Nuclear Instruments and Methods in Physics Research A - PDF document

NuclearInstrumentsandMethodsinPhysicsResearchA294(1990)179-192North-HollandC\tTARGETCALW.K.SAKUMOTO,P.DEBA\tARO,A.BODED,H.S.DepartmentofPhysicsandAstronomy,UniversityofRochester,Rochester,NY14627,USAF.S.MERRITT,M.J.OREGLIA,\t.SCELLMAN*and.A.SCHUMMEnricoFermiInstituteandDepartmentofPhysics,UniversityofChicago, Fig.1.TheLabE(CCFR)neutrinodetector.Thetargetcalorimeterconsistsofsixindependentlymovabletargetcarts,numbered1through6asindicated.Themuonspectrometerconsistsofthreetoroidalmagnetunits(A,BandC),andapairofdriftchamberstationsatthefardownstreamend.orneutralcurrents,theperformanceandcalibrationofthetargetcalorimeterandofthemuonspectrometershouldbeknownatthelevelofapercentorbetter[1].Section2givesadescriptionoftheCCFRtargetcalorimeter.Thenextsectiondescribesthegaincalibra-tionandmonitoringofthetarget'scalorimetrycountersandassociatedelectronics.Section4describesthepro-gramoftotalenergycalibrationwithtestbeamsandtheassociateddataanalysis.Section5givestheresultsofthecalibration.DRIFTCHAMBERS\tLIQUIDSCINTILLATION\tSTEELPLATESCOUNTERS'COUNTER3~JL~1)1.2,3,4-\tTUSESH!\tSakumotoetal./CalibrationoftheCCFRtargetcalorimeterLABENEUTRINODETECTOR690TONTARGET420TONTOROIDSYSTEM0VETOWALL\tSCALE4-r-3'n--lI\tCONTENTS1.28STEELPLATES(5.2cmTHIN(,3.05mx3.05m)2.14SCINTILLATIONCOUNTERS3.7DRIFTCHAMBERS2.TheC\ttargetcalorimeterThecalorimeterhasamassof690tons,alengthof17.7m,andanaveragedensityof4.2g/cm3.Thislargecalorimeterconsistsofsixidenticalmodulescalled"targetcarts".Eachcartsitsonseparateatraywhichisindependentlymovable.Acartconsistsoftwenty-eight3mx3mx5.2cmsteelplatesinterspersedwithfour-teenliquidscintillationcountersplacedevery10.3cmofsteelandsevendriftchamberstationsplacedeveryModuleI®--ii----3.05m--~~---®DriftChamber2Fig.2.(a)Aviewofacalorimetertargetcart(nottoscale)containing285.2cmthicksteelplates.14liquidscintillationcountersand7drift-chamberstations.Adriftchamberstationhasbothanxplaneandayplane.(b)Adetailedviewofatargetcart'scomponents.Themassdistributionofthesecomponentsaregivenintable1.tb~BeamAverage1thickness5.15cmCounterI6.48cmCounter2DriftChamberI}13.37cm.Module2Counter3Counter4 Table1Massdistributionofthecalorimeterelements.Thescintillatoroilhasamineraloilbaseandahydrogen-to-carbonratioof1.8.TheAlI-beambracesofthedriftchamberaredistributedwithinthedriftchambergasregion20.6cmofsteel.Figs.2aand2bshowtheplacementanddimensionsoftheseelementswithinatargetcart.Table1givestheamountofmaterialpresentedbytheseelements.Thescintillationcountershaveanactivevolumeof3mx3mx2.54cm.Thisvolumeisfilledwithmineraloildopedwithascintillatorandasecondaryfluorwhichproducesabluelight.Thewallsofthecontainerconsistof3.175mmthickplexiglascoveredwithopaqueMarvelguard[2).Topreventthewallsfromcollapsing,theinteriorofthiscontainerisstuddedwithmanyplexiglasribs.Tobalancethehydrostaticoilpressure,thecounterissandwichedbetweenmylarreinforcedpolyethylenebagsfilledwithwater.Thesebagspressagainsttheadjacent,anchoredsteelplates.Theedgesofthecontaineraremadeofclear,2.54cmthickplexiglas.Thereare1.27cmthickBBQK5]dopedplasticbarsalongeachedge(withanairgapbetween)thatabsorbthebluefluorescencelight,reemitlightandthentrans-portthislighttothefourcornersofthecounter(seefig.2a).RCA6342Aphototubesareusedtoreadoutthelight.Thesensitivitytoamuontraversingthroughthecenterofacounterisabout2.5photoelectronsperphototube.AnalogsignalsfromthephototubesaredigitizedbytheLeCroy4300FERAsystem.Thissystemhasfea-turessuitedforthehigh-rateenvironmentoftheTevatronquadrupole-tripletwidebandneutrinobeam:itisfastandbuffered(32-eventmemory;14psfordigitizing,buffering,andclearing).Theanalog-to-digitalconverters(ADCs)inthissystemhave11bitsofdy-namicrange.TheseADCsarelinearwiththegatewidthseiat240nsandwiththesensitivityat0.25pCperADCcount.Theleadingedgeofthegatepulseistimedtoarrive25nsbeforetheleadingedgeofthecorre-W.K.Sakumotoetal./CalibrationoftheCCFRtargetcalorimeterspondingphototubecurrentpulse;i.e.theintegrationtimeis215ns.Ifthisgateisshiftedtoarrive10nslater(earlier)thanthetimingdescribedpreviously,thedig-itizedphototubesignalincreases(decreases)by0.6%.Theanalogphototubesignalsfromasinglecounterarefannedoutintochannelsofdifferentsensitivityanddigitizedasshowninfig.3.ThepedestalsfortheseADCsaresetatapproximately30ADCcounts.DirectmeasurementsrevealthattheLOWchannelshaveatmost0.56nonlinearityovertheirfullrange.TheSU-PER-LOWchannels,whicharemixedandattenuatedLOWsignals(seethecaptionoffig.3),haveatmost0.2%nonlinearity.TheanalogsumoftheLOWs,theSUM-LO,isslightlynonlinearbecausetheFNALES7138"linear"fan-inmodulesareabout1.56nonlin-earovertherangeoftheSUM-LOchannel.Theusageofthesechannelsaredescribedinthenextsection.EventtimeinformationforeachcounterisfromadiscriminatedoutputontheSUM-HIchannel.Thediscriminatorproducesa100nswidelogicpulseiftheSUM-HIsignalislargerthanaquarteroftheminimumionizinglevel.Thetimingofthislogicpulseismeasuredusingthesametime-to-digitalconverters(TDCs)thatareusedforthedriftchambers.Whentiminginforma-tionfromallactivecountersarecombined,theeventtimeresolutionhasanrmsofabout2.4ns.AttheTevatron,thereare18.8nsbetweenacceleratorrfbuckets.LRSFAN-OUT(FNALES7137)\tFERAADCsFig.3.Thereadoutelectronicsforasinglescintillationcounter.Thefourphototubesignalsfromacounterarelabeledas:1,2,3,and4.The"SL"fan-outoutputofphototubeN(N=1,2,3,4)fromeverytenthconsecutivecounteraremixed,attenuatedbyafactorof12,andthendigitizedbyaFERAADC:thisisoneofeightSUPER-LOWADCchannelsforthephototubeNfromallcounters.ThethresholdoftheSBITdiscriminatorissetatthequarterofthesingleminimumionizinglevel.ElementSubelementLength[cm]Mass[g/CM2]Steelplate5.153240.555ScintillationScintillatoroil-2.542.032counterPlexiglas0.640.749Water-2.542.540Mylarandbag0.761.056DriftchamberAlHexcel[3]2.540.857AlI-beambraces-0.148G-100.6351.154Cuclad(G-10)0.00030.003Chambergas2.8580.004(50-50Ar--ethane)Deadspace(air)-2.840.003 