AS 1199.2 pdf download – Sampling procedures for inspection by attributes Part 2: Sampling plans indexed by limiting quality (LQ) for isolated lot inspection
Scope and field of application1.1 Scope
This part of lSO 2859 establishes LQ sampling plans and pro-cedures for Inspectlon by attributes compatible with IS0 2859/1that can be used when the switching rules given in iSo2859/1are not applied, for example, when lots are of an isolatednature.
The plans in this part of ISO 2859 are irdexed by a preferredseries of limiting qualities (LQ) with a consumer’s risk usuallybelow 10 %, but always below 13 %. This methnd of indexingpesnits the “ad-hoc” procedure 1/ mentioned in sub-clause12.6 of ISO 2859/1 to be implementod as a standard procedure.NOTE – The plans in iSo 2859/1 are indexed by a preferred series ofAQL values and inspection lovele.During the inepection of a continu-ing series of lots the application of switching rules helps to ensure thatthe proceos average in that acrics of lots is kept below the specifiedAoL. Limiting qualitv does not have the same direct relationship withthe proccas avorage lsce 3.5.1).
1.2 Field of application
Alternative procedures are provided in this part of ISO 2859 tocater for two situations often met in practice:
a)Procedure A, to be used when the supplicr and consumer both wish to regard the lot in isalatinn. The tahles arebased on random sampling fron finite lots for both con-eumer and produccr risks. This procedure shall be usedunless there is a specific instruction to use procedure B.b)Procedure B, to be used when the supplier regards thelot as one of continuing series but the consumer considersthe lot recelved in isolation.The tables are based on randomsampling from a finite lot for consumer’s risk at the limitingquality, but random campling from a prococc for tho producer’s risk and the tabulated values of tha operatingcharacteristic (OC)curves.The plans used are a selection ofthe plane available in Iso 2859/1 eo that a produccr canmaintain consistent procedures for culstomers whether ornot they receive individual lots or a continulng series of lots.This procedurc is suitablc for inclusion in product standardsor snecifications thatcontain samplingclauses. Themanufacturer will be concerned with all of his production.but the individual consumer only with the particular lotreceived.
The diefinitions given in lSO 2R59/1 shall apply.3Choice of sampling plan
3.1Specifying a sampling plan
when reference is made to this part of IS0 2858 in a contract orspecification,it is necessary to specify the procedure to beused (i.e. either A or B). In the abeence of any refercncc toeither procedure, then procedure A shall he used.
3.2Choico of sampling procedure
Although procedure A is based on the hypergeometric distribu-tion for sampling results, this distribution is well approximatodby the binomial distribution for plans with non-zero acceptancenumbers in procedure A. Hence the oc curves for these plansare well approximated by thc tabulatcd pcrformance of thesame plans in procedure R. However,procedure A uses planswith acceptance number zero and wlth sample sizes based onthe hypergcomctric distribution of sampling rcaults whilc pro-cedure Bexcludes accept zero plans，replacing them with100 % inspection.
The choice botween the procedures is dominated by the at-titude towards accept zero plans. Procedure A uses accept zeroplans linking the sarmplesize and the lot size by thehypergcometric distribution until thc proposcd plon cor-responds to the accept zero plan in Iso 2859/1 correspondingto that limiting quality.Thereafter the progresslon of samplingsize and acceptance number with lot size is based on the in-spection levels available in ISO 2859/1.The overall effect ofprocedure A resembles inspection level for limiting qualityless than 8 % and level ! for limiting quality greater than 8 %6with 8 % being intermediate between these levels.
Procedure B provides greater flexiblllty in the cholce of inspec-tion levels. The tabulated oc curves are based on the prob-ability of producing a nonconforming unit so they are approxi-mately correct for small sampling fractions but as the samplingfraction increases the curves (and tables) underestimate theprobability of acceptance for good quality and underestimatethe probability of rejection for poor quality. For sufficientlysmall tots, procedure Brequires 100 % inspection.
Both procedures treat the limiting quality (LO) as the acfualpercent nonconforrning units in the subnitted lots and theprobability of occeptance at this limiting quality can be found intable D1 for procedure A and tables B1 to B10 for procedure B.Procedure B is indicated as the appropriate procedure if themanufacturer is concerned with a continuing series of lots;whether or not these go to the same consumer. Procedure A isindicated as the apnrnpriate prncediure when the lnt is a singlelot and shall be used if accepi zero plans are required to be partof tho campling ochcmc.
3.3 Procedure A (use table A)
plan is idcntified by the lot size and the limiting quality (LQ).
With the specified lot size and lirniting quality as indexingvalues,the sampling sizc (n} and acccptoncc numbcr (Ac} aregiven in table A.