1、A Referencenumbe r ISO3084:1998 (E) INTERNATIONAL STANDARD ISO 3084 Thirdedition 19981215 IronoresExperimentalmethodsfor evaluationofqualityvariation MineraisdeferMthodesexprimentalespourlvaluationdela variationdequalitISO3084:1998(E) ISO1998 Allrightsreserved.Unlessotherwisespecified,nopartofthispu
2、blicationmaybereproducedorutilizedinanyformorbyan ymeans,electronic ormechanical,includingphotocopyingandmicrofilm,withoutpermissioninwritingfromthepublisher. InternationalOrganizationforStandardization Casepostale56 CH1211Genve20 Switzerland Internet isoiso.ch PrintedinSwitzerland ii Contents 1Scop
3、e 1 2Normativereferences . .1 3Definitions 1 4Generalconditions . .2 5Evaluationofqualityvariationusinginterleavedsampling . .2 6Estimationofqualityvariationfromthevariogram. .8 7Classificationofqualityvariation 10 AnnexA (informative) Examplesforcalculationsofstandarddeviation 12ISO ISO3084:1998(E)
4、 iii Foreword ISO(theInternationalOrganizationforStandardization)isaworldwidefederationofnationalstandardsbodies(ISO memberbodies).TheworkofpreparingInternationalStandardsisnormallycarriedoutthroughISOtechnical committees.Eachmemberbodyinterestedinasubjectforwhichatechnicalcommitteehasbeenestablishe
5、dhas therighttoberepresentedonthatcommittee.Internationalorganizations,governmentalandnongovernmental,in liaisonwithISO,alsotakepartinthework.ISOcollaboratescloselywiththeInternationalElectrotechnical Commission(IEC)onallmattersofelectrotechnicalstandardization. DraftInternationalStandardsadoptedbyt
6、hetechnicalcommitteesarecirculatedtothememberbodiesforvoting. PublicationasanInternationalStandardrequiresapprovalbyatleast75%ofthememberbodiescastingavote. InternationalStandardISO3084waspreparedbyTechnicalCommitteeISO/TC102, Ironores,subcommittee SC1, Sampling. Thisthirdeditioncancelsandreplacesth
7、esecondedition(ISO3084:1986)whichhasbeentechnicallyrevised. AnnexAofthisInternationalStandardisforinformationonly.INTERNATIONALSTANDARD ISO ISO3084:1998(E) 1 IronoresExperimentalmethodsforevaluationofquality variation 1Scope ThisInternationalStandardspecifiesexperimentalmethodsfortheevaluationofqual
8、ityvariationofironoresforeach typeofironorebeingtradedandforeachhandlingplant. Twodistinctapproachesarespecified.Thefirstistoanalyseinterleavedsamplescomposedofanumberofpaired incrementstakenandcombinedalternatelyfollowingstratifiedsamplingorsystematicsamplingasspecifiedin ISO 3082.Thesecondistocoll
9、ectandanalyseindividualincrementsandthentoanalysethedatausing variographicmethods. Usinginterleavedsamplesinvolveslesswork,butuseofvariogramsprovidesabetterestimateofqualityvariation andhenceabetterestimateofthesamplingvariance.Thevariogrammethodisusuallyusedtofinetuneasampling operation. NOTETheexp
10、erimentalmethodsmaybeappliedapproximatelytotimebasissamplingwhentheflowrateisalmostuniform. 2Normative references Thefollowingstandardscontainprovisionswhich,throughreferenceinthistext,constituteprovisionsofthis InternationalStandard.Atthetimeofpublication,theeditionsindicatedwerevalid.Allstandardsa
11、resubjectto revision,andpartiestoagreementsbasedonthisInternationalStandardareencouragedtoinvestigatethepossibility ofapplyingthemostrecenteditionsofthestandardsindicatedbelow.MembersofIECandISOmaintainregistersof currentlyvalidInternationalStandards. ISO3082: 1) , IronoresSamplingandsamplepreparati
12、onprocedures . ISO3085:1996, IronoresExperimentalmethodsforcheckingtheprecisionofsampling . ISO11323:1996, IronoresVocabulary . 3Definitions ForthepurposesofthisInternationalStandard,thedefinitionsgiveninISO11323andthefollowingapply. 3.1interleavedsamples samplesconstitutedbyplacingconsecutiveprimar