chdriftchamberstationhasanactiveareaof3X3n?with24horizontaland24verticalcells.Eachcallis12.7cmacross,3mIon,;,and1.9cmthick.Ithasaleft-rightambiguityresolvingthree-wirecellgeome-try:acentral0.13mmdiameterfieldwirewithtwoopposing0.03mmdiametersensewires2mmfromit.ehitinformationisprocessedbyabuffered,multihitT\tsystemwith4nsresolutionanda2pslivetime.'ssystemtakesabout12pstostoreanevent.Downstreamofthetargetcalorimeteristhe420ton,magnetizedsteel,muonspectrometer.Itconsistsofthreeseparatetoroidalmagnetswithdriftchamberstationsaftereachmagnetandtwoadditionaldriftchamberstationsatthefardownstreamend.Thesedriftcham-bershavethesamecellsizeasthoseinthetargetcalorimeterbuteachcellonlyhasasinglesensewire.Eachdrift-chamberstationconsistsoffivepairsofhorizontalandverticalplanes.Therearemuontriggercountersafterthefirsttwomagnetsand3.81cmthickacrylicscintillationcountersevery20cmofsteel.Thesecountersanddriftchamberscoveranareaof3X3n12.ThenitionmomentumresolutionislimitedbymultipleCoulombscatteringinthesteeltoroidsand8(p)/p=11%.Thetotaltransversemomentumkickforalongitu-dinaltrackis2.45C;eV/candtheangularresolutionoverthefulllengthofthespectrometer(17.8m)isabout0.3mrad.Theangular-resolutionpartofthemomentumresolutionbecomescomparablewiththemultipleCoulombscatteringpartatamomentumofabout1TeV/c.3.Counter\tcalibrationandmonitoringShowerenergydepositedinanyscintillation.counterismeasuredbyphototubesatitsfourcorners.Theanalogphototubesignals(seefig.3)aredigitizedsimul-taneouslybysevenADCchannelsofdifferingsensitiv-ity:aSUPER-LOW,fourLOWs,aSUM-LOandaSUM-HI.ThefourLOWchannelsvieweachofthefourphototubesindividually.ASUPER-LOWisessentiallyanattenuatedLOW.TheSUM-LO,HIchannelsareanalogsumsofthefourLOWs.Thiscollectionof11-bitADCswithvariouslevelsofinputgainisequivalenttoasingle19-bitSUM-HIchannel.Typically,aminimumionizingmuonthatpassesthroughthecenterofacounteryie:Is80ADCcountsintheSUM-HIADCchannel,8ADCcountsintheSUM-LOADCchannel,2ADCcountsinaLOWADCchannel,and0.2ADCcountsinaSUPER-LOWADCchannel.ThepurposeoftheSUM-HIchannelistoobserveminimumionizingmuonsclearly.Itisusedtonormalizethepositionalenergyresponsevariationofacounteroveritsactiveareaandtonormalizecounter-to-countergainvariations.Wecallthisthe"muoncalibration".ThepurposeoftheSUM-LOchannelistoi!\tSakumotoetal./CalibrationoftheCCFRtargetcalorimeteraccuratelyrelatetheminimumionizingmuonsignalmeasuredinaSUM-HIchanneltothecorrespondingLOWchannels.TheSUM-LO,HIchannelsareonlyusedforcalibrationpurposes.TheLOWandSUPER-LOWchannelsareusedforcalorimetryaftercountergainvariationsareremovedviathemuoncalibration.TheSUPER-LOWchannelisusedonlywhenaLOWchannelsaturates.OneortwoLOWchannelscansaturatewhenahadronshowersnearaphototubeorwhenanenergetichadroninexcessof400GeVshowersnearthecenterofthecalorimeter.ThesummedenergylossoftheSUM-HIchannelisusedforthemuoncalibrationorrelativeresponsenor-inalizationofthecounters.ItsequivalentcalculatedfromtheLOWADCchannelsisusedtomeasuretheshowerenergiesofhadronsandelectrons.ToconverttheenergymeasuredbythefourLOWchannelsofacounter(orbyaSUPER-LOWiftheLOWhassaturated)tothatwhichwouldbemeasuredbyalinear,nonsaturatingSUM-LO,HIchannel,weuseasetofinterchannelcalibrationconstants:Theconstantsajandßaremeasurabletoanaccuracybetterthan1%becauseofgoodoverlapinthedynamicrangeofthechannelsandbecauseweunderstandthesmallnonlinearitiesinourelectronics.(a,-0.8andß-12).Theseconstantsarefairlystableovertime.Majorchangesoccurwhenmalfunctioningelectronicmodulesarereplaced.Ouranalysissoftwaremaintainstablesofthesecalibrationconstantstoaccountforchangesovertime.Themuoncalibrationcalibratestheresponseofthecounterstoastandardsignal,themostprobablevalueofthemuondE/dxenergylossdistribution.Sinceeachcounterisuniformoveritsactiveareaandsinceallthecountersareidentical,thedE/dxenergydepositedbymuonsastheytraversethroughanyregionofanycounterisidentical.TheenergylossatthepeakofthedE/dxdistributionistakentobethestandardunitofenergydepositionbecausethemostprobableenergylossisexpectedtobeinsensitivetothemuonmome:_-tumwhilelargerenergylossesaredependentonit.TheSUM-HIADCchannelisusedtoobservetheintrinsicmuondE/dxenergylossdistribution.Atypi-calenergylossdistributionfromacounterisshowninfig.4.Becauseofdifferencesincountergain,themostprobablevalueoftheenergylossdistributionvariesfromonecountertoanother.Weestimatethemostprobablevalueofthesedistributionsbycalculatinga"truncated"meanabouttheirpeaks.Thetruncatedmeanisthemeanmuonenergylossofthedistributionbetween20and200%ofthemeanitself.OnlycleanlyaSUM-LOey;v_1:ajXLOWj,(3)j=1SUM-HI.q;=.8XSUM-LOq;,(4) 1030102UUmR100wWK.Sakwnotoetat./CalibrationoftheCCFRtargetcalorimeteri\tI\ti\tIf.i0100200300SUM-HI\tADCCountsFig.4.ThemuondE/dxenergylossdistributionfromtheSUM-HIADCchannelofcounternumber2?.Therearemanysmallplasticribsinsidetheliquidscintillationcounter.MuonsthatgothroughtheseribsproducethesmallpeaknearzeroADCcounts.identifiedsinglemuonsthataremomentumanalyzedbythetoroidspectrometerareusedinthesedistributions.Whiletheupperlimitonthemeancalculationreducesthesensitivitytotheunderlyingmuonmomentumdis-tribution,thereisstillasmalldependence.Asthemuonmomentumincreasesfrom10to300GeV/c,thetrun-catedmeanincreasesby3%.Individualenergylossentriesinthedistributionthatenterthetruncatedmeancalculationarecorrectedtoreducethismuon-momen-tumdependence.DetailsaregiveninappendixA.Duetotheattenuationoflightwithinacounteranddifferencesinitsphototubegains,itsenergyresponseisnotuniformovertheactivearea.Theresponseisnotnecessarilysymmetricamongthefourphototubesofacounternorisitnecessarilythesameamongdifferentcounters.Muonsareusedtocalibratethisposition-de-pendentenergyresponsebecausethemostprobableenergylossofitsdE/dxdistributioncorrespondstoastandardunitofenergydepositionandbecausethemuonfluxassociatedorproducedbytheacceleratorbeamisspreadthroughouttheactiveareaofthecoun-ters.ThemeasuredenergylossintheSUM-HIADCchannelcorrespondingtothemostprobabledE/dxofmuonsinterceptingthecounteratthelocation(x,y)isdenotedasDE,.