Simulation Project Leaders Meeting 27-June-2006 ==== Attendees John Apostolakis (JA), Paolo Bartalini (PB), Alfredo Ferrari (AF), Witek Pokorski (WP), Alberto Ribon (AR), Pere Mato (PM), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: John Harvey (JH) --------------------------------------------------------------------- ==== Minutes - Preparation for the 2nd quarter 2006 quarterly report will start in the next days. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Meetings: o There was a Physics Validation meeting held on June 21st where three presentations were made, one by Anthony Waugh on the ATLAS fast-simulation, where in collaboration with CMS they have implemented a framework based on G4-gflash: the results look promising and they're now trying to use it with real events (top, higgs, ...). There were also two talks on developments in Geant4 for EM physics and hadronics by Vladimir Ivantchenko and Gunter Folger respectively. - CMS HCAL analysis: it is expected an update of the analysis using version 7.1.p01 of Geant4, likely to happen at the next Physics Validation meeting. - ATLAS combined 2004: first results of the analysis are expected soon, likely in September. - ATLAS TileCal: o Completed the technical work for the geometry/digitizations setup, the work is now focussed on the physics issues. There has been considerable progress after the various meeting with the MC experts: the discrepancy which was observed for muons has been reduced in the tests made with Geant4 8.0.p01 (where more energy response for muons are expected) and after understanding the Birks quenching issue in Fluka (which instead has reduced the energy response). There is now better agreement between the two simulation engines. By using a simple block of matter and/or a simple scintillator there is now very good agreement (below 1% for muons and 1-2% for electrons). The full test-beam setup still shows discrepancies (4% for muons at 10GeV, and up to 10% for electrons) requiring understanding. It seems clear the discrepancy is originated mainly from the EM physics: results seem consistent in the sense that shooting protons where there is less EM contribution also the discrepancy is reduced. A possibility is that the multiple-scattering models are somehow sensitive to the geometry topology. An attempt will be made to change the shooting direction of the particles to be orthogonal to the scintillators, and see if the discrepancy in this case gets reduced. Other attempts which will be made are to run in "single-scattering" mode or switching off completely the multiple-scattering, or limiting the step to a reasonably small step. Another attempt is to shoot at different angles and study the effects. --------------------------------------------------------------------- ==== Generator services - Paolo - Events: o Hera-LHC Workshop, held on June 6-9th: Presentations of M.Kirsanov (LCG Generator, Physics Validation) and M.Bargiotti (Pythia validation). Emerged new important initiatives for MC Tuning / Validation for the LHC. A new joint CEDAR/LCG Generator software initiatives concerning cross-compatibility of the MC packages (GENSER usage in RunMC, HzTool, HzSteer, Rivet; request to support the same packages in LCG; common HepML solution for data format; etc...). o MC4LHC Workshop to be held on July 17-21st: The LCG Generator meetings will move in the w/s slots. o PB convener for both workshops above. o Seymour is currently advertising a new EU network for the development of C++ MCs (Pythia 8, Herwig++, Sherpa, Ariadne++) and for the training of experimentalists. We need to decide how LCG Generator should interact with the new network. AR thinks that the activity will start from next year, and, as interface to the experiments, we should be particularly active on the training part. o Given the large number of new (and quickly growing) initiatives, which in particular take place in summer, the review of the LCG Generator milestones has been postponed to August. Further information will be available in the quarterly report. - Shifts: o GENSER Integration: M.Kirsanov (summer). o MCDB development: S.Belov (soon replaced by L.Doudko). - GENSER: o R.Yaari already tested g95 for 2/3 of the GENSER packages. Good behaviour of the most popular generators is confirmed (thanks also to the previous activity of the GENSER team). Problems for few generators are currently being reported to the authors. o A new release (1_4_0) is in preparation. Pre-release available, with included: Changes in GENSER configuration: a) LCG_45. New versions of installed generators: a) PYTHIA 6_327, 6_400, 6_402, 6_403 b) PYTHIA 8 052, 053 c) ISAJET 7_74. Bug that prevented usage of LHAPDF fixed. d) FeynHiggs 2_3_2. e) PYQUEN 1_1, HYDJET 1_1. f) Herwig++ 2_0beta2 bugfix release. g) LHAPDF 5_1, 5_2_1 (5_2 is in CVS but not compiled), 5_2_1a (modified