CMS HCAL Test Beam Results and Comparison CMS HCAL Test Beam Results and Comparison with GEANT 4 Simulation with GEANT 4 Simulation Jordan Damgov Jordan Damgov on behalf of the on behalf of the CMS HCAL Collaboration CMS HCAL Collaboration CALOR '06 CALOR '06 Chicago June 5-9, 2006 Chicago June 5-9, 2006
HCAL CMS Calorimeter system CMS Calorimeter system ECAL ECAL: PbWO 4 crystals See Julie Whitmore's talk for details Sampling calorimeter Scintillator Brass (70%Cu,30%Zn) Wedge, 20 o 17 Layers 2 2 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Interaction point like geometry Interaction point like geometry The test beam setup is designed to preserve the interaction HO point-like geometry of CMS. HB HE EC Table IP=Pivot Beam 3 3 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Test beam setup Test beam setup 7x7 crystals ECAL 2 Hadron barrel wedges Material for Magnet 1 Hadron endcap wedge Movable table in η - φ plane Hadron outer calorimeter “Magnet” HO HE ECAL HB2 HB1 beam 4 4 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
ECAL and HO ECAL and HO Beam HB2 HO HB1 ECAL View from top VM Al BEAM ECAL is readout by PMTs. Light guides are attached Light 7x7 crystals=14x14 cm to the front face of the guides to crystals. PMTs 5 5 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
GEANT4 simulation GEANT4 simulation Trigger counters H C A L Wire Chambers O ECAL u H t C e M A r L C a ● Detailed HCAL geometry g a B n l a o with HB1&HB2 read-out e r r t r i schema. m e C l o e ● ECAL – PbWO 4 crystals, Al box t i l e and Al block behind ECAL. r ● Beam line - trigger HCAL Endcap counters and wire chambers 6 6 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
GEANT4 physics models validation GEANT4 physics models validation Physics lists tested against the test beam data : ● LHEP : L EP/ HEP parametrized models for inelastic scattering. ● QGSP : Q uark G luon S tring model for the “ P unch-through” interactions. ● QGSC : QGS P + C hiral invariant phase-space decay. ● FTFP : diffractive string excitation similar to that in FRITOF and Lund In this comparison is used Geant 4.6.2.p02 Parametrized: LHEP-3.7 Model based: QGSP-2.8 QGSC-2.9 and FTFP-2.8 produce very similar to QGSP- 2.8 results. 7 7 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Beam line with particle identification 2-300 GeV/c Beam line with particle identification 2-300 GeV/c WC A,B,C V3,V6 CK2 80 GeV/c ECAL ECAL VM CK3 HCAL HCAL SCI_VLE S1-S4 VLE tag against Beam punchthrough trigger muon counters Wire Chambers (WC A,B,C): single hit to reject interaction in beam line Available beam tunes: P-ID: pions 2-300 GeV Cerenkov counter (CK2) - electron muons 80/150 GeV Cerenkov counter (CK3) - pion / kaon / proton Scintillators (V3, V6, VM) – muon tagging electrons 9-100 GeV 8 8 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Beam contamination and cleaning Beam contamination and cleaning Beam contamination Beam cleaning: before the clean up: ● The particle Id counters are used for beam with P beam mu el- momentum from 2 to 15 GeV [GeV] [%] [%] ● Calorimeter based cuts: use the particle Id 300 0.7 0 capabilities of the calorimeters. 150 2 0 High energy muons a tagged by the muon veto 100 2 0 counters with 99% efficiency. Low energy muons 50 2.5 0 30 3.5 7 form pion decay are evaluated to be less then 1.5% 15 11 35 for 9 GeV and below. 10 7 70 Electrons are clearly identify 9 1.5 6 VLE beam line by ECAL/HCAL energy HCAL [GeV] 7 3.5 10 pi- 30GeV deposition. 5 5.5 6 Any remaining uncertainty 3 30 30 in the beam contamination 2 85 7 and interaction in the e- + interaction in the beam line is added to the beam line µ systematics of the 2 and 3 GeV are measurement. ECAL [GeV] not used 9 9 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Uniformity calibration and Energy Scale Uniformity calibration and Energy Scale Wire source Scintillator Reconstructed energy: See Mayda Velasco's talk E rec =a*E ECAL +b*E HCAL signal Energy scale: ECAL : 100 GeV e- backward forward HCAL : 50 GeV pi- with MIP in ECAL. The uniformity calibration is done with Co 60 , per-tower and per-layer 10 10 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Energy spectra ECAL+HCAL: data vs GEANT4 Energy spectra ECAL+HCAL: data vs GEANT4 50 GeV 300 GeV 10 GeV Good agreement 5 GeV 7 GeV between data and GEANT4 prediction. 