HAL Id: in2p3-00021539
http://hal.in2p3.fr/in2p3-00021539
Submitted on 14 Jan 2004
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Performance of the Atlas electromagnetic calorimeter barrel module 0
B. Aubert, J. Ballansat, A. Bazan, B. Beaugiraud, J. Boniface, François Chollet, J. Colas, P. Delebecque, L. Di Ciaccio, N. Dumont-Dayot, et al.
To cite this version:
B. Aubert, J. Ballansat, A. Bazan, B. Beaugiraud, J. Boniface, et al.. Performance of the Atlas
electromagnetic calorimeter barrel module 0. Nuclear Instruments and Methods in Physics Research
Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, Elsevier, 2003, 500,
pp.202-231. �in2p3-00021539�
CERN{EP/2002{087
November7,2002
Performance of the ATLAS electromagnetic
calorimeter barrel module 0
The ATLAS Electromagnetic LiquidArgon Calorimeter Group
ABSTRACT
The construction and performance of the barrel pre-series module 0 of the future ATLAS elec-
tromagnetic calorimeter at the LHC is described. The signal reconstruction and performance of
ATLAS-likeelectronicshasbeenstudied.Thesignalto noiseratiofor muonshasbeenfoundtobe
7.110.07.Anenergyresolutionofbetterthan9.5%GeV 1=2
= p
E(samplingterm)hasbeenobtained
withelectron beams ofup to 245 GeV. The uniformity of the response to electrons inanarea of
=1:20:075hasbeenmeasuredtobebetterthan0.8%.
(Submitted to Nucl. Instr. and Meth. A)
)Seenextpagesforthelistof authors.
L. Di Ciaccio, N. Dumont-Dayot, M. El Kacimi a
, O. Gaumer, P. Ghez, C. Girard, M. Gouanere, H.
Kambara,A.Jeremie,S.Jezequel,R.Lafaye,T.Leour,C.LeManer,J.Lesueur,N.Massol,M.Moynot,
L.Neukermans,P.Perrodo,G.Perrot,L.Poggioli,J.Prast,H.Przysiezniak,X.Riccadona,G.Sauvage,
J.Thion,I.Wingerter-Seez,R.Zitoun,Y.Zolnierowski
Laboratoire de Physique de Particules (LAPP), IN2P3-CNRS, F-74019 Annecy-le-Vieux Cedex,
France.
H. Chen, M. Citterio b
, J. Farrell, H. Gordon, B. Hackenburg, A. Homan, J. Kierstead, F. Lanni,
M.Leite c
,D.Lissauer,H.Ma,D.Makowiecki,V.Radeka,D.Rahm,S.Rajagopalan,S.Rescia,I.Stumer,
H. Takai,K.Yip
BrookhavenNationalLaboratory(BNL),Upton,NY11973-5000,USA.
D. Benchekroun,C.Driouichi,A.Hoummada,M.Hakimi
FacultedesSciencesAnChock,Casablanca,Morocco.
R.Stroynowski,J.Ye
SouthernMethodistUniversity,Dallas,Texas75275-0175,USA.
J.BeckHansen,A.Belymam,J.Bremer,J.L.Chevalley,P.Fassnacht,F.Gianotti,L.Hervas,C.P.Marin,
P.Pailler,P.SchillyW.Seidl,J.Vossebeld, V.Vuillemin
EuropeanLaboratoryforParticlePhysics(CERN),CH-1211Geneva23,Switzerland.
A. Clark,I.Efthymiopoulos,L.Moneta
UniversitedeGeneve,CH-1211Geneva4,Switzerland.
B. Belhorma, J. Collot, P. de Saintignon, D. Dzahini, A. Ferrari, M.L. Gallin-Martel, J.Y. Hostachy,
P.Martin,J.F.Muraz,F.Ohlsson-Malek,S.Saboumazrag
InstitutdesSciencesNucleaires,UniversiteJosephFourier,IN2P3-CNRS,F-38026Grenoble,France.
J.Ban,N.Cartiglia d
, H.Cunitz,J.Dodd,A. Gara,M.Leltchouk,S. Negroni,J.A. Parsons,M.Seman,
S. Simion,W.Sippach,W. Willis
NevisLaboratories,ColumbiaUniversity,Irvington,NY 10533,USA.
