~Seztl =nletl =s0Erl ~62E°T -enert -BHE'T ty ooore -500°0 -810%U ~860°0 -650°0 -990°C ~££0°0 e90ro 290s £9270 o6e70 enseo 22L°9 1289 9z6°0 LEOtT asttt SL2"l TOHt ocstt 299°T 96Ltt oerore ert OT oT zt ort 690 a0 400 940 S70 400 £10 210 Tr0 oro 000° -910°0 600°0 490°O Lateo aEz*0 s6z*0 65770 L0z*0 TET? ££0°0 -¥20°0 -21Z°0 -81270 -2HE0 -€0%°0 ~659°0 -805°O -145*0 ~nLg*0 -585*0 ~615*0 -295°0 ‘8 000°O 920° 620°O zT0°0 0400 =9721°0 ~652°0 -€0%*0 ~695°0 ~€89"0 -005"0 ~565*0 -994°0 -Ty90 -0Tb*0 ~€LE%0 -0¢e"0 -198270 ~622°0 ~ELT"O -STtt0 -150°0 000°0 g'0 = 13/0 ex0um zby + Oy = £3 *Aejo 10} uojAnjos aiid onseja Je0UN “s aGeL 000°0 -LTO*O -TE0"0 ~2n0°0 -LEO"O 000% 760°O o9zt0 94E*0 sent BESO Beso SELtO TEL? ezaro 2980 868°0 6z6°0 956° mL6°0 686° 16690 ooort 0000 700% 910%0. Seo ssor0 990°0 9H0*0 -070%0 -00T°0 ~6LTI0 -LL2*0 ~h6ErO ~zEsto -8090 -289O ~2LLr0 -098"0 -156°0 -940°T -2eEet -092°T ~6EET ~6En"L oo0%0 €00*0 -z00°0 ="T080 -LE0°0 -89080 -960°0 -E0Tt0 -060%0 -290°0 -LTO%0 6r0°0 THrt0 g6T°0 9290 9€6°0 Lust B05°O g09%0 LTL90 geno 5960 nOrrt o00%0 storo 190° s2teo Larco e02°0 n2t°o -880°0 -0TZ*0 -BrE°0 ~S64°0 -0n9e0 -ELLt0 -1e8"0 -098+0 -BT6*0 -796°0 -95690 26690 ~0€6°0 -698°0 -8z8*0 =S9L"0 7000°0 -220°0 -600°0 1e0%0 sOoTso g0z°0 $620 z1Er0 eezt0 92z°0 ToT 9z0°0 -S1Tt0 -£61'0 -942°0 -296°0 -6990 ~Se5"0 ~619°0 8690 ~69L"0 -0¢8"0 -618°0 a) ooo 2z0°0 o€0r0 9200 -900°0 -£80°0 -01z°0 -996°0 -9%40 Lind HHTSt0 -tes*o -£25°0 -L0S*0 -784"0 -z590 -Ttyto ~296°0 -S06°0 ~6£2°0 -s9T 0 -980°0 000°0 '@ ort = absyOy ooum 2 4 Oy = £3 “Ae]9 104 voHIN|as ajc onseIe eau “9 eIgeL 000°0 -ot0r ~Ez0°0 =8E0°0 -sn0°0 -sz0°0 2H0°0 O61 69220 o9E°0 0940 69970 11980 €2L°0 2LL°0 61970 e980 2060 Ge6r0 £9690 £8670 96670 ooo*t 0000 000° 600°0 720° sh0°0 790° 190°O g00°0 -040"0 -€0T"O -sat'o -982°0 -T1H70 -T8h70 -955°0 =5E9°0 ~6TL"0 =L08°0 - 6680 96690 -160°T -06T*T -062"1 8 o00°0 n00°O z00°0 -900°0 -€20°0 -1g00 -990°O -v010 -10TtO -180°O -650°O -z10°0 £50°O z201°0 estt0 £1z"0 19z°0 usero 2nnrd LES*O 1H9t0 Sgro 61a" 0000 stot 2000 -020°0 -SL0°0 -95T0 0420 -18z"0 -OLz"0 -eez"0 ~BL tt -060°0 oor zorto Tato L9z°0 oso asht0 295°0 69970 6LLIO0 o6eto ooo"t Nw o00%0 400°O 910°0 9z0°0 TE0*o L10°0 -9£0°0 ~6nT "0 -e1z"0 -982°0 -2LE"0 -7940 -6550 -909°O -eg9rd ~169°0 ~6ELt0 -LLLtO -11e'0 ~6E8"0 -198°0 -nL9"O -619°0oot ors sth Ory ste ore Gz orz ert ot at zt ort 670 aro 160 970 gto 460 €*0 o00%0 ~220°0 -120°0 100+0 0900 o9t*0 zez°0 9960 69670 ZHEr0 oez*0 9LTt0 920°0 -190°0 -TLT*O -882°0 ~yt9t0 -159"0 -169°0 -678"0 -900°T -99UT =L2e*t z= aby aroymn 2h 4 Oy = 83 ‘Avj9 40y uornyos ajid onseye seoUry “y aIgeL 000°0 -900*0 -910°0 62040 -€40°0 -TH0"0 100°o attto esto 49290 6960 nly 0650 69980 LOL? €9L70 LT8*O 9980 O16t0 L76°0 sL6°0 £66°0 oo0rt 0000 €00°0 €00"0 oo0+0 -1L0rO -2€0°O -€90°O -z60°0 -L60°O 560° -za0°0 -350°0 -600°0 €z0°0 190°0 LOTr0 6st*0 ozzr0 06270 69670 isnt gsg*0 €99°0 “a oooro -L00"0 600°0 €v0r0 $60°0 Lateo T9T*0 81070 