Simulation Project Leaders Meeting 24-April-2012 ==== Attendees John Apostolakis (JA), Andrea Dotti (AD), Alfredo Ferrari (AF), Witold Pokorski (WP), Rob Veenhof (RV), Gabriele Cosmo (GC). --------------------------------------------------------------------- ==== Minutes - For the 2nd year, the SFT group is participating to the Google Summmer of Code initiative. Two students have been allocated this year to work on simulation on the reordering of code execution in Geant4 and on GPU studies. The students (from Greece and from India respectively) will work for a total period of 3 months. In total, the SFT group received 85 proposals, and 7 slots have been allocated for SFT. --------------------------------------------------------------------- ==== Simulation physics validation & framework - Andrea - Meetings: o Participated at the MC4BSM workshop at Cornell University. Interesting exchange of experience on the subject of validation of LHC data. Two main topics of possible future collaborations have been discussed: use of Geant4 for specific BSM models with long-lived particles interacting with detectors; use of MCPLOTS scripts to improve the validation web-pages and validation web-applications. o Try to organize a meeting with experiments before the summer. Tentative date: first days of June - Activities: o Validation of the two recent Geant4 patch releases performed: 9.5.p01 and 9.4.p04. No particular findings or discrepancies to report o Consolidation of GRID tools to run validation on distributed resources: a list of reliable sites where to run the Geant4 validation suite has been identified (CERN: 35% resources, KEK: 35%, France: 20%, UK+Nikhef: rest). It should be noted that the tools and scripts, even if developed for Geant4 validation, could be adapted to any application which needs to run on the GRID and produces relatively small output (CPU bound). o Detailed documentation to run validation has been produced. Still revising and improving it, but all relevant steps (from obtaining a GRID certificate to submitting jobs and analysis of results) are covered. o A new test, initially developed by the LHCb collaboration, is now available in the Geant4 testing suite and is regularly run as part of the nightlies and CDash testing systems. A technical issue has been identified and will be corrected soon. --------------------------------------------------------------------- ==== Generator services - Witek - GENSER: o New generators installed: epos-1.99.crmc.v2975, pythia8-163, photos++-3.4, pythia6-426p12, rivet-1.8.0, agile-1.4.0, sherpa-1.4.0.2, thepeg-1.7.3 and lhapdf-5.8.7. - HepMC: o Two bug-fix releases 2.06.07 and 2.06.08 (fix for use of default units and introduction of CMake build system), requiring rebuild of the whole GENSER stack which is based on it. - Meeting: o The 2012 yearly review planning meeting will take place on April 25th; expecting some feedback from the experiments to also finalize the plan of work. No big changes foreseen. --------------------------------------------------------------------- ==== Fluka - Alfredo - Fluka: o The last public release is 2011.2, no yet decided when a new public version will be released, perhaps beginning of next year. o The FLAIR package now allows also to build the geometry interactively and import technical drawings; it embeds the capability to produce 3D displays of the geometries. o Currently working on improvements to the nuclear models, particularly in the low energy region, with three main goals: further improving processes of relevance for the ENVISION European program for hadrontherapy monitoring; improving photonuclear reactions particularly for underground studies; improving light nuclei reactions (alphas etc) at low energies. o Finalization of sophisticated scripts for generation of geometries out of accelerator optics descriptions. o Neutrino activities: heavily involved to provide calculations on the tests made at CERN and Gran Sasso, and in calculations about the changes in the neutrino spectra and the rates of pair Bremsstrahlung events in Gran Sasso in case of superluminal neutrinos. - Manpower: o Turn-over of fellows, asking for three fellows in substitution to others finishing their contracts; among these Anton Lechner who now got a staff position in the group. o Got a PhD student to work on improving and extending the physics of DPMJet-3, in collaboration with the Karlsruhe University; need to improve the description of p-p and heavy ions interactions. - Courses: o A course has been held foreseen