13、yincrementsalternatelyintotwosamplecontainers 1) Tobepublished.(RevisionofISO3081:1986,ISO3082:1987andISO3083:1986)ISO3084:1998(E) ISO 2 4General conditions 4.1Quality variation Thequalityvariationordegreeofheterogeneityofironoreshallbedeterminedintermsofthestandarddeviation. Thestandarddeviationofa
14、qualitycharacteristicbetweenincrementstakenfromwithinstrata,denotedby s w ,shall bedeterminedeitherbyestimatingthevariancebetweeninterleavedsamplesorbymeasuringindividualincrements anddeterminingtheslopeandinterceptofalinearfittoavariogramcorrectedbysubtractionofsamplepreparation andmeasurementvaria
15、nces.Inbothcases,wherecorrectionsforsamplepreparationandmeasurementvariances aremade(see5.6.2,note2,and6.1),itisessentialthatthesamplepreparationandmeasurementvariancesbe determinedatthesametimeastheexperimentstodeterminequalityvariationarecarriedout. 4.2Quality characteristics Thequalitycharacteris
16、ticchosenfordeterminingthequalityvariationisgenerallythetotalironcontent,butsilica content,aluminacontent,moisturecontent,sizedistributionandotherqualitycharacteristicsmayalsobechosen. Whenseparatesamplesaretakenforthedeterminationofchemicalcomposition,moisturecontent,sizedistribution, etc.,thequali
17、tyvariationfortheindividualcharacteristicsshallbeadopted.Whenthesampleisusedforthe determinationofmorethanonequalitycharacteristic,thelargestclassificationcategoryforqualityvariationamong thesecharacteristicsshallbeadopted. 4.3 Sampling,samplepreparationandmeasurement Samplingandsamplepreparationsha
18、llbecarriedoutinaccordancewithISO3082.Measurementofsamplesshall becarriedoutinaccordancewiththerelevantInternationalStandardsforchemicalanalysis,moisturecontentand sizeanalysisofironores. Thesamplingforevaluationofqualityvariationmaybeconductedinconjunctionwithroutinesamplingfor determinationofthequ
19、alityofthelot.Inotherwords,thesamplescollectedfromthelotmaybeusedforboth purposes. 5 Evaluationofqualityvariationusinginterleavedsampling 5.1General Theproceduresforevaluatingthestandarddeviationwithinstrata, s w ,applicabletobothstratifiedandsystematic sampling,aredescribedin5.2to5.7. 5.2 Typeofinv
20、estigation 5.2.1Type 1 Whenlotsarefrequentlydelivered,thequalityvariationmaybedeterminedfromalargenumberoflotsofalmost equalmassasfollows: a) treateachlotseparately; b) makeupapairofinterleavedsamplesforeachlotasshowninfigure1a)andexample1.ISO ISO3084:1998(E) 3 5.2.2Type 2 Whenlargelotsareinfrequent
21、lydelivered,thequalityvariationmaybedeterminedfromasinglelotasfollows: a) splitthelotintoatleast10partsofalmostequalmass; b) makeupapairofinterleavedsamplesforeachpartbycombiningtheincrementstakenfromeachpartas showninfigure1b)andexample2. 5.2.3Type 3 Whensmalllotsarefrequentlydelivered,thequalityva
22、riationmaybedeterminedfromseverallotsofalmostequal massasfollows: a) splitallthelotsinvolvedintoatleast10partsofalmostequalmass; b) makeupapairofinterleavedsamplesforeachpartbycombiningtheincrementstakenfromeachpartas showninfigure1c)andexample3. 5.2.4Type 4 Whensamplingawagonbornelotandwhenincremen
23、tsaretakenfromallwagonscomprisingthelot,thesampling schememayberegardedasstratifiedsampling.Whenlotsarefrequentlydelivered,theprocedurefordetermining qualityvariationisasfollows: a) treateachlotseparately; b) makeupapairofinterleavedsamplesforeachlotasshowninfigure1d). 5.3 Numberofincrementsandconst