(x,y).Thisresponsefunctionisex-tractedforeachcounterinthefollowingmanner.Theactiveareaofacounterisbinnedintoagridofsquares,andtheenergylossatthedE/dxpeakofmuonsinterceptingeachbinismeasured(viathetruncatedmean).Thesediscretemeasurementsoverthegridareparameterizedtoa15-term(c..d)polynomial:m+ne.4DE',(x,y)=\tE\tcm.nxmy".m+n=oThesizeofabinis23x23ccn?.Sincetherearere-sponsevariationswithinabin,wenormalizeeachen-ergylossentrythatgoesintoabin'sdE/dxdisuibutiontothatatthebincenterbeforefindingitsendlossatthedE/dxpeak.ThenormalizationfactorisAE,,(x',y)1AE,,(x,y),where(x,Y)isthebincenterand(x,y)isQmuon'sinterceptinthebin.Astheextractionprocessusesthefunctionitself,severalitera-tionsarerequired.Foranycounter,thefunctionisreasonablyflataboutacounter'scenterandsteepnearitsedges(seefig.5).SUM-HIequivalentenergymeasurements(eqs.(3)and(4)),whennormalizedwiththemuoncalibrationfunction,AE,,(x,y),arepresumablyindependentofintercounterandintercountergainvariations.Anor-malizedunitofthisSUM-HIenergydepositioniscalledthe"Fa4uivalentParticle".Themuoncalibrationfunc-tionistimedependentbecausetheattenuationlengthoflightinacounterand/orthephototubegainscanchange.WesymbolicallydenotethistimedependenceasAE,,(x,y,t).Changesinthespatialcomponentofthiscalibrationovertimehavebeenfoundtobemodest[6].Thuswemeasurethespatialdependenceofthesefunctionsatafewpointsintimeoverasix-monthneutrinoexposure.Wedenotethe"tmes)atwhichthesefunctionsaremeasuredas"t=0",andnowdiscusshowthesecalibrationfunctionsareusedatothertimes.TocalculateenergylossesinEquivalentParticleunitsusingDE,,(x,y,t=0),measurementstakenattimeTmustbeadjustedasiftheyweretakenatt=0:4SUM-HI.qi(t=0)=1B(0)Eaj(0)LOWl(t=0).,i-t0.5-1.0-1.5183-1.5-1-0.500.511.5x(m)Fig.5.ContoursoftherelativemuonresponsefunctionAE,(x,y)/AE,,(0,0)forcounternumber33.The(x,y)coor-dinatesarerelativetothecenterofthecounter. ere,a,;(0),\t(0)arethet=0interchannelcalibrationconstants(eqs.(3)and(4)),andLOWj(t=0)isamea-surementfromaLOWADCchanneltakenattimezwitha"t=Ttot=0"gainnormalization.Inprinciple,'snormalizationrequiresamuoncalibrationfunctione,&E,(x,y,t)exceptthataLOWADCchannelisu\ttomeasurethemuondE/dxinsteadoftheSUMHIA\tchannel.Thisresponsefunctionisforasinglephototuberatherthanthesumofacounter'sfourpototubes.Wedenoteitaslj(x,y,t),anddefineitas:aR(t)\tai(t)1,(x,y,t)=®E!,(x,y,t),\t(7)wherea,(t)andß(t)areinterchannelcalibrationcon-stants(eqs.(3)and(4))attimet.Tosimplifymatters,weassumethatacounter'stime-dependentgainvaria-tionisduetophototubegainchangesandisthereforedecoupledfrompositionalgainvariationsli(x,y,tà=f»,Agi(t).\t(8)Consequently,thet=,rtot=0gainnoimah-tionofaLOWADCchannelis:LoWi(t=0)=LoWi(t=z)gi(0),\t(9)g,(\t)wheregi(t)are"phototubegains"whicharemucheasiertomeasure.Measurementsofthesephototubegainsaremadeatthecenterofacounter,wherethestatisticsarelargest.Individualphototubegainsareextractedfromamea-surementofmuondE/dxattheorigin,®E,,(0,0,t),andfrommeasurementsofrelativegainsamongthephototubesofacounter.Indefiningtherelativegains,consideranidealsituationwherethereisshowerenergydepositedexactlyatacounter'scenter,andwheretherearenomeasurementfluctuationsfromnoiseorphotonstatistics.AnydifferencesofmeasuredenergylossamongthefourLOWADCchannels,LOW;,wouldbeduetodifferencesinphototubegains.Ourrelativegainwouldbedefinedas4LOW,/FLOW,,wherethesumoverjrunsoverthecounter'sfourphototubes.Astherearemeasurementfluctuations,therelativegainsaremeasuredonastatisticalbasis.Wedenotetheserelativegainsasr,(t).Therelationbetweeng,(t)ofeqs.(7)and(8)andthemeasurementsare:.8(t)\taj(t)fj(0,0)gj(t)=®Ee(0,0,1),\t(10)4fî(0"0)gi(t)f(010)9i(1)-ri(t),wherethesumsoverjrunoveracounter'sfourph&-tubes.Sincevaluesofboth®E,,(0,0,1)andr,(t)varyfromonescintillationcountertoanother,separatemea-surementsaremadeforeachcounter,andtheyaremadefrequently.Sakumotoetat/CalibrationoftheCCFRtargetcalorimeterAE1,(0,0,t)ismeasuredusingmuonsthatinterceptacounterwithinhalfameterofitscenter.Withinthisregion,thepositionaldependenceofacounter'senergyresponseisfairlyflat(seefig.5).Onaverage,thedifferenceinresponserelativetothecenterrangesfrom-296to1096.Wenormalizetheresponsetotheoriginusinga"bincentering''correction(seediscussionbe-neatheq.(5)).Theresponsefunctionusedinthenor-malizationistheoneextractedwit=0,AE(x,y,t=0).Sincethecorrectionissmall,usingthet=0responsefunctiondoesnotintroducesignificanterrorsinthiscalculation.®E,,(0,0,t)ismeasurableonaruntorunbasistoanaccuracyofseveralpercent,andittypicallydecreasesbyabout1096overthecourseofasix-monthneutrinoexposure.Priortoanyextensivedatataking,thephototubehighvoltageswereadjustedtoequalizetherelativegainsofacounter'sfourphototubes.A1MCi137CSsourcewaspositionedatthecenterofthecounter.Thehighvoltageswerethenadjustedtomakethedoanodecurrentsfromthefourphototubebasesequaltowithinf2596.Iftherelativegainsareideallybalancedbythishighvoltageadjustment,thefourphototubesofacoun-teroutputthesamesignalforexcitationsatthecenter.Duringdatataking,therelativegainsareextractedfromeventswithhadronsshoweringwithinaquartermeterofacounter'scenter.Eveninthissmallregion,thefractionoflightproducedbyshowerthatistransmittedtoaparticularphototubedependssignificantlyontheshower'sposition.Attheedgesoftheregion,thefrac-tionoftransmittedlightdiffersbyabout4096relativetothecenter.Weuseasimplemodelofenergyresponse(cf.f(x,y)ofeq.