Theory and Practice of Quality and Engineering in Asia Industry 1st Edition Cher Ming Tan Visit to download the full and correct content document: https://textbookfull.com/product/theory-and-practice-of-quality-and-reliability-engineeri ng-in-asia-industry-1st-edition-cher-ming-tan/
More products digital (pdf, epub, mobi) instant download maybe you interests ...
Reliability engineering theory and practice 8th Edition
Birolini A
https://textbookfull.com/product/reliability-engineering-theoryand-practice-8th-edition-birolini-a/
Quality of Teacher Education and Learning Theory and Practice 1st Edition Xudong Zhu
https://textbookfull.com/product/quality-of-teacher-educationand-learning-theory-and-practice-1st-edition-xudong-zhu/
Stimulating Student Interest in Language Learning: Theory, Research and Practice 1st Edition Tan Bee Tin (Auth.)
https://textbookfull.com/product/stimulating-student-interest-inlanguage-learning-theory-research-and-practice-1st-edition-tanbee-tin-auth/
Stochastic Models in Reliability Engineering (Advanced Research in Reliability and System Assurance Engineering) 1st Edition Lirong Cui (Editor)
https://textbookfull.com/product/stochastic-models-inreliability-engineering-advanced-research-in-reliability-andsystem-assurance-engineering-1st-edition-lirong-cui-editor/
Constitutional Foundings In South Asia Yl Tan https://textbookfull.com/product/constitutional-foundings-insouth-asia-kevin-yl-tan/
Long Term Reliability of Nanometer VLSI Systems Modeling Analysis and Optimization Sheldon Tan
https://textbookfull.com/product/long-term-reliability-ofnanometer-vlsi-systems-modeling-analysis-and-optimizationsheldon-tan/
Advances in Reliability and System Engineering 1st Edition Mangey Ram https://textbookfull.com/product/advances-in-reliability-andsystem-engineering-1st-edition-mangey-ram/
Handbook of Pharmaceutical Wet Granulation Theory and Practice in a Quality by Design Paradigm 1st Edition Ajit S. Narang
https://textbookfull.com/product/handbook-of-pharmaceutical-wetgranulation-theory-and-practice-in-a-quality-by-designparadigm-1st-edition-ajit-s-narang/
Development of Tourism and the Hospitality Industry in Southeast Asia 1st Edition Purnendu Mandal
https://textbookfull.com/product/development-of-tourism-and-thehospitality-industry-in-southeast-asia-1st-edition-purnendumandal/
Cher Ming Tan · Thong Ngee Goh Editors
Theory and Practice of Quality and Reliability Engineering in Asia Industry TheoryandPracticeofQualityandReliability
EngineeringinAsiaIndustry
• ThongNgeeGoh Editors
TheoryandPractice ofQualityandReliability EngineeringinAsiaIndustry CherMingTan
Editors CherMingTan CenterforReliabilityScience andTechnologies
ChangGungUniversity
Taoyuan Taiwan
ThongNgeeGoh DepartmentofIndustrialandSystems Engineering
NationalUniversityofSingapore
Singapore
Singapore
ISBN978-981-10-3288-2ISBN978-981-10-3290-5(eBook) DOI10.1007/978-981-10-3290-5
LibraryofCongressControlNumber:2016958465
© SpringerNatureSingaporePteLtd.2017
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart ofthematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations, recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmission orinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodologynowknownorhereafterdeveloped.
Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthis publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfrom therelevantprotectivelawsandregulationsandthereforefreeforgeneraluse.
Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthis bookarebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernorthe authorsortheeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinor foranyerrorsoromissionsthatmayhavebeenmade.
Printedonacid-freepaper
ThisSpringerimprintispublishedbySpringerNature TheregisteredcompanyisSpringerNatureSingaporePteLtd. Theregisteredcompanyaddressis:152BeachRoad,#22-06/08GatewayEast,Singapore189721,Singapore
Preface Theincreasinginstancesofproductrecallsglobally,withthehigh-profilecases oftheTanakaairbagsincarsandSamsungNote7smartphones,signifythe importanceandnecessityofmeasuresforQualityandReliabilityinindustries.This isespeciallysoinAsiaasmoreandmoremanufacturingactivitiesaretakingplace inthispartoftheworld,withacorrespondingincreaseinresearchanddevelopment initiatives.
ItcannotbeoveremphasizedthatQualityandReliabilityManagement,which focusesonorganization,planningandbehavioralmatters,withanorientation towardsanalyticaltoolsaswellastechnicalperformance,isdynamicinnatureas customerdemandandmarketconditionsevolve.Theteachingofpast “Quality Gurus” likeJuranandDemingprovidesabearingforbusinessleadersandoperationspersonnel,butastimepasses,newthinking,techniquesandanalyticalprocedureswillcontinuetodeveloptomeetfreshneedsandrequirementsasproduct variety,complexityandapplicationdomainsincrease.Thusglobalization,bigdata, customization,sustainability,instantandwidecommunications tonameafew majoritems arerealitieswhichtraditionalmainstreammethodologiesof StatisticalProcessControlandReliabilityTesting,forexample,havehardly addressed.Indeedprofessionalseverywhereareconstantlyfacedwithexpectedor unexpectedemergingneedsandrequirements.Thusthereisanincessantstreamof newinterestsandissuesfacedbyresearchersandpractitioners,makingQualityand Reliabilitystudiessuchfascinatingandimportantareasofendeavorwithinboth academiaandindustrialsectorsworldwide.
ThisbookiscontributedbyresearchersandpractitionersofQualityand ReliabilityManagementandEngineeringintherecentpast.Thepapersarebased ontheauthors’ presentationsatthe2014ANQ(AsianQualityNetwork)Congress inSingapore;theyhavebeenselectedbasedontheirrelevancetothepracticeof QualityandReliability,aswellasthedirectionstheyhavepointedoutinthe profession’sdevelopment.ThecompilationisnotmeanttobeatextbookonQuality andReliability,forwhichtherearealreadymanytitlesinthemarket,butisa snapshotofareaswithinQualityandReliabilitythatareofcurrentinterestto
professionalsinthis fi eld,andalso asexplicitlydescribedinsomeofthepapers societyingeneral.ThecontentsofthisbookwouldconstituteaveryusefulsupplementtousersoftraditionaltextbooksonQualityandReliabilityasthereare manyactualcasesandexamplesreflectingthelatestapplicationsinareassuchas servicesandtheenvironment.
WewouldliketothanktheANQGoverningBoardforitsapprovalofand supportforthepublicationofthisbook,andinparticulartoMrKennethLiang, PresidentoftheSingaporeQualityInstitute(SQI),whohadmobilizedresourcesfor thepublication’ssuccessfulpreparation.Wewouldalsoliketotakethisopportunity tothankallthosewhohaveinonewayoranothercontributedtothecompilation ofthepaperinthisvolume.Inparticular,wewouldliketomentionDrHoSiong Lin,MrTeoLipHong,MrEllsonBoey,MsSharonTayandMrKitsonLeewho helpedwiththeselectionandeditingofpapersinChapters 3, 4, 5, 7 and 8 respectively.Lastbutnotleast,wewouldliketothankMrTooMengKen,SQI publicationchairman,andallthereviewerswhohavehelpedinensuringthehigh qualityofthepapersincludedinthisbook.ThanksarealsoextendedtoMr.Vivek Sagawan,ResearchAssistantofProfessorTan,whoassistedinliaisingwith Springertoensuretheaccuracyofthecontentsofthesepapers.
Taoyuan,TaiwanCherMingTan Singapore,SingaporeThongNgeeGoh
TheImpactofASEANUniversityNetwork-QualityAssurance (AUN-QA)AssessmentontheQualityofEducationalProgrammes .... 87 OngCheeBinJohnson
TheEvaluationofServiceQualityforHigherEducationinTaiwan byUsingImportance-SatisfactionModel 99 Kuo-ChunTsai,PoTsangB.HuangandChing-ChowYang
QualityAssuranceInformationSystemofChulalongkornUniversity
ThongchaiPitimanuskul,GrittapatAkkaphuriwong,WiboolKhongdee andVimolvanPimpan
TotalQualityManagementinEducationalInstitutions
A.K.M.F.Hoque,M.S.HossainandM.A.Zaman
PartVQualityinHealthcare
AStudyontheFallsPreventionSysteminanAcuteHospital
MasatakaSanoandMasatoYoshida
AStudyontheMeasurementMethodofEffect
TakeyukiGoto,MasahikoMunechikaandChisatoKajihara PatientAssignmentandGroupingProblemonHomeHealthCare:
M.Y.Lin,K.S.ChinandK.L.Tsui
PartVIQualityinServiceIndustry
DevenderMalhotra TheRelationshipsBetweenInnovationOrientation,Customer
ChinHonTan,GuanghuaHanandThongNgeeGoh
PartVIIQualityForEnvironmentalSustainability
MinZhouandThongNgeeGoh
EnvironmentalQuality:ACaseStudyofPublicPerceptions
JingLi,ThongNgeeGohandMinZhou
PartVIIIReliabilityPractices QualityDecisionforOverchargedLi-IonBatteryfromReliability andSafetyPerspective ........................................
