Skip to content

LHCb optimisations of processes with optical photons

Michal Kreps requested to merge torchSim11 into master

In the past, we had private copy with optimisations of several Geant4 processes. Changes were designed or to optimise process to avoid optical photons which will never make it or will be thrown later due to efficiencies or to collect useful information for studies and understanding. Modifications I'm aware off were:

  • G4Cerenkov: Main changes used to be
    • Limit in which materials/volumes Cherenkov photons will be generated. This modification is implemented here.
    • Gather information about Cherenkov photon production for detailed studies. In order to improve maintainability, this is moved out of the physics process and will be implemented as separate stepping action.
    • Generate correlated random background. I'm considering to implement this as separate physics process to run after Cherenkov process. It would have small penalty, but at the same time make code structure cleaner and easier to maintain.
  • G4OpBoundaryProcess: This contains couple of modifications:
    • Store some information on what happened to photon on its way, especially whether it went through Rayleigh scattering or whether it was affected by surface roughness. I think this can be done in stepping action and will be tackled elsewhere.
    • Absorb photons on some boundaries where they would continue by reflection or transmission. This removes photons which will not make it to sensitive detector (or get there in completely wrong time). I do not have use case in Torch, there is case for this in Rich and we should look into how to implement it. Can be done here or in another MR.
    • Fiddle with transmission factor on some surfaces to avoid reflecting in situation when reflected photon would be lost. Again, no use case for Torch, but Rich might want to do this.
    • Need to double check that bug we discovered in Torch is already fixed here or backport in case it is not.
  • G4Scintillation: Never used it in Torch, so never looked through details, but I suspect we can handle needed changes in the same way as in G4Cerenkov
  • G4PhotoElectricEffect: Never used it in Torch, so do not know what has been done inside. Will need to look when there is use for it.
Edited by Michele Veltri

Merge request reports