24、itutionofinterleavedsamples 5.3.1 Numberofincrements Thenumberofincrementstobetakenfromoneorseverallotsmaybethesameasthatselectedforroutine sampling.However,whentheroutinesamplingisbasedontheclassificationcategoryof“small“qualityvariationand thenumberofincrementsisconsideredtobeinsufficienttoobtaina
25、reliablestandarddeviation,thenthenumberof incrementsshallbeincreased(seeISO3082). 5.3.1.1 Type1investigation Fortype1investigations,thenumberofincrements,n 1 ,tobetakenfromeachlotshallbeinaccordancewith ISO3082,andapairofinterleavedsamplesshallbeconstitutedforeachlotseefigure1a). 5.3.1.2 Type2invest
26、igation Fortype2investigations,thenumberofincrements, n 1 ,shallbedeterminedfromISO3082,andatleast10pairsof interleavedsamplesshallbeconstitutedseefigure1b). 5.3.1.3 Type3investigation Fortype3investigations,thenumberofincrements,n 1 ,tobetakenfromeachlotshallbeinaccordancewith ISO3082.Eachlotshallb
27、edividedintoanumberofstrata,andincrementsfromeachstratumshallbecombined intoapairofinterleavedsamplesseefigure1c). 5.3.1.4 Type4investigation Fortype4investigations,thenumberofincrements, n 1 ,beingcollectedfromeachlotshallbedeterminedfromtable 3ofISO3082,andthenumberofincrements,n w ,tobetakenfrome
28、achwagonshallbeinaccordancewithISO3084:1998(E) ISO 4 ISO3082.Ifthenumberisodd,itshallbeincreasedbyonetomakeiteven.Apairofinterleavedsamplesshallbe constitutedforeachlotseefigure1d). 5.3.2 Constitutionofinterleavedsamples Theinterleavedsamplesshallbemadeupaccordingtothefollowingprocedure: allocatease
29、rialnumbertotheincrementsfromeachlotorpartlotinorderofsampling; constitutepairsofinterleavedsamplesfromconsecutiveoddnumberedincrements(denotedbyinterleaved sampleA i )andconsecutiveevennumberedincrements(denotedbyinterleavedsampleB i )foreachlotorpart lot(seefigure2); foreachinvestigation,prepare n
30、setsofpairedinterleavedsamples. Eachinterleavedsampleshallbemadeupoftwoormoreincrements. 5.4 Preparationoftestsamplesandmeasurement Separatetestsamplesshallbepreparedfromtheinterleavedsamples,A i andB i . Chemicalanalysis,moisturedetermination,sizedeterminationorphysicaltestingshallbecarriedoutonthe
31、test samplesasrequired.ISO ISO3084:1998(E) 5 a)Type1investigation nlots:onelot=onepart Key:Therectangularboxindicatesonelot;eachdivisionoftheboxmadebybrokenlinesindicatesonepart;apairofcircles indicatesapairofinterleavedsamples.Thisalsoappliestoa)andc). b)Type2investigationOnelot(examplefor10parts)
32、c)Type3investigationSeverallots(exampleforthreelotsand12parts) Key:Eachboxindicatesawagon;pointsintheboxindicateincrements;circlesindicateinterleavedsamples. d)Type4investigationStratifiedsamplingofwagonbornelots.ISO3084:1998(E) ISO 6 Figure1Schematicdiagramforoneinvestigation NOTEThisdiagramisanexa
33、mpleforasinglelotof5000tto15000toforeof“large“qualityvariation.Inaccordancewi th ISO3082,therequiredminimumnumberofincrementsis100,and10pairsofinterleavedsamplesA i and B i ( i=1,2,.,10), eachcomprisingfiveincrements,areprepared. Figure2Exampleofschematicdiagramforconstitutionofpairsofinterleavedsam
34、ples(type2) 5.5 Numberofinvestigations Becausethestandarddeviation, s w ,ofaqualitycharacteristicwithinstratacannotbeestimatedverypreciselyfroma smallnumberofinvestigations,thefollowingminimumnumberofinvestigationsisrecommended: a) fortype2and3investigations,atleastfiveseparateinvestigations; b) for
35、type1and4investigations,atleast10separateinvestigations. 5.6 Calculationofstandarddeviationwithinstrata 5.6.1Data sheet Theexperimentaldatageneratedbychemicalanalysis,moisturedetermination,sizedeterminationorphysical testingofindividualtestsamplesshallberecordedonasuitableform(seeexamples1to3). 5.6.
