Simulation study of scintillatorbased calorimeter Hiroyuki Matsunaga (Tsukuba) For GLD-CAL & ACFA-SIM-J groups Main contributors: M. C. Chang, K. Fujii, T. Takeshita, S. Yamauchi, A. Nagano, S. Kim Simulation and Reconstruction session LCWS5 at Stanford
Motivation Basic studies of calorimeter behavior for single particle in simple (testbeam) configuration Energy resolution, linearity lateral/longitudinal shower profiles It is extremely important to understand lateral shower profile for particle flow analysis Comparison with previous testbeam results Understand the detector responses in detail Get detector effects which should be implemented in full simulator 25/3/21 H. Matsunaga, LCWS5 2
Outline Range cut study for Geant4 Tile-scintillator ECAL Energy resolution Strip-scintillator ECAL Energy resolution, linearity lateral / longitudinal shower profiles Summary and Outlook 25/3/21 H. Matsunaga, LCWS5 3
Range cut Geant4 parameter set by user How it works? When the range of the particle for the next step is calculated to be less than the range cut, GEANT4 kills the particle and deposits all of its energy there. A secondary particle is not created if its range is less than the range cut. 1mm by default Used Geant4 6.2p2 Problem in EM process for thin material; Should be fixed in 7. patch-1 (claimed by Geant4 team) 25/3/21 H. Matsunaga, LCWS5 4
Energy cuts vs. Range cuts in scintillator Default range cut (1mm) seems too large Default range cut Energy Cuts vs. Range Cuts Energy Cuts (kev) 35 3 25 2 e + / e - 15 1 1µm 5 1-3 1-2 -1 γ 1 1 Range Cuts (mm) 25/3/21 H. Matsunaga, LCWS5 5
Energy cuts vs. Range cuts for e +/ e - and γ Large difference between absorbers and active media at larger range cuts Energy Cuts vs. Range Cuts W Energy Cuts vs. Range Cuts W Energy Cuts (kev) 22 2 18 16 14 12 1 8 6 4 2 1-3 1-2 -1 Default range cut Default range cut 1 1 Range Cuts (mm) Pb Energy Cuts (kev) 1 8 6 1µm 1µm 4 CsI Sci 2 1-3 1-2 -1 1 1 Range Cuts (mm) 25/3/21 H. Matsunaga, LCWS5 6 Pb CsI Sci
Energy deposits vs. Range cuts Two configurations with same radiation length 2.5mm W / 1mm Sci 4mm Pb / 1mm Sci Threshold behavior seen from ~.3 to ~1 µm Gap ~7.5 MeV (Pb/Sci) More energy deposits in absorber at higher range cuts 27.5 MeV vs. 2MeV in scintillator (Pb/Sci) Sum of energy deposit is constant Abs Energy Deposit vs. Range Cuts 98 Energy Deposit (MeV) 978 976 974 972 97 968 966 964 962 96 1 1 W-Sci Pb-Sci -6-6 1 1-5 -5 1 1-4 -4 1 Sci Energy Deposit vs. Range Cuts Energy Deposit (MeV) 3 28 26 24 22 2 18 Pb-Sci W-Sci 1-3 -3 1µm 1 1-2 -2 ~7.5 MeV -1 1 1 Range Cuts (mm) ~7.5 MeV -1 1 1 Range Cuts (mm) 25/3/21 H. Matsunaga, LCWS5 7
Decompose contributions from Threshold effect comes mainly from multiple scattering Most frequent process mean-free-path for multiple scattering is around the threshold region Should set the range cut ~1 µm or less (limited by CPU) Will try to check again with 7.p1 physics effects Pb(4mm)/Sci(1mm) Energy deposit in scintillator Energy Deposit vs. Range Cuts Energy Deposit (MeV) 35 3 25 2 15 1 1 1 Range Cuts (mm) 25/3/21 H. Matsunaga, LCWS5 8 5 1 switch off: mutiple scattering switch off: Photo-electric effect all included switch off: Compton scattering switch off: electron Bremsstrahlung switch off: Gamma conversion (pair production) switch off: electron Ionisation -6 1-5 1-4 1-3 1-2 -1
ECAL Energy Resolution 4mm Pb / 1mm Sci., electron injection 16 %/ E with ~1 µm range cuts (.3, 1, 3 µm) Nearly expected values Energy Resolution vs. Beam Energy Energy Resolution: σ/e (%) 2 18 3µm 16 1µm 14.3µ m 12 1 8 6 Energy Resolution vs. Beam Energy Energy Resolution: σ/ E (%) 2 18 3µm 16 1µm 14.3µ m 12 1 8 6 4 2 4 2 1 1 2 1 Beam Energy (GeV) 1 1 2 1 Beam Energy (GeV) 25/3/21 H. Matsunaga, LCWS5 9
Energy Resolution for various Pb thicknesses Compare with testbeam results (T45 & T411 at KEK ) Scintillator thickness is 2mm (fixed), 1-4 GeV electron beam MC results are slightly better than data MC does not include detector effects, e.g. photo statistics Energy Resolution vs. Beam Energy σ/e (%) 4 35 3 25 2 15 Data 16:2 MC 14:2 12:2 1:2 8:2 6:2 4:2 Data MC lead:scintillator (in mm) Data MC Data MC 1 5.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Beam Energy (GeV) 25/3/21 H. Matsunaga, LCWS5 1