F.Barreiro,G.Garcia e
,L.Labarga,S.Rodier f
,J.delPeso
PhysicsDepartment,UniversidadAutonomadeMadrid,Spain.
C. Alexa g
, P. Barrillon,C. Benchouk, A. Chekhtman, BDinkespiler,F. Djama,P.Y. Duval,F. Henry-
couannier,L. Hinz h
,M.Jevaud,P.Karst,A.LeVanSuu,L.Martin,O.Martin,A.Mirea i
,E.Monnier,
E.Nagy,D.Nicod,C.Olivier,P.Pralavorio,B.Repetti,M.Raymond,D.Sauvage j
,S.Tisserant,J.Toth k
,
M.Wielers l
CentredePhysiquedesParticulesdeMarseille,Univ.Mediterranee,IN2P3-CNRS,F-13288Marseille,
France.
G.Battistoni,W.Bonivento m
,L.Carminati,D.Cavalli,G. Costa,M.Delmastro,M.Fanti,L.Mandelli,
M.Mazzanti,L.Perini,S. Resconi,G.F.Tartarelli
Dipartimentodi Fisicadell'Universitadi MilanoandINFN,I-20133Milano,Italy.
V. Aulchenko, V. Kazanin, G. Kolachev , V. Malyshev, A. Maslennikov,G. Pospelov, R. Snopkov, A.
Shousharo,A.Talyshev,Yu. Tikhonov
BudkerInstituteofNuclearPhysics,RU-630090Novosibirsk,Russia.
E. Auge, C. Bourdarios, D. Breton, P. Cros, C. de La Taille, I. Falleau, D. Fournier, G. Guilhem,
S. Hassani,Y. Jacquier,K.Kordas i
,G.Mace,B.Merkel, J.M.Noppe,G. Parrour,P.Petro,P. Puzo,
J.P.Richer n
, D.Rousseau,N. Seguin-Moreau,L.Serin,V. Tocut,J.J.Veillet,D. Zerwas
Laboratoiredel'AccelerateurLineaire,UniversitedeParis-Sud,IN2P3-CNRS,F-91898OrsayCedex,
France.
F.Astesan,W.Bertoli,A.Camard,B.Canton,S.Fichet,F.Hubaut,D.Imbault,D.Lacour,B.Laforge,
O.LeDortz,D. Martin,I.Nikolic-Audit,F.Orsini,F.Rossel,P.Schwemling
LaboratoiredePhysiqueNucleaireet deHautes Energies,UniversiteParisVIet VII,IN2P3-CNRS,
F-Paris,France.
W. Cleland,J.McDonald
DepartmentofPhysicsandAstronomy,UniversityofPittsburgh,Pittsburgh,PA15260,USA.
E.M.Abouelouafa,A. BenMansour,R. Cherkaoui,Y.ElMouahhidi, H.GhazlaneandA.Idrissi
Faculte des Sciences and Centre National de l'
Energie des Sciences et des Techniques Nucleaires,
Rabat,Morocco.
J.Belorgey,R.Bernard,M.Chalifour,A.LeCoroller,J.Ernwein,B.Mansoulie,J.F.Renardy,J.Schwin-
dling,J.-P.Taguet,J.Teiger
CEA,DAPNIA/ServicedePhysiquedesParticules,CE-Saclay,F-91191Gif-sur-YvetteCedex,France.
C.Clement,B.Lund-Jensen,J.Lundqvist,L.Megner,M.Pearce,S. Rydstrom
RoyalInstituteofTechnology,Stockholm,Sweden.
J.Egdemir,R.Engelmann,J.Homan,R. McCarthy,M.Rijssenbeek,J.Steens
StateUniversityofNewYork,StonyBrook,NewYork11794,USA.
(a)VisitorfromLPHEA,FSSM-Marrakech(Morroco).
(b)NowatDipartimentodiFisicadell'UniversitadiMilanoandINFN,I-20133Milano,Italy.
