Simulation Project Leaders Meeting 18-December-2007 ==== Attendees Alfredo Ferrari (AF), Pere Mato (PM), Witek Pokorski (WP), Alberto Ribon (AR), Gabriele Cosmo (GC). Excused: John Apostolakis (JA), Rob Veenhof (RV). --------------------------------------------------------------------- ==== Minutes - As from February 2008, the Generator Services sub-project will have a new leader, Alberto Ribon, who will replace Witek Pokorski who is moving to a new position in the CERN/IT department. I wish to congratulate with Witek for the excellent job made and excellent collaboration everybody in the Simulation Project could benefit from him in the past years. We all wish to Witek all the best for his new job and congratulate for the new position. - In view of the new scenario which will take place in 2008 concerning the available manpower, the Simulation Project will be reorganised and the scope of affected sub-projects reduced accordingly: o The GDML work-package, now mainly in maintenance phase, will become part of the general maintenance of the client tools which now implement internally the related plugins for import/export of GDML descriptions. o Direct involvement of the current manpower in physics validation benchmarks will be reduced and some of the existing milestones eventually delayed until new manpower will be assigned to the project. o Physics Validation and Simulation Framework will be merged in a single sub-project under the leadership of Alberto Ribon. - LHCC Comprehensive review results: positive; no particular issues reported and related to the Simulation Project. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Meetings: o The Physics Validation meeting was taken on November 21st; it was a busy meeting with 6 presentations. RV summarised the work done on the activation studies in CMS using Fluka, confirming the results obtained with the old analysis; Flugg was not used due to overlaps in the CMS geometry detector description available. The final results on the neutron TARC analysis were presented by A.Howard; two physics lists are good in describing the setup: QGSP_BERT_HP and QGSP_BIC_HP. M.Kossov presented the latest developments on ion-ion physics in CHIPS. V.Grichine on behalf of N.Starkov gave an overview on the latest developments on elastic-scattering in Geant4. P.Arce presented a summary on the evolution of the multiple-scattering process in the various releases of Geant4. Finally, S.Piperov on the CMS 2006 test-beam studies, where a clear disagreement on the crystal energy deposition is observed; here many details are under investigation (considering also that quenching in the experimental setup has never been measured!); recently, after several discussions, improvements were achieved, by including noise, Birks effect for BGO crystals, and same requirements for minimum ionizing particle in the simulation as for the data; investigations are going on Cerenkov effect; it was also identified room of improvement in Geant4. - Extension of ATLAS TileCal 2002 analysis with Fluka: o Generation of all samples is essentially finished; now completing the analysis. The results for the 20 GeV samples have already been presented. Currently working on the 50, 100, 180 and 350 GeV samples analysis. Reasonable agreement is observed so far. - Extension of ATLAS HEC analysis with Fluka: o For this new study planned for 2008 the evolution is not clear, given the shortage of manpower affecting from February the project. It is proposed to start anyhow the work already in January in collaboration with the group from Munich, and quantify the work necessary to put in place the setup with Flugg and start the samples production, considering that it should easier in terms of geometry to put in place the setup compared to the TileCal analysis. In the following months the situation needs to be clarified and it will be asked for having assigned a student. - Simple benchmarks: o For the 4th benchmark, the situation with the data needs to be investigated further; furthermore, proton-proton data is considered important for more complete investigation of the diffraction, therefore requiring additional analysis. The activity will be delayed until a student will be eventually assigned. --------------------------------------------------------------------- ==== Generator services - Witek - Manpower: o Leadership: Alberto Ribon will lead the sub-project from February 2008. o M.Kirsanov has prepared a plan of manpower for visits in 2008. The plan is satisfactory for the needs of GENSER. o L.Dudko has been giving a presentation in CMS for MCDB; it is not clear yet if the tool will be used in CMS. This needs to be clarified with people in CMS, also in view of the LCG resources assigned to the MCDB project. o There has been interaction with people in Desy (H.Jung); there seems to be availability of manpower for work related to MC generators; a contribution may come to GENSER both at technical level and for the validation aspects. - Meetings: o The project planning meeting was held at the end of November. Minutes are posted on the web: http://indico.cern.ch/conferenceDisplay.py?confId=24411 o In the monthly meeting taken on November 28th, there was the presentation of the new release of Herwig++ by P.Richardson, and a presentation on the new developments in Rivet by A.Buckley. - GENSER: o The new Herwig++ 2.1.1 bug-fix release has been added, together with a new version of Cascade (2.0) and more tests. --------------------------------------------------------------------- ==== Fluka - Alfredo - Busy time in CNGS measurements following the failure on the electronics which happened after two weeks of full beam operation. Radiation calculations are in agreement with the measurements. - Quite a lot of activity also for the neutron Time Of Flight