Too Many Decay Processes assigned to particle types

[andresailer]

• OS version: N/A
• Compiler version: N/A
• DD4hep version: master
• Reproduced by: ddsim with ExtraParticles Plugin or PhysicsLists.decays=True or both
• Input: link to input files if applicable
• Output: full build and run log output
• Goal: Proper life time distribution

There are two places where dd4hep adds the decay process for particles, which can lead to a lifetime 2 or 3 times lower than desired.

ExtraParticles adds decays to particles, after Geant4 also does

The Geant4ExtraParticles plugin can add new particles to Geant4 to simulate certain b hadrons or charginos, or anything. In a first step the particles are added to Geant4

https://github.com/AIDASoft/DD4hep/blob/4fc82f8530a14a0cbfba107003d6715310d04c84/DDG4/plugins/Geant4ExtraParticles.cpp#L134-L136

And then later processes are added to the new "extra" particles

https://github.com/AIDASoft/DD4hep/blob/4fc82f8530a14a0cbfba107003d6715310d04c84/DDG4/plugins/Geant4ExtraParticles.cpp#L177-L179

However, the Decay process is already added to particles by Geant4 in most (I guess) regular physics lists

https://github.com/Geant4/geant4/blob/f794fcf60b75ac24b3ce8b64873b50757732a52a/source/physics_lists/constructors/decay/src/G4DecayPhysics.cc#L100-L123

void G4DecayPhysics::ConstructProcess()
{
  if(wasActivated) { return; }
  wasActivated = true;

  G4PhysicsListHelper* ph = G4PhysicsListHelper::GetPhysicsListHelper();

  // Add Decay Process
  fDecayProcess = new G4Decay();
  aParticleIterator->reset();
  G4ParticleDefinition* particle=0;

  while( (*aParticleIterator)() ) {
    particle = aParticleIterator->value();
    if( fDecayProcess->IsApplicable(*particle) ) { 
      if(verbose > 1) {
        G4cout << "### Decays for " << particle->GetParticleName() << G4endl;
      }
      ph->RegisterProcess(fDecayProcess, particle);
    }}}

Which leads to the decay process being defined multiple times for one particle, which increases the width and reduced the particle lifetime, unless the particle to be simulated has a pre-assigned life-time from the generator.

I think this addition of the decay process can be removed from the Geant4ExtraParticles plugin, because Geant4 will add the decay process. (or it is done again in dd4hep see next point).

PhysicsLists adds decays:

In addition to the above case, if the "decays" option for the PhysicsList is set to True, https://github.com/AIDASoft/DD4hep/blob/4fc82f8530a14a0cbfba107003d6715310d04c84/DDG4/src/Geant4PhysicsList.cpp#L338

https://github.com/AIDASoft/DD4hep/blob/4fc82f8530a14a0cbfba107003d6715310d04c84/DDG4/src/Geant4PhysicsList.cpp#L408-L410

Decay processes are added for all "applicable" particles

https://github.com/AIDASoft/DD4hep/blob/4fc82f8530a14a0cbfba107003d6715310d04c84/DDG4/src/Geant4PhysicsList.cpp#L413-L431

Obviously decays should be set to False when using a pre-existing physics list, but maybe even more protections could be added for this?

[andresailer]

Untreated, one can see multiple decay processes with /particle/process/dump

Idle> /particle/select lambda
Idle> /particle/process/dump
G4ProcessManager:  particle[lambda]
[0]=== process[Transportation :Transportation] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                 -1:      -1:       0:       0:       5:       0:
  parameter             -1:      -1:       0:       0:       0:       0:
[1]=== process[Decay :Decay] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                  1:       0:      -1:      -1:       4:       1:
  parameter           1000:    1000:      -1:      -1:    1000:    1000:
[2]=== process[hadElastic :Hadronic] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                 -1:      -1:      -1:      -1:       3:       2:
  parameter             -1:      -1:      -1:      -1:    1000:    1000:
[3]=== process[lambdaInelastic :Hadronic] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                 -1:      -1:      -1:      -1:       2:       3:
  parameter             -1:      -1:      -1:      -1:    1000:    1000:
[4]=== process[UserSpecialCut :General] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                 -1:      -1:      -1:      -1:       1:       4:
  parameter             -1:      -1:      -1:      -1:    1000:    1000:
[5]=== process[Decay :Decay] Active
  Ordering::             AtRest             AlongStep          PostStep
                    GetPIL/    DoIt    GetPIL/    DoIt    GetPIL/    DoIt
  Ordering::
  index                  0:       1:      -1:      -1:       0:       5:
  parameter           1000:    1000:      -1:      -1:    1000:    1000: 

[petricm]

From Alberto Ribon

I have checked the Geant4 code regarding the EM physics processes of charged particles (i.e. multiple scattering and ionization) and it seems that it does NOT add automatically these processes for those particles that are unknown to Geant4. In other words, differently than for decays, you need to explicitly assign the EM processes to the extra particles, unknown to Geant4, that you create.