11 11 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Calorimeter response to pions Calorimeter response to pions ECAL+HCAL 30 100 3 5 10 Important for Jets GEANT4 models correctly the calorimeter response to pions in broad energy range. Correct representation of the single hadron response at low energy is important for simulation of the calorimeter response to jets. Some discontinuity is observed at 7-10 GeV in the GEANT4 prediction. 12 12 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
HCAL alone response to pions HCAL alone response to pions LHEP models better the high energy calorimeter HCAL alone response. QGSP has less leakage on the back due to shorter shower. 100 30 300 10 HCAL alone: MIP in ECAL is required. HO is not used in this measurement to compensate the HB leakage on the back 13 13 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Proton over pion response ratio Proton over pion response ratio Significant Protons have 15% Protons have 15% difference in the calorimeter lower response then lower response then response to pions at this pions at this protons with energies. energies. respect to pions is observed in the data and is well represented in the GEANT4 simulation. We will remeasure We will remeasure it this summer with it this summer with improved particle Id. improved particle Id. 14 14 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Longitudinal shower profile measurement Longitudinal shower profile measurement Modified read-out: redesigned optical Optically HB2 decoding units to masked out allow longitudinal shower measurement beam Electromagnetic e- 100 GeV shower profile in HCAL Very good agreement between test beam data and GEANT4 prediction. 15 15 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Longitudinal shower profile measurement (cont.) Longitudinal shower profile measurement (cont.) pi 10 GeV pi 30 GeV pi 5 GeV LHEP and QGSP show good agreement with test beam data at low and intermediate energies. 16 16 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Longitudinal shower profile measurement (cont.) Longitudinal shower profile measurement (cont.) QGSP physics list has shorter shower MIP in ECAL MIP in L0 profile for incident particles with high momentum. LHEP pi 300 GeV QGSP 17 17 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
HCAL - ECAL response to very low energy pion beam HCAL - ECAL response to very low energy pion beam HCAL [GeV] HCAL [GeV] 7 GeV 5 GeV 5 GeV TB TB TB ECAL [GeV] ECAL [GeV] ECAL [GeV] HCAL [GeV] 5 GeV 7 GeV 5 GeV HCAL [GeV] G4 G4 G4 ECAL [GeV] ECAL [GeV] ECAL [GeV] ECAL response is higher in GEANT4 at low energy: geometry or physics or ...? 18 18 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Cluster-based response compensation Cluster-based response compensation Fractional energy resolution for pions. Uses test beam data to fit the σ /E Dan Green intrinsic electron to hadron response (e/h) and the average neutral fraction f 0 of the ECAL and HCAL as a function of the raw total calorimeter energy, E + H. 70% Stochastic 8% Constant E [GeV] 19 19 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Conclusions and Outlook Conclusions and Outlook Calorimeter response for momentum range 5-300 GeV/c was measured with test beam in 2004. GEANT4 is in good overall agreement with the data LHEP shows best agreement We observed small discrepancy in the following quantities : 1) Longitudinal shower shape for 150-300 GeV/c pions, modeled by QGSP physics list 2) Discontinuity in the calorimeter response in 7-10 GeV/c range 3) ECAL response to very low energy pion beam is higher in the GEANT4 simulation . We plan to repeat the measurements this summer with ECAL production super-module and improved particle Id. 20 20 Jordan Damgov CALOR06 – CMS HCAL Test Beam Results
Recommend
More recommend