(c)AlsoatSaoPauloUniversity.WorkpartiallysupportedbyFAPESP/SaoPaulo(Brazil).
(d)NowatDipartimentodiFisicadell'UniversitadiTorino,I-10125Torino,Italy.
(e)Nowat"Instituto NicolasCabrera",U.A.M.Madrid.
(f)SupportedbytheTMR-MCurieProgramme,Brussels.
(g) Also at Institute of Atomics Physics,NationalInstitute for Physics and Nuclear EngineeringIFIN-HH, Bucharest,
Romania.
(h)NowatUniversitedeLausanne,FacultedesSciences,InstitutdePhysiquedesHautesEnergies.
(i)NowatUniversityofToronto,Dept.ofPhysics,Toronto,Canada.
(j)Deceased.
(k)AlsoatKFKI,Budapest,Hungary.SupportedbytheMAE,France,theHNCfTD(ContractF15-00)andtheHungarian
OTKA(ContractT037350).
(l)NowatTRIUMFlaboratory,Vancouver,Canada.
(m)AlsoatINFN,SezionediCagliari,Italy.
(n)NowatInstitutdesSciencesNucleaires,UniversiteJosephFourier.
TheelectromagneticcalorimeteroftheATLASexperimentatCERN'sfutureproton-protoncollider,
the LHC, is alead-liquid argonsampling calorimeter with accordionshaped absorbersand electrodes.
Liquidargoncalorimetryhasbeenchosenbecauseofitsintrinsiclinearbehavior,stabilityoftheresponse
andradiationtolerance.
Results of R&D work were reported in [1, 2, 3, 4]. In this paper the construction and tests of
the pre-series module 0of the barrelcalorimeter are described. The module has the same dimensions
as the 32 modules of the future barrel calorimeter of ATLAS. As a result of extensive Monte Carlo
optimizations [5], in particular the study of = 0
separation,the granularityof the presamplerand of
thecalorimetermodulewaschangedwithrespecttothepreshowerdetectorandtheprototypedescribed
in [4]. The granularity of the presampler is coarser, while the granularity in of the rst calorimeter
samplingisneranditsdepthisdecreased.Inadditionthelength()ofthemoduleiscoveredwithonly
twoelectrodesinsteadof vetominimize thedeadzones betweenelectrodes.Themodule 0electronics
havethefullfunctionalityrequiredforATLAS,withtheexceptionthatnotallcomponentsareresistant
to thehighradiation levelsexpectedattheLHC.
Themodule wasthelastprototypetobetestedbeforetheproductionoftheseriesmodulesofthe
ATLASbarrelcalorimeter.Itspurposewastovalidatetheconstructionchainandtostudytheelectronics
performance.Themodulewasexposedto electronandmuonbeamsintheNorthArea atCERN's SPS
in ordertostudytheenergyresolutionandtheuniformityofthecalorimeterresponse.
The paper is organized as follows. In Sections 2 and 3 the construction and quality control of
the calorimeter module and of the presampler are discussed. The electronics chain is then described
(Section 4), followed by the description of the beam setup (Section 5). The modeling of the detector
responseisdescribedinSection 6,thesignalreconstruction andperformancearediscussedin Section7.
TheresponsetomuonsandelectronsisdiscussedinSections8and9respectively,followedbyconclusions
andprospects.
2 CalorimeterModule0Construction
The ATLAS barrel electromagnetic calorimeter, described in detail in [6], is made of two half-
barrels, centered around the z-axis (ATLAS beam axis). One half-barrel covers z> 0 (pseudorapidity
> 0) and the other onez< 0( < 0), from jj = 0to jj = 1:475;the length of each half-barrel is
3.2m,theinnerandouterdiametersareabout2.8mand4mrespectively.Ahalf-barrelismadeof1024
accordionshapedabsorbers,interleavedwithreadoutelectrodes.
Theelectrodesarekeptinthemiddleofthegapbyspacersmadefromstripsofhoneycombmaterial,
joined into anaccordionshapedsheet by resin impregnatedthreads.The driftgap on each side of the
electrodeis2.1 mm,which correspondsto atotaldrifttimeof about450ns foranoperatingvoltage of
2000V.
