Simulation Project Leaders Meeting 4-November-2008 ==== Attendees John Apostolakis (JA), Alberto Ribon (AR), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: Alfredo Ferrari (AF), Pere Mato (PM). --------------------------------------------------------------------- ==== Minutes - A technical student has been assigned for work on the Geant4 geometry and maintenance/development on Flugg. The stage is supposed to start from February 2009. --------------------------------------------------------------------- ==== Simulation physics validation & framework - Alberto - Meetings: o In the meeting held on September 24th, there was a presentation by M.Simonyan on the ATLAS TileCal test-beam analysis at 90 degrees, showing a comparison between data and 9.1/9.2-beta releases of Geant4. The results of the analysis show QGSP-BERT performing 3-4% higher than the data in energy response and within roughly 10% in energy resolution; other lists (FTF_BIC and FTFP_BERT) show higher response (6-10%) for both pion and proton. For the lateral shower-shape of pions, reasonable agreement with all tested lists; for protons the FTF-based lists are best performing. For the longitudinal profile, all tested lists are within 10% agreement with pion data, whereas for protons QGSP-BERT is shorter and the FTF-based lists are in good agreement. It was also reported that 9.2-beta produces narrower showers than 9.1, in worse agreement with data, whereas for the longitudinal shower-shapes, QGSP_BERT is unchanged and the FTT-based lists produce shorter showers in better agreement with data. It also reported about the non-smooth energy response in beam energy scan, in particular, around 9-10 GeV for QGSP_BERT, and around 4-5 GeV for FTF-based lists, with the latter showing less discontinuities. o The next meeting will taken on December 3rd. --------------------------------------------------------------------- ==== Generator services - Alberto - Meetings: o The last Generator Services meeting was held on October 1st and was dedicated to MCTester, the tools for allowing the comparison of the decays of a specified particle between two simulation packages. The tool was originally designed for comparing all decays of tau and later extended also for b-decays. The idea is to study the invariant mass of all decay products and use it in GENSER for testing packages with non-trivial b,tau-decays. o Trying to organise a special meeting on EvtGen, for defining plans on how to update the package. The meeting will be held most likely in January. - GENSER: o Several new versions of generators introduced as usual. o The next GENSER meeting will be held on November 5th. - HepMC: o AR is preparing a proposal for better handling urgent requests for new features to the package. The proposal will be discussed at the project planning meeting at the end of November. --------------------------------------------------------------------- ==== Fluka - Alfredo - No report. --------------------------------------------------------------------- ==== Geant4 - John - Releases: o 9.1.p03 was released on September 19th; the patch includes the remaining fix to deal with problems reported by HARP/CDP (angular distribution within Pt range in parameterised models). o The testing phase for release 9.2, due in December has started this week. - Development highlights: o The new release will include different variants of multiple-scattering models: the frozen Urban model as in 9.1, a re-tuned version of the Urban model, a new model based on Wentzel and a new one from Omrane Kadri (NSNST, Tunis). o The particles mass table will be revised in release 9.2, consolidating it and making sure its usage is made in a consistent way by all involved physics processes. Particle masses and data will be upgraded to PDG-2008. o Release 9.2 will also include improvements of the Bremsstrahlung process for the LPM effect for energies above 1 GeV, comparison with SLAC measurements. o The low-energy photon-gamma processes are being migrated to the EM standard model design and will be included in release 9.2. - Feedback from experiments: o ATLAS is now testing 9.1.p03 as candidate production version. No crashes so far in their tests, it has been very stable. A problem was uncovered in the simulation when Cherenkov photons are tracked, which led to tracks being suspended. The problem was traced to multiple-scattering and part of its state by Z.Marshall (ATLAS). A fix has been provided and is being tested on top of 9.1.p03. o ATLAS also reports that their fast simulation using pre-computed showers has now reached the end of their validation and performs rather well. Tests showed it to be 10 times faster for a single e- at 62 GeV, and about 3 times for a Z to e+ e- event. o CMS is now regularly testing development releases of Geant4, the last one (9.1.ref08) including also some CPU improvements contributed by CMS and the FNAL group in the Cascade hadronic models. - Meetings: o IEEE/NSS in Dresden: large number of reports on using Geant4 in several fields; there have been several reports from space missions, such as the NASA/ESA proposed EXO mission developing a new x-ray telescope with 3 m2 area, using Geant4 as main simulator engine. The number of volumes used to describe it of order millions. There have been interactions with the GATE developers who reported their progress with Geant4; in particular a paper on Woodcock tracking, providing significant speedup for photon tracking in voxel geometries when used in combination with the regular navigation technique available in Geant4; several presentations also on radiotherapy applications. o Geant4 workshop in Kobe: 70 people attended. A few items to note: an overview on the use of Geant4 in LHC was given by S.Banerjee; revision of the FTF model and new multiple-scattering models; presentations on ILC-Japan using G4, proton and ion therapy. The hadronic group is focusing effort on improving the energy response in the transition regions between the different hadronic models, to address the feedback from CMS and ATLAS, aiming to meet their requirements. Next Geant4 Workshop will be on October 2009 in Catania (Italy). --------------------------------------------------------------------- ==== Garfield - Rob - Meetings: o RD51 in Paris, by now 50 institutes and about 150 people; project is taking shape and should be approved on December 5th. - Gas detector calculation with Geant4: o The summer student work has been taken over by a PhD student from the University of Aveiro, at CERN this week. The idea is to make the interface compatible with Geant4 conventions before generalising to other geometric configurations. This student will progressively move to work on the simulation of scintillation in Xe for medical imaging. o An EU-funded PhD student from Greece will start work in the same area from December. She will continue to work on neutron interactions in gases. - Avalanche modeling: o The PhD student has almost completed his study, the outcome of which is to be published, and he will now move on the simulation of silicon detectors. - Penning transfers: o Approaching a publication based on the work from the PhD student in Turkey. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for December 16th.