LCDG4: a Geant4-based detector simulator for the LCD Guilherme Lima

23 Slides245.00 KB

LCDG4: a Geant4-based detector simulator for the LCD Guilherme Lima for the NICADD simulations group Geant4 Tutorial Fermilab, Oct.28 2003

Outline NICADD Linear Collider and LCD Detector simulation Requirements Implementation details Event displays Status Oct. 28, 2003 2

NICADD Northern Illinois Center for Accelerator and Detector Development Research and Developments for the next Linear Collider (LC) Detectors: design, development, prototyping, testing and construction of a sampling, scintillator-based digital hadronic calorimeter Simulation and software developments by Dhiman Chakraborty (coordinator), Guilherme Lima, Jeremy McCormick, Vishnu Zutshi More information at http://nicadd.niu.edu Oct. 28, 2003 3

Linear Collider Electron-positron collisions at 500 to 1000 GeV Luminosities of 1034 cm-2 s-1 Physics motivation: Precision studies: Higgs, supersymmetry and other new phenomena Possible e-e-, e- and configurations Accelerator technology and site not yet chosen Start date: around 2015 Overlap with LHC is desirable Oct. 28, 2003 4

Detector geometry Some versions under study SDJan03 SD detector LD detector PD detector Detector components Oct. 28, 2003 VXD (vertex detector) Tracker EM calorimeter Hadronic calorimeter Luminosity detector Solenoid coil Muon system 5

Options under study: SD, LD, PD Oct. 28, 2003 6

A Digital Hadronic Calorimeter? Oct. 28, 2003 7

Simulation of a digital HCal High segmentation for better energy resolution: one bit (digital) or two bits (semidigital) per cell Questions to be addressed Cell shapes (squares, rectangles, hexagons) Cell dimensions (digital calorimeter) Projective vs. non-projective calorimeters Sampling fractions (number of layers, active to absorber ratios) Cost / performance optimization Oct. 28, 2003 8

Detector Simulation Support the detector design activities at NICADD Useful tool for the whole LCD collaboration Some basic requirements: Flexible geometry description Input from binary STDHEP Output in LCD-specific formats (sio, lcio) Hits should be associated to MC particles Oct. 28, 2003 9

Geant4 implementation G4RunManager UserInitialization UserAction PhysicsList LCDXML2G4DetConst (detector construction) LCDG4PrimaryGeneratorAction LCDG4EventAction LCDG4SteppingAction LCDG4TrackingAction G4VisManager G4UIManager Oct. 28, 2003 10

PhysicsList Basically inherited from Mokka (See http://polype.in2p3.fr/geant4/tesla/www/mokka/mokka.html) Oct. 28, 2003 GeneralPhysics charged, neutral geantinos, and all decays EMPhysics e , e-, , e, e HadronPhysicsLHEP G4 Meson, Barion and ShortLived constructors MuonPhysics muons and taus IonPhysics G4IonConstructor, Deuteron, Triton, Alpha, 3He 11

LCDXML2G4DetConst Detector construction from XML input file XML parsing is done by Sun’s xercesc library Limitation: cylinders or disks only Uses other classes to do the real work: Oct. 28, 2003 LCDXML2G4VXD LCDXML2G4Tracker LCDXML2G4Lum LCDXML2G4EMCal LCDXML2G4HDCal LCDXML2G4Muon LCDG4MaterialMgr LCDXML2G4Field LCDXML2G4InactiveVolume 12

XML detector description ?xml version "1.0" ? !-- SDJan03.xml - Standard Silicon Detector -- !DOCTYPE lcdparm SYSTEM "detParms.dtd" lcdparm global file "SDJan03.xml" / physical detector topology "silicon" id "SDJan03" -- Just an example: description of the HCal barrel -- volume id "HAD BARREL" rad len cm "1.133" inter len cm "0.1193" tube barrel dimensions inner r "144.0" outer z "286.0" / layering n "34" slice material "Stainless steel" width "2.0" / slice material "Polystyrene" width "1.0" sensitive "yes“ / /layering segmentation cos theta "600" phi "1200" / /tube calorimeter type "had" / /volume -- all other volumes dropped. -- /physical detector /lcdparm Oct. 28, 2003 13

LCDG4PrimaryGeneratorAction Inherits from G4VUserPrimaryGeneratorAction LCDG4readStdFile for binary StdHEP input file A Geant4 primary track is created for each MC particle in the input file LCDMcPart keeps track of parents and some additional information (one for each track, except for tracks produced inside a calorimeter shower) LCDG4Trk2McPart inherits from G4VUserTrackInformation and associates one or more G4Tracks to one LCDMcPart Standard G4ParticleGun can also be used Oct. 28, 2003 LCDG4PrimaryGeneratorMessenger can be used for simple interactive changes to the G4ParticleGun generator 14

LCDG4TrackingAction Oct. 28, 2003 Inherits from G4UserTrackingAction Builds the associations between parents and children, so that all hits can be traced back to a MC particle saved in output file It does this by creating an LCDG4Trk2McPart object for each secondary particle created by Geant during tracking 15

LCDG4SteppingAction Inherits from G4UserSteppingAction It complements LCDG4TrackingAction by creating LCDMcPart objects for non-primary particles (except for particles from calorimeter showers) It is also used for geantino scans, to calculate accumulated radiation and interaction lengths along straight lines Oct. 28, 2003 16

LCDG4EventAction Accesses all hit collections (G4HCofThisEvent) and streams hits and MC particles into output format Output formats are LCD-specific (sio and lcio) Data format keeps links between different objects, like hits associated to MC particles Some limited ASCII output can be used for debugging purposes Oct. 28, 2003 17

Hit Processing LCDG4HDCalSD inherits from G4VSensitiveDetector Determines which cell was hit based on the PreStepPoint coordinates (virtual cells!) LCDG4HDCalHit keeps track of hit information in each individual cell. It inherits from G4VHit Energy deposition is linked to the MC particle Energy deposition in absorbers are also kept for analysis Similar classes exist for each subcomponent Oct. 28, 2003 18

Projective vs. Non-projective An important parameter to be studied at the earlier stages of the design Mokka doesn’t currently have this functionality Projective Oct. 28, 2003 Non-projective 19

e e Z (SDJan03) - Oct. 28, 2003 - 20

e e tt event (SDJan03) - Oct. 28, 2003 21

e e Z qq event (SDJan03) - Oct. 28, 2003 22

Status Detailed comparisons between LCDG4 and Mokka under way Only cylinders and disks currently supported Plans to upgrade from XML to GDML Other Geant4 projects: NICADD test stand Test beam prototype (see Jeremy’s talk!) For more information, please check: http://nicadd.niu.edu/research/lcd Oct. 28, 2003 23

Back to top button