36、2Calculation Theestimatedstandarddeviationwithinstratashallbecalculatedfromequation(4). Therange, R i ,ofpairedmeasurementsisgivenbyequation(1): RAB iii =- (1)ISO ISO3084:1998(E) 7 where A i isthemeasuredqualitycharacteristic(suchas%Fe)ofthetestsamplepreparedfrominterleavedsample A i ; B i isthemeas
37、uredqualitycharacteristicofthetestsamplepreparedfrominterleavedsampleB i ,whichisfrom thesamepartlotasinterleavedsampleA i ; i isasubscriptdesignatingeachpartlot. Themean, R ,ofranges R i isgivenbyequation(2): R n R i = 1 4 (2) where n 4 isthenumberofranges, R i ,whichisthesameasthenumberofpartlotsi
38、ntheinvestigation. Themean, x i ,ofpairedmeasurementsforeachpartisgivenbyequation(3): xAB iii =+ 1 2 () (3) Theestimatedstandarddeviationwithinstrata, s w , isgivenbyequation(4): s w = n R d 5 2 (4) where n 5 isthenumberofincrementscomprisingeachinterleavedsampleA i orB i ; d 2 isthefactortoestimate
39、standarddeviationfromtherange;forpaireddata1/ d 2 =0,8862. NOTES 1Fortype3investigations,themeanvalueofthequalitycharacteristicforthe jthlot, x j ,maybeobtainedfromequation(5). x n x jj i = 1 6 (5) where x ji isthemeanofpairedmeasurementsforeachpartinlot j; n 6 isthenumberofpartsinthelot. 2The estim
40、ated standard deviation within strata, s w ,obtainedfromequation(4)isameasureofthecombinedstandard deviationofsampling,samplepreparationandmeasurement.Whilethestandarddeviationwithinstrataisoverestimated,this valuemaybeusedfortheclassificationinclause7(see5.7). Whenitisdesiredtoobtainanunbiasedestim
41、ateofthestandarddeviationwithinstrata,andwhentheestimatedstandard deviationofsamplepreparation,denotedby s P ,andtheestimatedstandarddeviationofmeasurement,denotedby s M ,are known,theestimatedstandarddeviationwithinstratashouldbecalculatedusingequation(6): () ss s w P 2 M =- - n R d 5 2 22 (6)ISO30
42、84:1998(E) ISO 8 3Ifthenumberofincrementsisdeterminedaccordingto5.3.1andthoseincrementsaretaken,thevariationamongthe numbersofincrementscomprisingthevariousinterleavedsampleswillbesmall.Ifthevariationis10%orless,equations(4) and(6)canbeappliedapproximatelybyusingthemeanvalueof n 5 . 5.7 Expressionofresults 5.7.1 Type2and3investigations Fortype2and3investigations,theestimatedmeanvalueofstandarddeviationwithinstrata, s w ,foraparticular ironoreandhandlingplant
