Database assessment of CPT-based design methods

DatabaseAssessmentofCPT-BasedDesignMethodsforAxialCapacityofDrivenPilesinSiliceousSandsJamesA.Schneider1;XiangtaoXu2;andBarryM.Lehane3Abstract:Numerousconepenetrationtest
CPT-basedmethodsexistforcalculationoftheaxialpilecapacityinsands,butnoclearguidanceispresentlyavailabletoassistdesignersintheselectionofthemostappropriatemethod.Toassistinthisregard,thispaperexaminesthepredictiveperformanceofarangeofpiledesignmethodsagainstanewlycompileddatabaseofstaticloadtestsondrivenpilesinsiliceoussandswithadjacentCPTprofiles.Sevendrivenpiledesignmethodsareconsidered,includingtheconventionalAmericanPetroleumInstitute
APIapproach,simplifiedCPTalphamethods,andfournewCPT-basedmethods,whicharenowpresentedinthecommentaryofthe22ndeditionoftheAPIrecommendations.Meanandstandarddeviationdatabasestatisticsforthedesignmethodsarepresentedfortheentire77piledatabase,aswellasforsmallersubsetdatabasesseparatedbypilematerial
steelandconcrete,endcondition
openversusclosed,anddirectionofloading
tensionversuscompression.Certainmethodsareseentoexhibitbiastowardlength,relativedensity,conetipresistance,andpileendcondition.Othermethodsdonotexhibitanyapparentbias
eventhoughtheirformulationsdiffersignificantlyduetothelimitedsizeofthedatabasesubsetsandthelargenumberoffactorsknowntoinfluencepilecapacityinsand.ThedatabasestatisticsforthebestperformingmethodsaresubstantiallybetterthanthosefortheAPIapproachandthesimplifiedalphamethods.Improvedpredictivereliabilitywillemergewithanextensionofthedatabaseandtheinclusionofadditionalimportantcontrollingfactorsaffectingcapacity.DOI:10.1061/
ASCE1090-0241
2008134:9
1227CEDatabasesubjectheadings:Drivenpiles;Conepenetrationtests;Foundationdesign;Compression;Sand;Pileloadtests;Databases.IntroductionThestaticaxialbearingcapacity
Qtofadeepfoundationisgivenasthesumoftheshaftcapacity
Qsandbasecapacity
Qb:Qt=Qs+Qb=Pztip−Lembztip
fdz+qbAb
1whereP=pileperimeter;
f=localultimateshaftfriction;qb=ultimateunitbaseresistance;Ab=pilebasearea;ztip=tipdepth;andLemb=embeddedpilelength.Thevalueofqb
whichiszeroforapileloadedintensionisnormallylimitedtothatmobilizedatapiletipsettlementof10%ofthepilediameter.Awiderangeofempiricalapproachesarecurrentlyusedtocalculate
fandqbfordrivenpilesincoarsegrainedsoils.Thisrangehasarisenbecauseoftheinadequacyofexistingtheoreticalmethodsforpredictionofthesoilresponseduringandafterpileinstallationcoupledwiththelargenumberofvariablesknowntoaffectpileaxialcapacitiesinsand
Randolph2003.Further,Den-nisandOlson
1983concludedthatwhilesoildensityinfluencesshaftfrictionandbasecapacityofopen-endedpiles,interpretationoftheinfluenceofdensityishighlyuncertainduetopoordefini-tionofsoilpropertiesalongthelengthofapile.Useoftheconepenetrationtest
CPTforsitecharacterizationprovidesalargeamountofrepeatableinformationontheverticalvariabilityofsoilstrengthandcompressibilityandthereforeleadstoincreasedde-signreliability
i.e.,BriaudandTucker1988.Thispaper,aswellasadditionalparametricstudiespresentedbyLehaneetal.
2005aandSchneider
2007seektoassistdesignersinmakinginformeddecisionsregardingaxialpilecapacitybasedonanun-derstandingofthebasisofagivenempiricalapproachandoftheassociatedrelativelevelofuncertaintyandbias.Althoughsitespecificstatic
anddynamicloadtestsreducethelevelofuncertaintyforonshorepractice,loadtestsarepro-hibitivelyexpensiveintheoffshoreenvironmentandrelianceisplacedonextrapolationofempiricalcorrelationsderivedforon-shoreapplications
McClellandetal.1969.Althoughthederiva-tionof
fandqbinEq.
1foroffshorepilesinsandhashistoricallybeenbasedon“visualclassificationmethods”placedwithinframeworksofbearingcapacitytheoryandanearthpres-sureapproach
McClellandetal.1969,fourCPT-basedmethodsarenowincludedinthecommentaryofthe22ndeditionoftheAmericanPetroleumInstitute
APIrecommendedpractice
RP2AforfixedoffshorestructuresAPI
2006.TheCPTisoftenthoughtofasamodelpileandhasalonghistoryofuseforestimationsofstaticaxialpilecapacity
Delft1936.Thereis,however,awidevariationofCPT-baseddesignmethodsinuse1Ph.D.Student,SchoolofCivilandResourceEngineering,TheUniv.ofWesternAustralia,Crawley,PerthWA6009,Australia
correspondingauthor.E-mail:[e-mail],SchoolofCivilandResourceEngineering,TheUniv.ofWesternAustralia,Crawley,PerthWA6009,Australia.3Professor,SchoolofCivilandResourceEngineering,TheUniv.ofWesternAustralia,Crawley,PerthWA6009,Australia.Note.DiscussionopenuntilFebruary1,2009.Separatediscussionsmustbesubmittedforindividualpapers.ThemanuscriptforthispaperwassubmittedforreviewandpossiblepublicationonJune15,2006;approvedonOctober29,2007.ThispaperispartoftheJournalofGeotechnicalandGeoenvironmentalEngineering,Vol.134,No.9,September1,2008.©ASCE,ISSN1090-0241/2008/9-1227–1244/$25.00.JOURNALOFGEOTECHNICALANDGEOENVIRONMENTALENGINEERING©ASCE/SEPTEMBER2008/1227worldwide
e.g.,DeCocketal.2003becausethefactorscontrol-lingaxialpilecapacityareinfluencedbymoresoilparametersthanthoseaffectingtheCPTtipresistance,qc,
e.g.,Lehaneetal.1993.Empiricaldesignmethodsrelyoncalibrationwithdatabasesofstaticload