Simulation: Ion trajectories seem to be sharper in air than in vacuum

[pgrusell]

Describe the bug
When running simulations of the propagation of the ions under the effect of our dipole GLAD magnetic field using R3BRoot, we see that the trajectories of the ions in vacuum are smoother than in air (as seen in the picture below). It seems to be a problem inherit from FairRoot, do you have any clue regarding the reason of this issue?

Logs / Screenshots

System information (please complete the following information):

• OS: AlmaLinux 9.3 (Shamrock Pampas Cat)
• Compiler: GCC 11.4.1
• Environment: FairSoft jan24

[karabowi]

I can confirm your observations. Tests underway.

[karabowi]

I was able to reproduce the problem in geant4_vmc.
The issue has been submitted.

[karabowi]

The visualization of the particle trajectories in FairRoot bases on particle transport in Geant, specifically depends on calls to a Stepping function, which is used by FairRoot to record particle positions for the track visualization. The frequency of the calls to this function depends on material properties, f.e. its density. When a particle travels through solid materials, the function may be called several times for every mm of transport, while in the air the particles are transported for meters without invoking the Stepping function. That saves transport time, but results in approximate trajectories in thin media, like air.
However,

In Geant4 VMC, there is by default set a step limit 10 cm for all materials with density lower than 0.001 g/cm3. This default setting can be overridden (...)

link

Hence, by default, the particle transport through vacuum calls the Stepping function every 10 cm resulting in smooth trajectories,
quite different from coarse trajectories in air.
This can be changed by setting any of the appropriate flags:

/mcDet/setLimitDensity 1.e-6 g/cm3
/mcDet/setMaxStepInLowDensityMaterials 1 mm
/mcDet/setIsMaxStepInLowDensityMaterials false

[pgrusell]

Hi, thank you very much for your help. Indeed, it worked. In our case I used:

/mcDet/setLimitDensity 2.e-3 g/cm3

So air is included as low density material.

[karabowi]

You should be warned however, that this change might have negative impact on the run time.