(8))tonormalizethesedifferences.Asinglefunctionisusedforallphototubes.Itisbasedonaphysicalmodeloflighttransmissionthroughthecountersthatusestwoattenuationlengths:oneforthescintillatoroilandotherforthelightpipes.Overthecourseofasix-monthneutrinoexposure,somecounter'srelativegainsdriftedbyasmuchas1096,whileothersremainedstable.4.TestbeamcalibrationTheCCFRtargetcalorimeterhasbeenabsolutelycalibratedtwicewithmomentum-analyzedhadronbeams.Thefirstcalibrationrunwasusingthespringof1984,ayearbeforetheE744neutrinorun.Thesecondcalibrationrunwasduringthewinterof1987,neartheendoftheE770neutrinorun.Bothofthesecalibrationrunswereinconjunctionwithmeasurementsofthemuonprods.^tionratefromhadronicshowers[7,8].TheNTW(neutrinotestwest)beamlineatFermilabisusedtocalibratethetotalenergyresponseoftheCCFRtargetcalorimeter.Sinceittransportsbeams directlyintotheCCFRexperimentalhall(LabE),vari-oustargetcartsofthecalorimeteraremovedhorizon-tallyintoandoutofthebeamforcalibrationwithoutuncabling.Vertically,thebeamispitchedontovariouslocations.Thishorizontalandverticaladjustmentofthebeampositiononthetargetcarts(i1maboutthecenter)permitscalibrationofnotonlythecenterofthecalorimeter,butalsoofitsouterregions.Thispoint-to-pointcalibrationgivestheresidual,positionaldepen-denceoftherelativenormalizationofenergytoEquiv-JentParticlesdescribedintheprevioussection.Inthecalibrationoftheouterregionsofthecalorimeter,weexpectnosignificantsideleakageofshowerparticlesbecausethepitchangleisunder25mradandbecausethereisatleasthalfameterofcalorimeterremainingtotheedge.Sincetherearetennuclearinteractionlengthstoasingletargetcart,high-energyhadronshowersmayleakoutthebackend[7].Therefore,thetargetcartsarealwayscalibratedasaset:carts2and1,carts3and2and1,etc.Thebeamconfigurationwherethebeamisdirectedintothecenteroftheleadingtargetcartisdesignatedascenteredbeam.Thebeamconfigurationwherethebeamisdirectedawayfromthecenteroftheleadingtargetcartisdesignatedaspitchedbeam.TheNTWbeamlinetransportssecondarychargedparticlesproducedfrominteractionsbetween800GeV/,7primaryprotonsfromtheTevatronandathick,38cmaluminumtarget.Secondaryparticlesarechargeandmomentumselectedbythebeamoptics.Pionsarethedominantcomponentofthesecondarybeam.Forpositivelychargedsecondarybeams,theprotonfractionabout2056at100GeV/candabout5056at300GeV/c.Atbeammomentaofabout50GeV/corless,theelectronfractionofthebeamisabout1056andincreaseswithdecreasingmomentum.Thekaonfractionis596[8]orless.Thefractionofmuonswithinthebeamaperture("beammuons")isontheorderofapercentanditincreaseswithdecreasingmomentum.Muonsoutsidethebeamaperture("halomuons")arealsotransportedtotheCCFRexperimentalhallbythefringefieldsoftheNTWbeam-lineoptics.Thesehalomuonsareofeitherchargeandofvariousmomenta.Themomentumofparticlesinthesecondarybeamisanalyzedbyamagneticspectrometerinanexperimentalhall(LabF)justupstreamoftheCCFRexperimentalhall.AdoubletofFermilabbeamtransportEPBdipoles[9)servesasthespectrometeranalyzingmagnet.4.1.1984calibrationrunAtthetimeofthiscalibrationrun,theCCFRneu-trinodetectorwasbeingupgradedfortheTevatronrun,E744,soitwaspartiallyinstrumented.Calorimetertargetcarts1,2,and3weresufficientlyinstrumentedforcalibrationswithgoodsystematiccrosschecks.TheW.K.Sakwnotoetal./CalibrationoftheCCFRtargetcalorimeter185calibrationoftargetcart2consistsof15,25,50GeV/ccenteredbeam.Thecalibrationoftargetcart3consistsof100,150,200,300,450GeV/ccenteredbeamand100GeV/cpitchedbeamontovariouslocationsuptot1mawayfromthecenter.Forthiscalibrationrun,theNTWbeamlineselectedandtransportedpositivelychargedparticles.Detailsofthebeam-linemagneticspectrometeraregivenelsewhere[7].Themeasuredmomentumspreadap/pofthebeamvariesfrom756at15GeV/ctoabout2%at100GeV/corhigher.Thefractionofbeampositronstransmittedbythisspectrometertothecalorimeterislargebecausetheamountofmaterialwithinthebeamaperturewassmall.Thesepositronsareeasilyidentifiedusingonlythecalorimeter.Forthe15GeV/ccentered-beamcalibration,a10cmlead"filter"wasplacedwithinthebeamapertureatapositionaheadofthespectrometermagnettoremovethesepositrons.Energycalibrationeventsweretakenonacoinci-dencebetweenabeamcrossingsignalandasignalofenergydepositionintheCCFRtargetcalorimeter.Asmallscintillationcounterjustinfrontoftheanalysismagnetprovidedthebeamcrossingsignal.Forthe15GeV/ccalibration,thetriggerwasjustthebeamcross-ing.Thebeamfluxwaskeptunder1000panicles/sforthecentered-beamcalibrationandunder5000par-ticles/sforthepitched-beamcalib-ation.Muoncalibra-tioneventsweretakenatseveraloccasionsduring'thiscalibrationrun(butnotsimultaneouslywiththeenergycalibrationpoints).Halomuonswereusedforthiscalibration.Themagneticfieldofthespectrometer'sanalysismagnetwasalsocalibrated.WemeasuredthecentralfieldofoneoftheEPBdipolesasafunctionofcurrentwithanNMR-calibratedHallprobe.4.2.1987calibrationrunDuringthiscalibrationrun,theCCFRneutrinode-tectorwasfullyfunctional.Thecalibration.oftargetcart2consistsof25,40,100GeV/ccenteredbeam.Thecalibrationoftargetcart3consistsof25,40,70,100,150,200GeV/ccenteredbeamand40,70,100GeV/cpitchedbeamontopoints(0m,1m)and(-1m,1m)fromthecenter.Thecalibrationoftargetcart4consistsof25,70,100,150,200GeV/ccenteredbeamand40,100GeV/cpitchedbeamontoasinglelocation(-1m,1m)fromthecenter.Thesespecificpitchedbeammeasurementswereselectedtocomplementthecalibra-tionmeasurementstakenin1984.Forthiscalibrationrun,theNTWbeamlinewassettoselectandtransportnegativelychargedparticlesatthestartofdatataking,andlaterswitchedtoselectandtransportpositivelychargedparticles.Duetocon-straintsonthepoweravailableforthetransportmag-nets,thebeammomentumwaslimitedto200GeV/c. 4.3.AnalysisyNTDC-A-~-DC-BDC-C---\'S4,S5DC-DMAGNET-S3\tS1,S2DC-EDC-F-(3mx3m).