Minutes of 20 March 2014 Meeting of Geant4 Technical Forum
Editor: A. Morsch, 20 March 2014
- Starting at 16:20
Highlights of release 10.0-patch01 - non-physics part
+ see slides
Highlights of release 10.0-patch01 - physics part
+ see slides
- Q: It is unusual that a patch changes the Multiple Scattering Model. Why has this been done ?
This is only a change in configuration for Penelope, Livermore, Option 3,4 physics lists used
mainly in space and medical application for which Urban96 shows the best performance.
- Q: What motivate the 10 eV cut for multiple scattering (slide 4) ?
Above 10 eV choice of step limits is not too low and electrons stop after a few steps.
Below 10 eV step limits are so low that electron performs many step for essentially no gain in physics precision.
- Q: Can the limit be changed for other applications (DNA, medical, ..) ?
The limit can be changed by the user.
2014 work plan - non-physics part
- Comment Jon: Improved voxel safety (slide 15) already works, what is driving it from the navigator can be improved.
- Comment Makoto: A discussion with users about the Multi Threading interface requirements will be scheduled for May 12/13.
2014 work plan - physics part
Q: On which basis additions of new developments are decided ?
It is a mixture of general model developments to improve the agreement with data and specific user requests.
Q: What are the specific requirements for physics lists used by the high intensity frontier experiments, in particular, difference wrt
HEP physics lists ?
- energy ranges for hadronic models
- fragmentation functions
Q: What is the field of application for mu-electronics ?
- solid state physics, semi-conductors (e-hole creation)
- simulation of interactions in crystals
Requirements from LPCC Simulation workshop (https://indico.cern.ch/event/279530/?ovw=True)
Comment Makoto: So far only one request for support of Windows Multi Threading. No resources available.
Will not work on this until C++11 delivers this for free.
Please send comments!
Feedback from MINERvA and other Fermilab Intensity Frontier Experiments
Exposing Geant4 model parameters, where appropriate, providing guidance about their meaning, and making it
simple for experiments to tune them can help experiments achieve these goals.
Alberto Ribon: Better do not provide parameters to users. Let developers perform the tuning.
Use different physics lists to estimate systematic errors.
Of course code is open source. However we do not recommend to change parameters.
Vladimir Ivantchenko: Flux calculation is a problem that is more difficult than standard Geant4 hadronic physics
since only a very limited part of phase space has to be simulated precisely whereas normally one averages over large
parts of phase space.
Federico Carminati: Some parameters are not meant to be changes. Vary parameters to evaluate sensitivity of the simulation.
However do not change them in the sense of a fit.
Alberto: Propose a meeting between G4 developers and Minerva to discuss this issue.
Thursday, March 20, 2014