Pr

1
Validation of GEANT4 by the CMS Experiment
ARCE Pedro1,2, ARCELLI Silvia3, BANERJEE
Sunanda4, BANERJEE Sudeshna4, BOCCALI Tommaso5,
DE LA CRUZ Begoña1, DE ROECK Albert2, DUTTA
Suchandra5, ELVIRA Daniel6, FANFANI Alessandra3,
INOCENTE Vincenzo2, JOSÁ Isabel1, LASSILA-PERINI
Kati7, LIENDL Martin2,8, MAZUMDAR Kajari2,4,
NEUMEISTER Norbert8, NIKINTENKO Alexandre9, RIZZI
Andrea10, SILVESTRIS Lucia5, STAVRIANAKOU Maya2,
STRAESSNER Arno2, WELLISCH Hans-Peter2, WYNHOFF
Stephan11, ZYCH Pawel12 1(CIEMAT, Madrid,
Spain), 2(CERN, Geneva, Switzerland), 3(INFN
Padova, Italy), 3(Tata Institute, Bombay, India),
5(INFN Bari, Italy), 6(Fermilab, Chicago, USA),
7(HIP, Helsinki, Finland), 8(HEPHY, Vienna,
Austria), 9(Imperial College, UK), 10(SNS, Pisa,
Italy), 11(Univ. Princeton, USA), 12(IEP. Warsaw,
Poland) (for the CMS Collaboration) CHEP03, La
Jolla, 24th March 2003
2
Index

• Status of GEANT4 simulation (OSCAR)
• Geometry checks
• Physics checks
• EM calorimeter
• Muon physics
• Hadronic calorimeter
• Release testing

3
GEANT4 Simulation (OSCAR)

• Geometry
• All CMS detectors
• Also several Testbeams
• Geometry described in XML
• Common to Simulation/Reconstruction/Visualisation
  
• All geometry converted from GEANT3 geometry -gt
  XML
• Detailed checking in progress
• Magnetic field
• 3D magnetic field implemented
• Review in progress to speed up its access

4
Sliced view of CMS barrel detectors
View of CMS muon system
View of CMS pixel detector
5
OSCAR status (II)

• Primary generator
• CMS particle gun
• Scan in ?, ?, energy
• Random (flat/gaussian) in ?, ?, energy
• Interface with PYTHIA ntuples and HEPEVT ASCII
  files
• Interface with Interface with HEPEVT and HepMC
• UserActions
• Several user actions of the same type loadable
  on demand
• Using COBRA Dispatcher-Observer pattern
• Persistency
• Hits for all detectors
• Primary generator particles
• Selected tracks
• Use CMS framework (COBRA)
• Use ROOT IO

6
OSCAR status (III)

• Physics
• Electromagnetic processes needed at LHC
• Hadronic processes needed at LHC
• CMS Physics Technical Design Report is planned to
  be written using OSCAR (due end 2005)
• OSCAR milestone June 2003
• Reproduce the physics results of the GEANT3
  simulation with similar performance
• Physics Cuts
• Production cuts per material
• Minimum kinetic energy cuts per material
• Still under investigation if we really need them
• Other User Limits Max. time of flight, Min.
  range, Max. step length, Max. track length

7
OSCAR - IGUANA

• Interactive visualisation is done with IGUANA
• Visualise any GEANT4 Geometry
• Select by logical volumes / by physical volumes
• Picking, slices
• Volume property window
• GEANT4 command line
• A wizard to guide through OSCAR settings
• Adding other extensions is trivial (e.g. GUIs
  like overlap detection)
• Visualise Tracks
• Visualise magnetic field
• Integrated with overlap detection
• Find overlaps, show result details in a list
• Highlight overlaps in 3D
• Hits, in the near future

8
OSCAR - IGUANA
Overlap detection
Example extension (a trivial wizard) Queried from
plug-in database, loaded on request and bound to
IGUANA G4 Run Manager
Control of arbitrary GEANT4 tree
Correlated Picking
IGUANA viewer displaying OpenInventor scene
9
OSCAR - IGUANA magnetic field

• 2D plane that
• can be rotated
• Field direction
• as an arrow
• Field intensity
• as colour

10
OSCAR status (IV)

• Framework
• COBRA Common generator interface for Simulation
  and Reconstruction
• COBRA/Mantis interface with GEANT4 (see talk at
  this conference - Maya Stavrianakou)
• Production tools
• Integrated with CMS production tools (see talk
  at this conference - Veronique Lefebure)
• OSCAR is ready for physics validation studies
• Detailed checks have to be done to allow the
  Data Challenge 04 production, starting july
  2003
• Handle the amount of data produced in one month
  at 5 50M events

11
Geometry debugging

• GEANT3 files in ZEBRA RZ format has been
  automatically converted to XML for simulation and
  reconstruction
• GEANT3 has always been the source of the
  geometry for simulation and reconstruction
• Detailed checks have been done to assure that
  the geometry is the same
• Check that you have the same objects(materials,
  solids, volumes, touchables)
• Check that touchable centres in global
  co-ordinates correspond to the same touchables in
  GEANT4 geometry
• Material budget
• Compare step by step if needed
• Still we have to hunt for overlaps (MANY is
  not allowed in GEANT4)
• Use CMS overlap detection tool (already
  integrated in GEANT4)
• (see talk at CHEP01 - Martin Liendl)

12
Checking physicsDe/Dx in Silicon Detector

• CMSIM 125 and ORCA_6_3_0
• OSCAR_2_1_1 (geant4.5.0 / CMSIM 127 Geometry) and
  ORCA_7_1_1
• Particle Types Muon, Pion, Kaon and Electron
• Momenta 2, 3, 4, 5 and 10 GeV
• Eta 0.8
• FED Dynamic Range 12 bit
• Secondary Particle Generation
• 1 MeV cut off in the Sensitive Material
  (optimized with PHOBOS/ALEPH Data)
• Very high cutoff in Insensitive Material (LOSS
  2) for CMSIM 125
• 10cm cutoff for OSCAR