\tS\toetaL/CalibrationoftheCCFRtargetcaloimeterESTBEAMZ=O2.37m4.75m8.66m13.05m37.42mCART4-r---\t44.51m44.51mCART3-;CART2-CART1-Fig.6.Planviewofthe1987testbeamsp%trometer.Itcontainstwobendingmagnets,sixdriftchambers(DC-AthroughDC-F)andbeam-definingcounters(SlthroughS5).Thesedriftchambershaveamassofabout3g/cm2.NotshownarethethresholdCherenkovcountersinfrontofDC-AandbehindDC-E.Thelayoutofthebeamlinemagneticspectrometerisshowninfig.6.Itisnotexactlythesamespectrometersystemusedinthe1984calibrations.Althoughtheirangularresolutionsaresimilar,thebendanglethroughtheanalysismagnetisdifferent:5mradin1984,and20mradin1987.Inaddition,theamountofmaterialwithinthebeamlineapertureismuchlarger.Themeasuredmomentumspread(ap/p)ofthebeamis2%at25GeV/c,dipsto1%at100GeV/c,andincreasesto1.5at200GeV/c.Theamountofbeamelectronsreachingthecalorimeterisverysmall.Energycalibrationeventsweretakenonacoinci-dencebetweenabeamcrossingsignalandasignalofenergydepositionintheCCFRtargetcalorimeter.Acoincidenceamongthesmallbeam-definingscintillationcountersadjacenttotheendsoftheanalysismagnetprovidedthebeamcrossingsignal.Forthe25GeV/ccalibration,thetriggerwasjustthebeamcrossing.Thebeamfluxwaskeptwellunder1000particles/s.Muoncalibrationeventsweretakensimultaneouslywiththeenergycalibration.BothbeammuonsandhalomuonsfromtheNTWbeamlinewereused.Aseriesofanalysiscuts,summarizedinappendixB,areappliedtotheenergycalibrationdatatoremoveunsuitableevents.Theyinsurethateventsareproperlyreconstructed,thatshowersarecontainedlongitudinallywithinthecalorimeter,thateventscontainasinglebeamparticle,andthatbeamparticlespassthroughtheupstreammagneticspectrometerwithoutinteracting.Themeasurementofshowerenergystartsfromthemostupstreamcounterinthebeamandstopsatthe"showerend".TheshowerendisthemostdownstreamscintillationcounterthathasactivityinitsSBITdis-cr:atorIfs--fig.3',.Thesediscriminatorshaveathresholdsettingcorrespondingto2596ofthesingleminimumionizinglevel.Showerenergiesmeasuredbycounterswithinthisregionaresummedtogetthetotalobservedenergy.Normalizedcounterenergies(inEquivalentParticles)areusedinthesummation.Wdenotethistotalobservedenergyas"E".TheEquivalent-ParticlenormalizationusescountergaincalibrationsobtainedfromthemuoncalibrationdataForthe1984data,thenormalizedcounterenergyiscalculatedasitwouldbeforneutrinointeractiondata(seesection3).Forthe1987data,asimplerbutequiv-alentcalculationofthenormalizedcounterenergyisused.WemeasurethemuondE/dxattheshowercentroidposition,AEK(x,y,t).anddirectlynormalizethecounterenergieswithit.Thisismeasuredwiththebeammuonsofthecorrespondingenergycalibrationpoint.(Beammuonsinterceptthecountersattheshowercentroids,andatthesametime).Wedenotethemomentummeasuredbytheup-streammagneticspectrometeras"p"anduseGeVunits(i.e.c=1).Thesystematicerrorinthemomentummeasurementresultingfromuncertaintiesinthemag-neticfieldoftheEPB,dipolespectroscopymagnetsis0.56.Measurementsofthecentralfieldtakeninthe1984calibrationrunhavea0.5`6error.Thesemeasure-mentsagreewiththoseoftheEPBdipolereference[9]atthe0.595level.Weinferfromthisthatthecurrentversusmagnetic-fieldexcitationcurveisuniformfromoneEPBdipoletoanotheratthe0.595level.Wedidnotmeasurethemagneticfieldduringthe1987calibrationrun.However,wedidcalibratethecurrentandusethisinconjunctionwithapreviousmeasurement[10]oftheexcitationcurvetogetthemagneticfield.Sincetheresponseofcalorimeterstoelectromagneticandhadronicshowersareinprincipledifferent,weseparateelectronsfromthehadronsinthebeam.Elec-trons(positrons)areidentifiedbyexaminingtheratiooftheshowerenergymeasuredinthethreemostup-streamcounterstothetotalobservedenergy.Thosefirstthreecounterscover1.7nuclearinteractionlengthsor16radiationlengths.Forelectrons,theenergyratioiscloseto1;forhadrons,itisdistributedbetween0and1(seefig.7).Thepitchedbeamcalibrationrevealsanonlinearityinthedigitizedsignalsfromphototubes.Thisismostclearlyseeninthe100GeVpitchedbeamcalibrationoftargetcarts3and4takenin1987.Thebeamentersthe 020w200ASYM=(LOW2+LOWS+LOW4)/LOW,,0.400.350.300.25tnd0.200.150.10.NoninteractingIC~ParticlesW.K.Sakumotoetal./CalibrationoftheCCFRtargetcalorimeterInteractingHadronsElectrons0.20.40.60.81E(first3)/EFig.7.Thedistributionoftheratiobetweentheenergyinthethreemostupstreamcountersandthetotalobservedshowerenergy.Thisisfromthe1984centered-beamcalibration.targetcartat(-1m,1m)fromitscenter.Forthisentrypoint,phototubenumber1(seefig.2a)seesabout806ofthelightfromtheshowercascadegoingthroughacounter.Foranycounter,theamountoflightseenvariesconsiderablybecauseofthelargelongitudinalfluctuationsinherentinhadronicshowers.Ifthedig-itizedresponsesofthephototubes(LOW;)arelinear,thentheirasymmetry,0100020003000LOW,\tADCCounts(12)Fig.8.LOW,vsASYM(eq.(12))forcounter41intargetcart3.Itisfroma100GeV/cpitched-beamcalibrationof1981.EntriesintheLOW,300ADCcountregionarenotshown.Thesolidcurveistheparametrizationofthescattergivenineq.(13).For300GeV/ccentered-beamhadrons,themeanvalueoftheLOWADCchannelsisabout860ADCcountsandthehighendtailofthedistributionextendstoaround1800ADCcounts.187shouldbeconstantagainstLOWi.ThisisnotobsefvSincethereadoutelectronics(LOW,SUPERLOWADCchannels)areknowntobelinearat0.5orbetter,thephototuberesponseispresumedtobenon-linear.Fig.8showsthisnonlinearbehaviorfromacounterintargetcart3thatisaboutonenuclearinter-actionlengthdeep.Maximumenergyfluctuationsareexpectedinthisregionwherethehadroniccascadejustbegins.Asimilarbehaviorisobservedfromthe1987pitched-beamcalibrationoftargetcart4andfromthe1984pitched-beamcalibrationoftargetcart3.Analysisofthecentered-beamcalibrationindicatethatthetotalenergyresponseisconsistentwithbeinglinearuptoabeamenergyof200GeV.Fromthis,wesurmisethatthephototube.responseseeninaLOWADCchannelislinearuptoacutofflevelof1000ADCcountsandnonlinearthereafter.Thustheriseinthepitched-beamasymmetryshowninfig.8isonlyfromthenonlinearbehaviorofphototubenumber1becauseitaloneisdrivenpastthecutofflevel.Wefitthisrisetoasingle-parameter"a"bentline;whenLOW,S1000ADCcounts,ASYM=a,andwhenLOW,�1000ADCcounts:ASYM=a[1+10-4(LOW,-1000).