ANSI ASTM D3244-2016 Standard Practice for Utilization of Test Data to Determine Conformance with Specifications.pdf

1、Designation: D3244 16 An American National StandardStandard Practice forUtilization of Test Data to Determine Conformance withSpecifications1This standard is issued under the fixed designation D3244; the number immediately following the designation indicates the year oforiginal adoption or, in the c

2、ase of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon () indicates an editorial change since the last revision or reapproval.INTRODUCTIONThe properties of commercial petroleum products are measured by standardized laboratory t

3、estmethods to assess their conformance to specifications. Two or more measurement results obtained byperforming the same test method for the same property of a specific sample usually will not benumerically identical. Therefore, the test methods generally include a paragraph on the precision ofresul

4、ts. This precision (or, a more appropriate term is imprecision) is an expression of the degree ofagreement that can be expected between the aforementioned measurements.Many difficulties that arise in assessing conformance to specifications are due to test imprecision.Because of this, a true value of

5、 a property can never be determined exactly; and it is necessary to inferfrom measured values the range within which the “true value” is likely to lie. The main purpose of thispractice is to indicate how test imprecision should be interpreted relative to specification limit values.1. Scope*1.1 This

6、practice covers guidelines and statistical method-ologies with which two parties, usually a supplier and areceiver, can compare and combine independently obtained testresults to obtain an Assigned Test Value (ATV) for the purposeof resolving a product quality dispute.1.2 This practice defines a tech

7、nique for comparing an ATVwith a specification limit.1.3 This practice applies only to those test methods whichspecifically state that the repeatability and reproducibilityvalues conform to the definitions herein.1.4 The statistical principles and methodology outlined inthis practice can also be use

8、d to obtain an ATV for specificationconformance decision when multiple results are obtained forthe same batch of product within a single laboratory. For thisapplication, site precision (R) as defined in Practice D6299shall be used in lieu of test method published reproducibility(R).2. Referenced Doc

9、uments2.1 ASTM Standards:2D1319 Test Method for Hydrocarbon Types in Liquid Petro-leum Products by Fluorescent Indicator AdsorptionD4057 Practice for Manual Sampling of Petroleum andPetroleum ProductsD4177 Practice for Automatic Sampling of Petroleum andPetroleum ProductsD6299 Practice for Applying

10、Statistical Quality Assuranceand Control Charting Techniques to Evaluate AnalyticalMeasurement System PerformanceD6300 Practice for Determination of Precision and BiasData for Use in Test Methods for Petroleum Products andLubricantsD6792 Practice for Quality System in Petroleum Productsand Lubricant

11、s Testing LaboratoriesD7372 Guide for Analysis and Interpretation of ProficiencyTest Program ResultsE29 Practice for Using Significant Digits in Test Data toDetermine Conformance with Specifications1This practice is under the jurisdiction of ASTM Committee D02 on PetroleumProducts, Liquid Fuels, and

12、 Lubricants and is the direct responsibility of D02.94 onCoordinating Subcommittee on Quality Assurance and Statistics.Current edition approved June 1, 2016. Published June 2016. Originallypublished as an appendix to the 1968 Annual Book of ASTM Standards, Part 18.Originally approved as a standard i

13、n 1974. Last previous edition approved in 2012as D3244 12. DOI: 10.1520/D3244-16.2For referenced ASTM standards, visit the ASTM website, www.astm.org, orcontact ASTM Customer Service at serviceastm.org. For Annual Book of ASTMStandards volume information, refer to the standards Document Summary page

14、 onthe ASTM website.*A Summary of Changes section appears at the end of this standardCopyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States12.2 ISO Standard:3ISO 4259 Determination and Application of Precision Datain Relation to Methods of T

15、est3. Terminology3.1 Definitions:3.1.1 acceptance limit (AL), na numerical value thatdefines the point between acceptable and unacceptable quality.3.1.1.1 DiscussionThe AL is not necessarily the specifi-cation limit. It is a value that takes into account the specifica-tion limit, the test method pre

16、cision, and the desired probabilityof product acceptance if the quality is at the specification limit.3.1.2 assigned test value (ATV), nthe average of all resultsobtained in the several laboratories which are consideredacceptable based on the reproducibility of the test method.3.1.3 determination, n

