關(guān)于FMEA的培訓(xùn)資料

1、Failure Mode & Effect AnalysisExplanation and Use of FMEAWhat is FMEA?FMEA is aproblemsolving processadoptedbySnap-on as akey componentoftheQualityForward System.Generally, FMEAisdonebyanon-site multifunctionalteam facilitated locallyorbyanSEQ Group Advanced QualityEngineer.Objectiveistoidentifyall

2、of thereasons whyquality of aprocessorproduct doesnot meetexpectations.Uses severity,occurrence anddetectionratingstodeterminecorrectiveactionpriorityfor eachpossiblefailurereason.When shoulda FMEAstudybeconducted?Aspart of aDesign/Development phase.When correctiveaction is neededduetounfavorableper

3、formancedata (performancebasedonQFSmetricsdoes notmeet goals).When aneedtomake incremental improvement is identifiedbyfacilitymanagementortheSEQ Group.FMEA TeamStart-UpTeam Selection donebylocalqualitycoordinator,businessunitmanager or,onrequest,SEQGroupAdvancedQualityEngineer.Problem Statement Acon

4、ciseandfocuseddescriptionofthequalityproblem or opportunity forimprovement,“whatwentwrong”or“whatneedsfixing”.Goal Statement adescriptionofwhatistobeaccomplishedbythe team, in quantifiableterms“reduce by”,“increase by”,“eliminate”, etc.Goodmetricstouseincludescrapcost,reworkcost,orDefectsPerMillionO

5、pportunities (DPMO),etc.For example, agoalstatementcouldbe“To reduceDPMO on productfamily Xfrom3000 to 2000byyearend”.CompleteFMEA TeamStart-UpWorksheet (example locatedinprocedure SEQ80.10,Figure 1).Problem Statement& GoalStatementProblem StatementandGoalStatementshould consider:What wentwrong?What

6、 is thegapbetweenthedesiredquality level andtheactual qualitylevel?Whichprocess or processesareinvolved?Whichdepartment or cellisthe deficiencyoccurring in?Istheproblemcyclic, canitbeattributed to specific events?What aretheappropriatemetricstomeasure?What is theimpactonoperationalfitnessandprofitab

7、ility?What is theimpactonthe bottomline (reduced,increased, eliminated)?Arethereconstraintsrecognizedearlyintheprocess?If so,have theybeenexploredandclarified withthe appropriate managementlevelFMEA ProcessFMEAsuncoverprocess problems relatedtothe manufacture of product.Examples:apieceofautomated as

8、sembly equipmentmaymisfeedpartsresultinginproductsnot being assembledcorrectly, or,ina chemical manufacturingprocess,temperatureormixing timecouldbesourcesofpotentialfailuresresulting in unusable product.Thinkintermsofthefiveelementsofa process:people, materials,equipment,methods andenvironment.With

9、thesefiveelementsinmind,ask“Howcanprocessfailure affecttheproduct,processingefficiency or safety?”Nine steps in FMEAprocess.Step #1 -Review theProcessEnsurethateveryoneontheFMEAteam hasthesameunderstanding of theprocess being workedon.Reviewblueprint/engineering drawingofproduct(ifconductingFMEA on

10、product) and/ora detailed flowchartoftheprocess.If these arenotavailable,theteamwill needtocreate one.Usingtheblueprint and/orflowchart, teammembersshouldfamiliarizethemselveswith theproduct or process.They shouldphysically seetheproductora prototypeoftheproduct,and physicallywalkthrough theprocess

11、exactlyasthe processflows.Itishelpful to havean“expert”onthe productorprocessavailabletoanswer anyquestionsthe teammighthave.Step #2 BrainstormPotentialFailureModesMethodofgeneratinganddocumentingideasofpotential failuremodes.Ifproduct or processiscomplex,aseriesofbrainstormingsessions,eachfocused o

12、n adifferent element(i.e.people,methods,equipment, materials,environment),maybeneeded to generate amorethoroughlist of potentialfailure modes.Facilitatorshouldwritedownallideasgenerated in brainstormingsessions.Once brainstormingiscomplete, ideas shouldbegrouped intolikecategories (asdetermined by t

13、heteam).Groupscanbetypeoffailure,whereontheproductfailure occurred,estimatedseriousnessofthe failure, etc.Failuremodesmay alsobecombinediftheyarethe sameorverysimilar.Once failuremodeshavebeen grouped/combined, theyshould be transferred to thePotentialFailureMode andEffectAnalysisSheet(exampleinSEQ

14、80.10,Figure2).Step #2 (continued) BrainstormingGuidelinesDonotmakeanyjudgmentonideas.Team membersshould notcomment on whetheranideaisgood or bad.Thinkingshouldbeunconventional, imaginative,oreven outrageous.Aimfor alargenumber of ideas in theshortestpossibletime.Team membersshould “hitchhike” on ot

15、her ideas,byexpandingthem,modifying them, or producingnewonesbyassociation.Allteammembers shouldparticipateinbrainstorming process.Step #3 ListPotentialEffectsofEach FailureModeAfterthefailuremodesarelisted on theFMEA worksheetform,theFMEAteam needs to revieweach failuremodeandidentifythe potentiale

16、ffects of thefailure shoulditoccur.Forsomefailure modes,theremaybeonlyoneeffect,whileotherfailure modes mayhave severaleffects.This stepmustbethorough thisinformationwilldetermineassignmentofrisk ratingsfor eachofthe failures.Itishelpful to think of thisstepasanif-thenprocess:Ifthefailureoccurs,then