Onceassembled,thereisnodiscontinuityalongtheazimuthalangle;butforeaseofconstruction,a
half-barrelisdividedinto16modules.Thetotalthicknessofamoduleisatleast22radiationlengths(X
0 ),
increasing from22X
0 to30X
0
betweenjj=0andjj =0:8, andfrom 24X
0
to 33X
0
betweenjj=0:8
andjj=1:3.
Amodule,showninFigure1,hasthreecompartmentsindepth(front,middle,back).Thereadout
granularityofthedierentcompartmentsisshowninTable1for<1:3(formodule0weset >0and
approximatevalues willbegivenfor insteadof theexactones,e.g., 0.025insteadof =128). Intotal
there are3428readoutcellspermodule.
2.1 Absorbers
Theabsorbersaremadeofleadsheets(thickness1.53mmfor<0:8,1.13mmfor>0:8),glued
betweentwo0.2mmthickstainlesssteel sheetsbyresin-impregnatedglassberfabric.Theberfabric
compensatesforthedierencein thicknessofthetwotypesof leadplatesso thatthenominal thickness
ofanabsorberis2.2mm.
A local non-uniformity of the thickness of the lead plates induces variations of the calorimeter
response[4,7].Itseectontheconstanttermoftheenergyresolutioncanbeapproximatedbytherelative
rms ofthedistributionofthesliding meanof thethicknessofveconsecutiveleadplatesmultipliedby
one-half [8, 9]. In order to keep the constant term induced by inhomogeneities below 0.3%, the lead
thicknesshas beenmeasured byradiography(using asetup developed withthe help of CEA-DAMRI)
duringthecoldrollingprocessatthefactory.Thisallowedforfastmachineadjustmentsandrejectionof
leadsheetsectionsoutoftolerance.Theprecisionoftheleadthicknessobtained,inanacceptancewindow
of60 m,isbetter than0:8%(rms)for <0:8(0:6% for>0:8).Theproductionprocedure improves
signicantly the precisionattainable by standard cold rolling alone. Plates were cut out of these rolls
algorithmwasusedtomatchleadplatesinordertocompensateforlocalnon-uniformities.Deducedfrom
the relativerms of the sliding meanthickness, acontribution to the constant termof 0.12% ( < 0:8)
and0.18%(>0:8)isexpected.
Theabsorbersareformedintoanaccordionshapeasdescribedin [6].Afterthegluingcycle,they
are equipped with precisely machined G10bars (maximal dispersionof 80 m for groupsof 16 bars
in R).Thenal geometry ismeasuredwitha3D coordinatemachineto givefeedbackon thebending
machinestability [11].Thedispersions oftheopening half-angles,varying from46.5 Æ
at smallradiusto
34 Æ
at large radius, are of the order of 0.15 Æ
to 0.5 Æ
. For the straight section lengths a dispersion of
0.02 mmis obtained,while forthe distance betweenthe precisionbarsthedispersionis 0.15mm. The
thicknessproleareshowninFigure2fortherawandslidingaverages.Thermsoftheabsorberthickness
distributionis19m(13m)fortheraw(sliding average)thicknesses.
2.2 Electrodes
Thereadoutelectrodes,275m thick,consistofthreeconductivecopperlayersseparatedbyinsu-
lating polyimide sheets.Thetwoouterlayersareat thehighvoltage potential,theinneroneis usedfor
readingoutthesignalthroughcapacitivecoupling.Thegranularityofthecalorimeterin andindepth
isobtainedbyetchedpatternsonthedierentlayers.Eachgapbetweentwoabsorbersisequippedwith
twoelectrodes:typeA(<0:8)and typeB(>0:8).
Themaindierence withrespect topreviousprototypesis theuseoflargesize electrodes.Before
bendingtheelectrodesare1:8m 0.8m.
Coppercladpolyimide(single-sidedwithpre-curedgluecoatinganddouble-sided)laminateisused
tomakethethreelayeredcircuits.Thebasepolyimideis(Dupont's)Kapton-Etype,whichhasathermal
expansioncoeÆcientclosetothatof copper.This leadstominimal geometricaldeformationsrelativeto
theabsorbersystematcryogenictemperatures.TheHVprotectionresistorpadsaresilkscreenedwithone
componentepoxybasedresistiveink 1)
.Nominal resistivityis1M=square(at roomtemperature)[12].