facility in working on designing the new target. The measurements of activation in the ToF chamber revealed some problems, with 5-10 times higher than expectations; the problem was understood and cross-checks made, it turned out to be due to the content of cobalt in the support of the target. The lesson learnt from this was that the composition of the steel is highly critical, a part over 10000 composition must be known. - Releases: o The new release of Fluka expected for December is delayed and will likely happen in mid-March. o Work undergoing in several areas: neutron cross-section library (in good shape, under control of a fellow on European funding); low energy ion-ion generator; new neutrino nucleus deep inelastic event generator; cosmic-rays benchmarks. o It is also planned to have included in the new release the new high-energy event generator - Meetings: o The Fluka Collaboration meeting was recently taken at CERN. o A Fluka course at CERN is foreseen in Spring. A second course in 2008 will be given in Paris. --------------------------------------------------------------------- ==== Simulation framework - Witek - Flugg: o A request has come from ATLAS to use Flugg for tracker studies; some problems were reported with the use of data cards together with the newest Fluka release. o Currently finishing the production of samples for the ATLAS TileCal validation study. - HepMC: o Implemented fixes in Root allowing for saving persistent HepMC objects; L.Sonnenschein is going to check this. - GDML: o Z.Torzsok has finished the integration of the GDML reader in Geant4, and it is now part of the latest Geant4 release, 9.1. The writer is the next task planned for the first months of 2008. --------------------------------------------------------------------- ==== Geant4 - John (offline report) - Releases: o Release 9.1 of Geant4 has been announced on December 14th as planned. Among the various fixes included, the new release also provides several new features, among which: a new GDML plugin for importing detector description setups; new command-based scoring capabilities, based on the parallel navigation feature; a new navigation technique for regular patterned geometries (phantoms) for use in medical physics; refinements in electron multiple scattering; first alpha versions of the Liege intranuclear cascade, a native QMD low-energy nucleus-nucleus model and re-scattering interface to the Binary Cascade; new physics lists options to enable the analysis of test beam data, including diffraction model for the description of the shower shapes at different energies and different combinations of models.; etc. Details can be found in the release notes published at: http://geant4.cern.ch/support/ReleaseNotes4.9.1.html The CPU performance should have some improvement in the hadronic part, of the order of 5 %. o Large scale testing of two candidate releases and the final release utilising several simple calorimeter setups with different physics lists, initial particle IDs and energies on the Grid identified a number of software issues, and tested in regression the physics performance. The validation of the hadronic physics has been done with 5000 jobs submitted on the Grid, all submitted on CERN resources; it is about half the number of jobs compared with those run for the previous release, mainly due to the reduced computing resources with SLC4 available for Geant4. o Problems regarding the reproducibility of Geant4 simulations in some hadronic models have been found. A fix for one identified issue has been undertaken and included in Release 9.1. Another, rare, issue is under active investigation. - Meetings: o Meetings last week with members of CMS regarding the HCAL and combined HCAL/ECAL test beam results have demonstrated progress. These preliminary results will need to be further confirmed. o Contacts are being undertaken with LHC experiments in order to establish a plan for concentrating the maintenance effort relevant to use of Geant4 in large scale productions. This is expected to involve minimizing the number of supported versions used for production in LHC experiments, including concentrating on a single version for an agreed period at a future date which would be agreed. A future standardisation on a single version will ensure that the effort available can concentrate on addressing any issues identified. o The Steering Board has met on December 11th, where an assessment on the developments planned for 2007 and the December release has been made. o The 5th Geant4 Space Users Workshop will be taken in Tokyo on February 13-15th. --------------------------------------------------------------------- ==== Garfield - Rob (offline report) - Garfield: o A proposal writing for RD51 is in progress, and exploratory work is underway to make the programs run with Root, in such a way as to ensure also a compatibility with Geant. Work is gradually starting on the full microscopic tracking of electrons - an area particularly relevant for the new generation of micro-pattern detectors. With some luck, this will be done in collaboration with an Austrian doctoral student. The survey of Penning transfer rates is advancing well, with a Turkish PhD student working on Ar mixtures with iC4H10 and other organic gases. In the context of the ATLAS upgrade study, a number of Micromegas mixtures having been analysed. - Activation: o Presentations were given at the physics validation meeting and at the CMS technical coordination meeting. Following this, CMS is likely to ask Steffen Mueller to continue working on the subject as a doctoral student, in a first stage to update the beam pipe layout. Hopefully, the Fluka group will take care of the water cooling activation calculation. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for February 5th.