17、what aretheconsequences.Step #4 Assign Severity RatingTheseverityrating is an estimate of howserious theeffects would be if agivenfailuredidoccur.Maybebasedonpastexperience,oranestimatebasedontheknowledge andexpertiseofthe teammembers.Because eachfailuremayhaveseveral differenteffects,andeacheffectm

18、ay haveadifferentlevelofseverity, theeffect, notthefailure,mustberated.Eacheffect shouldbegivenitsown severity rating.Ratings based on 10-point scale,with 1beingthe lowestratingand 10 being thehighest.General exampleonnextpage may be customizedbyteamfortheirspecificFMEAproject.Step #4 continued Seve

19、rity RatingScaleExampleRatingDescriptionDefinition10Dangerously HighFailure could injure the customer or an employee9Extremely HighFailure would create noncompliance with federal regulations8Very HighFailure renders the unit inoperable or unfit for use7HighFailure causes a high degree of customer di

20、ssatisfaction6ModerateFailure results in a subsystem or partial malfunction of the product5LowFailure creates enough of a performance loss to cause the customer to complain4Very LowFailure can be overcome with modifications to the customers process or product, but there is minor performance loss3Min

21、orFailure would create a minor nuisance to the customer, but the customer can overcome it in the process or product without performance loss.2Very MinorFailure would not be readily apparent to the customer, but would have minor effects on the customers process or product1NoneFailure would not be not

22、iceable to the customer and would not affect the customers process or product.Step #5 Assign OccurrenceRatingTheoccurrenceratingisameasure of howoftenanfailure modemay occur.Best methodfordetermine occurrencerating is to useactualdatafrom theprocess (forexample,failurelogs or processcapabilitydata).

23、Ifactualdataisnotavailable,theteamneedstoestimatehowoftenafailure modecouldoccur.Team canmake abetter estimate of howlikelyafailure modeistooccurand at whatfrequency by knowingthe potentialcausesoffailure.Ratings based on 10-point scale,with 1beingthe lowestratingand 10 being thehighest.General exam

24、pleonnextpage may be customizedbyteamfortheirspecificFMEAproject.Step #5 continued OccurrenceRating Scale ExampleRatingDescriptionPotential Failure Rate10Very High Failure is almost inevitableMore than one occurrence per day or a probability of more than three occurrences in 10 events (Cpk 0.33).9On

25、e occurrence every three to four days or a probability of three occurrences in 10 events (Cpk approximately equal to 0.33).8High Repeated failuresOne occurrence per week or a probability of 5 occurrences in 100 events (Cpk approximately equal to 0.67).7One occurrence every month or one occurrence in

26、 100 events (Cpk approximately equal to 0.83)6Moderate Occasional failuresOne occurrence every three months or three occurrences in 1,000 events (Cpk approximately equal to 1.00)5One occurrence every six months to one year or one occurrence in 10,000 events (Cpk approximately equal to 1.17)4One occu

27、rrence per year or six occurrences in 100,000 events (Cpk approximately equal to 1.33)3Low Relatively few failuresOne occurrence every one to three years or six occurrences in 10 million events (Cpk approximately equal to 1.67)2One occurrence every three to five years or two occurrences in one billi

28、on events (Cpk approximately equal to 2.00)1Remote Failure is unlikelyOne occurrence in greater than five years or less than two occurrences in one billion events (Cpk greater than 2.00)Step #6 Assign DetectionRatingThedetection ratinglooksathowlikely it is thatafailure or theeffectofafailure willbe

29、detected.Startbyidentifyingcurrent controls thatmay detecta failureoreffect of afailure(example SPCorin-process checks).More controls in place generatesa lownumber(better)rating;nocurrent controls means thelikelihood of detectionislowand thereforetheitemwouldreceive ahighnumber(worse)rating.Ratings

30、based on 10-point scale,with 1beingthe lowestratingand 10 being thehighest.General exampleonnextpage may be customizedbyteamfortheirspecificFMEAproject.Step #6 continued DetectionRatingScaleExampleRatingDescriptionDefinition10Absolute UncertaintyThe product is not inspected or the defect caused by f

31、ailure is not detectable.9Very RemoteProduct is sampled, inspected and released based on Acceptable Quality Level (AQL) sampling plans.8RemoteProduct is accepted based on new defectives in a sample.7Very LowProduct is 100% manually inspected in the process.6LowProduct is 100% manually inspected usin

32、g go/no go or other mistake-proofing gages5ModerateSome Statistical Process Control (SPC) is used in process and product is final inspected off-line4Moderately HighSPC is used and there is immediate reaction to out-of-control conditions3HighAn effective SPC program is in place with process capabilit

33、ies (Cpk) greater than 1.332Very HighAll product is 100% automatically inspected.1Almost CertainThe defect is obvious or there is 100% automatic inspection with regular calibration and preventive maintenance of the inspection equipment.Step #7 Calculate RiskPriorityNumberTheriskprioritynumber(RPN)is

34、calculatedbymultiplyingtheseverityrating times theoccurrence ratingtimesthedetection ratingforall of theitems.RPN=SeverityX OccurrenceXDetectionTotalRPNiscalculatedbyaddingall of theRPNnumbers.Thisnumber servesasa rankingtooltoprioritize which FailureModesare to be consideredaconcern andthereforeshould be addressedfirst.Step #8 PrioritizetheFailureModesforActionOnce theRPNiscalculated,the failuremodescan be prioritized by rankingtheminorderfrom t