Thepadsweremeasuredindividually foreachelectrodeondedicatedtestbenches.Eachsignalpathand
HVconnectionisequippedwithoneormoregoldplatedfemalecontacts,directlycrimped(andsoldered)
oncopperpadsonthecircuit.Theelectrodesareequippedwithspringsforthegroundreturnconnection
betweentheelectrodeandabsorber.
Several important production parameterswere obtainedwith the pre-series electrode production
for thismodule 0.Theglobal dimensionsofetched imageswithrespectto lmsaretypicallypreciseto
0:2mm.Thecopperetchingispreciseandwellcontrolledoverthewholearea.Formostelectrodesthe
alignmentbetweenlayersisbetterthan0:4mm,lessthanthe0:5mmleftasseparationbetweencells.
Thishasbeencheckedvisuallyforallelectrodes.
Ontheotherhand several defects andproblems were encountered. Thelmdimensions,notcon-
trolled before usage, were 0:7 mm smallerthan design. Aboutone-third of the electrodeshad to be
discarded,astheydidnottoleratethecuringcyclefortheresistiveinkduetobadgluingofthelaminates.
Forabouthalfoftheremainingelectrodesthecuringcycletemperaturehadtobelowered,yieldinglarge
valuesandalargedispersionoftheserigraphiedresistors.Onabout20electrodes,outputtraces(inthe
signallayer)weredamagedwhenetchingtheouterlayersleadingtoafewdeadcellsinthefrontsection.
Forpartoftheelectrodes,thepositioningofthecrimpswasdonemanuallyresultinginamisalignmentof
connectorsupto1mmbetweendierentgaps.Onequarteroftheelectrodessueredfrompositioning
defects betweenimageandthestackingholes(usedtoxtheelectrodestotheG10baroftheabsorber).
Mostofthesedefects were understood andwereduemostlytothelackofpropertoolingatthat time.
Theelectrodeswerebenttoaccordionshapewiththemachinedescribedindetailin[6].Allresistors
andblockingcapacitancesweremeasuredwithasemi-automaticsetup(lowvoltagetest).Ahighvoltage
test wasperformedbymeasuringtheleakagecurrentbetweentheelectrodeouterlayerskeptat2000V
andtheinnerlayerconnectedtoground.Thestandardtesthadadurationofonehour.Bothtestswere
performed before and after bending [13, 14]. Depending on the degree of curing of the resistive ink,
someelectrodesshoweddamagedresistorsafterbendingtheatcircuits.Thiswasseenin particularfor
resistorscloserthan10mmtothepeakofthebend (seeFigure 3).
In amodule, 64 A- and 64 B-electrodesare needed. In total 150 electrodes were produced 2)
, of
which 49 were unusable. 31 A-electrodes out of 47 and 22 B-electrodes out of 54 had high resistance
values with a large dispersion.The experience gained by the factories in the pre-series production for
module0andadditionalR&Dandsignicantdesignchangeshaveimprovedthesituationandledtothe
1)
DL1216ESL.
2)
CICOREL SA,8 route del'Europe, CH-2017 Boudryand MCBIndustrie, 107-11rue duMoulin Sarrazin,F-95100
Argenteuil.
beengreatlyimproved.
2.3 ModuleAssembly
With the limited numberof electrodes available, only the central region of the module could be
equippedwithelectrodeswithnormalresistancevalues.Thisregioncorrespondsto 0:2<<0:3in the
Aelectrodesand0:1<<0:3intheBelectrodes.Theotherelectrodeswereplacedaroundthesecentral
regions,sothat for <0:8theregion0:125<<0:4and for >0:8theregion0:025<<0:35was
equipped.
ThegeometryofthemoduleisdenedbytheexternalandinternalG10bars.Theirprismaticshape
is such that astackof 64absorbersmakesawedgewith anangle of 22.5 Æ
. Theouter barsarescrewed
onsupportstainless steelringpieceswhichdene theexternalradiusofthemodule.