400° =160°0 -L2270 BE -995*O -199°0 -19L00 -€98°0 -596"0 =990°T -E9TtT -962"T -2herl ett -eabeT oooro -800°O =910°0 -BT0°0 -1t0°o oz0r0 180° e610 zn2°0 T6z"0 Leero €Le*o n6E°0 96e°0 26°70 ogeto 6560 zero Bazto gezt0 TL1t0 £60°O 0000 oooro z00°0 gooro 910° "200 zz0°0 -£00°C ~zL0"0 -STTt0 ~691°0 -1ez"0 -20€°O ~6LEnO ~6Tn -859°0 -L65°0 =nEsto -B95°O -665°0 -629°G -949°0 -659°0 -£99°O ooore Toor -100°0 -L00°0 ~LTO*G -620°0 -960°6 ~020°0 -t00"0 L200 490°0 aztro gato azz°0 zL2°0 zero 1LE*0 9znt0 santo 9nst0 o1sto SL9°0 enLto000°0 -E10°0 -£20°0 -920°0 -200°0 4L0°0 aizro Too 2970 00570 6690 OnGr0 £060 g6rr0 £90°0 -760°0 -L92°0 =yTst0 =SLLto -ELOST =90%"T HSLLtT HLLUtZ 000°0 700°O €10°0 920° 4€0°0 6t0ro -150°0 -902°0 -€62°0 ~686°0 -064%*0 -785°O -099°O ~589°0 86970 -169°0 -LL9°0 ~Legto -EL5°0 -Taet0 -LSE*O -86T°0 0000 079 = abry0y o1aun zby + Oy «53 *Aej9 40} von njos aud onseje veaUTT “B aIaeL 000°0 100° -200°0 -210°0 -820°0 -€40°0 -8€0°O zzor0 2L0°0 Ovtro azz 9Ee°0 1997°0 azsr0 B65°0 aggro LeLeo e080 199° Bt6r0 09690 686°0 ooo't 000°0 100°O z00°0 €00°O 0000 -600°0 -620°0 -150°0 -990°O -910"0 =910°O -1L0°0 -050°0 -7€0°0 -21080 LT0*0 1s0°0 €60°0 nett z0z*0 O1z70 arero sent a 000°0 -800°0 -L00%O s0uro Leora 160°0 ert 9g1t0 ozo 1900 -140"0 -89T°O -L86%0 -905°0 ~869°0 -€aLt0 -0%6°0 -LOtrt ~eeztT ~S9beT -1S98T -LEat 60°72 000°0 s00°O 6000 oror 100" -0€0"0 -160°0 “LTO ~661°0 ~atzt0 -1zz"0 -861°0 “2et0 -160°0 -0n0"0 Te0°0 ettto ozzt0 OnEro LLatO €£9°0 Logo oo0'T oo0ro -To0"0 ~s00°0 -o10ro ="T0"G -390°O Tz0°6 980°O H2tto g9tto orzo es2*0 gezto ooera Loero g0c"0 10E"O Haz Lszto g1zt0 1910 060°0 0000 "y oo0ro -100°0 To0°o s00°0 Tt0"o Ltot0 store -1t0%o -2€0°0 ~090°O -860°O ~beT tO ~6610 -822°0 -862°0 -692°0 ~6TE*0 ~6HEO -916°0 -666°0 -8T4t0 -tento ~sE""0 oooro toore Tooro -100"0 -700°C -U1080 -020°C -220°0 -810°6 -600°0 900°0 ocoro, n90°6 9g0r0 ote, Leto Bolt loz" Bez LLZ70 atero o9era 7090 wonou er eao ue 7007.0 ANALYSIS OF DRILLED SHAFTS 7-1 General There are many common aspects in the behavior of drilled shafte and piles. Both Foundation types are used to transfer moment, shear, and axial load fron the structure to the soil medium at sone depth. Analysis of load-deformation behavior of a drilled shaft can essentially be treated similarly to that of a pile. However, the input parameters must reflect the differences in (1) member stiffnesses, (2) installation methods, and (3) dimensions and con- figurations. In general, the analysis procedure for piles described in section 6 could be used for drilled shafts. Discussions on the necessary modifications to account for the above mentioned differences are provided below. Differences in the load-transfer characteristics for axial and lateral loading considerations arise from the following: (1) the diameter of the drilled shaft (up to 10 £t (3 m)) Le usually much bigger than the typical piles (usually less than 2 ft (0.6 m)) and (2) the aspect ratio (Length