SPC统计过程控制(英文版).pptx
《SPC统计过程控制(英文版).pptx》由会员分享,可在线阅读,更多相关《SPC统计过程控制(英文版).pptx(23页珍藏版)》请在冰点文库上搜索。
StatisticalProcessControl,StatisticalProcessControl,Variabilityexistsineverything-people,machinesandnaturePurpose-tobesurethatprocessesareperforminginanacceptablemannermonitorovertimeforecastifwillstillbewellinfuturedeterminefactorsthatmakeitoutofcontrol,StatisticalProcessControl,Designqualitychecksintotheprocess-eliminateneedforinspectionbuildbetterproductduringdesignmakeproductwithstandvariation-TaguchiMethod,History,PriortoIRcraftsmenwereresponsibleforentireprocess-prideofworkmanshipDuring1920sW.ShewartofBellLabsbuildControlCharts-ASQformedDuring1950sDemingintroducedQCtechniquestoJapanese1961FeigenbaumwroteTotalQualityControl,StatisticalProcessControl,TwomainsubjectsAcceptanceSampling-testrandomsamplesoffinishedgoodsandacceptorrejectwholelotProcessControl-sampleprocessoutputatdifferentpointsintime-seeifprocessisincontrol-determinecauseofout-of-controlbehavior,StatisticalProcessControl,Monitortheproductionprocessinoneofthreeplaces-before,duringorafterPhysicallyexaminesomeoftheunitsBasicissues;wheredoweinspecthowmuchandhowoftendoweinspectdoweinspectvariablesorattributesBalancetwocosts:
inspectionandservice,Rulesofthumbforwheretoinspect;,1.Inputs-donotpayforgoodsthatdonotmeetstandards2.FinishedGoods-protectyourreputation3.Beforeacostlyoperation-donotwastetimeandlabor4.Beforeanirreversibleoperation-reworkonlyuptoacertainpoint5.Beforeacoveringoperation-paintingorassembly,TypesofSampling,Defn:
AttributeSampling-checkforthepresenceorabsenceofaproductcharacteristicDefn:
VariableSampling-productcharacteristicsthatcanbemeasuredonacontinuousscale,StatisticalProcessControl,usedtoevaluatetheoutputofaprocessmostprocessesexhibitsomekindofvariationrandomvariation-naturalvariation,uncontrollableassignablevariation-canidentifycausedeterminewhethernon-randomsourcesofvariationareresponsibleforproblems,ControlCharts,Preventproductionofexcessivenumberofdefects-lookforcausesofvariabilityAtime-orderedplotofsamplestatisticsControllimits-determinethedifferencebetweenrandomandassignablevariation-usually3Eachnewsamplestatisticisplottedandcomparedtothecontrollimits,StatisticalProcessControl,TypesofControlCharts,1.VariableControlchartsA.MeansChart-monitorscentraltendencyB.RangeChart-monitorsdispersion-bothcanbeusedtomonitorsameprocess-use20-25samplestobuildinitiallimits-takesamplesof5-9observations-reasonableamountoftimeandcostsless,TypesofControlCharts(cont),2.AttributesControlChartsA.P-Chart-countnumberofdefectsinsampleB.C-Chart-countnumberofdefectsperunit-buildingcontrolchartsaddstothetimeandcostofproduction-attributesamplingiseasierthanvariable-needlargersamplesizes,ProcessCapability,SpecifyrangesforcertainproductdimensionsduringdesignTolerancesMaxandminacceptablelevelofoutputBearingforrotatingshaft-diameteris1.25inches0.005Processhas=0.02inches-a3-controlchartlimitsarewiderthantolerances,CapabilityIndex,ShowshowwellpartsfitintotherangespecifiedbydesignlimitsIfdesignlimitsarewiderthanprocesslimits,theprocesscanshiftandstillproducegoodpartsCpk=,CapabilityIndex,Example-lightbulbsareproducedwithanaveragelifeof90hours.Thetargetdesignlifeis100hours.TheUTL=120andtheLTL=80hours.If=4.8whatistheCpk?
4,AcceptanceSampling,PurposesDeterminepolicyforacceptingorrejectingaproductionlotofmaterialsorcomponentsSamplequalityisusedtomakedecisionAdvantagesLesshandlingdamageFewerinspectorsApplicabilitytodestructivetestingEntirelotrejection(motivationforimprovement),5,AcceptanceSampling,DisadvantagesRisksofaccepting“bad”lotsandrejecting“good”lotsAddedplanninganddocumentationSampleprovideslessinformationthan100-percentinspection,AcceptancePlans,1.Single-SamplePlans-onlyonesampledefinenandcOperatingCharacteristicCurve-graphstheperformanceofplanAcceptableQualityLevel-definesgoodlotsLotTolerancePercentDefective-definesbadlotsAverageOutgoingQuality-thepercentdefectsinlotsleavinginspection,AcceptancePlans(cont),-ProducersRisk()-probabilityofrejectinggoodlot-ConsumersRisk()-probabilityofacceptingabadlot-canreduceerrorsbynotbiasingthesampleandtakingalargerone-usethePoissondistributiontocomputeprobabilityofacceptance,9,OperatingCharacteristicCurve,AcceptancePlans,Plansconsiderstheinteractionamongtheabovecriteria-itconsiderstheobjectivesofthesupplierandconsumerTheproducerwouldlikelowprobabilityofrejectingagoodlotandconsumerwantstonotacceptabadlotHPandJapanesevendor,AcceptancePlans(cont),2.Double-SamplePlansA.DrawonesampleB.ifnumberofdefectsislessthanalowerlimit,acceptlotifnumberofdefectsisgreaterthansomeupperlimitrejectlotC.ifnumberofdefectsisbetweenthelimits,takeanothersample,