Title: TRD Response Simulation
1TRD Response Simulation
TRD Software Meeting Oct 23, 2003 Ken Oyama,
Heidelberg
2- Done
- Eth was different between GEANT(10 keV) and
StepManager(50 keV) and it was causing overlap. - Now both DCUTE50 keV and CUTELE50 keV.
- This suppresses number of secondary electrons
generated in TRD gas to be 1/6 (180 to 30 of
primary). - Checked what happen if gMC-gtEdep() (destep) is
used instead of Ermilova. - Checked what happen if Poisson distribution is
used for determining Q. - On Going
- Implementation of Ermilova (or something better)
into GEANT. - Check whether fluctuation is simulated
correctly or not. - Check about GEANT parameterization. Simulate it
in StepManager.
3Momentum Dependence
Absolute value in the simulation is higher than
test beam data.
AliRoot
4Bethe Bloch and Step Size
In TPC, step size is typically order of 1 mm. In
TRD, 0.1 mm.
Step size cm for 2 GeV/c pion
Number of primary collision /cm
TPC
TRD
5Delta Electron Spectra Simulated in Our
StepManager
Red Ermilova function (default) Blue Geant
Edep (restricted continuous energy loss)
Threshold 10 keV
Threshold 50 keV
6dE/dx Simulation Method in Geant3 and StepManager
GEANT delta electrons of energy ? gt Eth
gas volume
primary collision
E
E- ?
Step size given by Bethe Bloch dNprim/dx
StepManager virtually generates delta electron
with energy Edelta according to Ermilova. Number
of charge generated by a primary collision is
calculated by Q (Edelta Kpot) / W 1. Then a
hit with Q at rest is generated ? dE/dx
ltEdeltagt dNprim/dx.
7Particles Generated in the TRD Delta Electron
Spectra
Factor 6 smaller delta electron yield for 50 keV
threshold.
These shapes seamlessly connect to the Geant Edep
shape.
DCUTE10 keV
DCUTE50 keV
8Q calculation
K 12 eV ltWgt 22 eV
Original method Q (int) 1 (Edelta K) /
ltWgt. Modified method Q 1 P(Edelta ltWgt)
/ ltWgt where P(m) is discrete Poisson
distribution with mean m.
9Comparison
Hit charge sum times 22 eV / 6 chambers energy
deposit in one chamber
Same as a plot in previous page