May 31th, 2007 LCWS 2007 - C. Gatto 1 Tracking Studies in the 4 th Concept On behalf of 4th Concept Software Group D. Barbareschi V. Di Benedetto E. Cavallo

Tracking Studiesin the4th Concept

On behalf of 4th Concept Software Group

D. BarbareschiV. Di BenedettoE. Cavallo F. Ignatov A MazzacaneG. TassielliG. Terracciano

LCWS 2007 - C. Gatto

Central TrackerStudies


Central Tracker IssuesMany interesting results (mostly simulations) at the Tracking Review in BeijingHowever, simulations were not on equal footing We decided to compare TPC vs DCH vs Si Tracker central trackers within the same framework and with comparable cheatGoal is to find the detector that best matches the Dual Readout Calorimeters and the Muon Spectrometer


TPC vs DCH Material BudgetDrift Chamber Gas [He-C4H10/90-10]: 0.15%Wires: 0.4%Vessel: Inner wall: 0.1% X/XoOuter wall: 2% X/XoEndcaps (wires, pads,electronics & services included): 8% X/XoTPC Gas[Ar-CF4/97-3]: 1.3% Vessel: Inner wall + cage: 0.29% X/XoOuter wall: 1.2% X/XoEndcaps (wires, pads,electronics & services included): 35-54% X/XoBeam Pipe: 0.18% X/XoVXD:Detector & support: 0.8% X/XoOuter shield: 0.16% X/XoCentral Tracker:Vessel: 23-150 cmActive volume: 37-136 cm


DCH Studies PlannerAxial Cell DCHIntended for resolution studies onlyEasily interfaceable with existing Seed Finding and Parallel Kalman Filter

Full Stereo DCHIntended for Occupancy related studiesCannot be easily adapted to existing reconstruction codeDedicated Pattern recognition (exploiting the vectorial informations provided by the Cluster Counting)Dedicated Track Fitter (not necessarily Kalman Filter)Short TermLong Term


e+e- -> tt -> 6 jets with DCH

ECM=500 GeVLoopersHits per cell vs layerOccupancy vs layer


e+e- -> HHZ -> 4 jets+2muons with DCH

ECM=500 GeVLoopersHits per cell vs layerOccupancy vs layer


LoopersHits per cell vs layerOccupancy vs layere+e- -> HoZo -> 2 jets+2muons with DCH

ECM=230 GeV


LoopersHits per cell vs layerOccupancy vs layere+e- -> W+W- -> 4 jetswith DCH

ECM=500 GeV


TPC vs DCH SimulationDrift Chamber Gaussian smearing: s(d) = 100 mm s(z) = 3000 mm TPC Gaussian smearing: Mumegass(d) = 55/sqrt(12) mmPadsVXD Full digitization (CCD assumed):Pixel size: 20 mm x 20 mmVeryconservative


Performance with Physics Several channels simulated10 muons [0.5,100] GeVe+e- -> W+W- -> 4 jets @ Ecm = 500 GeVe+e- -> tt -> 6 jets @ Ecm = 500 GeVe+e- -> HoZo -> 2 jets 2 muons @ Ecm = 230 GeVe+e- -> HoHoZo -> 4 jets 2 muons @ Ecm = 500 GeVResults not available yet for the Drift Chamber


D(1/pt) vs Pm [0.5,100] GeVe+e- -> W+W-e+e- -> tte+e- -> HoZoTPC


D(1/pt)m [0.5,100] GeVe+e- -> W+W-e+e- -> tte+e- -> HoZoTPC


DDom [0.5,100] GeVe+e- -> W+W-e+e- -> tte+e- -> HoZoTPC


Djm [0.5,100] GeVe+e- -> W+W-e+e- -> HoZoTPC


Tracking Efficiencym [0.5,100] GeVe+e- -> W+W-e+e- -> tte+e- -> HoZoTPC


Tracking EfficiencyTPCe+e- -> tt -> 6 jets


e+e- -> HoZo @ Ecm = 230 GeV AnalysisDMrecoil = 380 MeV (90% RMS)DMrecoil = 250 MeV (gauss fit)

TPC


Summary of PerformanceTracking efficiency: ereco > 90 % above 100 MeV ereco = 99.7 % above 1.5 Gev

TPC + VXD resolution: s(1/pt) = 6 x10-5 up to 8 x10-4 s(j) = 0.05 up to 1 mrads(d) = 2.6 mm up to 15 mms(z) = 3.9 mm up to 18 mm

Totally dominated by MSe+e- -> HoZo @ Ecm = 230 GeV is the worse case

TPC


Plans for Central Tracking SimulationTPC simulation frozenDCHGaussian smearing response: June 2007Digitizaton: August 2007Si trackerWill start mid June 2007Expect first results in September 2007Start with Gaussian smearingDigitization by end of 2007


