Dario Barberis Evaluation of GEANT4 Electromagnetic and Hadronic Physics in ATLAS LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 1
The ATLAS detector LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 2
Electromagnetic Physics G4 Physics studied using separate test beam simulations (simple geometries) for TRT, SCT, Pixel detectors G3 and G4 simulations compared directly to test beam data Detector response simulation very important for overall agreement between simulated and real data Physics models tested (or under test): Standard energy loss PAI (Photon Absorption and Ionisation) energy loss (relevant for the TRT) Transition radiation production and absorption LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 3
Transition Radiation Tracker support and alignment plate radiator matrix straws carbon fiber shell zoom of module end plate LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 4
TRT: Energy Loss in Straws 300 GeV muons 20 GeV pions 20 GeV electrons Energy loss measured in ATLAS test beam compared to Geant-3 and to Geant-4 simulations (PAI model) including effects of detector and electronics (K.A.Assamagan): spectra match reasonably for different particles and energies some more checks needed for electrons LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 5
Events TRT: Transition Radiation Deposited energy (kev) Spectrum of energy deposited by 20 GeV electrons in TRT straws, with and without foil radiator in front of the detector (V.Mitsou): PAI = Geant-4 with PAI model DATA = 1999 TRT test beam data Transition Radiation is produced in foam and foil radiators placed between the straws The fraction of hits above a given threshold (5 kev) is used to discriminate electrons from hadrons and muons Test beam data can only measure the convolution of energy loss by ionisation, emission and absorption of TR photons Geant-4 offers several ways of describing the radiator (as a more or a less regular medium) So far none of them reproduce the test beam spectra (but V.Grichine produces new models faster than we can test them!) More work is needed! LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 6
Silicon Detectors Standard ionisation model compared (by M.Klute) to PAI model for 100 GeV pions crossing a Pixel detector module (280 µm thick silicon): distribution around peak identical PAI model does not link properly to δ-ray production PAI model in any case not really applicable for ATLAS silicon detectors more important is the correct spatial distribution of ionisation energy loss: range cut should match detector resolution (10 µm for Pixels) LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 7
Silicon Detectors Expected: 78.9 28.2 (from data on Si detectors) Variation of Landau width with range cut somewhat disturbing Need small r-cut as detector resolution ~10 µm Investigating alternative geometry descriptions of Pixel module (~50k pixels/module, 1750 modules in ATLAS...) LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 8
SCT: Detector Response Detector response model very important for comparisons between simulation and test beam Strictly not part of Geant-4 (or Geant-3) but cannot be factorised away in real test beam data! SCT efficiency vs threshold for different depletion bias voltages: points are data, lines are Geant-4 simulation plus detector response model (from S.Gadomski) SCT Geant-4 test beam simulation with correct detector response model reproduces very well available test beam data LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 9
Hadronic Physics The Pixel detector can measure the energy released around the interaction point with a very fine granularity A first attempt at extracting hadronic interaction events from test beam data did not bring much information as the events were scattered amongst many different runs under different operating conditions In August 2001 we took some dedicated runs in the Pixel test beam with an interaction trigger data are being analysed right now Comparisons will be possible between test beam data and the hadronic interaction models avaliable in Geant-4: parametric model (old Gheisha) and theoretical (parton-string) model Observables are: local energy deposit and cluster size/shape multiplicity and angular distribution of (forward) outgoing tracks their correlations LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 10
Calorimetry LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 11
Liquid Argon EM Calorimetry LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 12
Liquid Argon EM Calorimetry muons in LAr EM Barrel electrons in LAr EM Barrel LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 13
Liquid Argon EM Calorimetry Geant4 describes better than Geant3 energy deposits as measured with muon test beam data agreement G4-test beam is within 1% Geant4 (as well as Geant3) describes well the linearity of electron response Geant4 predicts larger energy resolution for electrons that Geant3. Direct comparison with test beam data still in progress LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 14
Liquid Argon Hadronic Calorimetry electron energy resolution in LAr Hadronic End Cap pion shower in LAr Hadronic End Cap LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 15
Liquid Argon Hadronic Calorimetry pion energy resolution in LAr Hadronic End Cap Electrons: Geant4 predicts less visible energy in LAr than Geant3 (~3%) and more energy in absorber (~0.1%). Total energy is the same energy resolution well reproduced by Geant3: Geant4 gives too good resolution Pions: first results of simulation with Geant4 look reasonable more detailed comparisons with test beam data in progress open questions being discussed with Geant4 people LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 16
Tile Calorimeter: electrons Electron energy resolution somewhat too good: sampling term 16% instead of 24% (was the same for Geant-3) Visible energy vs impact point has the correct shape but amplitude of variations and energy dependence do not match test beam data Electron energy resolution Visible energy vs impact position for 100 GeV and 20 GeV electrons LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 17
Tile Calorimeter: muons Energy loss distribution fatter than Geant-3 for both Fe and scintillator: therefore it does not match perfectly test beam data but remember: energy loss distribution in silicon narrower than in real data! LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 18
Tile Calorimeter: pions GHEISHA in GEANT4 is similar to GHEISHA in GEANT3 LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 19
Tile Calorimeter: Fluka vs G4 Test beam data from 20 to 300 GeV: σ/e E = (43.5 ± 2.5)% / E + (2.2 ± 1.2)% e/h = 1.30 ± 0.1 LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 20
Muon Detector EM shower production by muons in absorber: extra hits in Muon Drift Tubes Transverse distance of extra hits from muon track in Geant-4 broadly reproduces test beam data Detailed agreement better for lighter absorber material LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 21
Conclusions Large progress in the last year in understanding electromagnetic processes, both in tracking and calorimetry. Interplay between geometry and physics processes being addressed. Work is continuing on both sides (ATLAS and G4) to improve understanding and produce optimised geometries as well as PhysicsLists. Possibility to set different cuts and use different PhysicsLists for each detector (by Logical Volume) will help considerably. There are still issues that remain to be clarified, but the gradient is positive! Collaboration between ATLAS and G4 people on a very good level, there could be faster progress if there was more manpower (especially on our side!). LC Workshop, CERN, 15 Nov 2001 Dario Barberis Genova University/INFN 22