Simulation Project Leaders Meeting 30-January-2007 ==== Attendees John Apostolakis (JA), Witek Pokorski (WP), Alberto Ribon (AR), Pere Mato (PM), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: John Harvey (JH). --------------------------------------------------------------------- ==== Minutes - Summer students: definition of projects for summer students must be advanced by end of January. One student is being asked for a project in the Geant4 geometry module. A student will be also asked for GENSER and one for GDML. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Meetings: o At the last meeting on January 24th there have been two presentations from ATLAS, one on a study for very low energy pions in the combined test-beam 2004 setup (1,2,3,5 and 9 GeV pion); good agreement with data especially in using the Geant4 cascade models (both Binary and Bertini) was reported; a discrepancy was observed, which could be explained with the missing application of Birks recombination; a new report on this analysis will follow at the next meeting. The second ATLAS presentation was concerning a study on 20-350 GeV muons in the combined test-beam (part of a PHD thesis under T.Carli supervision), proposing a new calibration technique for hadronic calorimeters, based on the detailed simulation of muons; a precision of few percent seems feasible. The final presentation was by M.Kossov on the recent improvements on inelastic cross-sections parameterisation in CHIPS. - Shower shapes: o Activity in the attempt of trying the INCL-ABLA model for the shower shapes (the INCL-ABLA model is an alternative intra-nuclear cascade model created by the Liege team); preliminary results show no significant visible improvement in the shower shape analysis. o Work is in progress in testing an improved version of the FRITIOF model (in the FTFP physics-list). o Studies of the impact of pions and kaons below 10 GeV in the hadronic shower shapes have started. o A report is being written (currently at 40% stage of the draft). - 3rd simple benchmark: o The benchmark and analysis are completed, samples have been regenerated using a new high-energy model included as optional in Fluka, no relevant differences are observed and the agreement with data both of Geant4 and Fluka is good. Currently in the process to complete the final note draft. o It is also under discussion a proposal for a new 4th benchmark, related to 450 GeV proton on beryllium target, an analysis initially of interest for neutrino studies (SPY, NA56) which provide a precise set of data for testing forward physics in the simulation. - Extension of ATLAS TileCal 2002 analysis with Fluka: o Started weekly meetings with the Pisa group, production is ongoing. Some improvements and fixes done in the smearing and Birks application in the analysis. M.Simonyan from ATLAS is at CERN for two months and will be involved in the study related to the 90 degrees setup, where the technical problems in applying Flugg have not been yet solved. --------------------------------------------------------------------- ==== Generator services - Witek - MCDB: o Ready to use. No much to report. Reduced manpower effort of 50 %. - HepMC: o New contributor L.Sonnenschein from MC-Net; will be at CERN for two years and will contribute to the maintenance and development of HepMC. Lars will also work on the Rivet tool. Immediate tasks will be to test version 2.0.0 of HepMC and verify the root-I/O schema evolution trying to convert HepMC 1.X to 2.X formats. = GENSER: o All generators requested by experiments are now installed in the new structure. The CMT interface is available for all packages and ready to be used by experiments interested. All tests have been moved to the new structure; new tests for Sherpa, Herwig++ and Pythia-8 have also been added; some minor bugs were found and reported to the authors. o The web page site (http://lcgapp.cern.ch/project/simu/generator/) has been redesigned. A list of the available generators is present, as well as links to GENSER instructions and validation page. For the latter, the idea is to make tables of results of the performed tests run against the most popular generators. o GENSER 2.0.0 will be announced next Wednesday, including the validation table of the tests. o Experiments and authors have expressed satisfaction so far. - Manpower: o It was asked to have permanently at least two russian collaborators at CERN for the completion of the new structure. This was fine for the past months, but appears not to be possible for the next month. Most of the work has been anyhow completed. - Meetings: o Next Generator Services meeting on February 7th. --------------------------------------------------------------------- ==== Fluka - Alfredo - No report. --------------------------------------------------------------------- ==== Simulation framework - Witek - GDML: o in the last weeks, the activity being carried out by B.Lloyd has been focussing in debugging reflection of solids for the Root GDML binding. o It is foreseen next the implementation of an installation script for GDML to aid users in the installation process of the library. - Geometry object persistency: o Still an issue in Root to be solved concerning persistency of C-arrays of pointers. --------------------------------------------------------------------- ==== Geant4 - John - Release 8.2 of Geant4: o Released, as scheduled, on December 15th. o Main features included are: coupled navigation in parallel geometries for charged particles, also in EM field; revised elastic scattering: completed new QElastic process for protons and neutrons; corrections to x-section above 100 GeV for protons; revised scaling of interpolated pion cross sections to A^(3/4) ; integration of physics lists to kernel libraries (CMS request). - Development/fixes: o The improved FRITIOF model, created by Vladimir Uzinskiy is planned to be included in the development release of February. o Fixes for energy conservation in String models are under testing. o The Binary cascade extensions on pions are being improved up to 1.5 GeV. - Experiments feedback: o Significant problems have been reported in the test production of 8.2 by ATLAS; when using QGSP that includes the new MSC in 8.2, about 75% of events abort due to stuck tracks. CMS is reporting similar problems in test productions using release 8.1 and QGSP. They report no problems with the more computing performant physics list QGSP_EMV. Some issues have been identified in the interface between MSC and the geometry; a candidate revision in the field has reduced the failure incidence from 70 to 10% (assumptions that the safety is a single-valued field become false in the presense of overlaps, and the fix avoids caching internally its value between tracks). The problem however was reproduced in a simple setup with field and massive geometry overlaps, indicating that the new Geant4 releases have become more sensitive to cases of ill geometry definitions. In addition to the Geant4 kernel improvements which are in progress, we have been in touch with user and recommended a careful debugging of the parts of a geometry setup in which such problems are found. - Meetings: o The 17th Technical Forum meeting takes place on January 31st at CERN (http://indico.cern.ch/conferenceDisplay.py?confId=11643). Among the items under discussion we recall: CPU performance profiling and G4 version comparisons; reports on shower shapes, new requirements/requests and priorities for 2007. A request has been advanced from neutrino experiments to improve low energy neutron capture. o The 12th Geant4 Collaboration Workshop will take place the 13-19 September in Hebden Bridge (UK) near Manchester, hosted by the Manchester University. o The Geant4 review will take place from 16th to 20th April 2007, hosted at CERN. It is expected to focus on physics and computing performance and on usability. Review panel members will begin polling users in the near future on open issues, experiences. --------------------------------------------------------------------- ==== Garfield - Rob - Garfield activities: o Mainly activity on Micromegas detectors. o SRIM interface stable and in use, further improvements expected. o Numerous small corrections and additions, including a 1500-fold increase in the processing speed of macros. - CMS activation studies with Fluka: o M.Brugger has converted M.Huhtinen's file to the Fluka 2006 format. o The Fluka group is working (or is expected soon to work on) producing a source routine. Once this is complete, a test run is scheduled to compare with the INB report. One candidate test case is the neutron flux in the cavern. o Currently going through the Fluka setup for adjusting to the new geometry, in particular for what concerns the beam pipe, shielding and silicon detectors. o Meeting with the technical coordination of CMS foreseen to discuss the areas to which the Fluka file is to be applied, and which will determine the weighting and scoring settings to use. - NA60: o Thesis on K0s, lambda and anti-lambda production to be defended in April. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for March 13th.