AcceptableQualityLevelAQL

AcceptableQualityLevel.AQL.AlsoknownasAssuredQualityLevel.
Thispublicationestablisheslotorbatchsamplingplansandproceduresforquality
inspectionbyattributes.Thesamplingplansdescribedinthisstandardareapplicableto
AQLof0.01%orhigherandarethereforenotsuitableforapplicationswherequalitylevel
inthedefectivepartspermillionrangecabberealized.

Theseplansareintendedprimarilytobeusedforacontinuingseriesoflotorbatch.The
planmayalsobeusedforthetheinspectionofisolatedlotorbatch.

MILSTD105EKeywordlisting

AcceptableQualityLevelAQL
Whenacontinuousseriesoflotsisconsidered,theAQListhequalitylevelwhich,forthepurposes
ofsamplinginspection,isthelimitofasatisfactoryprocessaverage.
AsamplingplanandanAQLarechoseninaccordancewiththeriskassumed.Useofthevalueof
AQLforcertaindefectorgroupofdefectsindicatesthatthesamplingplanwillacceptthegreat
majorityofthelotorbatchprovidedtheprocessaveragelevelofpercentagedefectiveintheselotor
batchbenogreaterthanthedesignatedvalueofAQL.

AverageoutgoingqualityAOQ
Foraparticularprocessaverage,theAOQistheaveragequalityofoutgoingproductincludingall
acceptedlotorbatch,plusallrejectedlotorbatchaftertherejectedlotorbatchhavebeen
effectively100%inspectedandalldefectivesreplacedbynondefectives.

ClassificationofDefects
Aclassificationofdefectsistheenumerationofpossibledefectsoftheunitofproductclassified
accordingtotheirseriousness.

Lotorbatch
Thetermlotorbatchshallmean"inspectionlot"or"Inspectionbatch".Thelotorbatchsizeisthe
numberofunitsofproductinalotorbatch.

ProcessAverage
Theprocessaverageistheaveragepercentdefectiveoraveragenumberofdefectsperhundred
unitsofproductcommittedbythesupplierfororiginalinspecting.Originalinspectionisthefirst
inspectionofaparticularquantityofproductasdistinguishedfromtheinspectionofproductwhich
hasbeenresubmittedafterpriorrejection.

Sample
Asampleconsistsofoneormoreunitsofproductdrawnfromalotorbatch,theunitofthesample
beingselectedatrandomwithoutregardtotheirquality.Thenumberofunitsofproductinthe
sampleisthesamplesize.

SamplingPlan
Asamplingplanindicatesthenumberofunitofproductfromeachlotorbatchwhicharetobe
insppectedandthecriteriafordeterminingtheacceptabilityofthelotorbatch.

Unitofproduct
Theunitofproductisthethinginspectedinordertodetermineitsclassicationasdefectiveor
nondefectiveortocountthenumberofdefects.Itmaybeasinglearticle,apair,aset,alengthor
theendproductitself.Theunitofproductmayormaynotbethesameastheunitofpurchase,
supplyorshipment.

Samplingplan,LevelI,LevelII,LevelIII
Samplingplans
Inspectionlevel
Theinspectionleveldeterminestherelationshipbetweenthelotorbatchsizeandthe
samplesize.Theinspectionleveltobeusedforanyparticularrequirementwillbeas
prescribedbythecontractor'swrittenprocedure.
Threeinspectionlevel:I,II,IIIaregivenintable1forgeneraluse.Normally,inspector
levelIIisused.However,inspectionlevelImaybeusedwhenlessdiscriminationis
needed,orlevelIIImaybeusedforgreaterdiscrimination.Fouradditionalspeciallevels:
S1.S2,S3,S4aregiveninthesametableandmaybeusedwhererelativelysmall
samplesizesarenecessaryandlargesamplingriskscanormustbetolerated.
Codeletters
Samplesizesaredesignatedbycodeletters,Table1shallbeusetofindtheapplicablecodeletter
fortheparticularlotorbatchsizeandtheprescribedinspectionlevel.

Obtainingsamplingplan.
TheAQLandthecodelettershallbeusedtoobtainthesamplingplanfrommastertable,Whenno
samplingplanisavailableforagivencombinationofAQLandcodeletter,thetablesdirecttheuser
todifferenceletter.Thesamplingsizetobeusedisgivenbythenewcodeletter,notbytheoriginal
letter.Iftheprocedureleadstodifferentsamplesizesfordifferenceclassesofdefects,thecode
lettercorrespondingtothelargestsamplesizederivedmaybeusedforallclassesofdefects,Asan
alternativetoasinglesamplingplanwithanacceptancenumberof0,theplanwithanacceptance
numberof1withitscorrespondinglylargersamplesizeforadesignatedAQL,maybeused.

