Simulation Project Leaders Meeting 29-September-2009 ==== Attendees John Apostolakis (JA), Alberto Ribon (AR), Paola Sala (PS), Pere Mato (PM) Gabriele Cosmo (GC). Excused: Rob Veenhof (RV). --------------------------------------------------------------------- ==== Minutes - Andrea Dotti has started the fellowship in Geant4 as of July 1st, to work on physics validation with A.Ribon. --------------------------------------------------------------------- ==== Simulation physics validation & framework - Alberto - Meetings: o Last physics validation meeting was held on June 10th: the first presentation was about CALICE by Erika Garutti from Desy, it has been mainly an introductory talk about the facility being set up, where several analyses of the test-beams with a very granular calorimeter will provide interesting observables; the next presentation was by S.Banerjee on the validation studies in CMS for the new physics-lists (QGSP_FTFP_BERT and QGSC_CHIPS) provided in Geant4, results look interesting and promising although still in experimental stage, and significantly reduce the use of the LEP parameterised model. o Next physics validation meeting will be on October 7th. o AR attended to the CALICE collaboration meeting in Lyon on September 16-18th; a talk from CALICE will be organized at the forthcoming Geant4 Collaboration workshop in Catania. - Validation studies: o Extensions and improvements have been introduced recently to CHIPS, by extending the model applicability to all particles, energies and targets. o Improvements and extensions have also been applied to Fritiof, where it has been added the possibility to couple Fritiof with a Reggeon Cascade; this extension is available as an option. o Tests with thin-target data have been performed by S.Banerjee and V.Uzhinskiy for validation of models in the range of 1-20 GeV. o Continued studies in Geant4 at model-level with simplified calorimeter quantities. --------------------------------------------------------------------- ==== Generator services - Alberto - Meetings: o The meeting on July 8th July was dedicated to HepMC. Release 2.05 has been made available on June 5th; it finally clarifies the meaning and definition of the status-code, includes improvement in the streaming I/O operators and addition of cross-section operators. The new release 2.06 is expected in early 2010. o In the last meeting of September 23rd, the new releases of Herwig++ 2.4.0 and Pythia8 130 were presented. It was reported that experiments (mainly ATLAS and CMS) are now starting using them and providing feedback. o Next meeting is planned in November. - MCDB: o Is now in use in CMS. Some problems are reported concerning CASTOR, but not at the level of preventing the usage of MCDB. - GENSER: o The migration to SLC5 has been now completed for all generators. o Almost all generators have migrated their installation to autotools. The new build system, together with a script for installing on AFS and choose the configuration will be part of GENSER-3. Feedback from the experiments will be required, especially for those building the libraries from the sources and distributing them. --------------------------------------------------------------------- ==== Fluka - Paola - FLUGG: o Implemented possibility to handle geometry volumes through names, and material assignments. o Currently working together with the ATLAS radiation working-group for radiation background studies; having a prototype geometry which is now working. - Fluka: o The porting to gfortran is almost completed; this will allow to build and run on 64-bits systems, and port the code to MacOSX. o Improvements to ionisation energy losses are under development, related to corrections of Z to the 3rd and 4th order; the effective charge of ions at low energy has been refitted. These improvements will be available in the next release of Fluka. o Most of the development team now busy with LHC tasks. o A technical student is now at CERN for web site and the GUI interface related work. o Growing user community: 3000 users registered from the official web-site; quite a lot of support load through the mailing list. o A course will take place in Mumbai at end of October. - Neutrino facility at Gran Sasso: the measurements made so far are consistent with the simulation. --------------------------------------------------------------------- ==== Geant4 - John - Releases: o A new patch release 9.2.p02 was announced in August. It includes various fixes, among which: a fix for running with parallel geometries, of interest for caverns studies; an enhancement in GDML for the treatment of nested Boolean setups, of interest for ATLAS. - Developments: o Some of the new physics lists provided with 9.3-beta were also adapted for use with 9.2 as additional physics configurations (only physics lists which use no new modeling, but change transition between models). o Improvements in recent development releases include tuning of the FTF model's binary channels, and correction of the interface from FTF to the Binary Cascade. o Improvements to CHIPS, where improved neutron cross-sections were developed and are used in the QGSC_CHIPS physics list. o First comparisons of simple setups at 1-20 GeV between Geant4 models including Bertini, LEP, FTF, QGS. Preliminary comparisons with results of Dubna cascade and Fluka. Large deviations noted for all particle yields of LEP parameterised models in case of incident protons (action to move to lists without LEP). Indications of excess proton yield above 4 GeV by Bertini and differences in pi-zero yield are being investigated. o Improvements underway in the memory use of Bertini Cascade model, in response to feedback from ATLAS. - Feedback from experiments: o Trials of 9.2.p02 by experiments are ongoing, with initial positive results. o Reported from ATLAS: a problem with their latest test runs 0.1% to 1% jobs get a track which is stuck (charged track in a magnetic field at a volume interface). A first fix has been identified, by raising the threshold distance for recognising when a track is not making progress. This identifies steps which are stuck, and enables the existing recovery action to occur. If this fails, a stuck track is abandoned. More investigation is needed, also concerning the level of accuracy requested, i.e. the value of the geometrical tolerance currently adopted. - Meetings: o At the Geant4 Collaboration and Users Workshop scheduled for October in Catania there are more than 30 users registered so far. o The training course at CERN has been postponed to January/February 2010. Other courses are taking place in November at SLAC, Gran Sasso and in December at KEK. --------------------------------------------------------------------- ==== Garfield - Rob (report offline) - Semiconductors: refinements in the ionisation model; integration with the boundary element method (both 2d and 3d) in progress; various electron and hole transport techniques being studied - Gas: ATLAS upgrade note being completed, 2nd Penning study under way and paper largely written, to be followed by an avalanche fluctuation paper; measurements in Saclay and NIKHEF to be scheduled - Fields: two new releases of the nearly exact boundary element method. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for November 10th.