\t(13)Thisparameterizationnominallyaccountsforthetypi-calbehaviorofASYMseeninothercounters.Itisthususedtocorrectthenonlinearityofeveryphototubeofallcounters;whenLOW�1000ADCcounts:LOW#,=Low[1+10-4(LOW-1000)],\t(14)otherwise,thereisnocorrection.Theoveralleffectofthisphototubenormalizationissmall.Itincreasestheaveragetotalenergyofthe100GeVpitchedbeamcalibrationbyabout1.5%.5.Results5.1.HadroncalibrationTheaccuracyofthemuoncalibrationaffectstheshowerenergynormalizationintoEquivalentParticleunits.Forthetestbeamcalibration,themeasurementerrorontheEquivalentParticleunitis2-3%.Thisintroducesabouta1%erroronthetotalenergymea-surementofhadronshowers.Inhadronshowers,backscatteredsecondariesocca-sionallyescapethroughthefrontofthecalorimeter.Hadronsthatpenetratebeforeinteractingshouldonaveragehaveslightlylargertotalenergiesbecauseofbettercontainment.Toseeifthisaffectsthecalibration,wemeasurethetotalenergyforhadronswithnopenetrationrequirementandforhadronsthatinteractafterpenetratingbeyondthemostupstreamcounter.PenetratinghadronsarerequiredtohaveadE/dxloss 4.854.60IW4.754.704.65'.!.._L!I177518001825185018751900RunNumberFig.9.Theenergyresponseoftargetcarts2,3and4to25to200GeVhadronsfromthe1987centered-beamcalibration.totalhadronshowerenergyEisinEquivalentParticle(EP)units.Therunnumberscorrespondingtoeachtargetcartc4'"brationare:cart2(1770-1810);cart3(1811-1851,1875-1900);cart4(1852-1874).Onlystatisticalerrorsareshown.consistentwithminimumionizing:under2.5EquivalentParticlespercounter.Weobservenosignificantdif-ferencesfor25or100GeVcentered-beamcalibrations.ThetotalenergyresponseandresolutionoftheCCFRcalorimeterareobtainedfromthecentered-beamcalibrations.Fortheresponse,weplot)asafunctionof)orofthetargetcartbeingcalibrated.Here,---)representmeansofdistributionsfromthevariouscentered-beamcalibrations,"P"isthemomen-tumofabeamparticleasmeasurer:bytheupstreammagneticspectrometer,and"E"isthetotalenergymeasuredinthecalorimeterforthatparticle.Iftheenergyresponseislinear,then)isconstant.Fig.9showstheenergyresponseasafunctionofthetargetcartbeingcalibrated.Thereisnodifferenceinthevaluesofthe)amongtheidenticaltargetcartsatthe$1%level.Fig.10isthetotalenergyresponseasafunctionof&#xP000;obtainedfromthe1987centered-beamcalibration.Fig.11isthetotalenergyresponseasafunctionof)obtainedfromthe1984centered-beamcalibration.Thesefiguresdemonstratethattheresponsebetween15and450GeVislinearattheonepercentlevel.Theremaybeadroopintheresponseat450GeV(seefig.11),butitisattheonepercentlevel.Inaddition,thesefiguresdemonstratethatthetotalenergyresponseisthesameforanyexperimentperformedontheCCFRtargetcalorimeteriftheEquivalentParticlecounterenergynormalizationisused.ecalorimeter'stotalenergydistributionsaremorePoissoninshapethanGaussian,especiallyatlowbeamenergies.Intheparameterizationofthesedistributions,:\tSolr\toetal/CalibrationoftheCCFRtagetcalorimeter4.854.80fsl4.75nawv4.70I\tI\tI\tI\tI050100150200&#xP000;GeVFig.10.TheenergyresponseoftheCCFRcalorimeterusinghadronsfromthe1987centered-beamcalibration.Thetotalhadronshowerenergy,E,isinEquivalentParticle(EP)units.Allthecalibrationpointsshowninfig.9areintheplot.Thehorizontallineistheweightedmean(4.737EPs/GeV)usingthestatisticalerrorindicatedoneachpoint.Thefractionalrmsdispersionaboutthemeanis0.7`6.4.65weuseaPoissondensityextendedtoaccommodatenonintegralarguments:wheretheargumentsareunitlessandxisthemean.Thefitusestwofreeparameters:ascalingparameters4.854.80W4.75naw4.704.65(15)100200300400500&#xP000;GeVFig.11.Theenergyresponseusinghadronsfromthe1984centered-beamcalibration[111.Theshowerenergy,E,isinEquivalentParticle(EP)units.Thehorizontallineisthemeanvalueofthe1987calibrationshowninfig.10.Theweightedmeanusingthestatisticalerrorindicatedoneachpointis4.721EPs/GeV.Thedifferenceiswithinthe0.56systematicerrorfromthemagnetic-fieldintegrals.Thefractionalrmsdispersionabouteithermeanis0.8%. 103102101w10010-1Fig.12.Thetotalenergydistributionsof25and200GeVhadronsfromthe1987centered-beamcalibration.ThetotalshowerenergyEisinEquivalentParticle(EP)units.ThesolidcurvesarethePoisson-likeparameterizationsofthedistributions.Forthe25GeVdistribution,eventsinthehigh-Ere-gionswerescannedandfoundtobeacceptable.thattransformsthetotalenergytotheunitlessvariablex=E/sandthemean.Fig.12showsthe25and200GeVtotalenergydistributionsfromthe1987centered-beamcalibration.Thestandarddeviationsandmeansfromthesefits,a=s(x)t12andE=sx,areusedtoobtainthefractionalsamplingresolutiona/E,whichisplottedasafunctionofthebeammomentum(P)infig.13.Thisresolutionisparameterizedas:(0.847f0.015)+(0.297±0.115)\t16)wherePisinGeV.The1/Ptermisanoisetermanditisconsistentwithindependentlymeasuredbeam-relatednoiseinthecalorimetrycounters.Thepitched-beamcalibrationsrevealthattheEquiv-alentParticlenormalizedtotalenergyresponseisslightlydependentonposition.Inthesecalibrations,thehadronbeamisdirectedtovariouslocationsaboutthecenterof1-1w1.000.95b0.900.850.80WKSaktunotoetal./CalibrationoftheCCFRtargetcalorimeter50\t150\t250'\t800100012001400E(EPs)o=1984Data_-1987Datal\tI\tI0100200300400500PGeVFig.13.ThehadronshowerenergyresolutionoftheCURcalorimeterfrom25to450GeVcentered-beamcalibrations.Thecurveistheparameterizationgivenbyeq.