223 FengLeng,CherMingTan,RachidYazami,KenzaMaher andRobertWang
StatisticalInferenceforGroupedFieldFailureData
233 PiaoChenandZhi-ShengYe
TheCorrelationBetweenDeviceAgingandSystemDegradation 249 SongLanandCherMingTan
LifetimePredictionofVehicleComponentsUsingOnline MonitoringData ............................................. 261 ChiharuKumazaki,WataluYamamotoandKazuyukiSuzuki
EstimationofLifetimeDistributionwithCovariatesUsing OnlineMonitoring ............................................ 279 MasahiroYokoyama,WataluYamamotoandKazuyukiSuzuki
AnalysisofWindow-CensoredRepairLogsandItsApplication toMaintenance .............................................. 291 WataluYamamoto,LuJinandKazuyukiSuzuki
Contributors GrittapatAkkaphuriwong OfficeofStrategyManagementandBudgeting, ChulalongkornUniversity,Pathumwan,Bangkok,Thailand
CharlesAubrey AsiaPaci ficQualityOrganization,GrandJunction,MI,USA; InternationalAcademyforQuality,GrandJunction,MI,USA;AubreyPartners, GrandJunction,MI,USA;AndersonPackaging,GrandJunction,MI,USA
PiaoChen DepartmentofIndustrialandSystemsEngineering(ISE),National UniversityofSingapore(NUS),Singapore,Singapore
Chuen-ShengCheng DepartmentofIndustrialEngineeringandManagement, Yuan-ZeUniversity,Chung-Li,Taoyuan,Taiwan,ROC
K.S.Chin CityUniversityofHongKong,Kowloon,HongKong
FrancisChua MasterinEducation,ColemanCollege,Singapore,Singapore
ThongNgeeGoh DepartmentofIndustrialandSystemsEngineering,National UniversityofSingapore,Singapore,Singapore
TakeyukiGoto WasedaUniversity,Okubo,Shinjuku,Tokyo,Japan
GuanghuaHan DepartmentofIndustrialandSystemsEngineering,National UniversityofSingapore,Singapore,Singapore
A.K.M.F.Hoque PhysicsDepartment,TejgaonCollege,Farmgate,Dhaka, Bangladesh
M.S.Hossain PhysicsDepartment,NationalUniversity,Gazipur,Bangladesh
PoTsangB.Huang DepartmentofIndustrialandSystemsEngineering,Chung YuanChristianUniversity(Taiwan),ChungLi,Taiwan
LuJin DepartmentofInformatics,UniversityofElectro-Communications,Chofu, Tokyo,Japan
OngCheeBinJohnson AUN-QAExpert,EducationQualityInternational, Singapore,Singapore
UkJung SchoolofBusiness,DonggukUniversity,Seoul,Korea
ChisatoKajihara WasedaUniversity,Okubo,Shinjuku,Tokyo,Japan
WiboolKhongdee OfficeofStrategyManagementandBudgeting,Chulalongkorn University,Pathumwan,Bangkok,Thailand
Gye-SooKim DepartmentofBusinessAdministration,SemyungUniversity, Jechon,Chungbuk,Korea
ChiharuKumazaki DepartmentofInformatics,Universityof Electro-Communications,Chofu,Tokyo,Japan;DepartmentofPowerTrain Experiment,HinoMotors,Ltd.,Hino,Tokyo,Japan
SamLam MasterinBusinessAdministration,SingaporeQualityInstitute, Singapore,Singapore
SongLan DepartmentofElectricalandElectronicEngineering,Nanyang TechnologicalUniversity,Singapore,Singapore
Hung-TingLee DepartmentofIndustrialEngineeringandManagement,Yuan-Ze University,Chung-Li,Taoyuan,Taiwan,ROC
FengLeng SchoolofElectricalElectronicsEngineering,NanyangTechnological University,Singapore,Singapore;TUMCREATEPTELTD,Singapore,Singapore
JingLi MechanicalEngineeringDepartment,NationalUniversityofSingapore, Singapore,Singapore
M.Y.Lin CityUniversityofHongKong,Kowloon,HongKong
KenzaMaher SchoolofMaterialsScienceandEngineeringandResearchInstitute atNanyang(ERIAN),NanyangTechnologicalUniversity,Singapore,Singapore
DevenderMalhotra WiproITServices,Bangalore,India
MasahikoMunechika WasedaUniversity,Okubo,Shinjuku,Tokyo,Japan
VimolvanPimpan ChulalongkornUniversity,Pathumwan,Thailand
ThongchaiPitimanuskul OfficeofStrategyManagementandBudgeting, ChulalongkornUniversity,Pathumwan,Bangkok,Thailand
MasatakaSano TokyoUniversityofScience,Tokyo,Japan
KazuyukiSuzuki DepartmentofInformatics,Universityof Electro-Communications,Chofu,Tokyo,Japan;JSQC,Universityof Electro-Communications,Chōfu,Japan
CherMingTan CenterforReliabilityScienceandTechnologies,ChangGung University,Taoyuan,Taiwan;DepartmentofElectronicEngineering,ChangGung University,TaoyuanCity,Taiwan
ChinHonTan DepartmentofIndustrialandSystemsEngineering,National UniversityofSingapore,Singapore,Singapore
SharonTay SengkangHealth,Singapore,Singapore
Kuo-ChunTsai DepartmentofIndustrialandSystemsEngineering,ChungYuan ChristianUniversity(Taiwan),ChungLi,Taiwan
K.L.Tsui CityUniversityofHongKong,Kowloon,HongKong
RobertWang SchoolofElectricalElectronicsEngineering,Nanyang TechnologicalUniversity,Singapore,Singapore
WataluYamamoto DepartmentofInformatics,UniversityofElectro-Communications, Chofu,Tokyo,Japan;JSQC,UniversityofElectro-Communications,Chōfu,Japan
Ching-ChowYang DepartmentofIndustrialandSystemsEngineering,Chung YuanChristianUniversity(Taiwan),ChungLi,Taiwan
RachidYazami TUMCREATEPTELTD,Singapore,Singapore;Schoolof MaterialsScienceandEngineeringandResearchInstituteatNanyang(ERIAN), NanyangTechnologicalUniversity,Singapore,Singapore
Zhi-ShengYe DepartmentofIndustrialandSystemsEngineering(ISE),National UniversityofSingapore(NUS),Singapore,Singapore
MasahiroYokoyama JSQC,PolytechnicUniversity,Kodaira,Tokyo,Japan
MasatoYoshida TokyoUniversityofScience,Tokyo,Japan
M.A.Zaman PhysicsDepartment,JahangirnagarUniversity,Savar,Dhaka, Bangladesh
MinZhou IndustrialandSystemsEngineeringDepartment,NationalUniversityof Singapore,Singapore,Singapore
QualityEngineeringToday T.N.Goh
TomanyintheQualityprofession,thetermsQualityManagement(QM)and QualityEngineering(QE)tendtobeusedinterchangeably,oratleasttherehasbeen littleofficialattempttodistinguishthem.Asrecentas2014,ASQ(American SocietyforQuality) http://asq.org/glossary/q.html (seenonApril29,2014) describesQEas “Theanalysisofamanufacturingsystematallstagestomaximize thequalityoftheprocessitselfandtheproductsitproduces.” (ASQ2014).Thefact is,Qualitytodayisnotjustaboutmanufacturing(asqualityofservicehasgained considerableattentionnowadays)andQE,unlikemanytraditionalengineering disciplines,cannotbeeffectivewithoutmanagementelements andthatiswhyone needsthewisdomofbothShewhartandDeming:isitasurprisethattheP-D-C/S-A cycleforqualityimprovementisattributedtobothShwehartandDeming,arguably fromtheengineeringandmanagementperspectives,respectively?
WhilethereisnocommonlyrecognizeddefinitivedelineationofwhatQMis meanttobeandwhatQEpreciselyis,itmaybesaidthattheformertouchesmore onconceptualandbehavioralmattersaswellasapproachestoaccomplishing businessobjectives,andthelattertheapplicationofgenericandusuallyrigorous analyticaltoolscomprisingtechnicallyjusti fiablemethodologiesdevoidofideologiesorsubjectivejudgments.Itmaybenotedthattherecentyearshasseenan increaseduseoftheterm “engineering” withaconnotationofplanningand endeavor asin “socialengineering.” Whatevertheinterpretation,theimpetusfor researchandapplicationofgoodtechniquescontinuesunabated,anddeveloping economiesoftheworldsuchasthoseinAsia,arewherethegrowthoftheQuality fieldisparticularlynoticeable.