a02m) h) ALPGEN 2_0_6 i) CASCADE 1_2_10 j) TESTS 1_02 - 13 packages are now tested. Bug fixes: a) Bug found in the PYTHIA interface to FeynHiggs. b) Bug in ISAJET 7_74 fixed in order to enable its usage with the LHAPDF library. c) CASCADE 1_2_09 in 1_3_0 did not work: missing -fno-automatic in BuildFile. New in examples: a) Authors' examples are adopted for Pythia8. b) Examples of LHApdf and HepMC usage are added for pythia and herwig. - MCDB, HEPML: o By now LCG MCDB is properly operating, well advertised and ready to be used. The system is accessible via web (http://mcdb.cern.ch). All MC event files accommodated in LCG MCDB are available over GridFTP or directly from CASTOR. The web-interface was tested and demonstrated to work properly with the most popular browsers as Mozilla, Firefox, Internet Explorer and Opera. o Entries in the data base are growing and documentation is currently being improved. o Hector Naves (CMS, LCG) and the MCDB developers are currently developing an API to access the data base from the generator level simulation framework. This will be the first example of integration of the major packages supported by LCG-Generator (i.e. GENSER, MCDB and the generator level simulation framework). o HEPML: unified XML format of event samples and its documentation plus conception and goals is under discussion and described in https://twiki.cern.ch/twiki/bin/view/Main/HepML Ongoing collaboration with the CEDAR project. - Generator level simulation framework: o The software is now in a "usable" status: it is possible to produce events with Pythia6. It is however still limited to the CMS environment (tutorials available). Although the development of the generator framework for CMS has to be considered of high priority (for GENSER integration, for example), it was stressed the importance of performing the last small effort consisting in removing the unnecessary dependencies on the CMS software, making available the framework also to the rest of the LHC community (for example to run GENSER tests and examples). - CMS Tutorials: o Pythia6Interface Tutorial: (under construction) https://twiki.cern.ch/twiki/bin/view/CMS/Pythia6Interface o CMS CPT Tutorial: (Howto produce gene+simu... covers GeneratorInterface) https://uimon.cern.ch/twiki/bin/view/CMS/May06CPTweekTutorials - MC Validation: o Marianne Bargiotti is currently concentrating on the validation of NRQCD models in Pythia 6.3. Impact on quarkonia physics (mosly LHCb) and on the muon trigger thresholds (all LHC experiments). Tests have been performed with Pythia 6.324. Predicted J/psi and Y production x-section and PT spectra at Tevatron have been compared to the CDF data. Agreement has been achieved following a cut off scheme which is inspired by the treatment of other QCD processes. These important new results are currently being reported in major Workshops (Hera/LHC at CERN) and at QWG Workshop at Brookhaven. After an additional effort dedicated to extract the LHC predictions, Marianne will move to the general Pythia 8 validation. --------------------------------------------------------------------- ==== Fluka - Alfredo - The intermediate energy generator PEANUT has been extended to high energy. It is not sure the new generator will be included in the next version of Fluka planned for mid July, but it may be included as an option. On most data the difference will be negligible. It was proposed to have a presentation at a physics-validation meeting. - Wish to cleanup cross-sections from 4-5 GeV on, to better distinguish between quasi-elastic and particle production contributions at high energy; since there exists a well-known theoretical framework based on the Glauber model, the idea is to extend this to extract precise cross section data. This will allow to reduce the uncertainty and also a smooth transition from the low-energy regime, and can be important for cosmic-rays physics and also (marginally) calorimetry. If any good data will come from HARP in the 5-10 GeV regime would help a lot in better understanding the onset of the Glauber cascade (multiple primary collision regime). - Working on heavy ions at low-energy. There will be a new model for very-low energy and one based on QMD (Quantum Molecular Dynamics). All this is aimed for usage mostly in hadron therapy, where nice results for protons are coming from the Massachusetts General Hospital and other projects in Italy. The work of comparing with data for therapy is being done mainly by people based in GSI and Heidelberg. - Two new staffs starting on July 1st. --------------------------------------------------------------------- ==== Simulation framework - Witek - GDML: o Following the CHEP06 conference, a paper has been submitted paper to TNS and accepted for publication. o The GDML->Root converter will be also part of Root, and the geometry modeler in Root has been extended