Scintillator-Strip ECAL Using 1cm-wide scintillator strips 2-dimensional array 1cm effective granularity 4mm Pb + 2 x 2mm sci. in a layer 24 layers (6 superlayers) Beam tests were carried out at KEK in 22 and 24 MA PMT MA PMT clear fiber WLS fiber Beam scintillator strip lead plate 1 layer 24 layers = 6 super layers 25/3/21 H. Matsunaga, LCWS5 11
Linearity MC takes account for light leakages between strips, noises, photo statistics effect, etc. Good agreement with data Energy Deposit [MIPs] Even for absolute energy deposits (in unit of MIPs) 3 25 2 15 1 5 X strips : 27.35 ±.5 MIPs/GeV Y strips : 27.23 ±.5 MIPs/GeV all strips : 54.65 ±.9 MIPs/GeV Simulation (X strips): 27.47 ±.1 MIPs/GeV Simulation (Y strips): 26.37 ±.1 MIPs/GeV Simulation (all strips): 53.89 ±.1 MIPs/GeV 1 2 3 4 5 Beam Momentum [GeV/c] -2.5 1 2 3 4 5 Beam Momentum [GeV/c] 25/3/21 H. Matsunaga, LCWS5 12 Deviation [%] 2.5 2 1.5 1.5 -.5-1 -1.5-2 X strips Y strips all strips Simulation (X strips) Simulation (Y strips) Simulation (all strips)
Energy Resolution Using energy deposits from all strips : Stochastic term ~13% Constant term ~% Good agreement between data and MC Energy Resolution [%] 2 18 16 14 12 1 8 6 4 2 Simulation: X : σ sto σ const Y : σ sto σ const all : σ sto σ const σ = E E σ +σ E sto 2 const 2 X : σ sto σ const Y : σ sto σ const all : σ sto = 14.75 ±.3 +.33 =. -. = 14.8 ±.1 = 1.3 ±.39 = 12.97 ±.3 +.4 =. -. σ const X strips Y strips all strips Simulation (X strips) Simulation (Y strips) Simulation (all strips) = 14.74 ±.12 +.83 =. -. = 14.62 ±.16 + 1.43 =. -. = 13.1 ±.12 +.73 =. -. 1 2 3 4 5 Beam Momentum [GeV/c] 25/3/21 H. Matsunaga, LCWS5 13
Longitudinal Shower Profile Longitudinal shower profile agrees with data Absolute values are well reproduced Shower-maximum ~ 2 nd superlayer Energy Deposit ( MIPs ) 1 2 1 1 GeV 2 GeV 3 GeV 4 GeV 1 2 4 6 8 1 12 14 16 18 Depth ( X ) 25/3/21 H. Matsunaga, LCWS5 14
Lateral shower profile Introduce energy fraction I(x) : I(x)= x - dx ' PH dx ' PH X = X i X dc X dc : incident position determined by drift chambers X i : position of i-th strip I() =.5 / - Pulse height (MIPs) Hit-Cluster X i x X dc = 25/3/21 H. Matsunaga, LCWS5 15
Integrated lateral shower profile 4GeV electron (each superlayer), MIP Width for 9% containment is ~1.7cm for 2 nd superlayer (shower maximum) Most MIP spread originates from light leakage between adjacent strips and cross-talks in MA-PMTs I(x) 1 1-1.5 1-2 MIP 1st Superlayer 2nd Superlayer 3rd Superlayer 4th Superlayer 5th Superlayer 6th Superlayer MIP.5 1 1.5 2 2.5 3 3.5 4 x (cm) 25/3/21 H. Matsunaga, LCWS5 16
Smeared spread function GEANT3-based shower simulation shows smaller width than data Some detector effects (such as light leakage) are not implemented in simulation Smeared function using MIP signal spread agrees with data very well! 1-3 25/3/21 H. Matsunaga, LCWS5 17 I (x) 1-1 1-2 MIP 4 GeV e - 2nd Superlayer Smeared function Simulation.5 1 1.5 2 2.5 3 3.5 4 x (cm)
RMS of lateral shower profile Checked RMS of hit cluster in each superlayer Implemented light leakage and cross-talks in simulation Slightly Narrower cluster widths in MC, especially for 1~3 superlayers : Narrow clusters Large pulse heights More detector effects? Number of Events Number of Events Number of Events 8 6 4 2 8 6 4 2 8 6 4 2 Data, Simulation 1st Superlayer 1 2 3 RMS of shower cluster ( cm ) 3rd Superlayer 1 2 3 RMS of shower cluster ( cm ) 5th Superlayer 1 2 3 RMS of shower cluster ( cm ) Number of Events Number of Events Number of Events 8 6 4 2 8 6 4 2 8 6 4 2 2nd Superlayer 1 2 3 RMS of shower cluster ( cm ) 4th Superlayer 1 2 3 RMS of shower cluster ( cm ) 6th Superlayer 1 2 3 RMS of shower cluster ( cm ) 25/3/21 H. Matsunaga, LCWS5 18
Summary and outlook Geant4 Range cut value should be less than ~1 µm Sci-Tile calorimeter Energy resolution agrees with data Sci-Strip calorimeter Energy deposits and resolution are consistent with data Good agreement for longitudinal profile Lateral shower profile is almost understood Outlook Study for hadrons in progress Need to understand hadronic processes in GEANT4 SLAC Geant4 team provides physics list for LC Performance study of optional HCAL digital calorimeter PFA study with simple CAL detector and with full simulator 25/3/21 H. Matsunaga, LCWS5 19