T.N.Goh(&)
NationalUniversityofSingapore,Singapore,Singapore e-mail:isegohtn@nus.edu.sg;tngoh2008@gmail.com
© SpringerNatureSingaporePteLtd.2017
C.M.TanandT.N.Goh(eds.), TheoryandPracticeofQuality andReliabilityEngineeringinAsiaIndustry, DOI10.1007/978-981-10-3290-5_1
Generally,professionalandacademicconferencesarethemostcommonplatformfornoteworthyqualitytechniquestobepresentedanddiscussed.Thus,this volumecontainsselectedpapersfromthe2014CongressoftheAsianNetworkfor Quality(ANQ)thatservetoreflectawidevarietyofapproachestoquality improvement.Readers,regardlessofindividualexperienceandmaturity,would catchaglimpseofsomelatestthoughtsandapplicationsintheQuality field.This collectionreflectsyetanotherstep,albeitcomingmostlyfromAsia,intheardent, globaljourneytowardqualityexcellenceinall fi eldsofhumanendeavor.
AuthorBiography GohThongNgee (吴桐毅),PE,isAcademicianoftheInternationalAcademyforQuality(IAQ), FellowoftheAmericanSocietyforQuality(ASQ)andHonoraryMemberandAdvisorofthe SingaporeQualityInstitute(SQI).HeisarecipientofIEEEEngineeringManagementSociety “EducatoroftheYear” awardin2005,IAQMasingMedalin2006(forthebook “SixSigma: AdvancedToolsforBlackBeltsandMasterBlackBelts”,WileyUK),ASQStatisticsDivision WilliamGHunterawardin2007,Asia-Paci ficQualityOrganizationHarrington-IshikawaMedal in2010,andASQEugeneLGrantMedalin2012.
ThecitationforProfGoh’sASQGrantMedalreads: “Foranoutstandingcareerdedicatedtothe educationandresearchofconceptsandtechniquesforqualityexcellenceandperformance improvement;andforexceptionalleadershipinpromotingthelearningandapplicationofeffective qualitytoolsthroughouttheworld.”
OverviewofReliabilityEngineering CherMingTan
Abstract Theevolutionofreliabilitysincethelate1940till2000sisdescribed. ThisevolutionmainlycamefromUnitedStates,andtheevolutionisfromsystem levelreliabilityandmaintenancetocomponentsreliabilitythroughstatistical methods.Exponentialdistributionwasusedinitially,anditwasfoundnotrealistic inthe60s.Physicsoffailurebeganinthe90s,andreliabilityevaluationwasthen shiftedfromstatisticalmethodstophysicsoffailure.Since2000,thehybridphysics–statisticalapproachevolveduetotheneedtocombinebothoftheminorderto evaluatecomponentandsystemreliabilityrealistically.Whilereliabilityisevolving andmaturinginUSA,theconceptofreliabilitywasintroducedtoAsiaonlyinthe late70sintheformofqualificationasUSmanufacturerswereshiftingtheirproductiontoAsia,andtheyneededtoqualifythefactoriesinAsia.Unfortunately,the growthofreliabilitymethodologyisveryslowinAsia,andstillmanyareusing exponentialdistributiontoevaluatereliability.Ontheotherhand,physicsoffailure approachisgrowingfastinAsia,dueprobablytotheneedtotroubleshootfailures manufacturedinAsia.Thischapterhelpsthereaderstohaveaborderviewof reliabilityandthenecessityforitsevolution.
Keywords Historicalevolution Statisticalapproach PhysicsofFailure approach Hybridapproach MonteCarlomodeling Systemreliability Eponentialdistribution
BeforeWorldWarII(WWII),reliabilityasawordcametomeandependabilityor repeatability.SuchmeaningstillexiststomanyinAsiatoday.Themodernusewas rede finedbytheU.S.militaryinthe1940sandevolvedtothepresent.Bythe 1940s,reliabilityandreliabilityengineeringstilldidnotexist.
Thecurrentmeaningconnotesanumberofadditionalattributesthatspanproducts,serviceapplications,softwarepackagesorhumanactivity.Theseattributesnow
C.M.Tan(&)
CenterforReliabilityScienceandTechnologies, ChangGungUniversity,Taoyuan,Taiwan e-mail:cmtan@mail.cgu.edu.tw
© SpringerNatureSingaporePteLtd.2017
C.M.TanandT.N.Goh(eds.), TheoryandPracticeofQuality andReliabilityEngineeringinAsiaIndustry, DOI10.1007/978-981-10-3290-5_2
pervadeeveryaspectofourpresentdaytechnologicallyintensiveworld.Letus followthejourneyofreliabilityengineeringfromtheearlydaystopresent,where mostoftheinformationdescribeduptoyear2000sarefromReferences(McLinn 2010;Denson 1998;AzarkhailandModarres 2012)unlessotherwisereferenced.
1TheLate1940s ThedemandsofWWIIintroducedmanynewelectronicsproductsintoUSmilitary. Attheonsetofthewar,itwasdiscoveredthatover50%oftheairborneelectronics equipmentinstoragewasunabletomeettherequirementsoftheAirCoreand Navy.Thiswasbecausethemainelectronicdeviceswerestillthevacuumtubes, whetheritwasinradarsystemsorotherelectronics.Thesesystemshadproved problematicandcostlyduringthewar.Forshipboardequipmentafterthewar,it wasestimatedthathalfoftheelectronicequipmentwasdownatanygiventime. Vacuumtubesinsocketswereanaturalcauseofsystemintermittentproblems. Theycouldnotaffordtohavehalfoftheiressentialequipmentnon-functionalallof thetime.Theoperationalandlogisticscostswouldbecomeastronomicalifthis situationwasnotsoonrecti fi ed.
Ontheotherhand,muchofthereliabilityworkduringthisperiodhadtodowith testingnewmaterialsandfatigueofmaterials.M.A.Minerpublishedtheseminal papertitled “CumulativeDamageinFatigue” in1945inanASMEJournal.
B.Epsteinpublished “StatisticalAspectsofFractureProblems” intheJournalof AppliedPhysicsinFebruary1948.Thesewereobviouslyinsuffi cienttomeetthe reliabilityneedonelectronics,andIEEEformedtheReliabilitySocietyin1948 withRichardRollmanasthe fi rstpresident.Z.W.BirnbaumfoundedtheLaboratory ofStatisticalResearchattheUniversityofWashingtoninthesameyear,and throughhislongassociationwiththeOfficeofNavalResearch,thelabservedto strengthenandexpandtheuseofstatisticsinreliability.
2The1950s Astudygroupwasinitiatedin1950.ThisgroupwascalledtheAdvisoryGroupon theReliabilityofElectronicEquipment,AGREEforshort.By1952,aninitial reportbythisgrouprecommendedthefollowingthreeitemsforthecreationof reliablesystems:
1.Therewasaneedtodevelopbettercomponentsandmoreconsistencyfrom suppliers.
2.Themilitaryshouldestablishqualityandreliabilityrequirementsforcomponent suppliers.
3.Actual fielddatashouldbecollectedoncomponentsinordertoestablishthe rootcausesofproblems.
In1957,a finalreportwasgeneratedbytheAGREEcommitteeanditsuggested thefollowing:
1.Mostvacuumtuberadiosystemsfollowedabathtub-typecurve.
2.Itwaseasiertodevelopreplaceableelectronicmodulestoquicklyrestorea failedsystem,andthustheyemphasizedmodularityofdesign.
3.Recommendtorunformaldemonstrationtestswithstatisticalconfidencefor products.
4.Recommendtorunlongerandharsherenvironmentalteststhatincludedtemperatureextremesandvibration
TherecommendationscametobeknownasAGREEtestingandeventually turnedintoMilitaryStandard781.AnotheritemprovidedbytheAGREEreport wastheclassicdefinitionofreliability.Thereportstatedthatthedefinitionis “the probabilityofaproductperformingaspeci fiedfunctionwithoutfailureundergiven conditionsforaspecifiedperiodoftime” .
Inparallel,RomeAirDevelopmentCenter(RADC)wasestablishedinRome, NewYorkin1951tostudyreliabilityissueswiththeAirForce.In1955,RADC issued “ReliabilityFactorsforGroundElectronicEquipment”.Thiswasauthored byJosephNaresky.ThisdecadeendedwithRCApublishinginformationin TR1100onthefailureratesofsomemilitarycomponents.RADCpickedthisup anditbecamethebasisforMil-StdHandbook217(MH-217)in1962.
