Simulation Project Leaders Meeting 27-September-2011 ==== Attendees John Apostolakis (JA), Andrea Dotti (AD), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: Alfredo Ferrari, Pere Mato. --------------------------------------------------------------------- ==== Minutes - LPCC (LHC Physics Center at CERN) is organizing a 2 days workshop on the Simulation of LHC detectors on October 6-7th. The workshop will feature presentations from the experiments and from Monte Carlo code developers. The workshop will contain a session to discuss requirements from LHC experiments that will drive the future work of the Physics Validation Project. More information can be found at the LPCC website: http://lpcc.web.cern.ch/ The agenda of the workshop is available here: https://indico.cern.ch/conferenceDisplay.py?confId=144956 --------------------------------------------------------------------- ==== Simulation physics validation & framework - Andrea - Meetings o No meetings recently in summer. The Workshop next week will substitute the September-October meeting, where will try to understand what is the best way for organizing the meetings in future; the format of the meeting will be defined there. - Activities o The simplified calorimeter test is now able to compare directly with some test beam data, specifically, the ATLAS TileCal and HEC. The idea is to extend this to other comparisons too, as the application itself is detached from the framework of the experiments. o A note for the HELIOS diffraction is now available from the AA document area and project website. The 4th benchmark can be considered now completed. --------------------------------------------------------------------- ==== Generator services - Witek - GENSER: o New generators installed: pythia8-153, vincia-1.0.26, lhapdf-5.8.6, agile-1.3.0, rivet-1.6.0, tauola++-1.0.5, hydjet-1.8, phojet-1.12-HepMC, sherpa-1.2.3.2, herwig++-2.5.1, pythia6-425.2, herwig-6.520.2, cascade-2.2.04, blackmax-2.02.0, alpgen-2.1.4, thepeg-1.7.1, rivet-1.5.1. o Some new tests implemented. - HepMC: o Last meeting was on HEPMC, next meeting will be scheduled in November. o Experiments finally moved to the 06 series. The transition was quite smooth. Quite some discussion especially with LHCb for the CMT interface. - MCSDB: o CMS officially declared that they will not use MCDB in future. As CMS was the only customer for MCDB so far, it was decided to keep the project alive until the end of the year, and close it after that. --------------------------------------------------------------------- ==== Fluka - Alfredo - No report. --------------------------------------------------------------------- ==== Geant4 - John - Meetings: o The 2011 Geant4 workshop was held at SLAC from 19-23 September. Among the topics presented were: * a new feature for layering mass geometries. This overlays additional geometry setups and undertakes tracking in parallel in each one. This can be used to present geometries of different detail to different particle types. It is well suited for the realization of support structures in HEP experiments. * the latest kernel developments, including the re-engineering of the costly GetVelocity() method to boost CPU performance; * the multi-threaded prototype of Geant4 (Geant4-MT) for which test measurements were reported. Also presented new ideas for parallel processing, in the context of a new parallel application framework. * guidelines for making code thread-safe were described. Tools for CPU profiling and their use in code improvement were demonstrated. * the new CMake build infrastructure and its integration with CPack tool for creating native installers on different systems, the use of the CTest test execution system and the CDash tool for presenting the results of test runs; * results of the latest code improvements to the Bertini cascade, where the memory churn due to Bertini. Once a dominant source of churn, this has now been reduced of a factor 100; * many physics improvements, including the extension of FTF to anti-nucleon and light anti-ion projectiles, and a fix to the kaon cross-sections used in Bertini allowing it to interact kaons above 6-7 GeV; * the improved FAST profiling tool from the Fermilab team. This is the first version of the tool which is available for download. * the latest improvements in reducing memory churn, achieved by revising the implementation and use of physics vectors, relevant for multi-process/thread applications. Several additional topics were discussed, including: * the need for a tool to check coding rules, and its potential adoption; * the ways to widen the use of performance profiling and tools, in order to improve performance. Proposal brought to automate the profiling of key benchmark applications, by using one or more of the existing tools (igprof, FAST and native tools.) Plans were presented to develop pi- capture and gamma-nuclear processes in Bertini as potential replacements of the only current production implementation, which is based on CHIPS. Presented also was the new capability to simulate phonons in silicon detectors. - Ongoing developments: o Improvements to the Bremsstrahlung model are going on, with the goal to address the discrepancy in the EM lateral shower shape reported by ATLAS and CALICE. - Experiments feedback: o Several issues reported by ATLAS have been investigated: for monopole simulations and propagation in magnetic field, where the correct integration stepper was needed for the equation (Classical-RK4); issues were identified in the extruded-solid shape, causing particles to exit the detector, the problems causing this were identified and fixes were prepared and included in the last Geant4 development release. o The reason particles became stuck at volume boundaries (reported by Atlas) is now understood and a fix is being prepared to improve the intersection of curved tracks with surface boundaries. o ALICE reported they have run productions with Geant4 9.4.p02 and 9.5-beta. o LHCb is testing an experimental physics-list using a refined model of multiple-scattering, which provides precise displacement in thin layers. The same list was tested also subsequently by CMS, which has reported some issues of robustness. --------------------------------------------------------------------- ==== Garfield - Rob Papers: - Electroluminescence paper published (Phys. Lett. B) 10.1016/j.physletb.2011.07.081 - GEM gain paper in preparation, Penning rates in these devices seem to match those calculated from measurements at much higher fields in tubes Gases: - Trimethyl amine (light emitter, Penning dopant) cross-sections added - low energy Ar excitation updated, small discrepancy found with recent data. - Krypton updated and revised. - Xe ionisation updated. - Xe excitation cross-sections under scrutiny. People: - Numerous and excellent summer students this year, also people from others groups joined. - Heinrich Schindler now working for LHCb. - Ozkan Sahin due to get his PhD in 1 month from now. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for November 29th.