Table1.samplesizecodeletters

Specialinspectionlevel Generalinspectionlevel
LotSize
S1 S2 S3 S4 I II III
2to8 A A A A A A B
9to15 A A A A A B C
16to25 A A B B B C D
26to50 A B B C C D E
51to90 B B C C C E F
91to150 B B C D D F G
151to280 B C D E E G H
281to500 B C D E F H J
501to1200 C C E F G J K
1201to3200 C D E G H K L
3201to10000 C D F G J L M
10001to35000 C D F H K M N
35001to150000 D E G J L N P
150001to500000 D E G J M P Q
500001over D E H K N Q R

PECTIONLEVELGeneralLevelIII
SAMPLINGPLANSingleTightened
SamplingSizesandRejectLevels
AcceptableQualityLevel(AQL)

Lotsize Codeletter Samplesize 0.65 1.0 1.5 2.5 4.0 6.5 10 15

2to8 A 2
9to15 B 3 1 V
16to25 C 5 1 V 2
26to50 D 8 1 V 2 3
51to90 E 13 1 V 2 3 4
91to150 F 20 1 V 2 3 4 6
151to280 G 32 1 V 2 3 4 6 9
281to500 H 50 V 2 3 4 6 9 13
501to1200 J 80 V 2 3 4 6 9 13 19
1201to3200 K 125 2 3 4 6 9 13 19 ^
3201to10000 L 200 3 4 6 9 13 19 ^
10001to35000 M 315 4 6 9 13 19 ^
35001to15000 N 500 6 9 13 19 ^
15001to500000 P 800 9 13 19 ^
500001over Q 1250 13 19 ^

INSPECTIONLEVELGeneralLevelI
SAMPLINGPLANSingleReduced
SamplingSizesandRejectLevels
AcceptableQualityLevel(AQL)
 Lotsize Codeletter Samplesize 0.65 1.0 1.5 2.5 4.0 6.5 10 15
2to8 A 2 1 ^ V
9to15 B 3 1 ^ V 2
16to25 C 5 1 ^ V 2 3
26to50 D 8 1 ^ V 2 3 4
51to90 E 13 1 ^ V 2 3 4 5
91to150 F 20 1 ^ V 2 3 4 5 6
151to280 G 32 ^ V 2 3 4 5 6 8
281to500 H 50 V 2 3 4 5 6 8 10
501to1200 J 80 2 3 4 513 6 8 10
1201to3200 K 125 3 4 5 6 8 10 13 ^
3201to10000 L 200 4 5 6 8 10 13 ^
10001to35000 M 315 5 6 8 10 13 ^
35001to15000 N 500 6 8 10 13 ^
15001to500000 P 800 8 10 13 ^
500001over Q 1250 10 13 ^

\/=usefirstsamplingplanbelowarrow
/\=usefirstsamplingplanabovearrow

AQLAcceptableQualityLevel

TheAQLofasamplingplanisalevelofqualityroutinelyacceptedbythesamplingplan.Itis
generallydefinedasthepercentdefective(defectivesperhundredunitsX100%)thatthe
samplingplanwillaccept95%ofthetime.ThismeanslotsatorbetterthantheAQLare
acceptedatleast95%ofthetimeandrejectedatmost5%ofthetime.

TheAQLcanbedeterminedusingtheoperatingcharacteristic(OC)curvebyfindingthatquality
levelonthebottomaxisthatcorrespondstoaprobabilityofacceptanceof0.95(95%)onthe
 leftaxis.

AssociatedwiththeAQLisaconfidencestatementonecanmake.Ifthelotpassesthe
samplingplan,onecanstatewith95%confidencethatthequalitylevelofthelotisequaltoor
betterthantheAQL(i.e.,thedefectiverateofthelot<AQL).Ontheotherhand,ifthelotfails
thesamplingplan,onecanstatewith95%confidencethatthequalitylevelofthelotisworse
thantheAQL.

TheAQLdescribeswhatthesamplingplanwillaccept,butitisalsoimportanttoknowwhatthe
samplingplanwillreject.Informationonwhatthesamplingplanwillrejectisprovidedbythe
LTPDofthesamplingplan.

Table2.AQLSamplingTablebasedontheMilSTD105D
AcceptableQualityLevelsforNormalInspection(%defective)
LotSize/
.01 .015 .025 .04 .065 .10 .15 .25 .40 .65 1.0 1.5 2.5 4.0
SampleSize
LotAcceptance(Ac)Number(maximumnumberofrejectstoacceptthelot)
LS=2to8
0 0 0 0 0 0 0 0 0 0 0 0 0 0
SS=2
LS=9to15
0 0 0 0 0 0 0 0 0 0 0 0 0 0
SS=3
LS=16to25
0 0 0 0 0 0 0 0 0 0 0 0 0 0
SS=5
LS=26to50
0 0 0 0 0 0 0 0 0 0 0 0 1 1
SS=8
LS=51to90
0 0 0 0 0 0 0 0 0 0 0 0 1 1
SS=13
91to150
0 0 0 0 0 0 0 0 0 0 0 0 1 2
SS=20
151to280
0 0 0 0 0 0 0 0 0 0 0 1 2 3
SS=32
281to500
0 0 0 0 0 0 0 0 0 0 1 2 3 4
SS=50
501to1200
0 0 0 0 0 0 0 0 0 1 2 3 5 7
SS=80
1201to3.2K
0 0 0 0 0 0 0 0 1 2 3 5 7 10
SS=125
3201to10K
0 0 0 0 0 0 1 1 2 3 5 7 10 14
SS=200
1000135K
0 0 0 0 0 0 1 2 3 5 7 10 14 21
SS=315
35001150K
0 0 0 1 1 1 2 3 5 7 10 14 21
SS=500
150001500K
0 0 0 1 1 2 3 5 7 10 14 21 21
SS=800
.01 .015 .025 .04 .065 .10 .15 .25 .40 .65 1.0 1.5 2.5 4.0