in Prague at the beginning of December; one at Jefferson Lab next week and one in Triumf in mid September. o A course on radiation damage calculations with FLUKA has been held in Montpellier one month ago. --------------------------------------------------------------------- ==== Geant4 - John, Gabriele - Releases: o A new patch, 9.5.p01 was released at the end of March, including important fixes for issues also reported by ATLAS. Fixes were back-ported also to the 9.4 series in a separate patch 9.4.p04, released last week. These include a protection to the Urban multiple-scattering model to avoid observed artificial scattering of big angles for high-energy e+-; fix to sampling at small steps (below 10 um) for fluctuation in heavy media, to correct shift in range of 3 MeV muons; fix to the WentzelVI multiple- scattering model for the sampling of single scattering; a check in CHIPS capture-at-rest process to identify cases of massive energy conservation violations and correct them. Patch 9.5.p01 also includes a fix in the hadron-elastic process for a problem of memory growth observed in release 9.5 of last December. All details for 9.4.p04 and 9.5.p01 patches can be found respectively in: http://geant4.cern.ch/support/Patch4.9.4-4.txt http://geant4.cern.ch/support/Patch4.9.5-1.txt - Experiments/users feedback: o Three issues identified by ATLAS with Geant4 9.4: unphysical TeV particles from anti-proton annihilation, large deflections in GeV electrons in a percent of tracks, and incorrect bending of muons from the new multiple scattering model in presence of field and ionization. All issues are patched in 9.4.p04 and 9.5.p01, as reported above. o Identification of improved sections for tritons, requested by ALICE. Comparisons found that Shen cross-sections are best; a physics-list that uses these has been provided. o On-demand parallelism at event level requested by CMS in the context of Geant4-MT, which will require adaptation. - Meetings & courses: o The last Technical Forum was hold on March 27th. The development plan for 2012 has been presented and discussed. Details on the 2012 planned developments can be found at: http://geant4.cern.ch/support/planned_features.shtml It was proposed to keep supporting release 9.4 series until end 2012; also proposed to drop support for 32-bits systems (only 64-bits). There will be a minor release 9.6 of Geant4 in 2012 to be released end of November; 9.6 will be the last release of the "9" series. A major release is foreseen in 2013 incorporating multi-threading. A new release of the Geant4-MT prototype is being prepared, based on release 9.5.p01. o A tutorial on Geant5 took place and organized by NASA/MSFC, Alabama (USA) on April 15-19th, restricted to NASA employees. o New courses will take place at Porto Conte, Alghero (Italy) end of May as part of the 9th Seminar for Nuclear, Sub-nuclear and Applied Physics, and at Jefferson Lab, VI (USA) on July 9-13th. - Makoto Asai (SLAC) has been re-elected spokesperson of the Geant4 Collaboration for a new mandate; Marc Verderi (IN2P3/LLR) acting as deputy. --------------------------------------------------------------------- ==== Garfield - Rob - Publications: o SRIM paper published (10.1016/j.nima.2012.03.020). o GEM gain paper in progress. - GEMs: o The gain in GEM detectors increases over a period of e few 100 s from switching on, before the gain stabilizes. New transients appear on each change in rate. This can flaw dE/dx measurements. o Measurements and calculations are being performed to better understand the underlying mechanisms: initial movement of light ions, including protons from polyamic acid, followed by electron conduction. The latter is hindered because conjugation is broken at the ODA ether bond, but vd Waals links may provide sufficient cross-chain conductivity. o Excitation transfer probabilities in GEMs seem to agree with those derived from tube measurements, once charging-up has been corrected for. - Fields: o neBEM has introduced SVD influence matrix solving for near-degenerate cases as encountered in Micromegas detectors, and to a lesser extent GEMs. o Stating interest in automatic surface panel generation for near arbitrary geometries. - Gases: o CF4 reported on to CMS. o Dissociation reactions in alkanes. o Non-equilibrium effects in avalanches. - Plans: o Dielectric transients. o Photo-layer TGEM. - Meetings: o Industrial application workshop going on in Annecy. o Schools in Firenze and Bodrum. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for June 26th.