SetParticleMomentumDirection for GPS is not working properly

Particles, Track, Event, Run and Biasing
1

Hi,

I’m trying to create a General Particle Source which I define in a configuration file.
I’m using this code for PosDist and AngDist:

        auto particle_source = std::make_unique<G4GeneralParticleSource>();
        auto single_source = particle_source_->GetCurrentSource();

        single_source->GetPosDist()->SetCentreCoords(config.get<G4ThreeVector>("source_position"));
        single_source->GetPosDist()->SetPosDisType("Beam");
        single_source->GetPosDist()->SetBeamSigmaInR(config.get<double>("beam_size", 0));

        single_source->GetAngDist()->SetAngDistType("beam2d");
        single_source->GetAngDist()->DefineAngRefAxes("angref1", G4ThreeVector(1., 0, 0));
        G4TwoVector divergence = config.get<G4TwoVector>("beam_divergence", G4TwoVector(0., 0.));
        single_source->GetAngDist()->SetBeamSigmaInAngX(divergence.x());
        single_source->GetAngDist()->SetBeamSigmaInAngY(divergence.y());

        G4ThreeVector direction = config.get<G4ThreeVector>("beam_direction");
        single_source->GetAngDist()->SetParticleMomentumDirection(direction);

The code compiles fine, but when I look at the MomentumDirection by using GetParticleMomentumDirection() I get the following direction, regardless of the G4ThreeVector direction used:
single_source->GetParticleMomentumDirection() = (1,0,0) (which is default)
particle_sourc->GetParticleMomentumDirection() = (0,0,-1) (which is default)

Could you tell me why the “single_source->GetAngDist()->SetParticleMomentumDirection(direction)” does not affect the direction?

For some reason, the “SetParticleMomentumDirection(direction)” function works if I add the following piece of code in the following position.

       ...
       single_source->GetAngDist()->SetAngDistType("beam2d");

       G4UImanager* UI = G4UImanager::GetUIpointer();
       UI->ApplyCommand("/gps/direction 1 0 1);

       single_source->GetAngDist()->DefineAngRefAxes("angref1", G4ThreeVector(1., 0, 0));
       ...

However, this gave me a slightly different beam compared to the beam without the added UI->ApplyCommand (probably because the “beam2d properties” are lost, but I’m not sure)
This something I don’t want so would like to solve the problem without adding the UI->ApplyCommand.

Cheers,

Koen

2

Hello,

The GPS code is “dense” and is more thought as to be used through command lines, ie macros, to avoid having to recompile your code.
Is using macros something limiting your applications ?

Marc

3

Hi Marc,

No need for macros anymore.
I solved the problem by rotating the coordinate system using “angref1” and “angref2” such that the negative z direction is aligned with the defined direction, which fixes the issue.

Cheers,
Koen