Simulation Project Leaders Meeting 15-December-2009 ==== Attendees John Apostolakis (JA), Pere Mato (PM), Alberto Ribon (AR), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: Alfredo Ferrari. --------------------------------------------------------------------- ==== Minutes - Nothing to report for line management. --------------------------------------------------------------------- ==== Simulation physics validation & framework - Alberto - Meetings: o At the meeting on November 25th, there was a presentation by P.Strizenec about validation of 9.3-beta on the ATLAS-HEC; both QGSP_BERT and QGSC_CHIPS lists provide good response and resolution whereas FTF_BIC shows too high response and resolution too good. Another presentation was by V.Grichine on an alternative approach on simulation of multiple scattering. The new developments in CHIPS with new CHIPS physics-list were instead presented by M.Kossov; the new list allows the use of the CHIPS as single model, with its own "QGS" implementation, providing a natural way to match this 1-D high-energy model with the 3-D CHIPS fragmentation; the new CHIPS physics list will be in the new Geant4 release 9.3. o Next meeting scheduled for February 10th. - Validation studies: o Tests have been performed on the new CHIPS physics list; it showed to be quite stable, 50-60% slower in CPU speed than QGSP_BERT; provides rather large energy resolution and wide shower, with similar longitudinal shower extension and energy response. The nuclei fragmentation implemented in the CHIPS model seems to help in achieving these results. o Also tests have been carried out on the upgraded Fritiof model which will be distributed in release 9.3, particularly concerning the energy transitions. --------------------------------------------------------------------- ==== Generator services - Alberto - Meetings: o The meeting on November 18th was dedicated on Sherpa, with the last release 1.2.0, which is the first full featured version with documentation now available, strongly recommended for use. A new version of Sherpa, 1.3.0, is expected soon, including an extra model for minimum-bias. o Feedback from the experiments on the use of the new C++ generators has been positive so far, indicating that the new generators (Herwig++, Sherpa and Pythia-8) are become more and more popular and being used. o The 2nd planning meeting of 2009 took place on November 27th, slides and minutes are on the web. AR reported on the status of the project, which is in good shape; also reported about the work done for using autotools in the installation of GENSER, and the migration to SLC5 which has been completed; the future plans are to concentrate on the deployment and debugging of the new build-system, improve testing by reviewing all tests and require feedback from the authors, include new tests, implement more histogram-based tests and integrate the Rivet tool for use as regression testing. - HepMC: o A meeting planned for February 3rd will be dedicated on a discussion for the new release 2.06. o A request from the Linear Collider community has been advanced, to allow for binary root-I/O persistency. --------------------------------------------------------------------- ==== Fluka - Alfredo - No report. --------------------------------------------------------------------- ==== Geant4 - John - Releases: o 9.3 is almost ready, be out on Friday. o Integration testing now reports no error or warnings in several platforms. o Support of release 9.2 for LHC experiment use cases is planned until December 2010. In process of creating a catalog of fixes that will be included in a future patch to release 9.2, 9.2.p03. - Developments (everything included in 9.3): o Improvements in hadronic physics models: FTF: new Reggeon cascade added to improve the low energy behavior (pion production at 3-6 GeV). Improved binary model channels. CHIPS: extended CHIPS model for hadronic interaction (blend of string model and absorption with CHIPS fragmentation/decay). Unique features: very large energy resolution and showers are very wide. o EM refined multiple-scattering model (available as option), providing separate tunings for e, mu and hadrons. o Physics lists: new CHIPS physics list; new lists from 9.3-Beta kept as experimental/emerging (QGSP_FTFP_BERT, FTFP_BERT_TRV); several lists labelled as obsolete, including QGSP. o Fixes: energy conservation fixed in interface between FTF and Binary cascade models; a correction in CHIPS for a case o memory corruption leading to a crash, reported by ATLAS. Many other fixes, found in integration, regression, physics testing... - Feedback from experiments: o Candidate release was provided to all experiments; CMS provided positive feedback and some first comparisons. o Effort to assess CPU performance of ATLAS Geant4-based simulation with report expected by January 30th 2010; participating are A.Dotti (SFT) and A.Buckley (Univ of Edinburgh, ATLAS). - Meetings: o Asia-Pacific Workshop and Tutorial in Tsukuba (JP), 7-11 December: several presentations from users in HEP and medical physics, among which a presentation by Super-Belle. A lot of discussion with users; to mention in particular a study on the effect of electric/magnetic field on muon spin precession. Also a meeting among developers was organised to discuss the preparation of the the architecture review which is planned for 2010. Key targets for improvement will performance and maintainability. Planning further meetings at CERN and Kobe as part of review process. --------------------------------------------------------------------- ==== Garfield - Rob - Meetings: o RD51 collaboration meeting: visit of Steve Biagi, author of the Magboltz program to calculates electron transport in gases. Updates to the excitations of noble gases were discussed as well as the splitting functions. Steve gave also a detector seminar. o Forthcoming meetings: Paris meeting on discharges in micromegas; Vienne conference on instrumentation; Freiburg RD51 collaboration meeting. - Developments/activities: o Measurement campaigns are planned for gain fluctuations (Paris, Jan 2010) and excitation transfers (Paris, tentatively Mar/Apr 2010). --------------------------------------------------------------------- ==== AOB Next meeting scheduled for February 23rd.