Theapplicationsoftheprobabilisticnotionsofreliabilityaroundthistimewere widelyrepresentedbytheexponentialdistribution.Oneofthemaindrivingforces forthispopularitywasthesimplicityofthecorrespondingreliabilityfunctions. Havinglimitedcomputationalresources,theearlyreliabilitypractitionerswere evidentlyseekingasimplereliabilitymodelwithastraightforwardmathematical representation.Acombinationofthesefactorsmadetheexponentialdistributionthe dominantmodelinearlyreliabilityassessments.Thissimplicityacceleratedmany improvementsintraditionalstatisticalandprobabilisticapproachestomeasuring, predictingandtestingofitemreliabilityinthe1950s.Ontheotherhand,Wallodi Weibullpublishedhis firstpaperfortheASMEJournalofAppliedMechanicsin English.Itwastitled “AStatisticalDistributionFunctionofWideApplicability” in 1951inviewofthelimitationsoftheexponentialdistributionthatheunderstood. Lateron,Weibullproduced “StatisticalEvaluationofDatafromFatigueand CreepRuptureTests:FundamentalConceptsandGeneralMethods” asaWrightAir DevelopmentCenterReport59-400fortheUSmilitaryin1959.Birnbaumalso madesignificantcontributionstoprobabilisticinequalities(i.e.Chebychev), non-parametricstatistics,reliabilityofcomplexsystems,cumulativedamage models,competingrisk,survivaldistributionsandmortalityratesduringthisdecade.By1956,ASQCwasofferingpapersonreliabilityaspartoftheirAmerican QualityCongress.Theradioengineers,ASME,ASTMandtheJournalofApplied Statisticswerecontributingresearchpapers.TheIREwasalreadyholdinga
conferenceandpublishingproceedingstitled “TransactiononReliabilityand QualityControlinElectronics”.Thisbeganin1954andcontinueduntilthisconferencemergedwithanIEEEReliabilityconferenceandbecametheReliabilityand MaintainabilitySymposium.
Anothermajorreporton “PredictingReliability” in1957wasthatbyRobert LusserofRedstoneArsenal,wherehepointedoutthat60%ofthefailuresofone Armymissilesystemwereduetocomponents.Heshowedthatcurrentmethodsfor obtainingqualityandreliabilityforelectroniccomponentswereinadequateandthat somethingmorewasneeded.
In1955,aconferenceonelectricalcontactsandconnectorswasstarted, emphasizingreliabilityphysicsandunderstandingfailuremechanisms.Other conferencesbeganinthe1950stofocusonsomeoftheseimportantreliability topics.
Thisdecadeendedwithalotofpromiseandactivities.Itwastheadventofthe reliabilityengineeringdiscipline,andseveraldifferentmethodswereintroducedto achievethegoalofhigherreliability.Twobranchesofreliabilitybegantoemerge. Onebranchexistedforinvestigationoffailuresandtheotherforpredictions.
3The1960s Thedemandsofthemilitaryrangingfrommissilestoairplanes,helicoptersand submarineapplicationsdroveavarietyoftechnologies.Thestudyoftheeffectsof EMConsystemswasinitiatedatRADCandthisproducedmanydevelopmentsin the1960s.
Bynow,thereliabilitydisciplinewasworkingunderthetenetthatreliabilitywas aquantitativedisciplinethatneededquantitativedatasourcestosupportitsmany statisticallybasedtechniques,suchasallocationandredundancymodelling. Anotherbranchofthereliabilitydisciplinefocusedonthephysicalprocessesby whichcomponentswerefailing.The firstsymposiumdevotedtothistopicwasthe “PhysicsofFailureinElectronics ” symposiumsponsoredbytheRADCandIIT ResearchInstitute(IITRI)in1962.RichardNelsonofRADCproducedthedocument “QualityandReliabilityAssuranceProceduresforMonolithicMicrocircuits,” whicheventuallybecameMil-Std883andMil-M38510.Thissymposiumlater becamethe “InternationalReliabilityPhysicsSymposium(IRPS)”.Unfortunately, theelectronicindustrycontinuedusingthelottolerancepercentagedefective (LTPD)statistictoaddressfailuresofelectronics.Itwasnotuntillate1980sthat PoFwasrevisitedasaseriousalternative.
In1962,G.A.DodsonandB.T.HowardofBellLabspublished “HighStress AgingtoFailureofSemiconductorDevices” intheProceedingsofthe7thNational SymposiumofReliabilityandQualityControl.ThispaperjustifiedtheArrhenius modelforsemiconductors.Lotsofotherpapersatthisconferencelookedatother componentsforimprovement.
1962wasakeyyearwiththe firstissueofMH-217bytheNavy.Onceissued, MH-217quicklybecamethestandardbywhichreliabilitypredictionswereperformed,andothersourcesoffailureratesgraduallydisappeared.Partofthereason forthedemiseofothersourceswasthefactthatMH-217wasoftenacontractually citeddocumentanddefencecontractorsdidnothavetheoptionofusingother sourcesofdata.
Bythe1960s,theexponentialdistributionturnedouttobenotsopracticalfor manyapplicationsandsensitivetodeparturefromtheinitialassumptions.The applicationofthismodelforcomponentswithhighreliabilitytargetscouldresultin unrealisticmean-time-to-failure(MTTF).Further,thismodelbasicallyignoredany aginganddegradationinthecomponentandhadnomemorytokeeptrackofthe damagebeingaccumulatedintheitem.
Aftersuchdisappointments,reliabilitypractitionersmadeanattempttocapture someofthephysicalcharacteristicsoffailureintotheirmodellingbyusingother availabletraditionaldistributions,suchastheWeibullandlognormaldistributions. Duringthisdecade,anumberofpeoplebegantouse,andcontributetothegrowth anddevelopmentof,theWeibullfunction,thecommonuseoftheWeibullgraph, andthepropagationofWeibullanalysismethodsandapplications.
In1969,BirnbaumandSaundersdescribedalifedistributionmodelthatcould bederivedfromaphysicalfatigueprocesswherecrackgrowthcausesfailure.
Humanreliabilityhadnowbeenrecognizedandstudied,whichresultedina paperbySwainonthetechniquesforhumanerrorrateprediction(THERP).
Duringthedecade,astrongcommitmentofUSGovernmenttospaceexplorationresultedintheestablishmentofNASA,andthisprovidedadrivingforcefor improvedreliabilityofcomponentsandsystems.Thedecadeendedwithalanding onthemoonshowinghowfarreliabilityhadprogressedinonly10years.
However,thetwobranchesofreliabilityengineeringseemedtobedivergingby theendofthedecade,withthe system engineersdevotedtothetasksofspecifying, allocating,predicting,anddemonstratingreliability,whilethephysics-of-failure engineersandscientistsweredevotingtheireffortstoidentifyingandmodellingthe physicalcausesoffailure.Bothbrancheswereintegralpartsofthereliability discipline.
4The1970s Intheearly1970s,theresponsibilityforpreparingMH-217wastransferredto RADC,whopublishedrevisionBin1974.However,otherthanthetransitionto RADC,the1970smaintainedthestatusquoinreliabilityprediction.MH-217was updatedtoreflectthetechnologyatthattime,andtherewasashiftinthecomplexityofthemodelsbeingdevelopedforMH-217.
Infact,therewereseveraleffortstodevelopnewinnovativemodelsforreliabilityprediction.Theresultsoftheseeffortswereextremelycomplexmodelsthat mighthavebeentechnicallysound,butwerecriticizedbytheusercommunityas
beingtoocomplex,toocostly,andunrealistic.ThesemodelswereneverincorporatedintoMH-217.
The1970smarkedthebirthoffaulttreeanalysis,whichwasmotivatedbythe needforsafetyassessmentintheaerospaceindustryandlaterfornuclearpower plants.Uptothispoint,mostreliabilityengineeringeffortswerefocusedonreliabilityofcomponentsanddevices.Nevertheless,therewasintenseinterestin system-levelsafety,riskandreliabilityindifferentapplicationssuchasthegas,oil, andchemicalindustries,andaboveall,innuclearpowerapplications.These challengeswereimmenselyappealingtoreliabilitycommunityinthe1970s.
TheNavyMaterialCommandbroughtinWillisWilloughbyfromNASAtohelp improvemilitaryreliabilityacrossavarietyofplatforms.DuringtheApollospace program,Willoughbyhadbeenresponsibleformakingsurethatthespacecraft workedreliablyallthewaytothemoonandback.IncomingtotheNavy,hewas determinedtopreventunreliability.Heinsistedthatallcontractscontainspeci ficationsforreliabilityandmaintainabilityinsteadofjustperformancerequirements. Willoughby’seffortsweresuccessfulbecauseheattackedthebasicsandworked uponabroadfront.
5The1980s The1980sdemonstratedprogressinreliabilityacrossanumberoffrontsfrom militarytoautomotiveandtelecommunicationstobiomedical.Thiswasbecause televisionshadbecomeallsemiconductor.Automobilesrapidlyincreasedtheiruse ofsemiconductorswithavarietyofmicrocomputersunderthehoodandinthedash. Largeairconditioningsystemsdevelopedelectroniccontrollers,ashadmicrowave ovensandavarietyofotherappliances.Communicationssystemsbegantoadopt electronicstoreplaceoldermechanicalswitchingsystems.RADCpublishedtheir firstReliabilityToolKit.
