Simulation Project Leaders Meeting 28-March-2006 ==== Attendees John Apostolakis (JA), Mikhail Kirsanov (MK), Witek Pokorski (WP), Alberto Ribon (AR), Pere Mato (PM), Rob Veenhof (RV), Gabriele Cosmo (GC). Excused: Paolo Bartalini (PB), Alfredo Ferrari (AF) --------------------------------------------------------------------- ==== Minutes - Quarterly reports: updates on milestones should be provided by April 5th. - Internal Application Area review is scheduled for September 18-20th followed by the LHCC Comprehensive review on September 25-26th. --------------------------------------------------------------------- ==== Simulation physics validation - Alberto - Combined Tile Calorimeter 2002: a lot of progress achieved in the past weeks. Preliminary results have been presented in internal meetings and first comparisons are under analysis in order to understand the physics; issues both in Geant4 and Fluka need to be discussed with the MC experts, particularly concerning the treatment of neutrons which seem to have a relevant impact in the Fluka simulation. Version used were Geant4-7.1.p01 and Fluka-2005.6. The Birk's law is used in the analysis; investigations are going on to understand its impact on the physics observables. A complete understanding of the shower shapes is required, as well as testing a comprehensive set of physics lists in Geant4. Final results are expected for June. - For the next Simulation Validation meeting, scheduled for April 5th, it is foreseen an update for the CMS HCAL analysis and ATLAS combined EMEC+HEC analysis. - No progress on the 3rd simple benchmark yet. --------------------------------------------------------------------- ==== Generator services - Mikhail - GENSER + Release 1_3_0 was announced on March 21st, although the lack of integrators in the period of January was causing some problems. It has been mainly the result of the work performed in December. The new release was presented at the last LCG Generators meeting. Main features included are the introduction of the new Pythia8 and Herwig++ C++ generators with ThePEG, and the introduction of two heavy-ions MC generators. The interface of Pythia8 and HepMC is now completed; some problems were found in Pythia8 and reported to T.Sjostrand who be back at CERN as from April 2nd. + A problem with "light" (with symlinks) GENSER release 1_2_1 at one remote ATLAS site was reported, although the tarball with the links resolved exists; problem to be understood. Usage of release 1_3_0 is recommended. + A discussion was taken at the last LCG Generators meeting concerning the evolution of the standalone HepMC package in the usage of vectors. No agreement is reached for the various proposals yet. One incompatibility of ThePEG with the standalone HepMC exists; it is however limited to one method (GenParticle::setGeneratedMass in GenEventConverter.cc), which probably can be commented out. For the moment this file is not put into the GENSER library ThePEG. + Current shifter is M.Kirsanov who will be replaced from beginning of April. - MCDB + The tool is now in production. There's the need for properly advertising it and collect users. A presentation with demo will be organised for one of the next AA meetings. + The tool is mainly of interest for the CMS and ATLAS experiments. The migration of CMS from CMS DB to MCBD is planned to June. - Meetings + The next meeting will be focused on bug-tracing with Savannah; a clear definition of the project is required in the attempt of helping reducing the amount of direct questions/e-mails to authors. --------------------------------------------------------------------- ==== Fluka - Alfredo No reports. --------------------------------------------------------------------- ==== Simulation framework - Witek - TileCal 2002 framework + The technical part is now completed; the translation of the geometry is now fully automated through GDML and thanks to the Flugg extensions, now minimal work is required from the user side for setting up sensitive detectors and mimic them through the MGDRAW routine in Fluka. Documentation has been added from the Simulation Framework web page: http://lcgapp.cern.ch/project/simu/framework + Some modified versions of routines for implementation of the Birk's law are expected soon from the Fluka authors. - MCTruth handling + A meeting with representatives from CMS was taken, there's interest in reusing solutions developed in this context. LHCb expressed also interest in it. WP implemented an example for Geant4, interfaced with HepMC and showing how to get MCTruth from Geant4 and store the output through HepMC. The example could be introduced as one of the official Geant4 extended examples. - Python interface to Geant4 + The solution provided in Geant4 and implemented by Koichi Murakami (KEK) has been evaluated; there's need to get in touch with him in order to have feedback about the adoption of Reflex in place of Boost. - GDML + A Summer student and a technical student assigned starting this summer. + A new release (2.5.0) was announced on March 9th; among the main features, is the support