to diameter) is usually smaller for drilled shafts. As a result of the Large diameter and smaller length to diameter ratio for drilled shafts, there is some evidence that the use of p-y curves alone may not be adequate to characterize soil reactions on a drilled shaft for lateral loading. As shown in figure 34, the following forns of soll resistance could be identified on a pile or a drilled shaft: 1. The lateral translational resistance which is simulated by the p-y support curves. 2. The axial frictional resistance along the side which is simulated by the tz support curves. 3. The axial translational end-bearing resistance at the tip of the shaft. 4. The torsional rotation about the axis of the member. 101~~ 7 NONLINEAR LATERAL SORT olz 3|3 2/2 3 8 la By ala Sle 6 ROTATION SUPPORT FOR 4, SHEAR REACTION MOMENT (tt) i y | % a2 x tle se % 32 Bel a x als ae g a a2 gel =e a 312 cui 2 z\2 gel a ale Be | tt} 4 Lg s IDEALIZED AXIAL ++} TIP REACTION BUE TOFIER ROTATION ay PROTOTYPE (8) BEAM—COLUMN MODEL Figure 34, Problem description of a drilled shaft. 702The rotational (tilting) resistance along the length of the shaft caused by the differential in frictional resistance on the shaft wall induced by the rotation of the shaft. 6. The rotational (tilting) resistance at the tip of the shaft. 7. The lateral translational resistance from the shearing at the tip of the shaft. In pile design, the first three modes of soll support are simulated by prescribing p-y and t-z curves along the length and an end-bearing axial Q-2 curve at the tip of the pile. The remaining components of soil resistance are small and can be neglected. For drilled shafts, there has been experimental evidence that some of the above modes of soil support (items 5 through 7) could not be ignored. Comments regarding the load-deformation behavior of a drilled shaft under lateral (shear and moment) and axial loading conditions are provided below. 7.2 Lateral Load-Deformation Behavior Due to a smaller length to diameter ratio of a typical drilled shaft, the magnitude of rotation of the shaft can be more significant than a pile. When this is compounded by the larger diameter (which acts as a aoment arm) of the drilled shaft, the rotational resistance along the length and at the tip of the shaft could be very significant as compared to the lateral passive resistance. A backfitting analysis was conducted to compare analytical solutions with available drilled shaft load test data-(60) The diameter of the drilled shaft used in the load test is 5 ft (1.5 m). In the first series, the site con- dition consists of predominantly clay (cohesive) soil (see figure 35). In the second series, the site condition is predominantly sands, gravels and silts (cohesionless) (see figure 36). In each series, the various components of soil supports for lateral loading are modelled, including: 1. The translational p-y support. 