(16).DE,,(x,y,t=0)AEm(0,0,t\t0)F,AE'189thecalorimeter.Thistestshowwellthemuoncalibra-tionfunction,AE,,(x,y,t=0),normalizestheposi-tionalenergyresponsevariationofthecounterstomuchlargerenergydepositionsfromhadronshowers.WefindthattheEquivalentParticlenormalizedtotalenergyEislowerthanexpectedawayfromthecenterofthecalorimeter,wheretherelativemuondE/dxresponseofthecounters:(17)getslarge(seefig.5).Thatis,themuoncalibrationfunctiontendstoovercompensatetheresponseofthecounterstohadronshowersthatareoutsideofthecalorimeter'scentralregion.Thisovercompensationisrelativetothehadronicresponseatthecenterofthecalorimeter,wheretheresponsehasbeendeterminedfromthecentered-beamcalibration.Toquantifythisovercompensation,weuseaparameterwhichisap-proximatelytheratioofthehadronicshowerresponseatlocation(x,y)relativetothecenter:F®E'R'(x,y)"\t(18)Here,AE'istheEquivalentParticlenormalizedenergyinacounter,R,,(x,y)istherelativemuonresponseofacounter(eq.(17)),andthesumsrunoverallcounterscontainingtheshower.Fig.14givesthefractionaldevi-ationofthetotalenergyfromwhatisexpected,8E/E,Fig.i4.Thefractionaldeviationofthetotalhadronenergymeasuredbythecalorimeterfromthatwhichisexpected[121asafunctionof1/R(x,y).HereR(x,y)istheestimatedhadronicresponseofatargetcartatlocation(x,y)relativetoitscenter(seeeq.(18)).Thebentlineisaparameterizationofthedatagivenbyeq.(19).ThedistributionofthepointsaboutthelineisconsistentwithbeingGaussian,andthemeanandrmsare-0.296and1.296,respectively. ôU0.97A(D0.96w1.005.2.Electroncalibrationin;\tSak\toetal/CalibrationoftheCCFRtargetcalorimetero=15GeVa=25GeV0=50GeV+=100GeVA=200GeV=300GeV=450GeV0.95rAl.1015202530CalorimeterLength(counters)Fig.15.Thefractionofhadronicshowerenergycontainedasafunctionofthelengthofthecalorimeter.Thedataarefromthe1984centered-beamcalibration.Showersarerequiredtostartwithinthetwomostupsteamcountersinthebeam.Theerrorsareabout0.5\tatalengthoftencountersanddecreasetoabout0.1\tat20counters.Ncountersare(11.74N-6.6)cmofsteelequivalent.asafunctionof1/R(x,y).Weparameterizethiswitha\ttline;whenR(x,y)�0.82-':8=0.161~R\t-0.82,\t(19)(x,y)otherwisethedeviationiszero.Thisnormalizestheresidualpositiondependenceofthetotalenergyre-sponseforshowersawayfromthecenterofthecalorim-etertothatatitscenter.Withthisfinal\ton,thehadronicenergyresponseofthecalorimeterisuniformwithpositionandlinearatthe±1%level.Thehadronicenergyresponseisthusrepresentedbysinglenumber,thecalibrationconstantbetweenEquivalentParticles(EPs)andGeV:4.730i0.018EPsperGeV[13].Here,resultsfromthe1984and1987centeredbeamhadroncalibrationshavencombinedandtheerrorincludessystematicerrors.Theenergyresponseandresolutionpresentedarefor100%longitudinalcontainmentofthehadronshower.Partialcon'\tentissimulatedbyremovingcountersattheendofthecalorimeterintheanalysissoftware.eresultofthis\t'alcontainmentsimulationisshowninfig.15.Fromtheelectroncomponentofthe25and50GeVbeamsofthe1984centeredbeamcalibration,wehaveeasuredtheEquivalentParticleandGeVcalibrationconstantforelectronshowers.Thisisforshowersthatemiddleofthetwosteelplatesbetweencounters(seefig.2a),anditisrelatedtothecalibrationconstantforelectromagneticshowersinducedbyhigh-energymuons.Suchshowersareproduceduniformlyoverthetwosteelplatesbetweencounters.Whiletheresponseofthecalorimeterdependsonwhereinthesteelthemuon-inducedshowerbegins,theresponseaveragedoverthetwosteelplatesbetweenthecountersshouldequalthemeasuredelectronresponse.Inaddition,wehavealsoextractedthecalibrationconstantforminimumionizingmuons.Neutrino-in-ducedcharged-currentreactions(eq.(1))wherethefi-nal-statemuonrangesoutinthecalorimeter(fromtheE744run)areused.Toavoidanypunclhthroughfromtheshowerofthefinalstatehadrons,themuonenergylossismeasuredinaregionsufficientlyfarfromtheendofthehadronshowerwherethereisonlythemuon.Withinthisregion,thetotalenergydepositedbythemuoninthecaalorimetrycounters,E,andthepathlengthoverwhichthisenergyisdeposited,L,aremea-sured.ThepathlengthsLofthesemoonsarebetween1.1and6.1mofsteelequivalent.ThemuonenergyinGeViscalculatedusingthepathlengthLandrange-energytables[14].Wedenotethisenergy!obtainedfromLandtherange-energytablesasP,andtheyarebetween1.5and8.7GeV.ThedistributionofE/Pfortherangeoutmuonsisshowninfig.16.TheresponseoftheCCFRtargetcalorimetertohadronicshowers(%),electromagneticshowers(e),andminimumionizingmuons(p)isdifferent.IntermsofthecalibrationconstantsbetweenEquivalentParticles(EPs)andGeV,theyare:it=4.73f0.02EPs/GeV,e=5.25i0.10EPs/GeV,\t(20)pi=6.33i0.17EPs/GeV,50.0°5.0s.1.00.5ff24S810E/P(EPs/GeV)Fig.16.TheE/rPdistributionforminimumionizingrangeoutmoons.Thetotalenergymeasuredfromthecalorimetrycoun-ters,E,isinEquivalentParticle(EP)units.Themuonenergyobtainedfromthepathlengthandtherange-energytables,P,isinGeV.Themeanandrmsofthedistributionare6.33and1.04EPs/GeV,respectively.ThecurveisaGaussianpara-meterizationofthedata. whereerrorsincludebothstatisticaland(dominant)systematicerrors.Thehadronicenergyresolutionisgivenbyeq.