17、the process of carrying out theseries of operations specified in the test method whereby asingle value is obtained.3.1.4 dispute, nwhen there is a question as to productquality conformance to specification because a test valueobtained falls outside the specification limit(s).3.1.5 operator, na perso

18、n who normally and regularlycarries out a particular test.3.1.6 precision, nthe degree of agreement between two ormore test results on the same property obtained using the sametest method on identical test material. In this practice, precisionstatements are framed in terms of the repeatability and r

19、epro-ducibility of the test method.3.1.7 receiver, nany individual or organization who re-ceives or accepts the product delivered by the supplier.3.1.8 receivers risk, nthe probability of accepting a prod-uct that fails to meet the specification.3.1.9 repeatability (r), nquantitative expression of t

20、herandom error associated with a single operator in a givenlaboratory obtaining replicate results with the same apparatusunder constant operating conditions on identical test materialwithin a short period of time. It is defined (3.1.9.1) as thatdifference between two such single results as would bee

21、xceeded in the long run in only 1 case in 20 in the normal andcorrect operation of the test method (3.1.9.3). (This is known asthe 95 % confidence level.)3.1.9.1 DiscussionThe repeatability and reproducibilityvalues should have been determined according to the methodsdescribed in ASTM Research Repor

22、t RR:D02-1007, Manualon Determining Precision data for ASTM Methods of Petro-leum Products and Lubricants,4Practice D6300, or ISO 4259.3.1.9.2 DiscussionNot all standards organizations definerepeatability and reproducibility in precisely these same terms,and attention should always be paid to defini

23、tions beforecomparing precision values quoted.3.1.9.3 DiscussionThis difference is related to the repeat-ability or the reproducibility standard deviation but is not thestandard deviation.3.1.10 reproducibility (R), nquantitative expression of therandom error associated with operators working in dif

24、ferentlaboratories, each obtaining single results on identical testmaterial when applying the same method. It is defined (3.1.9.1)as that difference between two such single and independentresults as would be exceeded in the long run in only 1 case in20 in the normal and correct operation of the test

25、 method. See3.1.9.3.3.1.11 result, nthe value obtained by following the com-plete set of instructions of a test method. It may be obtainedfrom a single determination or several determinations, depend-ing on the instruction of the test method.3.1.12 supplier, nany individual or organization respon-si

26、ble for the quality of a product just before it is taken over bythe receiver.3.1.13 suppliers risk, nthe probability of rejecting aproduct that meets the specification.3.1.14 test sample, na portion of the product taken at theplace where the product is exchanged, that is, where theresponsibility for

27、 the product quality passes from the supplierto the receiver. In the event that this is not possible, a suitablesampling location should be mutually agreed upon.3.1.15 true value (), nfor practical purposes, the valuetowards which the average of single results obtained by Nlaboratories using the sam

28、e standard test method tends, when Nbecomes very large. Consequently, this definition of true valueis associated with the particular test method employed.4. Significance and Use4.1 This practice provides a means whereby the parties to atransaction can resolve potential quality disputes over thosepro

29、duct properties which can be tested and expressed numeri-cally.4.1.1 This practice can be used to ensure that such proper-ties are correctly stated on labels or in other descriptions of theproduct.4.1.2 This practice can be implemented in those cases wherea supplier uses an in-house or a commercial

30、testing laboratoryto sample and test a product prior to releasing the product to ashipper (intermediate receiver) and the ultimate receiver alsouses an in-house or commercial testing laboratory to sampleand test the product upon arrival at the destination. The ATVwould still be determined according

31、to 8.3.4.2 This practice can assist in the determination of toler-ances from specification limits which will ensure that the truevalue of a property is sufficiently close to the specificationvalue with a mutually agreed probability so that the product isacceptable to the receiver. Such tolerances ar

32、e bounded by anacceptance limit (AL).IftheATV value determined by applyingthis practice falls on the AL or on the acceptable side of the AL,the product can be accepted; otherwise it shall be deemed tohave failed the product acceptance requirement established byapplying this practice.3Available from

33、American National Standards Institute (ANSI), 25 W. 43rd St.,4th Floor, New York, NY 10036, http:/www.ansi.org.4Supporting data have been filed at ASTM International Headquarters and maybe obtained by requesting Research Report RR:D02-1007.D3244 1624.3 Application of this practice requires the AL be