WhileMH-217wasupdatedseveraltimes,otheragenciesweredeveloping reliabilitypredictionmodelsuniquetotheirindustries.Asanexample,theautomotiveindustry,undertheauspicesoftheSocietyofAutomotiveEngineers (SAE)ReliabilityStandardsCommittee,developedasetofmodelsspeci ficto automotiveelectronics,SAE870050.TheSAEcommitteebelievedthattherewere noexistingpredictionmethodologiesthatappliedtothespecifi cqualitylevelsand environmentsofautomotiveapplications.
TheBellcorereliabilitypredictionstandardisanotherexampleofaspeci fic industrydevelopingmethodologiesfortheiruniqueconditionsandequipment.It originallywasdevelopedbymodifyingMH-217toreflectbettertheconditionsof interesttothetelecommunicationindustry.Ithassincetakenonitsownidentity withmodelsderivedfromtelecommunicationequipmentandisnowusedwidely withinthatindustry.
Duringthisdecade,thefailurerateofmanycomponentsdroppedbyafactorof 10.Softwarebecameimportanttothereliabilityofsystems;thisdisciplinerapidly
advancedwithworkatRADC.Complexsoftware-controlledrepairablesystems begantouseavailabilityasameasureofsuccess.Repairsonthe flyorquickrepairs tokeepasystemoperatingwouldbeacceptable.
TheVeryHighSpeedIntegratedCircuit(VHSIC)programwastheUS Government’sattempttoleveragefromthetechnologicaladvancementsofthe commercialindustryandatthesametimeproducecircuitscapableofmeetingthe uniquerequirementsofmilitaryapplications.FromtheVHSICprogramcame theQuali fiedManufacturersList(QML) aqualifi cationmethodologythatqualifiesanICmanufacturingline,unlikethetraditionalqualificationofspeci ficparts. TheUSGovernmentrealizedthatitneededaQML-likeprocessifitwereto leveragefromtheadvancementsincommercialtechnologies,andatthesametime haveatimelyandeffectivequalificationschemeformilitaryparts.Atthispoint, reliabilityconceptwasintroducedtoAsiaasaformofqualificationassomeUS electronicsmanufacturersweremovingsomeoftheirmanufacturingsitestosome AsiacountriessuchasSingaporeandTaiwansincethe1970s.
Astechnologyhasadvanced,thegateortransistorcountbecamesohighthatit couldnolongereffectivelybeusedasthemeasureofcomplexityinareliability model.Furthermore,transistororgatecountdatawereoftendifficultorimpossible toobtain.Therefore,themodeldevelopedforVHSICmicrocircuitsneededanother measureofcomplexityonwhichtobasethemodel.Thebestmeasures,andthe onesmosthighlystatisticallycorrelatedtoreliabilityaredefectdensityandthedie areaapplicabletospeci ficICfeatures,e.g.,metalizationandoxide.Thefailurerate (forsmallcumulativefractionfailure)isdirectlyproportionaltotheproductofthe areaanddefectdensity.Anotherfactorthatisalsohighlystatisticallycorrelated withdefectdensityandareaistheyieldofthedie,orthefractionofdiethatis functionaluponmanufacture.
However,theprobleminusingthesefactorsinamodelisthattheyarehighly sensitiveparametersfromamarketcompetitionviewpointandthereforearerarely releasedbythemanufacturers,renderedinaccuracyofthemodels.Theconflict betweentheusabilityofamodelanditsaccuracyhasalwaysbeenadiffi cult compromisetoaddressformodeldevelopers.
Thetraditionalapproachindevelopingalifemodelforsuchcomponentswasto collectasmuch fi eldfailuredataaspossibletobuildastatisticalmodelforthe componentlife.Becauseofthedecreasingbudgetandresources,andalsodueto fastertrendsinmassproduction,greatemphasiswasplacedoncapturingthe neededinformationwithmuchlesseffort.Asaresult,designandassessment methodologiesthataddressedtherootcausesoffailureandotheroperatingconditionsemergedaspowerfulcostsavingtechniques.Theacceleratedlifemodelling approachwasadirectoutcomeofthismovement.
Acceleratedlifemodelstookintoaccountsomeoftheoperationalconditionsand wereaprimaryattemptbyreliabilitypractitionerstomakethelifemodelsmore flexible.Inthe firststepofthisapproachastressagent,whichcouldbeanaggregate effectofmanyphysicalandoperationalconditions,wasintroduced.Inthenextstep thisagentwasaddedtothestatisticaldistributionofTTFtoformarobustand
generallifemodel.Suchmodelshadmore flexibility,yetneededmuchlessreliability(failure)data.
However,itisusuallyverycomplicated(ifpossibleatall)tointroduceoneor twostressagentstoreplacetheaggregateeffectofallinfluentialfactorsinacceleratedlifetestingapproach.Manyassumptionsandsimplifi cationsneedtobe made,whichcouldunacceptablylimittherelevanceoftheoutcomes.Yet,thiswas nottheonlychallenge,sincetheacceleratedlifemodels,likeallother statistical-basedapproaches,neededdataforvalidation,anddatacollectionmeant timeandresourcesthatweredecisivelytightformoststart-upsandfast-growing businesses.Inaddition,therewerenodataavailableforaproductinthedesign stageorforahighlyreliableproductthatwashardtobreak.Insuchcases,ifthe modellerwaslucky,reliabilitymodelscouldbeconstructedbasedonsomegeneric datafromthehistoryofsimilarproducts.Thisdatacouldbeupdatedlaterwith expertjudgmentsorothersoftdatawithstatisticalinferencetechniques.The uncertaintyboundsofsuchpredictionsdependedupondirectfailuredata,if available.Therefore,moredependablereliabilitytechniquesneededtobedevelopedtoaddressreliabilitychallengesimposedbyemergingmassmanufacturing technologies.
ThislatterchallengeledtothegrowingapplicationofBayesianmethodin probabilisticdataanalysisinthe1980s.Usingthisapproach,engineersutilizeddata availableingenerichandbooksandfromexpertopinionsandanypreviousexperiencewithsimilarproductstomakeaprobabilitydensityreferredtoasaprior distribution.TheBayesianframeworkmadeitpossibletoupdatethisprior knowledgelaterandwithjustafewavailabledata,anupgradedposteriorstateof knowledgecanbemade.Theapplicationsofthisapproach,however,wereoriginallylimitedtosimplereliabilitymodelsduetothemathematicalcomplexityof involvedalgorithms.TheintegralsnecessaryfornormalizationatBayesianconditionalprobabilitycalculationscanbeverycomplexwhendealingwith multi-parameterreliabilitymodels.Thisremainedoneofthetwomostimportant constrainingelementsofthisapproach(theotherwasdevelopingaproperlikelihoodfunctionrepresentingreliabilitydata)untilrecently,whenadvancedcomputationaltoolsandtechniquesbecameavailableafterrevolutionaryimprovementin thecomputationalpowerofpersonalcomputers.
The1980salsomarkedthedevelopmentofinitiativesformodellingdependenciesatthesystemlevel.Mostoftheseeffortstackledthecommoncausefailures asfrequentdependencyproblemsinsystems.Thecommoncausefailure(CCF), whichisthefailureofmorethanonecomponentduetoasharedrootcause,is classi fiedasadependentfailure.Inthe1980smanyimplicitandexplicitmethods weredevelopedtoincorporatecommoncauseintothesystemfailureanalysis (McLinn 2010).InearlyattemptstomodelCCFatthesystemlevel,anewindependentfailureeventwithaspeci ficprobabilitywasusuallyaddedtothesystem model.Theprobabilityofthiseventwasestimatedusing fi elddataavailableonthe dependentfailuresofcomponents.
ContributionsbyWilliamMeeker,GeraldHahn,RichardBarlowandFrank Proschanindevelopingtheirstatisticalmodelsforwear,degradationandsystem reliabilitymadeanimpactonreliabilityinthisdecadeandsubsequent.
6The1990s Bythe1990s,thepaceofICdevelopmentwaspickingup.Newcompaniesbuilt morespecializedcircuits.Wideruseofstand-alonemicrocomputerswascommon andthePCmarkethelpedkeepICdensitiesfollowingMoore’sLawanddoubling aboutevery18months.Itquicklybecameclearthathighvolumecommercial componentsoftenexceededthequalityandreliabilityofthesmallbatchspecially screenedmilitaryversions,andthemovetowardsCommercialofftheShelf(COTS) componentsgainedmomentum.Manyofthemilitaryspeci ficationsbecame obsoleteandbestcommercialpracticeswereoftenadopted.RADCupdatedtheir ReliabilityToolKitforCOTSapplicationsinthisdecade.
Withtheendofthecoldwar,themilitaryreliabilitychangedquickly.MH-217 endedin1991atrevisionF2.Newresearchdevelopedfailureratemodelsbased uponintrinsicdefectsthatreplacedsomeofthecomplexity-drivenfailureratesthat dominatedfromthe1960sthroughthe1980s.