13
Energy Loss by 5 GeV Muon
Checking physicsDe/Dx in Silicon Detector
14
dE/dX vs Momentum
Checking physicsDe/Dx in Silicon Detector
15
Checking physics EM calorimeter

• SETUP
• CMS electromagnetic calorimeter lead tungstate
  crystals pointing to interaction point
• Photon of E 30 GeV enters front face of
  crystal
• No magnetic field
• What was compared
• E in hottest crystal, in 3x3, in 5x5 matrix
• E total
• dNhits/dt, for 1 ns time slices
• GEANT3.21 vs GEANT4.4.0.p02

16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
Bremsstrahlung
10 GeV
100 GeV
1000 GeV
20
Pair Production
10 GeV
ee- production
100 GeV
1000 GeV
21
Multiple scattering
10 GeV
LEP I Z??? events in L3 detector
100 GeV
1000 GeV
22
Checking Physics EM calor. Hadr. calor.
testbeam

• 1996 Testbeam Setup for CMS HCAL
• A test module of copper absorber plates with
  scintillator tile sampling
• 28 scintillator plates with absorber of varying
  thickness in between
• A prototype lead tungstate crystal
  electromagnetic calorimeter
• Energies 225 GeV ? (for calibration), 10-300
  GeV ?, 10-300 GeV e-
• Magnetic Field 0, 0.75, 1.5, 3 tesla (direction
  parallel to the face of the scintillator)
• Configuration Only HCAL or ECAL HCAL
• ?5000 events were taken for each setup
• Response of the Calorimeter was studied as a
  function of
• Magnetic field effect on scintillator
• Absorber thickness optimisation of resolution
  versus containment
• Absorber depth energy containment
• Electromagnetic Calorimeter contribution e/?
  effects

23
Checking Physics EM calor. Hadr. calor.
testbeam
GEANT4 4.0.p02
100 GeV pi 0 Tesla ECALHCAL
24
Checking Physics EM calor. Hadr. calor.
testbeam
100 GeV e- 0 Tesla HCAL
25
Checking Physics EM calor. Hadr. calor.
testbeam
pi 0 Tesla ECALHCAL
Events with shower in ECAL
Events with MIP in ECAL
26
Checking Physics EM calor. Hadr. calor.
testbeam
pi 0 Tesla ECALHCAL
Events with shower in ECAL
Events with MIP in ECAL
27
Release testing

• The quality of each release should be checked
  before making it public
• Also each production centre has to test the
  release quality after installation
• Not only that it compiles and runs, but also
  that differences on results are understood
• Tool must be easy to use and allow to compare
  results automatically, without help from expert
• ? A tool has been developed a tool for
  regression testing in CMS OVAL
• See talk in this conference - David Chamont
• For physics we cannot do but an statistical
  testing of some distributions
• We are investigating the use of the GEANT4
  statistical testing suite (see talk at this
  conference M. Grazia Pia)

28
Current release testing in OSCAR

• Geometry
• Test the number of objects constructed
• OVAL Number of G4VSolid's 3190
• OVAL Number of G4LogicalVolume's 3190
• OVAL Number of G4VPhysicalVolume's 18988
• OVAL Number of Touchable's 1171167
• OVAL Number of G4Material's 341
• Send 1000 tracks randomly distributed in eta and
  phi and check material budget
• OVAL MaterialBudget 0 -3.52742 112.678 335.693
• Magnetic field propagation
• Send charged geantinos of energies 1, 10, 100,
  1000 GeV
• Check end point position w.r.t. straight line
  prediction
• OVAL DEVIATION 0 -575.577 708.6
• Check change in kinetic energy and deviation in
  direction
• OVAL MOMENTUM 0 2.11234e-07 -0.0276428 0.021342

29
Current release testing in OSCAR

• Hits
• Test that the same hits are stored in DB at the
  same positions
• OVAL TrkMuon Hit MBarrel hit 5 ID 1644952
• OVAL TrkMuon Hit MBarrel globalR3 438.533
• OVAL TrkMuon Hit MBarrel globalETA 0.0435001
• OVAL TrkMuon Hit MBarrel globalPHI 0.0436332
• OVAL TrkMuon Hit MBarrel TOF 14.65
• OVAL TrkMuon Hit MBarrel PABS 98.0501
• OVAL TrkMuon Hit MBarrel entryPtX 19.2179
• OVAL TrkMuon Hit MBarrel entryPtY 20.2891
• OVAL TrkMuon Hit MBarrel entryPtZ -6.57598
• OVAL TrkMuon Hit MBarrel exitPtX 19.1677
• OVAL TrkMuon Hit MBarrel exitPtY 20.3392
• OVAL TrkMuon Hit MBarrel exitPtZ -7.72598

30
Current release testing in OSCAR

• Physics
• First test check energy deposition and
• shower shape in calorimeter
• Not yet OVALised. Investigating GEANT4
• validation tool
• Other tests in progress
• Generator particles (tracking by GEANT4 and DB
  checking)
• Persistent tracks
• Monitor performance
• More comprehensive physics tests

31
Summary

• CMS GEANT4 simulation is ready for detailed
  physics comparison
• Data Chalenge 04 (50M events) starts July03
• OSCAR physics has to be validated in detail
• First physics tests show GEANT4 is as good or
  better as GEANT3
• OSCAR performance has to be similar to GEANT3
  simulation
• A comprehensive suite of OVAL tests is being
  built to assure the quality of each release