Vertex DetectorStudies


Vertex Detector StudiesDigitization and clusterization implemented for VXDVery parametric implemetationSeveral tecnologies can be representedFTD simulated in IlcRoot (help from Valencias group)Gaussian smearing onlyKalman filter needs to be optimized


New VXD Digitization & ClusterizationFollow the path of the track inside the Si in steps of 1 mmconvert the energy deposited into chargespreads the charge asymmetrically across several pixels:Include also: capacitive pixel coupling random noiseelectronic thresholdClusterization:divide the initial cluster in smaller NxN clusters (to be optimized) Kalman filter picks up the best clusters


SiD/4th VXD


FTD


VXD vs FTD LayoutFTDZ=200, 320, 440, 550, 800, 1050, 1300 mmRin=38, 48,59, 68, 90, 111, 132 mmRout=140, 140, 210, 270 , 290 , 290 , 290 mmAzimuth coverage: >5.8Thickness: 100 mmThermal shield: r=29

VXD Endcaps z=76, 95, 125, 180 mmRin=16, 16, 20, 20mmRout= 75, 75, 75, 75 mmAzimuth coverage: >6.5Thickness: 100 +100 mmThermal shield: r=12

VXD Barrel Dz=125 mmR=1.40-1.52, 2.20-2.31, 3.50-3.59, 4.76-4.84 and 6.00 6.08 cmAzimuth coverage: >11Thickness: 100+100 mm


VXD vs FTD SimulationsFTD (gaussian smearing):

Barrel: srf = 5 mm sz = 6 mm

Pixel Endcaps srf = 14 mm sz = 60 mm

Strips Endcaps (20mrad stereo angle)srf = 14 mm sz = 50 mm/sqrt(12)/tan(40 mrad) VXD (full digitization):

Pixel size: 20 mm x 20 mm


Track ResolutionsVXD


e+e- -> tt -> 6 jetsVXD


Efficiency vs qVXD


VXD vs FTD (tt-> 6 jets)FTD

Tracking efficiency: Need optimization

Params resolution s(j) = ~1.3 mrads(l) = ~1 mrads(d) = 21 mms(z) = 17 mm

VXD

Tracking efficiency: ereco = 98 %ereco = 99 % above 4 Gev

Params resolution : s(j) = ~1.3 mrads(l) = ~1 mrads(d) = 18 mms(z) = 16 mm


Beam Pair Background Studywith FTD , VXDand TPCNew studies with Guinea Pig (in collaboration with M. Vos and C. Marias)

Beam Parameters (Nominal Settings):Ebeam = 250 GeVNbeam = 2.05 1010Bunch Crossing = 140


VXDVXD Beam Background


FTD Beam Background


TPC Beam Background


VXD Tracks Reconstruction


FTD Tracks Reconstruction


TPC Tracks Reconstruction


SummaryAbout 13,000 hits found in the VXD (14,000 in FTD+Barrel VXD)About 278,000 hits in TPC aloneAbout 163 reconstructed tracks (SA Vertex Tracker)VXD average clusters/track = 4.3About 176 reconstructed tracks in VXD +TPC443 particles entering the TPCTPC average clusters/track = 19.6Very dangerous background when integrating many bunch crossings


SummaryFirst DCH occupancy studies presentedOccupancy is tolerableOnly 1 beam crossing is integrated with current cell designPreliminary resolution results in few weeksExpect improved performance vs TPC for slow momentum tracks in forward directionMultiple scattering lower in HeResolution is not dependent on z (no wt effect)VXD and FTD studies are in progressSA Kalman filter needs to be optimized for FTDStrip digitization ready in the next 2-3 months Background studies: need more studiesAlbedo neutronsMore beam generated background


ConclusionsLarge effort from the 4th Concept Simulation group to understand optimal tracking Other people are getting interested (Valencia, Paris VI (?))Growing interest at Fermilab on Tracking and Calorimetry simulationILCroot is also available at Fermilab 4th Concept also uses Grid Computing at Fermilab Central Tracker studies and background will take center stage in the next few months (TPC vs Si vs DCH)


Backup Slides


e+e- -> tt -> 6 jets


Marcel VossTalkOrsay2007


e+e- -> tt -> 6 jets (VXD+TPC)Barrel onlyBarrel+EC


Kalman Fit Clusters Usage (Guinea Pig 140 BX)FTDVXD


Pulls from Kalman FilterTPCe+e- -> tt -> 6 jets


Beam Background Cluster/Track


Documents

May 31th, 2007 LCWS 2007 - C. Gatto 1 Tracking Studies in the 4 th Concept On behalf of 4th Concept Software Group D. Barbareschi V. Di Benedetto E. Cavallo