RACissuedasixsetBlueprintforEstablishingEffectiveReliabilityProgramsin 1996.Theinternetshowedthatonesinglesoftwaremodelwouldnotworkacross thewiderangeofworldwideapplications,whichnowincludeswireless.Also, networkavailabilitygoalsbecame “fi ve9sor5minannually” todescribethe expectedperformanceintelecommunication,raisingthestandardofreliability. Newapproacheswererequiredsuchassoftwaremirroring,rollingupgrades,hot swapping,self-healingandarchitecturechanges.
Withthespreadofreliabilityrequirementstocommercialworld,newreliability trainingopportunitiesandbooksbecameavailabletothepractitioners.ASQmadea majorupdatetoitsprofessionalcerti ficationexam(i.e.certi fi edreliabilityengineers certi fication)tokeeppacewiththechangesevident.ISO9000addedreliability measuresaspartofthedesignanddevelopmentportionofthecerti fication.
Newtechnologiessuchasmicroelectromechanicalsystems(MEMS),hand-held GPS,andhand-helddevicesthatcombinedcellphonesandcomputersallrepresent challengestomaintainreliability.Productdevelopmenttimecontinuedtoshorten throughthisdecadeandwhathadbeendoneinthreeyearswasnowdonein 18months.Thismeantreliabilitytoolsandtasksmustbemorecloselytiedtothe developmentprocessitself.Consumershavebecomemoreawareofreliability failuresandthecosttothem.
Inthisdecade,severalvocalcriticsofMH-217havecomplainedaboutitsutility asaneffectivemethodforassessingreliability.Whilethefaultfi ndersclaimthatitis inaccurateandcostly,therewasnoviablereplacementinthepublicdomainthen.
Table1 Failure-cause distributionanddefinitionfor electronicsystems
• Parts(22%):Failureresultingfromapart(e.g.,microcircuit, transistor,resistor,connector)failingtoperformitsintended function
• Nodefect(20%):Perceivedfailuresthatcannotbe reproduceduponfurthertesting.Thesemayormaynotbean actualfailure;however,theyareremovalsandthereforecount towardsthelogisticreplacementrate(whichisoften incorrectlyreferredtoasfailurerate)
• Manufacturing(15%):Failuresresultingfromanomaliesin themanufacturingprocess,(e.g.,facultysolderjoints, inadequatewireroutingresultinginchafing,bentconnector pins)
• Induced(12%):Failuresresultingfromanexternallyapplied stresssuchaselectricaloverstressandmaintenanceinduced failures(e.g.,dropping,bendingpins)
• Design(9%):Failuresresultingfromaninadequatedesign (e.g.,tolerancestack-up,unanticipatedlogicconditions,a non-robustdesignforgivenenvironmentalstresses)
• Wearout(9%):Failuresresultingfromwearout-related failuremechanisms(e.g.,electrolyticcapacitors,solderjoints, tubes(TWT),andswitchandrelaycontacts)
• Software(9%):Failureofsystemtoperformitsintended functionduetothemanifestationofasoftwarefault
• Systemmanagement(4%):Failurestointerpretsystem requirements,orfailuretoprovidetheresourcesrequiredto designandbuildareliablesystem
Thefractionforeachfailurecauseisgivenin(),andlistedfrom mosttoleast
Dataareasof2012
Thepremiseoftraditionalmethods,suchasMH-217,isthatthefailurerateis primarilydeterminedbycomponentscomprisingthesystem.Thiswasa not-unreasonablepremiseinthe1960sand1970swhencomponentshadhigher failurerates,andwhensystemswerelesscomplexthantheyaretoday.Also, MH-217wasoriginallycreatedtosupporttheweaponsystemacquisitionprocess. Assuch,itwasrequiredthatareliabilityestimatebeavailabletosystemprogram offices(SPO)atcertainmilestonesintheDoDacquisitioncycle.
Inmanycases,anestimateofreliabilitywasrequiredbeforethepartslistwas evencomplete.Modeldevelopershavelongknownthatmanyofthefactorswhich hadamajorinfluenceonthereliabilityoftheendproductwerenotincludedin MH-217,butundertheconstraintsofhandbookusers,nobettersolutionwas available.
Increasedsystemcomplexityandcomponentqualityinthe1990shaveresulted inashiftofsystemfailurecausesawayfromcomponentstomoresystem-level factors,includingmanufacturing,design,systemrequirements,interface,and software.Table 1 showsthefailurecausedistributionofelectronicsystemsas
Table2 Digitalcircuitboardfailurerates(failuresper106 part-hours)
GroundbenignGround fixed Temperature(°C)10701070
Stress(%)1050105010501050
ALCATEL6.5910.1813.3019.8922.0829.7932.5147.27
BellcoreIssue45.727.0931.6435.438.5610.6347.4653.14
BellcoreIssue58.479.25134.45137.8516.9418.49268.90275.70
BrTelecomHDR46.726.726.726.729.849.849.849.84
BrTelecomHDR52.592.592.592.592.592.592.592.59
MH-217ENotice110.9220.2094.37111.3636.3856.04128.98165.91
MH-217FNotice19.3218.3820.1535.4028.3148.7845.4479.46
MH-217FNotice26.419.8318.3126.7624.7440.1573.63119.21
RACdata3.3
Br ) British
collectedbyRAC.Historically,thesefactorshavenotbeenexplicitlyaddressedin predictionmethods.
Therefore,agoodmodelshouldreflectsstate-of-the-arttechnology,anditshould includestheassessmentofprocessesusedinthedesignandmanufactureofthe system,includingfactorscontributingtofailurecausesasfollows:
• Parts
• Design
• Manufacturing
• Systemmanagement
• Induced
• Wearout
• Nodefectfound
• Software
Theprobabilityoffailureisacomplexinteractionbetweenthedefectdensity, defectseverity,andstressesincurredinoperation.Failureratespredictedusing empiricalmodelsarethereforetypicalfailureratesandrepresenttypicaldefect rates,design,anduseconditions.
Table 2 containsthepredictedfailurerateofthevariousempiricalmethodologiesforadigitalcircuitboard.Thefailureratesinthistablewerecalculatedforeach combinationofenvironment,temperatureandstress.Thesedatashowthattherecan beimportantdifferencesbetweenthepredictedfailureratevalues,dependingonthe methodused.
Differencesareanticipatedbecauseeachmethodologyisbasedonunique assumptionsanddata.TheRACdatainthelastrowofthetableisbasedon observedcomponentfailureratesinagroundbenignapplication.
Furthermore,theempiricalmodelsareusuallydevelopedfromtheanalysisof fielddata,whichtaketimetocollect.Thefasterthegrowth,themoredifficultitisto deriveanaccuratemodel.
Asaresult,the1990smarkedtherebirthandwidespreaddevelopmentofthe PoFapproach.WhiletherootsofPoFmethodscanbefoundinfracturemechanics, whichbyitselfstartedthroughtheworkofAlanA.GriffithjustafterWorldWarI andwasdevelopedbyGeorgeR.Irwininthe1950s,andthePoFapproachstarted throughaseriesoffoursymposiafrom1962to1966,buttheconceptlostitsdrive untilthelate1980s.Enormousadvancementsincomputationaltoolsandfaster personalcomputersononesideandemerging,advancedtestingtechnologiesin materialscience,ontheotherside,acceleratedthistrend.Inthisapproach,facts fromrootcausephysical/chemicalfailureprocessesareusedtopreventthefailure oftheproductsbyrobustdesignandbettermanufacturingpractices.Becauseofthe competitiveenvironmentinproductionofconsumerproductsandlimitedbudget andresources,greatemphasiswasplacedoncapturingtheneededinformationmore quicklyandwithlesseffort.Asaresult,designandassessmentmethodologiesthat addresstherootcausesoffailurehaveemergedaspowerfulcostsavingtechniques.
Intheearly1990s,theUSArmyandAirForceinitiatedtworeliabilityphysics relatedprograms.In1992,theArmyauthorizedtheElectronicEquipmentPoF projectstopromoteamorescientificapproachtoreliabilityassessmentofelectronic equipment.
ThePoFapproachtoreliabilityutilizedscientifi cknowledgeofdegradation processesandtheloadprofi leappliedtoanitem,itsarchitecture,materialproperties andenvironmentalconditionstoidentifypotentialfailuremechanismsthatindividuallyorincombinationleadtotheitem’sfailure.ThePoFmodels,once developedandvalidated,wouldbeusedtoestimatelifeexpendedandexpected. UseofPoFreducedtheneedforasubstantialamountoflifedatatoarriveata reliabilitymodel,sincePoFemploystheavailablewell-developedknowledgeabout theprocessoffailure.Suchknowledgemodelshowandwhytheitemfailsand reducestheneedforlargequantitiesoflifedata.