34、 deter-mined prior to actual commencement of testing. Therefore, thedegree of criticality of the specification, as determined by theProbability ofAcceptance (Pvalue) that is required to calculatethe AL, shall have been mutually agreed upon between bothparties prior to execution of actual product tes

35、ting.4.3.1 This agreement should include a decision as towhether the ATV is to be determined by the absolute orrounding-off method of Practice E29, as therein defined.4.3.1.1 If the rounding-off method is to be used, the numberof significant digits to be retained must also be agreed upon.4.3.1.2 The

36、se decisions must also be made in the case whereonly one party is involved, as in the case of a label.4.3.1.3 In the absence of such an agreement, this practicerecommends the ATV be rounded in accordance with therounding-off method in Practice E29 to the number of signifi-cant digits that are specif

37、ied in the governing specification.4.4 This practice is designed to be suitable for reference incontracts governing the transfer of petroleum products andlubricants from a supplier to a receiver.4.5 As a prerequisite for acceptance for lab test results to beused in this practice, the following condi

38、tions shall be satisfied:4.5.1 Site precision (R) as defined in Practice D6299 for theappropriate test method(s) from each lab, as substantiated byin-house quality control programs, for property typical of theproduct in dispute, should have a TPI 1.2 (see PracticeD6792 for TPI explanation), but at a

39、 minimum shall be betterthan the published method reproducibility (R).4.5.2 Each lab shall be able to demonstrate, by way ofresults from interlaboratory exchange programs, a lack of asystemic bias relative to exchange averages for the appropriatetest method(s) as per methodology outlined in Guide D7

40、372.4.5.3 In the event that the site precision of laboratories fromtwo parties are statistically different as confirmed by the F-test(see Annex A4), then, for the purpose of establishing the ATV,each laboratorys test result shall be inversely weighted inaccordance with laboratorys demonstrated varia

41、nce.4.6 It is recommended that this practice be conducted underthe guidance of a qualified statistician.5. Sampling5.1 Sampling should be carried out as specified in accor-dance with the referenced test method, contract, or specifica-tion for the petroleum product under test, such as PracticeD4057,

42、or Practice D4177, or other referenced standards asappropriate. Obtain enough sample to allow for all requireddeterminations to be made by supplier, receiver, and a possiblethird party.NOTE 1In the event the contractual or specification-referenced sam-pling procedure differs from that outlined in th

43、e relevant test method,supplier and receiver need to ensure the correct sampling procedure isused.6. Applying Test Method Precision Data to Accept orReject Test Results6.1 This section describes procedures in which the precisionlimits of test methods can be used as a decision criterion toaccept or r

44、eject test results.6.2 Significance of Repeatability (r):6.2.1 Acceptance of ResultsWhen only two results areobtained under conditions of repeatability and the difference isequal to or less than the repeatability of the method, theoperator may report the average of the two results as beingapplicable

45、 to the sample tested.6.2.2 Rejection of ResultsWhen two results are obtainedthat differ by more than the repeatability of the method, bothshould be rejected. Obtain two additional results immediatelyunder conditions of repeatability. If the difference betweenthese two results is equal to or less th

46、an the repeatability of themethod, the operator should report the average of the two asbeing applicable to the sample tested. If, however, the differ-ence so obtained again exceeds the repeatability, reject theresults and investigate the application of the method.6.3 Significance of Reproducibility

47、(R):6.3.1 Acceptance of ResultsWhen two results are obtainedand comprise one result from each laboratory (Note 2), if thedifference is equal to or less than the reproducibility of themethod, then both results should be considered acceptable.NOTE 2When a comparison for reproducibility is made between

48、results from two laboratories, it is a common practice that single resultsfrom each will be compared. If each of the laboratories has produced morethan a single result, see 6.4.6.3.2 Rejection of ResultsWhen the results from twolaboratories differ by more than the reproducibility of themethod, rejec

49、t both results and each laboratory should repeatthe test on the retained sample. If the difference is now equalto or less than the reproducibility, both results should beconsidered acceptable. If, however, the difference betweenthese results is still greater than the reproducibility, reject theresults and investigate the application of the method at eachlaboratory, sampling, sample preparation and storage and allother factors which can contribute to the variance.6.4 Significance of Reduced Reproducibility (R_reduced)from Multiple TestingIf the numbe