(16).Forelectromagneticshowers,theresolutionLQe/E=0.60/E[GeV].Fortheminimumionizing,rangeoutmuons,theresolutionisa,/E=0.17.6.STheenergycalibrationoftheCCFRtargetcalorime-terconsistsoftherelativegainnormalizationamongthescintillationcounterswithmuonsandthemeasurementofitsenergyresponseandresolutionwithatestbeam.Thehadronicenergyresponseandresolutionhasbeenmeasuredforbeamenergiesbetween15and450GeV.Theleveloferroronourknowledgeoftheresponseis1%.Therearenodifferencesinresponseamongthetargetcartsofthecalorimeterthatweretested.Thisisexpectedbecausethesixcomponenttargetcartsofthecalorimeterareidentical.Inaddition,thehadronicen-ergyresponseandresolutionobtainedfromthetwotestbeamcalibrationsareconsistentwithoneanother.Thatis,theenergycalibrationisthesameandcanbeusedwithoutadjustmentforthetwoneutrinodataruns,E744andE770,whichweretakenbetweenthetwotestbeamcalibrations.Wehavemeasuredtheenergyresponseandresolu-tionfor25and50GeVelectronsandwehavemeasuredtheenergyresponseoflow-energy(lessthan9GeV)minimumionizingmuons.Theirresponseisdifferentfromthatofhadronsandwefindthatthe%randIL/qrratiosare1.11and1.34,respectively.AcknowledgementsWegratefullyacknowledgethecontributionsofJeanBirkenmaierandKenGrayfortheirinvaluabletechni-calassistance.We.1%lsothanktheFermilabstaff;espe-ciallytheExperimentalAreaspersonnel,fortheirsup-port.Inparticular,wethankGordonKoizumiforassis-tanceintheoperationoftheNTWbeamline.ThisworkissupportedinpartbytheU.S.DepartmentofEnergy.AppendixAEnergydefinitions(1)MuondE/dx:thetruncatedmeanThetruncatedmeanisthemeanofthemuonenergylossdistributionbetween20and20096ofthemean.Sincethemean'slimitsaredependentonthemeanitself,thetruncatedmeaniscalculatediteratively.IntheW.K.SakwnotoetaL/CalibrationoftheCCFRtargetcalorimetercalculation,individualenergylossesinthedistribution,E,areshiftedtoaccountforthemomentumPdenceofthehighendoftheenergylossdistribution=0.00967In77[GeVc/]AE=SUM-HIgw/DE,,(x,y,t),AppendixAnalysiscutsß(t)~a~(t)LOW~(t)G~(0)/G~(t)DE=\tAEI,(0,0,t)R,,(x,y)\t'Gi(t)=4a;(t)r:(t)/Ta.(t)r;(t),Onaveraging,thisfurtherreducestheresidual\t-momentumdependencyofthetruncatedmean.Thetruncatedmeanisabout1496largerthanthemostprobablyenergyloss.Betweenthelimitsofthetrun-catedmean,thermsaboutthemeanisapproximately4096ofthetruncatedmean.(2)EquivalentParticleenergyTheenergylossmeasuredbyascintillationcounterexpressedinEquivalentParticleunitsis:whereSUM-HIgwisdefinedbyeqs.(3)and(4).Sinceweuseacounter'st=0muoncalibration,AE,,(x,y,t=0),thecorrespondingSUM-HI,,;(t=0)fromeqs.(6)and(9)isused.Thephototubegains,g;(t),aredefinedb,,-,eqs.(10)and(11),andareobtainedfromthemeasuredmuondE/dxatthecenterofthecounterandtherelativegainsofthecounter.Combiningthemgives:whereallsumsrunoverthefourphototubesofacounter.R~,(x,y)istherelativeresponseofacountertomuons(eq.(17)).(1)CleanbeamcutsOnlyeventsconsistentwithasinglebeamparticlewithintheupstreammagneticspectrometerareallowed.TheremustonlybeasinglehitintheTDCviewingthebeamcrossingsignal.Upstreamshowersandbeamas-sociatedhaloarerejectedbyrequiringonlyonehitwithinthelargeupstreamdriftchamber(ineitherthexoryview),andbyrequiringthatitslocationbecon-sistentwiththatofthebeamposition.Thisdriftcham-berisDC-Finfig.6.(2)MomentumreconstructionThetailsofthereconstructedmomentumdistribu-tionareremovedbyapplyingaf10%cutaboutthemean.Af2096cutisappliedifthenominalbeammomentumisunder30GeV/c.Duringthe1987 ofso\tdataweretakenwhilethecurrenttoyEPdipoleswasbeingramptoitsisnotused.(,3)\tgi\t'nalshower\tdoriese1\tforshowerstructureabovethesingleminimumionizinglevelthatisintimewiththeeventtrigger.'sstructureisestablishedusingtheSBITdiscriminatorsofthescintillationcounters(\tfig.3).Thesediscriminatorsfirewhentheenergydepositionincounterislargerthan25\toftheminimumionizinglevel.\tetimeofarrivalofthesehitsareviewedbytheT\tIftherearenoSBIThits,oriftheSBIThittimeofacounterisnotwithini2rfbuckets(18.8nsperrfbucket)ofthatforthetriggeringbeamparticle,thecounterisconsideredtobeinactive.Thisgivesalongi-tudinalmapoftLeenergydepositionwhichiscorre-latedtotheshoweringbeamparticle.Theensureshowerenergycontainment,aneventisrejectedifparticlespenetratetowithinthreecountersoftheendofthecalorimeter.Penetratingmuonsarealsorejectedbythiscut-WY-Sak\toetal./CalibnuionoftheCCFRtargetcalorimeter(4)ElectrwrsDefineR3astheshowerenergyinthethreemostupscountersdividedbythetotalobservedenergy(seefig.7).EventswithR3zt0.96areidentifiedaselectrons.Fcrthe1984calibrationrun,theelectronsamplehasabouta10\thadroncontamination.Forthe1987calibrationrun,theelectronsamplehasmorethan80\thadroncontaminationandisnotused.[1)F.L.Navarria,Nucl.Instr.andMeth.212(1983)125.[2]MarvelguardEC1Aisanopaque,conductiveanti-stat(MIL-B-81705),LudlowLaminatingandCoating,Homer,LA,USA.[3]Hexcelstructuralpanel,HexcelCorp.,Dublin,CA,USA.[4]P.Lasenetal.,Nud.Instr.andMeth.185(1981)67.[5]B.Barishetal.,IEEETrans.Nucl.Sci.NS-25(1978)532.[6]Therelativemuonresponsefunction,®E,,(x,y,t)/0E,,(0,0,t)changesslowlywithtime;nearthecorners,(f1.3m,f1.3m),theworst-casecounterhadchangesoff30%overasix-monthneutrinoexposure.[7]F.S.Merrittetal.,Nucl.Instr.andMeth.A245(1986)27.[8]P.H.Sandleretal.,HadronShowerPunchthroughandMuonProductionbyHadronsfor40,70,and100GeV,Univ.ofWisc.,Madisonpreprint:WISC-EX89-306,tobepublishedinPhys.Rev.D.[9)R.Juhala,FenmilabTM-434(1973).[10]Ref.[9],SecondGroup,tableI,appendixB.[ll]Forthe15GeVcalibration,anadditional2%normaliza-tionerrormustbeincludedinthestatisticalerrorshown.Thisisduetoamuonnormalizationuncertaintyfrominfrequentmuoncalibrationsandamalfunctioningscintil-lationcounter(counter26atshowermaximum).The150GeV,oneofthe300GeV,andtwoofthe450GeVpointsaremeasurementsathigherbeamintensitiesofabout60000particlespersixsecondspill.[12]Forthe1984calibrationpoints,thebeammomentumis100GeV.Theexpectedenergyistheaveragetotalenergyofbeamintothecenteroftargetcart3.Forthe1987calibration,theexpectedenergyisbasedonthebeammomentumandthetotalenergyconversionof4.737EquivalentParticlesperGeV.[13)NotethatthedefinitionofanEquivalentParticleisbasedonthetruncatedmeanestimate(seeappendixA)ofthemostprobablemuonenergylossofacounter.Asthistruncatedmeanisabout1496largerthanthetruemostprobableenergyloss,thecorrespondingcalibrationcon-stantbetweenEquivalentParticlesbasedonthetruemostprobablemuonenergylossandGeVisabout5.4(=4.73x1.14).[14]G.Koizumi,FernWabTM-786(1978).Indeterminingthemuone&.-gyfromtheenergy-rangetables,wehaveused11.49cmofFeequivalentasthepathlengthofironbetweenthecalorimeteycounters.