Simulation Project Leaders Meeting
10-November-2009

==== Attendees

John Apostolakis (JA), Alberto Ribon (AR), Alfredo Ferrari (AF),
Rob Veenhof (RV), Gabriele Cosmo (GC).

---------------------------------------------------------------------
==== Minutes

- Witold Pokorski is joining back PH/SFT; Witek will replace A.Ribon in
  leading the Generator Services project starting from January; he will
  also be involved in the Physics Validation activity.

---------------------------------------------------------------------
==== Simulation physics validation & framework - Alberto

- Meetings:
  o No meetings held in the last month; the next meeting is scheduled
    for November 25th.
- Validation studies:
  o Nothing to report, most of the time in the last month was spent for
    preparation of the Geant4 Workshop and the IEEE/NSS conference in
    Orlando (Florida), where a poster on the recent validation activities
    was presented.
  o The last development release of Geant4 includes the first version of
    the 'CHIPS' physics-list which adopts CHIPS as unique hadronic model,
    where it is expected a smoother energy transition between energy
    ranges. First tests have been successful so far, also showing good
    stability.

---------------------------------------------------------------------
==== Generator services - Alberto

- Meetings:
  o The next monthly meeting will be held on November 18th and will be
    dedicated to the en release of Sherpa, 1.2.0.
  o The planning meeting is scheduled for November 27th.
- GENSER:
  o Activity currently going on for moving to use autotools, the
    migrations has been almost completed.

---------------------------------------------------------------------
==== Fluka - Alfredo

- Fluka:
  o Currently starting the validation end testing phase of the new code
    implementing damage predictions of displacement per atom (DPA).
  o A considerable effort is going on for porting the code to gfortran,
    and is almost completed; divergences between different versions/
    platforms 32/64-bits are being understood and mostly due to rounding
    differences.
    Some compiler options seems not to work well (-unreachable), g95 finds
    uninitialised variables rather efficiently, while gfortran is not as
    reliable (JA suggest to use implicit-none option).
    First porting works anyhow, but is not yet fully validated.
  o In negotiation with Siemens for the usage of the code (mainly in
    hadron-therapy facilities, to make use of the DB generated by the MC),
    license issues for a scientific vs. commercial agreement.
  o In the next month two IC boards for positions in the team will take
    place.
  o No new releases planned before Spring next year.
- Meetings:
  o The course in Mumbai went well, 40 people in total, 10 people from
    other countries.

---------------------------------------------------------------------
==== Geant4 - John

- Releases:
  o Started the integration and testing phase for release 9.3 due in December.
- Developments:
  o Fixes for performance and robustness issues identified in ATLAS and
    CMS (description below).
  o Improvements in the FTF physics model include tuning of the binary
    channels at energies up to 20 GeV, and the introduction of a simple
    Reggeon cascade for use with FTF.
  o New CHIPS physics processes for hadron projectiles, with the melding of
    string-like and absorption interactions.
  o New physics-lists being provided, including the first release of
    the CHIPS-single model physics-list in beta version; the new
    physics-lists QGSP_FTFP_BERT and FTFP_BERT_TRV which limit the use of
    the parameterised LEP models are labelled emerging, i.e. proposed for
    evaluation; several old lists are being made obsolete.
  o Among various developments from external parties is a new
    visualization driver, gMocren, which interfaces to a renderer that
    displays 3-D data. Its primary purpose is use in medical physics to
    display volume DICOM data from a patient's CT scan; it can visualize
    also dose distributions, and for this purpose can be reused in other
    application areas including HEP.
- Feedback from experiments:
  o ATLAS, CMS reported on memory allocation issues; thanks also to the
    study on the optimizations being implemented for the multi-threaded
    prototype, a fix was identified for reducing memory fragmentation
    and calls to malloc(), giving a measured 5% CPU improvement for
    a Geant4 example which simulates single particles in the CMS geometry.
- Meetings:
  o The Users workshop in Catania was very successful, with 35 users
    attending. There were three sessions on medical applications, and
    two on HEP including low background and rare event applications.
    Lots of discussion, also at the Collaboration Workshop. Notable progress
    reported by the EM validation team on validation testing tools; a new web
    tool developed for monitoring physics performance can provide comparison
    plots for any measured quantity between any set of development releases.
  o A first schedule of the next architecture review has been agreed, with
    the target to make recommendations to the Collaboration Workshop 2010,
    to be held in late September or early October.
  o A Workshop (Asia-Pacific Geant4 Workshop) and Training Course will
    take place in Tsukuba (Japan) on December 7-11th.
  o A training for the MC-PAD ITN project will take place at DESY in January,
    and be repeated in March for the Helmholtz alliance of universities and
    institutions. The three day event will provide a short introduction to
    Monte Carlo, Geant4 and analysis in the context of a tracker detector.
  o A Geant4 tutorial at CERN is planned for 15-19th February 2010, and
    will be announced shortly.

---------------------------------------------------------------------
==== Garfield - Rob

- Meetings:
  o RD51 meetings were held in Khania (Kriti), both an MPGD conference and
    an RD51 Collaboration meeting, and at CERN.
  o The next RD51 Collaboration meeting will be held at CERN at the end
    of November.
- Developments/activities over the last few months:
  o neBEM: new release which includes periodicities and an improved handling
    of far-field elements
  o Penning: a fairly simple, yet apparently complete, model of the
    principal processes (direct, exchange, homonuclear associative
    ionisation, radiative ionisation, excimer formation) has been made and
    the model parameters have been measured in 7 Ar-based mixtures (using
    gain curves from the literature), as function of pressure and
    concentration; this has permitted a detailed reconstruction of the
    mechanisms that are at play, quantitatively and in time.
  o Magboltz: new version 8.6 with includes far more detailed information
    regarding the excited states of e.g. Ar.
    Excimers and scintillation: gradually taking advantage of this newer
    version of Magboltz.

---------------------------------------------------------------------
==== AOB

Next meeting scheduled for December 15th.