Bythemiddleofthe1990s,criticismagainsttheapplicationofgenericdatain generalandMH-217,inparticular,becameincreasinglyintense,andthebasicidea ofusingfailureratedatagatheredinsuchdatabaseswasseriouslyquestioned. However,thecriticswhowantedtoabandonthedataprovidedinhandbooksfor beingirrelevantanduselessinmanyapplicationshaddifficultiesprovingthatthe PoFapproachcoulddoanybetterinreliabilitypredictions.Infact,mostofthePoF modelsstronglydependedonlifeortestdatainonewayoranother.Thequestion was,ifthereisenoughdataavailabletoevaluateaPoF-basedmodel,whynotuse thesamedataforstatisticalinferenceandtakethetraditionalfailureratemodelling pathagain?
Tohelpidentifythemannerinwhichreliabilitypredictionsareusedbyreliabilitypractitioners,asurveywasissuedattheendofthedecade,andapproximately 60non-DoDcompaniesresponded.Fromthesedata,the3predominantpurposes forperformingreliabilityassessments,inorderofimportancewerereportedas:
1.Determiningfeasibilityinachievingareliabilitygoalorrequirement.
2.Aidinginachievingareliabledesign,derating,componentselection,environmentalprecautions,inputtoFMEAandFaultTrees.
3.Predictingwarrantycostsandmaintenancesupportrequirements.
Surveyrespondentswerealsoaskedtoidentifythemethodologiestheyusewhen predictionsareperformed.MH-217wasreportedtobebyfarthemostuniversally appliedfailureratepredictionmethodology,althoughitsupdatedceasedin1991.
7The2000s InviewofthecriticsagainstMH-217andtheneedtoprovetheusefulnessofthe PoFapproachesasmentionedearlier,thereliabilitycommunitymovedtowardsan integrateduseofbothapproaches.WherethePoF-basedapproachcouldsavetime andmoneybyaddressingtherootcausesoffailureandreducetheburdenof gatheringasubstantialamountofdata,thetraditionalstatisticalfailureratescould beusefulinprobabilisticreliabilitypredictionsconsideringuncertaintiesinvolved. However,theuncertaintyboundswereoftensowideastomaketheresultalmost worthlessindecision-makingprocesses.Inordertobettermanageuncertaintyand makepracticalengineeringdecisions,twofactorsneededtobeconsidered.The fi rst importantelementwasindeedrelianceonmoredata,forwhichacceleratedlife testing,step-stresstesting,expertjudgmentandmanydifferentresourceswere exhausted.Thesecondelement,whichwasconsideredasimportantasthe first,was anappropriatecomputationalframeworkthatallowsnewdatatobeeasilyaddedto theanalysis.Theclassicalmaximumlikelihoodestimation(MLE)methodintroducedbyFisherwasoneofthepossiblechoices.
TheMLEmethodusedthelikelihoodfunctionoftheavailabledataforthe modelparameterestimation.Thismethodprovidednomeanstoincorporateprior knowledgeavailableforthemodelparameters.Thismethodwouldmathematically collapsewhennocompletefailuredatawasavailable.Thiswasalmostalwaysthe casewithnew,highlyreliablecomponentsandsystems.Theuncertaintybounds providedbythelocalFisherinformationmatrixarenotusefulwhendealingwith smallsamplesizes.
IncontrastwithMLEmethods,theBayesianapproachprovidedmanyuseful features,includingapowerfulmeanstoincorporatepriorknowledge,dealingwith thewholedistributionofthelikelihoodfunction,faircoverageofuncertainties,and finallythepossibilityofusingmanydifferentformsofdata(exact,censored,fuzzy, partiallyrelevantandexpertjudgments).Nevertheless,oneofthelimitingfactorsof theBayesianinferencemethodsinpracticalreliabilityanalysiswasthemathematicalcomplexityoftheproblem.
Multidimensionaljointdistributionsaregenerallyhardtodealwith.Laterinthe 2000s,thenumericalandcomputationaladvancementsinBayesianstatistical methods,suchasMarkovChainMonteCarlo(MCMC)simulationsandother
sampling-basedmethodologies,combinedwithadvancementsincomputational toolsanddevelopmentofpowerfulprogrammingplatforms,madetheBayesian inferencetechniquesacommonreliablepractice.
EmployingBayesiananalysis,reliabilitypractitionerscombineddifferenttypes ofdata,includingsimplefailureratesfromtraditionalhandbooks,engineering expertjudgments,simulatedresultsofsophisticatedPoFmodelsanddirecttest results,inahybridplatform.Withavailabilityoffastcomputing,hybridmethodologybecamewidelyavailableandpractical.Thesetechniquescouldrelyonthe physicaland,toalesserextent,chemicalphenomenathatdrivedegradationand failures.Alongwithsmall(accelerated)testsand fieldorexpertjudgmentdata,such hybridmodelsbecamethesourceofindustry-speci ficreliabilitydataandanalytical modelsneededtoassessthelifeandsafetyofhighlyreliableconsumerproductsand othercomplexengineeringsystemsinthe2000s.
Fromtheabovedescriptionofthehistoricalpathsofreliabilityengineering,we canseethattheadvancementofreliabilityengineeringisdrivenfromUSmilitary andtheweakreliabilityofelectronicscomponents.Overthedecade,reliability engineeringhasexpandedtocommercialworldoutsideUSandreliabilityof electroniccomponentsarealsogettingmuchhigher.Todaythetwoprimarypurposesforperformingaquantitativereliabilityassessmentofsystemsaretwofolds, namelyto
1.assessthecapabilityofthepartsanddesigntooperatereliablyinagiven application;
2.estimatethe ‘numberof fieldfailures ’ orthe ‘probabilityofmissionsuccess’
Fig.1 Historyofmethodsdevelopinginreliabilityengineering
EmergingPoFmodellingapproachesinreliabilityassesments
Purpose#1doesnotrequirestatisticallybaseddataormodels,and physics-of-failureapproacheshavemerit.
Purpose#2,requiresempiricaldataandmodelsderivedfromthosedata;thisis because fieldcomponentfailuresarepredominantlycausedbycomponentand manufacturingdefectswhichcanonlybequanti fiedthroughthestatisticalanalysis ofempiricaldata.
Forsystemacquisition,thereisclearlyatimelinesselement.Thepredictionmust beavailableearlyenoughtoinfluencethedesignand/orselectionofthedesignfor thesystem.Yet,theearlierthepredictionisneeded,thelessdetailedtechnical informationisavailabletosupportthepredictionitself.Also,thereisacost/bene fit
Fig.2
Another random document with no related content on Scribd:
The Project Gutenberg eBook of In the volcano's mouth; or, A boy against an army This ebook is for the use of anyone anywhere in the United States and most other parts of the world at no cost and with almost no restrictions whatsoever. You may copy it, give it away or re-use it under the terms of the Project Gutenberg License included with this ebook or online at www.gutenberg.org. If you are not located in the United States, you will have to check the laws of the country where you are located before using this eBook.
Title: In the volcano's mouth; or, A boy against an army
Author: Frank Sheridan
Release date: May 24, 2022 [eBook #68164]
Language: English
Original publication: United States: Street & Smith, 1890
Credits: Demian Katz, Craig Kirkwood, and the Online Distributed Proofreading Team at https://www.pgdp.net (Images courtesy of the Digital Library@Villanova University.) *** START OF THE PROJECT GUTENBERG EBOOK IN THE VOLCANO'S MOUTH; OR, A BOY AGAINST AN ARMY ***
Transcriber’s Notes: The Table of Contents was created by the transcriber and placed in the public domain.
Additional Transcriber’s Notes are at the end.
CONTENTS Chapter I. Madcap Max.
Chapter II. Emin Bey’s Escape.
Chapter III. In a Desert Tomb.
Chapter IV. Under the Pyramid.
Chapter V. Girzilla.
Chapter VI. Was It an Echo?
Chapter VII. Splendid Heroism.
Chapter VIII. Sherif El Habib.
Chapter IX. Ibrahim and Max.
Chapter X. The Petrified Forest.
Chapter XI. The Tribe of Klatch.
Chapter XII. “What Says Girzilla?”
Chapter XIII. Dangerous Jests.
Chapter XIV. The Subterranean River.
Chapter XV. In the Volcano’s Mouth.
Chapter XVI. Beyond Human Imagination.
Chapter XVII. The Rainmaker.
Chapter XVIII. Why Our Heroes Desert.
Chapter XIX. Mohammed.
Chapter XX. “Where Is Girzilla?”
Chapter XXI. The Mahdi.
Chapter XXII. Trick or Miracle.
Chapter XXIII. Under the Mahdi.
Chapter XXIV. Counting Chickens.
Chapter XXV. Victory.
Chapter XXVI. A Plan of Campaign.
Chapter XXVII. Sowing the Seed.
Chapter XXVIII. An Unexpected Bath.
Chapter XXIX. Saved!
Chapter XXX. The Mahdi’s Justice.