2. The rotational support along the length of the shaft. 1033) agwivuann (44) 19 13m Jann 104 aNI-¥e01 x Se°ae 813“aIyjoud pues e 40} 1803 japow aoeysnsqns pazweop| “9 aunb1y ZINN 6'Le = 3914 (5) ZINN 0069 = #4 (H) tN C510 = 3994 (e) AN £8 = UPN L (Zh WeO= ¥I LH) 'SBLON $l Te 460 _ ae ze or aNyS aSN30 “03K $761 oe-o = sr ate ay anys 39N30 ot as 860 a se art ay as O01 es 860 _ ae ort ay ee Pe 26 85°0 — ae eo" veers eeu Sc os a r6-0 a ae a0" oir oo anys Altis son] ze se zs or — ze $90 ol VRAvUa ONY GNYS nan aovul L118 38001 rr asH) ‘e) cg (34) woly4s "9 ah lusitevra a No14d142530 (4) wo119n038 | HO1S3HO | yol 3th, as 4H9138 03Z 1783439 daa -— - monauss | ogwrvuann | rvaint | sminoon LIN a5 705 zor3. The rotational support at the tip. 4, The lateral shear resistance at the tip of the shaft. ‘There is no well established procedure to estimate the monent-rotation soil support curves in item 2 and 3 and the load-deflection curve in item 4. Therefore, they are developed using subjective engineering judgments. In the case of the rotational supports in item 2, the moment-rotation curves are related to the skin friction (t-z) curve on the pile shaft. The beam-column model used in the backfitting analyses is presented in figure 34. The com parisons of analyses with test data are presented in figures 37 and 38 for the clay and sand site, respectively. The general findings are summarized below: 1. Tn general, modeling soil supports by p-y curves alone (see Section 6 for procedure) would be conservative and would lead to an over-estimate of the lateral and rotational displacements of the shaft (see figures 37 and 38). 2. In both cases, the rotational resistance at the tip of a straight shaft could be ignored. For a bell shaped shaft at the pile tip, the above conclusion may not be valid. For a clay eite (see figure 37), the rotational support along the length of the shaft provided a significant amount of rotational soil resistance to resist moment loading. 4. For a sandy site (cohesionless soil), the rotational soil resistance along the length and at the tip of the shaft is negligible. However, the lateral shear resistance at the tip of the shaft which primarily depends on the axial load on the shaft may need to be accounted for. ‘The rotational resistance along the length of the shaft appears to be very significant in affecting the load~ deformation behavior of the drilled shaft for the clay site, vhereas the significance of the rotational soil resistance greatly diminishes for the sandy sites. The rotational resistance along the length of the shaft is related to the skin friction characteristics (t-z curves). Therefore, it is of interest to compare the unit skin friction capacity with the ultimate resistance of the p-y curve (py). The ratio of £D/py 1s plotted in figure 39 for both the clay _ site and the sand site. It could be noted that the ratio of skin friction to ultimate p-y resistance (fD/py) is auch 106