外文翻译基于钢混凝土粘接组合梁可靠度优化设计的概率和非概率不确定性文档格式.docx

外文翻译基于钢混凝土粘接组合梁可靠度优化设计的概率和非概率不确定性文档格式.docx

《外文翻译基于钢混凝土粘接组合梁可靠度优化设计的概率和非概率不确定性文档格式.docx》由会员分享，可在线阅读，更多相关《外文翻译基于钢混凝土粘接组合梁可靠度优化设计的概率和非概率不确定性文档格式.docx（8页珍藏版）》请在冰豆网上搜索。

外文翻译基于钢混凝土粘接组合梁可靠度优化设计的概率和非概率不确定性@#@1英文原文@#@Reliability-baseddesignoptimizationofadhesivebondedsteel-Cconcretcompositebeamswithprobabilisticandnon-probabilisticuncertainties@#@ABSTRACT:

@#@Itismeaningfultoaccountforvariousuncertaintiesintheoptimizationdesignoftheadhesivebondedsteel-cconcretecompositebeam.Basedonthedefinitionofthemixedreliabilityindexforstructuralsafetyevaluationwithprobabilisticandnon-probabilisticuncertainties,thereliability-basedoptimizationincorporatingsuchmixedreliabilityconstraintsaremathematicallyformulatedasanestedproblem.The@#@performancemeasureapproachisemployedtoimprovetheconvergenceandthestabilityinsolvingtheinner-loop.Moreover,thedouble-loopoptimizationproblemistransformedintoaseriesofapproximatedeterministicproblemsbyincorporatingthesequentialapproximateprogrammingandtheiterationscheme,whichgreatlyreducestheburdensomecomputationworkloadsinseekingtheoptimaldesign.Thevalidityoftheproposedformulationaswellastheefficiencyofthepresentednumericaltechniquesisdemonstratedbyamathematicalexample.Finally,reliability-basedoptimizationdesignsofasinglespanadhesivebondedsteel-cconcretecompositebeamwithdifferentloadingcasesareachievedthrougintegratingthepresentsystematicmethod,thefiniteelementanalysisandtheoptimizationpackage.@#@1.Introduction@#@Thesteel-cconcretecompositebeam,whichintegratesthehightensilestrengthofsteelandthehighcompressivestrengthofconcrete,hasbeenwidelyusedinmulti-storeybuildingsandbridgesallovertheworld.Atthebeginningofthe1960s,anefficientadhesivebondingtechnique[1,2]wasintroducedtoconnecttheConcreteslabandthesteelgirderbyanadhesivejoint,notbytheconventionalmetallicshearconnectors.Thisso-calledadhesivebondedsteel-concretecompositebeamisconsideredtobeaveryprospectivealternatestructurebecauseithastheadvantagesofrelievingstressconcentration,avoidingsitewelding,andusingtheprefabricatedconcreteslab.Recently,anumberofstudiesontheexperimentaltestsandnumericalsimulationofadhesivebondedsteel-concretecompositebeamshavebeenpresentedinliteratures[3-5].@#@Withtheeverincreasingcomputationalpower,thepasttwodecadeshaveseenarapiddevelopmentofstructuraloptimizationinboththeoriesandengineeringapplications.Inparticular,thenon-deterministicoptimaldesignofsteelorconcretebeamsincorporatingstochasticuncertaintieshasbeenintensivelystudiedbyusingthereliability-baseddesignoptimization（RBDO）method[6,7].Basedontheclassicalprobabilitytheory,thisconventionalRBDOmethoddescribesuncertaintiesinstructuralsystemsasstochasticvariablesorrandomfieldswithcertainprobabilitydistributionandthusprovidesaneffectivetoolfordeterminingthebestdesignsolutionwhileexplicitlyconsideringtheunavoidableeffectsofparametervariations[8].Asthemostmaturenon-deterministicdesignapproach,theRBDOhasbeensuccessfullyusedinmanyreal-lifeengineeringapplications[9,10].However,theprimarychallengetoapplytheconventionalRBDOinpracticalapplicationsistheavailabilityoftheprecisestatisticalcharacteristics,whicharecrucialforasuccessfulprobabilisticreliabilityanalysisanddesign.Unfortunately,theseaccuratedatausuallycannotbeobtainedinsomepracticalapplicationswhereonlyalimitednumberofsamplesareavailable.@#@Theearlytreatment[11,12]forinsufficientuncertaintiesistoconstructionaclosestuniformprobabilisticdistributionbyusingtheprincipleofmaximumentropy.In1990s,Elishakoff[13,14]exploredthatasmallerrorinconstructingtheprobabilisticdensityfunctionforinputuncertaintiesmayleadtomisleadingassessmentoftheprobabilisticreliabilityinparticularcases.Thisconclusionilluminatesthatusingthetraditionalprobabilisticapproachtodealwiththoseproblemsinvolvingincompletetheinformationmightbeinconvincible.Consequently,analternativecategory,namelythenon-probabilisticapproach[15],hasbeenrapiddevelopedfordescribinguncertaintywithincompletestatisticalinformationbyafuzzysetoraconvexset.Inthefuzzysetmethod[16,17],thefuzzyfailureprobabilityofstructuresisassessedbasedonmembershipfunctionrepresentationoftheobserved/measuredinputs.Intheconvexsetmethod[18-20],allpossiblevaluesoftheuncertaintiesareboundedwithinahyper-boxorhyper-ellipsoidwithoutassuminganyinnerprobabilitydistributions.Non-probabilisticmodelshavebeenregardedasattractivesupplementstothetraditionalprobabilisticmodelinthereliabilitydesignofstructuralengineering.Theinterestedreadersarereferredtoresearchpapersbye.g.MoensandVandepitte[21],MollerandBeer[22],ElishakoffandOhsaki[23].@#@Inapracticalengineeringproblemofadhesivebondedsteel-concretecompo