Chapter XXXI. Victory All Along the Line.
Chapter XXXII. “All’s Well That Ends Well.” BOUND-TO-WIN LIBRARY IN THE VOLCANO’S MOUTH BY FRANK SHERIDAN
STREET & SMITH · PUBLISHERS · NEW YORK THE BOUND TO WIN LIBRARY We called this new line of high-class copyrighted stories of adventure for boys by this name because we felt assured that it was “bound to win” its way into the heart of every true American lad. The stories are exceptionally bright, clean and interesting. The writers had the interest of our boys at heart when they wrote the stories, and have not failed to show what a pure-minded lad with courage and mettle can do. Remember, that these stories are copyrighted and cannot be had in any other series. We give herewith a list of those already published and those scheduled for publication.
To be Published During September 136—Spider and Stump By Bracebridge Hemyng
135—The Creature of the Pines By John De Morgan
134—In the Volcano’s Mouth By Frank Sheridan
133—Muscles of Steel By Weldon J. Cobb
To be Published During August
132—Home Base By Bracebridge Hemyng
131—The Jewel of Florida By Cornelius Shea
130—The Boys’ Revolt By Harrie Irving Hancock
129—The Mystic Isle By Fred Thorpe
128—With the Mad Mullah By Weldon J. Cobb
To be Published During July
127—A Humble Hero By John De Morgan
Sea Hemyng 104—Fatherless Bob By Bracebridge Hemyng
103—Hank the Hustler By Fred Thorpe
102—Dick Stanhope Afloat By Harrie Irving Hancock
101—The Golden Harpoon By Weldon J. Cobb
100—Mischievous Matt’s Pranks By Bracebridge Hemyng
99—Mischievous Matt By Bracebridge Hemyng
98—Bert Chipley By John De Morgan
97—Down-East Dave By Fred Thorpe
96—The Young Diplomat By Harrie Irving Hancock
95—The Fool of the Family By Bracebridge Hemyng
94—Slam, Bang & Co By Weldon J. Cobb
93—On the Road By Stanley Norris
92—The Blood-Red Hand By John De Morgan
91—The Diamond King By Cornelius Shea
90—The Double-Faced Mystery By Fred Thorpe
89—The Young Theatrical Manager By Stanley Norris
88—The Young West-Pointer By Harrie Irving Hancock
87—Held for Ransom By Weldon J. Cobb
86—Boot-Black Bob By John De Morgan
85—Engineer Tom By Cornelius Shea
84—The Mascot of Hoodooville By Fred Thorpe
In the Volcano’s Mouth OR A BOY AGAINST AN ARMY By FRANK SHERIDAN, author
of “Bert Fairfax,” “Through Flame to Fame,” “Life-Line Larry,” “Lion-Hearted Jack,” etc.
STREET AND SMITH, PUBLISHERS
79-89 SEVENTH AVENUE, NEW YORK
Copyright, 1890
By Norman L. Munro
In the Volcano’s Mouth IN THE VOLCANO’S MOUTH. “All aboard!”
CHAPTER I. MADCAP MAX. “All but passengers ashore.”
The loud, stentorian voices of the officers of the magnificent palace steamer L’Orient, of the Peninsular and Oriental Line, sounded all along the Southampton docks, up the streets to the old gates, and even penetrated into some of the business houses of the quaint old English town.
The shout, so commonplace to the citizens of Southampton, was one of serious import to those gathered on the deck of the steamer.
Parting is never pleasant, and when the journey is a long one, and it is known the absence is for years, the last words are always tearful. On the deck stood two men, alone.
Not one had come to bid them good-by or a godspeed on their journey.
And yet tears filled the eyes of both.
The elder was a bronzed veteran, his face as dark as that of any mulatto, his long, white mustache standing out in startling contrast to the color of his skin.
He was sixty years of age, but his strong body, his hard muscles, and firm walk, would rather betoken a man of forty.
By his side stood his son, a youth almost effeminate in appearance, but perhaps only because of the contrast to his father; there was a brightness in his eyes which betokens an active spirit, and although so effeminate-looking, when he clinched his hand one could see the strong muscle rising beneath the sleeve.
The elder man is Maximilian Gordon, of the mercantile firm of Gordon, Welter & Maxwell, of New York.
The son is Maximilian Gordon, also, but always called Max by those who are intimate with him, and “Madcap Max” by his closest companions.
Gordon, Welter & Maxwell were interested in Egyptian produce, and for many years Maximilian Gordon had been a resident of Alexandria.
His wife, sickly and delicate at all times, had been compelled to live in England, where young Max had been educated.
The elder man paid a yearly visit to his family, and had just completed arrangements for them to return to Egypt with him when cholera broke out, and he arrived home only just in time to close his wife’s eyes in death and see her body committed to its eternal resting place.
Hence it was that, as father and son looked at the English coast, which was by this time fast receding, their eyes were filled with tears, for they were leaving a plot of earth hallowed and sacred, because it was a wife’s and mother’s grave.
Youth is ever buoyant, and before the steamer had left the English Channel, Max was the happy, light-hearted lad once again, laughing, chatting and larking with everyone he came in contact with.
His father could not hide his grief so easily, but showed by his manner how nearly broken was his heart and ruined his life.
When the troubled waters of the Bay of Biscay were reached, Max had given plentiful evidence of his love of practical joking, and showed that he fully deserved his sobriquet of Madcap.
One of the passengers had on board an African monkey.
This little, frolicsome animal became very fond of Max, and was easily induced to adapt itself to the ways of the fun-loving youth.
One night Max took Jocko and dressed him in a lady’s nightcap, which he had obtained from a stewardess, and told Jocko he must
lie in a certain bed.
The stateroom was occupied by a snarling old bachelor, who declared that women and children were a nuisance.
When the old fellow entered his room he saw, to his utter astonishment, a head resting on his pillow.
Without staying to investigate, he rushed out of his room, shouting “Steward!” at the top of his voice.
“What is it, Mr. Lawrence?” asked the first officer, startled by the frantic shouting.
“Some one has placed a nigger baby in my bed.”
“Nonsense, Mr. Lawrence!”
“I say they have, and I’ll report every officer of the vessel if the offender is not punished.”
“I will see that the matter is investigated,” said Officer Tunley.
“Of course—but when? Why, in a week’s time, when everyone will have easily forgotten—no, sir, come at once.”
“I will do so; but allow me to suggest, Mr. Lawrence, that it may have been the extra bottle of Bass’ ale——”
“Do you dare, officer, to insinuate——”
“Nothing, save that Welsh rarebit, highly seasoned, and three bottles of strong ale, are likely to disturb the vision.”
“I’ll report you, sir—mark me, I’ll report you. Come, now, to my room, and if there is not a nigger baby there I’ll eat my hat.”
“Very well, sir, I will come with you.”
By the time the stateroom was reached, Jocko had fled the room, and Max had stripped the cap from its head.
The monkey sat on the table in the saloon, grinning, as if it enjoyed the joke.
The officer and Mr. Lawrence entered the stateroom.
“By Jove!” exclaimed Lawrence, as he looked at his bed.
“I was afraid you were romancing, sir,” said the officer, with proud indignation. “Take care, sir, that it does not occur again.”
The passenger was speechless.
Another day, when the steamer L’Orient was being tossed about in the most fantastic manner, sometimes taking a swift pitch forward, then curving and twisting in a way which would bring joy to the heart of a baseball pitcher, Madcap Max thought the time had come for a pleasant diversion.
A drove of pigs, with other animals, was on board, to enable the company to provide fresh meat for the passengers.
Max quietly released the pigs from their quarters, and saw them, with one accord, make for the saloon.
That was just what he wanted.
A lady was tossed off her bed to the floor, but to her horror she fell on the back of a pig, who set up such a squeaking and squealing that, although the passengers were feeling sick, they were compelled to laugh.
After a voyage of fourteen days the city of Alexandria was sighted.
“Thank goodness!” exclaimed an old Indian nabob. “I am glad I have to stay at Alexandria, for L’Orient is the worst disciplined ship I was ever in.”
The verdict was concurred in by nearly everyone on board.
And yet it was not the officers’ fault, for nine-tenths of the trouble was caused by the pranks of Madcap Max.
“Do we land here?” asked Max.
“Yes, Max. We shall finish our journey overland.”
“Our journey?” repeated Max, opening his bright eyes still wider with astonishment.
“Yes, Max. We go to Cairo before we settle down at Alexandria.”
“I am so glad.”
Several scores of boats surrounded L’Orient, manned by swarthy and not too-much dressed Arabs; a dozen or so seized upon Max and his father and literally dragged them to a boat.
On the way from the steamer to the landing dock, Mr. Gordon whispered to Max:
“No jokes with these fellows, or your life is not your own.”
“All right, dad; I’ll be as sober as a judge and as full of fun as an undertaker.”
“For your own sake be careful.”
“I will, dad. That is, as careful as I can be.”
