Simulation Project Leaders Meeting 6-November-2007 ==== Attendees John Apostolakis (JA), Witek Pokorski (WP), Alberto Ribon (AR), Paola Sala (PS), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: Alfredo Ferrari (AF), Pere Mato (PM). --------------------------------------------------------------------- ==== Minutes - The next LHCC Comprehensive review is scheduled for November 19th, slides must be read one week before. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Meetings: o In the Physics Validation of October 17th meeting it was presented an update from the ATLAS HEC group where Geant4 9.0 was used and compared against previous versions. Results show an improvement for the QGSP Physics List, which is now closer to the data than in version 8.2, whereas QGSP_BERT is getting farther from the data in e/pi and energy resolution. There was also a presentation on the ATLAS TileCal at 90 degrees analysis using both Geant4 8.3 and 9.0 versions: QGSP_BERT Physics List looks fine, in energy resolution, e/pi and shower shapes, except when applied for protons; it shows anyhow an improvement compared to the older versions; the FTFP Physics List was tested as well, here showing good results for protons, but not as good for pions. It is an important issue under investigation the fact that QGSP gives the best energy resolution and e/pi for ATLAS HEC but not for ATLAS TileCal. o The next Physics Validation meeting is scheduled for November 21st; there will be a presentation from RV on the recently completed CMS activation studies and a presentation from Alex Howard on the TARC neutron benchmark. - Simple benchmarks: o For the 4th benchmark (diffraction in proton-nucleus collisions based on the Helios data) AR has met with Geant4 and Fluka experts; it turned out that the analysis in the published paper included the acceptance corrections to the data which were based on an older simulation code (Ranft 1988) that did not include resonances in the low diffractive mass region (these resonances are instead simulated in Fluka). So the data published cannot be used directly for the benchmark analysis; AF is trying to contact the authors to get the raw data before the correction; alternatively, one could try to un-correct the data using the acceptance curves described in the paper. It would be anyhow useful to complement this benchmark with p-p or p-pbar diffractive data, where the systematic uncertainties are much smaller. - Extension of ATLAS TileCal 2002 analysis with Fluka: o WP and MG are now generating the last samples, 50 GeV currently under investigation; the analysis should be completed by December in order to start with the analogous study with the ATLAS HEC setup. --------------------------------------------------------------------- ==== Generator services - Witek - Manpower: o Two new people joining the team, currently under training. o Concern is expressed about the low efficiency and consequent slow progress, due too much frequent turn-around of people involved and expertise. - Meetings: o In the Generator Services monthly meeting last week there was the announcement of Pythia 8.1 by T.Sjostrand, experiments are encouraged to use this ew version for production. There was also a presentation on MCDB by L.Doudko encouraging the use of the tool in CMS. A presentation by L.Sonnenschein was also reporting about progress in the Root-I/O interface for HepMC; few problems have been fixed in both sides, and necessary changes have been introduced in HepMC already. o The collaboration with the Rivet developers has been discussed at the Hera meeting last week. The goal is to have Rivet-based analysis for validation soon available in GENSER. o The next meeting will be on November 28th, it is scheduled a presentation by P.Richardson on the new Herwig++ production release. o The next Project Planning meeting will be on November 30th. - GENSER: o New versions of generators, among which Pythia 8.1 presented at last meeting (8.1). o Added few more tests for Hydjet and Pyquen generators. --------------------------------------------------------------------- ==== Fluka - Paola - Releases: o Currently working on a new release; the major new feature will be a new neutron library, where more groups (132 instead of 72) have been included; the thermal group has been subdivided in more than one. The new release will also include better spectra of emitted protons, improvements on the possibility to work in a more analogic way and more levels of self shielding. o Some minor bug fixes are included on the current version distributed on the web. o A new release of the Flair interface is available on the web. - Meetings: o A Fluka course has been taken in Legnaro (Italy) last October; it saw the participation of 30 students, unfortunately only 12 participated to the final questionnaire, resulting to be ayhow very useful. o There has been a proposal from INFN and CERN/PH to organise a course for beginners/intermediate students at CERN in Spring. It would be useful to know from the LHC experiments who are interested in participating. The course could be given in February or March. o Regular Fluka users meetings at taken at CERN for advanced users, every 2-3 weeks on Thursdays, interested people should contact PS so that they can be included in the distribution mailing list; the meetings in general consist of a lecture on some features of Fluka and presentations from users. --------------------------------------------------------------------- ==== Simulation framework - Witek - Flugg: o Currently being used for the TileCal analysis; 50K events generated so far with no real problems. It is planned to be used for the next ATLAS HEC validation study as well. o Among the future improvements, it would be nice to extend it in order to explit the new Fluka feature for using region names. - GDML: o The technical student Z.Torzsok is concentrated in working on the new GDML parser to be integrated in Geant4. The goal is to have it included in the new December release 9.1 of Geant4. --------------------------------------------------------------------- ==== Geant4 - John - Releases: o The new Geant4 release 9.1 is scheduled for December 14th. It will see the inclusion of the first implementation of the Liege Cascade model, applicable for 200MeV-3GeV energy range on carbon/uranium targets; it will be a beta release eventually not yet optimised to achieve the physics performance of the original model. The new release will also see the inclusion of a specialised navigator for voxel phantoms geometries; the new navigator will be fast and require little memory, it will also provide the possibility to merge regions with same materials for further optimisation; a gain in speed of a factor varying from 2 to 5 applied to phantoms with a reasonable number of materials has been measured compared to the traditional optimised navigation. Physics list excluding Geisha models for protons, neutron and pions will be also provided; tests are undergoing. - Meetings: o Geant4 Oversight Board on November 6th. o Public Technical Forum at CERN on November 7th: http://indico.cern.ch/conferenceDisplay.py?confId=22985 o The next Space Users Workshop will be in Tokyo (Japan) in February: http://www.astro.isas.jaxa.jp/g4space5/ . --------------------------------------------------------------------- ==== Garfield - Rob - Garfield: o RD51 seems to be gathering speed. A draft programme for the upgrade of gas-based detector simulation has been submitted. Two lines of work can be distinguished: refinement of algorithms and a new user interface. o A new version of Heed has become available and needs to be interfaced. o An alternative program to simulate ionisation processes, exclusively based on experimental data, is being written by Steve Biagi. o Weighting functions for RPC signal induction have been computed and need to be implemented. o Penning transfers in avalanches need to be made available to users. o Ongoing reduction in size of the detectors calls for yet a more detailed description of the diffusion process since the characteristic dimensions of some detector components is now sometimes smaller than the mean free path of electrons. o On the user interface side, Root seems to be the most popular option, both for development and for users. An important consideration is to do this in such a manner that an interface with Geant4 can be developed too. It looks plausible that one or more graduate students will take part in this activity. o Ongoing work concerns an interface with Ansys, and for the first time elements with curved sides, and intermediate level diffusion processes, which was part of the master's thesis of Gabriele Croci who will continue to work on this. - Activation: o The CMS description for use with Fluka that has been produced earlier, mostly by Markus Brugger, has over the last few months been used by CMS. At first, this concerned a careful comparisons of the track length spectra for neutrons, charged particles and photons (Steffen Mueller). The next step will be a study of the Beam Conditions Monitor. Further, the activation (including isotope composition) of the cavern air as computed by Fluka has been compared with that published in CERN/TIS-RP/96-06 (Stefan Roesler). Only minor discrepancies have been found thus far. The comparisons are subject of a talk in SFT on Nov 21st and CMS on Nov 22nd. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for December 11th.