ALLEGRO migration to ddsim
Juraj Smieško
FCC Detector Full Sim Working Meeting
06 December 2023
DDSim vs. K4SimGeant4
FCC Study Report: One simulation tool to rule them all
DDSim
- Part of DD4hep (able to compartmentalize subdetectors)
- Wrapping around Geant4 — Geant4 controls event loop
- Configured with standalone Python steering
- Used by multiple experiments (including currently running ones)
- Larger feature set
k4SimGeant4
- Set of Gaudi algorithms
- Event loop controlled by Gaudi
- Ability to be integrated into large Gaudi steerings
- Uses DD4hep to compartmentalize subdetectors
Simulation details
DDSim + digitization algorithms from k4SimGeant4
vs.particle gun + k4SimGeant4 Geant4 Gaudi algorithm and digitization
-
No changes to the k4SimGeant4 algorithms
- geantservice.physicslist = "SimG4FtfpBert"
- geantservice.g4PreInitCommands += ["/run/setCut 0.1 mm"]
-
Default DDSim steering
- SIM.physicsList = None
- SIM.physics.list = "FTFP_BERT"
- SIM.physics.rangecut = 0.7
- Compact files from FCCDetectors
-
Number of events:
- electron @ 50 GeV: 10k
- pion @ 50 GeV: 10k
ECal
Geant4 Output
Electron @ 50 GeV
ECal Hits
Electron @ 50 GeV
ECal Hit Contributions
Electron @ 50 GeV
ECal Cells
Electron @ 50 GeV
ECal Positioned cells
Electron @ 50 GeV
ECal Cluster
Electron @ 50 GeV
ECal Corr. cluster
Pion @ 50 GeV
ECal Hits
Pion @ 50 GeV
ECal Hit Contributions
Pion @ 50 GeV
ECal Cells
Pion @ 50 GeV
ECal Positioned cells
Pion @ 50 GeV
ECal Cluster
Pion @ 50 GeV
ECal Corr. cluster
HCal
Electron @ 50 GeV
HCal Hits
Electron @ 50 GeV
HCal Hit Contributions
Electron @ 50 GeV
HCal Cells
Pion @ 50 GeV
HCal Hits
Pion @ 50 GeV
HCal Hit Contributions
Pion @ 50 GeV
HCal Cells
Conclusions / Open questions
- Running DDSim with default settings will get you close
- Investigate sensitive actions
-
Differences in simulation settings?
- Step length? Filters? Mag. field?
- Effects of hit deduplication?
- Can we drop positioning tools?