Designation:G172−02(Reapproved2010)´1StandardGuideforStatisticalAnalysisofAcceleratedServiceLifeData1ThisstandardisissuedunderthefixeddesignationG172;thenumberimmediatelyfollowingthedesignationindicatestheyearoforiginaladoptionor,inthecaseofrevision,theyearoflastrevision.Anumberinparenthesesindicatestheyearoflastreapproval.Asuperscriptepsilon(´)indicatesaneditorialchangesincethelastrevisionorreapproval.ε1NOTE—Editoriallycorrecteddesignationandfootnote1inNovember20131.Scope1.1Thisguidebrieflypresentssomegenerallyacceptedmethodsofstatisticalanalysesthatareusefulintheinterpre-tationofacceleratedservicelifedata.Itisintendedtoproduceacommonterminologyaswellasdevelopingacommonmethodologyandquantitativeexpressionsrelatingtoservicelifeestimation.1.2ThisguidecoverstheapplicationoftheArrheniusequationtoservicelifedata.Itservesasageneralmodelfordeterminingratesatusageconditions,suchastemperature.Itservesasageneralguidefordeterminingservicelifedistribu-tionatusagecondition.Italsocoversapplicationswheremorethanonevariableactsimultaneouslytoaffecttheservicelife.Forthepurposesofthisguide,theaccelerationmodelusedformultiplestressvariablesistheEyringModel.Thismodelwasderivedfromthefundamentallawsofthermodynamicsandhasbeenshowntobeusefulformodelingsometwovariableacceleratedservicelifedata.Itcanbeextendedtomorethantwovariables.1.3Onlythosestatisticalmethodsthathavefoundwideacceptanceinservicelifedataanalyseshavebeenconsideredinthisguide.1.4TheWeibulllifedistributionisemphasizedinthisguideandexamplecalculationsofsituationscommonlyencounteredinanalysisofservicelifedataarecoveredindetail.ItistheintentionofthisguidethatitbeusedinconjunctionwithGuideG166.1.5Theaccuracyofthemodelbecomesmorecriticalasthenumberofvariablesincreasesand/ortheextentofextrapola-tionfromtheacceleratedstresslevelstotheusagelevelincreases.Themodelsandmethodologyusedinthisguideareshownforthepurposeofdataanalysistechniquesonly.Thefundamentalrequirementsofpropervariableselectionandmeasurementmuststillbemetforame...