to support GDML and import GDML geometries. o A summer student Daniele Kruse is working on a converter for STEP-Viewer intermediate files to GDML. The work was inspired by the associated ESA project in Geant4. Files from CAD systems can now be converted to GDML through the Geant4 tessellated solids. The tool will be part of the GDML processor. o A new technical student is expected for July 1st. - Python: o A presentation for the Python interface to Geant4 is planned for the Europython conference at CERN in July. --------------------------------------------------------------------- ==== Geant4 - John - Release 8.1 of Geant4 will be out on June 30th as planned. Among the features included: o Fixes in physics models for energy conservation due to infrequent double-counting of particles in an interface and issues in treatment of anti-nucleons. o Additional issues in the string models addressed. o Revision of the cross-sections and models for nucleon elastic scattering: new detailed cross-sections and models for proton, deuterium and Helium targets, and improved treatment of low energy projectiles for all target nuclei. o Reintroduced option for sampling delta electrons production below threshold when these are generated near volume boundaries - which is called the 'sub-cutoff' option. o Physics-lists: incorporated revision of elastic and stopping processes improving treatment of energy deposition in plastic scintillators, using theoretical models for important interactions. Energy deposit of recoil is now modelled, below a threshold. The models utilised for stopping particles are based on CHIPS, have been shown to agree well with data. They are now available in the majority of physics-lists (QGSP and FTF based lists, but not LHEP based). Typically an extra cost of 5-10% increase in CPU speed results. Kaons also now use the Bertini cascade in QGSP_BERT and derived physics lists. In QGSP_BIC binary cascade is now activated for light ions. o Physics lists using the older Geant4 version (typically 7.x) of the multiple scattering process are available for applications with large CPU requirements which are potentially less sensitive to the details of energy deposition. Choice of QGSP_EMV and the experimental QGSP_EMX which generate sample secondaries below the threshold near boundaries. o EM physics: new option for modeling the struggling and new tuned multiple-scattering. Regression and validation tests on simplified calorimeter setups clearly show the effects of the improvement in the elastic-scattering. - Release 8.1 will be covered by an EGEE-derived license 'Geant4 Software License (hybrid between LGPL and GPL), available at: http://cern.ch/geant4/license - Future meetings: o Physics Validation Workshop: July 17th-19th at CERN. o Meeting for Hadronic showers, Fermilab, September 6-8th. o Geant4 Collaboration Workshop in Lisbon, October 6-10th. o Geant4 Space Users Workshop at the NASA Jet Propulsion Laboratory (JPL) in Pasadena, November, 6-10th. --------------------------------------------------------------------- ==== Garfield - Rob - Development in Garfield: o Corrections for Argon at high electron energy o Reproducibility of random number sequences ensured; Heed random number generator replaced to ensure independence of long runs o Modifications with a view for compiling on Windows. o Various finite element interfaces. - Geant4: o LHEP: rapidity pecularities have in fact been known for 20 years, and have been reported by the author of Gheisha in 1985; anomalies in many other parameters have been reported, in particular by the Fluka team in 1996 ... Working on an overview of more appropriate generators. - ATLAS: o Activities are currently related to contaminations by various gases such as methane, ethane, nitroge, water vapour ... + the alignment with a magnetic field. Their calculations are highly CPU intensive and are done using the BOINC facility. - HARP: o Currently working on the corrections related to the water contamination of the TPC gas. Earlier, they worked on space charge effects and cross talk. There are other corrections to be applied like a potential offset between inner and outer field cage and wandering of the beam. Regrettably, the official part of HARP doesn't seem to be much interested in such corrections - at any rate, it was never asked for information on such topics. - NA60: o Analysis of re-processed data in progress (better alignments). o Anti-lambda baryon seen sufficiently well to measure lambda/anti-lambda ratios as function of centrality and transverse momentum. o Temperature of lambda and anti-lambda measured, acceptance correction still in progress. - Others: o Future involvement in Cloud, which needs field calculations for the field cage. --------------------------------------------------------------------- ==== AOB Next meeting scheduled for September 12th.