Introduction to Geant4

Geant4 is a complete redesign and rewrite of the Geant3 simulation package using C++ and object oriented technology. The project (CERN RD44) was initiated in 1994 by the CERN Geant team led by Simone Giani and is being undertaken by an international collaboration of over 70 physicists representing most major current and upcoming HEP experiments. Geant4 is being continuously compared through its development with the frozen Geant3.21, with the objective of producing by the end of 1998 a validated public release that betters Geant 3.21 in terms of physics, performance and capabilities. The complementary RD45 Project is developing a persistent object manager for HEP and will provide Geant4's object persistency (I/O). Geant4 is allied with the LHC++/CLHEP project to develop a standard suite of C++ classes and development/analysis tools for HEP.

Geant4 Background and Related Information

• Geant4: Simulation for the Next Generation of HEP Experiments (ps): HEP '95 Proceedings. Geant4 general overview. (S. Giani)
• 1995 Status Report (ps): Fall '95 prototype benchmarks comparisons with Geant3 (which were excellent). (S. Giani). More benchmark results.
• Status presentation at first-year review: November 1995. OOAD, class diagrams, benchmark results, 1996 activities. (S. Giani)
• Project subdomains and working groups
• Requirements document (ps)
• Class diagrams, top level class category diagram
• Milestones for 1995 (met) and 1996-97 (due in early '97)
• 8/96 TRIUMF workshop
• 11/96 RD45 workshop and Geant4 Objectivity benchmarks
• 12/96 CLHEP workshop
• CHEP97 papers
  • Geant4
  • Geometry in Geant4
  • Hadronic interactions in Geant4
  • ODBMS for LHC detector simulation
  • Geant4 Visualization
  • OPACS and Geant4 (visualization)
  • DAWN for Geant4 visualization
• 1997 Status Report ps (5.3MB), ps.gz (0.7MB)
• SLAC workshop 10/97
• Browse a recent code snapshot of the complete Geant4 code repository
• Related projects: RD45, LHC++

Geant4 user's documentation

Geant4 Geometry

Geant3 to Geant4 Migration via g3tog4

• g3tog4 documentation (Somewhat dated, but conveys what g3tog4 does)

STAR Activity and Plans

• Three G4 collaborators in STAR: Peter Jacobs, Thomas Ullrich, Torre Wenaus
• Thomas Ullrich is actively working with G4 for STAR
• Import and build G4 alpha release 2, done
• g3tog4 conversion of detailed GSTAR geometry, in progress
• Use Geant4 with converted geometry as testbed to study capabilities and performance
• Timescale for migration is long; post-startup
• Possible areas for future Geant4 development in STAR are being explored. Peter Jacobs has enumerated some good candidates...
  • Event size: Our heavy ion events are simply too large to be held in memory at once and it is necessary to break them into pieces (whose size needs to be optimised for performance but is otherwise irrelevant) and reassembled on output in a way that is transparent to the user, as is done now with gstar/g2t. This is handled by G4.
  • Event pileup: especially for pp running, we will have pileup in the TPC and SVT from events prior to and following the trigger event, due to the long drift time of the devices (the amount fo this obviosuly goes as the luminosity). An explicit mechanism for this is in G4.
  • Parameterisation and fast simulation: trading off simulation detail for speed for coarse high-statistics studies. G4 is developing powerful facilities for parameterised simulation. See this page for parameterisation information.