3. Process Capability study

ProcessCapabilityStudiesProcessCapabilityStudyFlow.DataAnalysisLogicalSequenceforProcessCapabilityStudyDataAnalysisFlowIstheprocessinstatisticalcontrolInStatisticalcontrolRecentertheprocessWorkontheprocesstobringitintostatisticalcontrolbyremovingundesirablespecialcauses,onebyoneConsiderthemanagementalternatives:changethespecs,changetheprocess;livewithitoroptimizethemean;stopmakingtheproductYESYESNONOStep1Step2Step6Step5Step3Step4Istheprocesscapable?CollectionofdataCAStandardizationCalculationofcontrollimitsAnalysisandinterpretationofpatternsApplicationofCACAEffectivenessVerificationLettheprocesstellyouaboutitself.Usecontrolchart.Otherunnaturalpatterns:1.Cycles-Shorttrendsinrepeatedpatterns-Specialcausesthatoccurinamoreorlessregularbasis051015202530147101316192225XXFig.163.Cycles.Examplecauses:Xbarchart•Regulardifferencebetweendayandnightshifts.•Rotationofpeopleonjob•VoltagefluctuationRchart•Maintenanceschedules•Rotationoffixturesandgages•Wearoftoolordiep-chart•Sortingpractices•Samplingpractices•Regulardifferencesbetweensuppliers.2.Freaks→Singleunit/measurementgreatlydifferentfromtheothers-5-3-1135791113151719212325272931333537394143454749515355147101316192225XXXFig.165.FreaksExamplecauses:XbarandRchart•Errorinmeasurement•Incompleteoperation•Omittedoperation•Breakdownoffacilities.p-chart•Variationsinsamplesize•Occasionalverygoodorverybadlot.Otherunnaturalpatterns:3.GradualChangeinLevel→trend”thatlateronsettlesatsomenewlevel-5051015202530147101316192225XXXXXFig.170.Gradualchangeinlevel.XXXXXXXExamplecauses:Xbarchart•Gradualintroductionofnewmaterial•bettersupervision•greaterskillonthepartoftheoperatorRchart•Betterfixturesmethods•worseningoffixturesormethodsp-chart•Relaxationortighteningofstandards.Otherunnaturalpatterns:4.“Grouping”or“Bunching”→clustertogetherinanonrandomfashion-10-505101520253014710131619222528XXXXFig.172.GroupingorBunchingExamplecauses:Xbarchart•Changeinthecalibrationofatestsetormeasuringinstrument•DifferentpersonmakingthemeasurementsRchart•Mixtureofdistributions•Freaksinthedatap-chart•Changeintechniqueofclassification•changesinassortmentofproduct.Otherunnaturalpatterns:5.Instability→Unnaturallylargefluctuations→Pointsoutsidethecontrollimits0510152025301471013161922XXXXXXXXXXFig.174.InstabilityonanXchart.Examplecauses:Xbarchart•Overadjustmentofamachine•mixtureofprocessvariables•Fixturesorholdersnotholdingtheworkinposition•ErraticbehaviorofautomaticcontrolsRchart•Toomuchplayinpositioningorholdingfixture•Problemwithtestset•Machineinneedofrepair•Workholderdistortedp-chart•Variationsinsamplesize•Operatorcarelessness•ImpropercheckingOtherunnaturalpatterns:6.Mixtures→Pointsavoidcenterline→Toomanypointsnearthecontrollimits0510152025301357911131517XXXXXXXXXXFig.178MixtureFig.180SeparateclusterofpointsExamplecauses:XbarandRchart•Consistentdifferencesinmaterials•Differenceintestsetsorgagesp-chart•Non-randomsamplingtechnique•Differenceincheckers,testsets,etc.Otherunnaturalpatterns:7.Stratification→Unnaturallysmallfluctuations→Nopointsnearcontrollimits051015202530147101316192225XXXXXFig.187StratificationChangeinthedegreeofstratification.Examplecauses:Xbarchart•Incorrectcalculationofcontrollimits•Displacingofdecimalpoints.Rchart•Consistentdifferencesinmaterials,testsets,etc.p-chart•Non-randomsamplingtechnique.Ex.Selectingsamplesbytakingonepartfromeachoperator,oronepartfromeachmachine,etc.Otherunnaturalpatterns:8.SuddenShiftinLevel→Changeinthedirectionofchart05101520147101316192225XXXXXXXXXXXFig.189SuddenshiftinlevelExamplecauses:Xbarchart•Newoperator,machine,machinesetting,material,etc.Rchart•Changeinmotivationofoperators•changeinprocedures/method.p-chart•Tighteningoradditionofrequirements.Otherunnaturalpatterns:9.SystematicVariables→Patternispredictable→High,low,high,low~“seesaw”effect024681012141618202224135791113Fig.192SystematicVariableExamplecauses:XbarandRchart•Differencebetweenshifts,testsets,assemblylines,etc.,•Systematicmannerofdividingordrawingsamplesfromdifferentsources.p-chart•Systematicmannerofdividingordrawingsamplesfromdifferentsources.Otherunnaturalpatterns:10.Trends→Continuousmovementupordown→Seriesofpointsplottedwithoutachangeofdirection02468101214161820222426283032147101316192225XXXXXXXXXXFig.196TrendExamplecauses:Xbarchart•Toolwear•wearofthreads,holdingdevicesorgages.Rchart•Inadequatemaintenanceontestsetsormachineparts•Gradualimprovementorworseningofoperators,machineparts,etc.p-chart•Operatorfatigue•RelaxationofstandardsOtherunnaturalpatterns:1.CapabilityStudy–CTLWYKODshiftDshiftDshiftEshiftEshiftEshiftPresenceofcyclesintheRangechartRChartMaintenanceschedules.Operatorfatigue.Rotationoffixturesorgages.Regulardifferencebetweendayandnightshifts.Wearortoolordie(causingexcessiveplay).Toolinneedofsharpening(causingburrs,etc.).Measuringinstrumentusedtomeasuretheprofile,(WYKO#0)isMANUALLYOPERATEDOperatorsareallowedtousebetween2~3fringesduringmeasurement(PIC-155E1(6.7.1.3.3),Dshiftisusing3FringeswhileEshiftuses2Fringes3fringesmethodgivessignificantlydifferentTCvaluereadingscomparedto2fring