The stacking is done on a rotating jig in a clean room (class ISO 8) with the relative humidity
controlledto50%5%.Themoduleisstackedinthehorizontalposition,but tighteningtheabsorberto
theprecedingoneandtotheringpiece isdoneinaverticalposition toavoidsaggingoftheabsorbers.
The stacking procedure is the following: the set of 6 ring-pieces xed to its module assembly
backboneisalignedonthemodule assemblyjig.Thentherstabsorberisxedin placewithreferences
in all axes (to about 0.1 mm). A rst spacer plane, the electrode plane, a second spacer plane and a
new absorber arestacked. Dowelpins position the electrode accurately in r and z with respect to the
preceding absorber;other dowel pins position the new absorberin r, its z position being dened by a
stopat z= 0.After rotatingto thevertical position,the newabsorberis pressedagainstthepreceding
one with pneumatic jacks actingon the outer and inner G10bars. The screwslinking theabsorberto
the preceding one and to thering-piece are tightened to the desired torque. This stacking sequence is
repeated64times.Typicallyfour absorberscanbestackedperday.
Regularly, every2 or4 detector gaps, two electrical tests are performed. A lowfrequency signal
(1 to 6Hz) isinjectedon theHVlines and theinducedsignalsonthe signaloutputsarerecorded[15].
This testchecksthecontinuity oftheelectricalcircuitandthe electrodeconnectionsincludingthehigh
voltagedistribution.Then aHV testis performed :1800V isapplied tothe electrodesand theleakage
currentsarerecorded.
Every4gaps, thethicknessofthestackismeasuredattheinnerandouterabsorberG10barsand
atseverallocationsinzinordertocontrolthenaldimensionofthemoduleandalsothethicknessofthe
gaps. Theimperfectshape,due to adeviation fromthe nominal foldingangle of thebent electrodes of
module0,introducedadeformationofthemodulegeometrywithasconsequenceasmall over-thickness
attheinnerandexternalradii.Theevolutionofthemoduleover-thicknessattheexternalradiusisshown
in Figure4.Afterthelast absorberisstacked,anincreaseofabout0.2mmisobserved.
Theaimwastoobtainanargongapdispersionoftheorderof50mtokeepthecontributiontothe
constantterm ofthe energyresolution below 0.15%. Inmodule 0, thegap capacitanceswere measured
andtheirdispersiontranslatedinto adispersionofthegapthicknessof82m.
2.4 Cold Electronicsand Cabling
Thefrontsection is read outat the inner radius, whereasmiddle and back sectionsare read out
from the back (outerradius). Summing boards areconnected to theelectrode connectors to groupthe
signalsin tothedesiredreadout granularity(Table1): 16electrodesare groupedintoonecellforthe
front section, 4 for themiddle and backsections. The boards are 10 layerPCBs with 15 strip-lines
whoselengthsareequalizedinin ordertoensureauniforminductance.
The summing boards are connected to the motherboards. These boards route the outputs to
the readoutcables through"lowprole" connectors andinclude precisioninjectionresistors(0.1% and
70ppm/
Æ
C)forthecalibrationsystem.Thereadoutcablesareminiaturepolyimidecoaxialcables 3)
[16]:
25forthebackandmiddlesections,50forthefrontsectiontooptimizethenoisecontribution.The
readout cables are groupedin bundles of 64channelsand connectedat the end of the module to high
densityD 100connectors, tting 64channelsand 32ground connectionsin a1030 mm 2
area.The
connectorsarexedonapatch panel.
Thehigh voltage is supplied to electrodes through boards connected at the back of the module.
Thetwosidesoftheelectrodesarefedbydierenthighvoltagelines,thusprovidingasafetymargin.In
themodule isdividedintosevenHVsectorsof=0:2;intheHVsectorsizeis=0:2.
Aftercabling,theHVconnectivityischeckedwithalowfrequencytest.Eachcellispulsedandits
signalisrecorded.Ahighvoltagetest isthenperformed.All testsareperformedat roomtemperature.
3)
AXONCABLESA,RoutedeChalons-en-Champagne,51210Montmirail,France.
