Minutes2012 03 27

Minutes of 27 March 2012 Meeting of Geant4 Technical Forum

Editor: A. Morsch, 2 April 2012

- Starting at 16:00

Geant4 9.5-patch01 release
+ See slides

Physics issues addressed in 9.5-patch01
+ See slides

2012 development plan
+ See slides
- Experiments were requested to communicate with Geant4 developers
if they want version 9.4 to be supported beyond 2012.
- The presentation was followed by a short discussion on the support
of gmake for the built (being replaced by cmake).
Gabriele suggests that gmake will be dropped at the next major release.

Updates on open requirements
+ See slides
On Slide 5
The "triton issue" (higher tracking efficiency for anti-tritons as compared to triton)
was indeed a particular problem with the Triton cross-section implementation and not related
to the general systematic error on the cross-section.
The systematic error is mainly due to the lack of data.

Requirements from ATLAS
+ See slides
Comment to Slide 6:
The computing performance task force is still active but the web page has been moved to a different url.

Requirements from CMS
Status of CMS Full Simulation
+ See slides

Envisioning Use of Geant4-MT in a Multithreaded CMS Application
+ See slides
Makoto comments on slide 4 (Time Scale):
The presented time line fits well the major release plans (G4 vX).
He proposes close communication with frequent feedback from CMS.

Requirements from LHCb
Nigel Watson reports:

We did validation in autumn 2011 for our main production running,
using G4, v9.4.patch02. (Most recent production prior to that used 9.2.patch03.)

We were particularly interested to improve modeling of our IP
resolution, as reported in LHC Sim. workshop in Oct. 2011, and compared
various EM physics lists in our pre-production validations.
These included EmStandardPhysics_option1, _option3 and G4-provided
(July 2011) list EmStandardPhysics_option1LHCb, in which WentzelVI model
was used for all particle species (not just muons).

In our validations, we saw no significant improvement from using any of
these (and some unclear dE/dx behaviour) compared to our previous
production choice of EmStandardPhysics_option1, so conservatively
continue with this list for our current production run, started Dec.
2011. We currently suspect that the IP discrepancy may be due to
imperfect modeling of the very complicated RF foil for VELO; more
complete geometry is being investigated via CAD/STL-> GDML -> G4 geometry.

We anticipate MC production at a rate of ~150M/month in 2012.

Now that 9.5.patch01 has been released, we will start validation with
this, "no harm" tests initially, revisit EM list/Multi. Scattering
studies and progress to hadronic lists (currently still use LHEP).

In feedback to issues raised as requirements:
3201: Uniform Signature of Physics Builders, proposed solution of a
constructor with a single string as argument would be OK for us, when
could this be implemented?

2701: Cross-sections for K+/K-, will look into this as part of our 9.5
pre-production validation tests.

2901: Lateral displacement in large volume, the principle of the
proposed solution seems like a workaround to a wider problem, addressing
as it is only the case of muons?

2902: Displacement in thin volumes, investigation of VELO RF shield in
progress, will report results when studies conclude.

Requirements from ALICE
+ See slides
Slides 3 and 4
The plots on the secondary background comparing G3 to G4 should be remade
separating hadrons and leptons/photons.

Requirements from MU2E
+ See slides

Requirements from NOvA
+ See slides

Meeting Date: 

Tuesday, March 27, 2012