Fix Geant4 step limiter when ionization is disabled #920
Conversation
… just the interaction length) when ionization is disabled in Geant4
There was a problem hiding this comment.
Great catch @amandalund ! I will wait on @mrguilima or @stognini to compare the results before merging.
| @@ -85,6 +86,7 @@ GeantPhysicsList::GeantPhysicsList(Options const& options) : options_(options) | |||
| em_parameters.SetMscEnergyLimit(100 * CLHEP::TeV); | |||
| } | |||
| em_parameters.SetApplyCuts(options.apply_cuts); | |||
| this->SetDefaultCutValue(options.default_cutoff * CLHEP::cm); | |||
There was a problem hiding this comment.
Wouldn't this forcefully override any cutoff value defined originally by the gdml?
There was a problem hiding this comment.
I'm not positive about that... but at least I don't think it will change the current behavior: G4RunManagerKernel::InitializePhysics() calls physicsList->SetCuts(). The default SetCuts() method in G4VUserPhysicsList calls SetDefaultCutValue(), so since we don't have our own SetCuts() method it will be called either way. So all this should do is possibly change the default from 1 mm to something else.
There was a problem hiding this comment.
So that means that we have to add the SetCuts() function in the near future, when we actually start using the secondary production thresholds correctly.
|
@stognini would this change have to be adapted to the separate validation code? |
|
Here is the summary after some testings today and discussing the results with @amandalund:
|
Yes. I am pushing that update to the |
sethrj
changed the title
sethrj
added
performance
I believe the difference @mrguilima is seeing between Geant4 and Celeritas in the bremsstrahlung step length is because the step is only being limited by the distance to interaction (and not the range) in our Geant4 apps. When a process is registered to the physics list, the default process ordering is used (this specifies in what order the processes' at rest, along-step, and post-step
GPIL/DoItmethods are called by the stepping manager, and whether any of them are inactive). The along-step methods for ionization are active and for bremsstrahlung are inactive by default, since the ionization dE/dx and range tables will contain the contributions from all energy loss processes. When ionization is missing, we need to activate the along-step for brems—that is done here in ourGeantPhysicsList. I also added an option to set the default cutoff value (mainly for debugging).@mrguilima could you update the geant validation app with this change and see if the brems results improve?