for Geant4 reflections, divisions and replicas. + A test of splitting GDML files in smaller files (using 'entity' statement) was made and demonstrated to work; this however cannot be the final solution for the modularisation of GDML files, being the requirement of having each GDML file as a self-consistent description which can be loaded independently. + A meeting in Brussels with ESA will be taken on April 6th to discuss about extensions to GDML in the context of binding to CAD systems. + Some effort is being put for porting of GDML on MacOSX; there're currently problems when attempting installation of shared libraries. --------------------------------------------------------------------- ==== Geant4 - John - A public patch to release 8.0 was released on February 10th, including mainly fixes in hadronics, geometry and transportation. - Hadronic physics + Main area of activity are currently: o Improvements to the elastic scattering process, and revision of cross-sections for the various models: a progress report in an internal meeting in February was made about Hydrogen; more to come on Deuterium and Helium. o Understanding the discrepancies in shower shapes: comparisons are being carried out between the available physics lists, investigating energy deposition and neutron transport. This is linked also to the TileCal study. o Simulation of the TARC experiment as a benchmark of neutron simulation. + A problem reported one year ago (and fixed in 8.0) was discovered recently in CMS productions using release 6.2.p02. Spurious energy gammas (>20 GeV) were found (at a rate of ~20 per 100K min bias events) coming from neutron capture. The problem was due to a unit error in the boost. Further investigation showed that neutrons captured in light material created higher energy gammas when the photons were emitted colinearly with the boost. The effect produced large energies (>10 GeV) only for the lightest target nuclei (in particular H, d, He). To cover experiment needs the following was provided: o Back-ported fixed versions of the Capture process for Geant4 7.x and 6.x. o Instrumented revisions of the buggy version to report gammas produced, in order to identify the frequency and properties of spurious ones. o A special build (7.1.p01a) based on release 7.1.p01 installed on AFS at CERN including the fix, for use in new productions by ATLAS and LHCb. - Meetings: + A public Technical Forum is held on March 30th. The the main item of discussion are the planned developments for 2006 (recently published from the Geant4 web site). Topics of interest from user responses include precision neutron simulation, polarised EM processes and CAD input (CMS). + The first OB (Oversight Board) meeting of the new Collaboration Agreement is being planned for May or early June. + At the last SB (Steering Board) meeting there was a proposal to have a face-to-face discussion in order to resolve issues regarding the license and Organisation regulations. --------------------------------------------------------------------- ==== Garfield - Rob - Garfield + Lecture series in York (22/2 to 27/2) have been taken for graduate students and work going on for Isolde and Triumf experiments. New lecture scheduled in Cagliari in the next weeks. + In collaboration with Ignacio Reguero (IT), a BOINC version of Garfield has been made and this is currently producing gas files with a precision and a level of detail that would otherwise have required years to produce. At the moment, ATLAS is taking advantage of this facility (for electron transport in the magnetic field), but it is anticipated other experiments will follow. + Usual maintenance ongoing. - Validation (calorimetry) + The estimate of the required statistics was presented. The conclusion confirms the approach taken by the Physics Validation project. + Apart from large areas where there is agreement, there are also areas of serious discrepancy between MC and data. Most noteably, this concerns the shower profiles. This discrepancy goes unnoticed in the resolution and response measurements of fully contained showers, but seriously affect the study of high energy showers which are not fully contained. There's need to get in touch with the models' developers to discuss these issues and understand in the physics what is at the origin of such discrepancy. + It would be timely to answer the questions which R.Wigmans repeatedly asks concerning the sampling fraction and pion/proton response. - NA60 + Good progress in the search for K0s and lambda. For both, statistics has vastly increased and both have now been located not only from production in upstream material (Si) but also from the target (In). A search is going on to find the peaks in the converter (Pb). Since temperatures are now routinely measured, studies of A-dependence and of centrality-dependence are within reach. Acceptance and efficiency studies are in progress. + No news regarding the anti-baryons, they will be addressed later on. --------------------------------------------------------------------- ==== AOB Next meeting to be scheduled for May 12th.