Geant4 [1] is a
general-purpose toolkit for the simulation of the passage of particles through
matter. It provides functionalities for all the typical domains of simulation:
geometry, tracking, detector response, run, event and track management,
Particle Data Group compliant particle management, visualization and user
interface, as well as a large variety of physics processes and models.
The wide spectrum of
functionalities provided makes it particularly suitable for the simulation of
astroparticle experiments. Thanks to
the OO technology, Geant4 provides a
high degree of flexibility and openness to evolution, as well as to the
integration of external tools.
The Geant4 Toolkit
provides an abundant set of physics processes to describe the interactions of
particles with matter. A variety of
underlying models is often available for the same physics process, depending on
the energy range, particle type or material.
Geant4
Electromagnetic Physics [3] manages electron and positron, photon, and muon
interactions, as well as the electromagnetic interactions of hadrons and ions.
The validity range
of all the muon processes, based on theoretical models, scales up to the 1000
PeV region, allowing the simulation of ultra-high energy and cosmic ray
physics. A set of physics processes is also available in Geant4, extending the
validity range of electromagnetic interactions down to low energies < 1 keV
[4].
Geant4 Hadronic
Physics [5] offers both parameterization-driven models and a variety of
theory-driven models, as well as treatment of low energy neutron transport. However,
most of the hadronic physics package needs serious testing & validation.
Long-term
radioactivity produced by nuclear interactions represents an important
contribution to background levels in space-borne gamma-ray and X-ray
instruments, as the ionization events that result often occur outside the
time-scales of any veto pulse; it is also essential to background studies in
underground experiments. A Radioactive
Decay Module [2] is available in Geant4.
The GEANT4 validation
The Geant4
collaboration provides user support and takes care of validation of physics
processes and other toolkit facilities (see the web page
http://geant4.web.cern.ch/geant4/ for more information). However, tests “on the
field” are useful.
Various HEP experiments
have started the migration from old Monte Carlo tools (like GEANT3) to GEANT4
and for all these experiments (and GLAST should not be an exception) the
physics reliability of the simulation is a high priority. This reflects on a
huge effort by big experiments (see for example [11] for ATLAS) on the physics
validation of GEANT4. Again GLAST should not be an exception, and the
overlapping with the range in which the validation of ATLAS is crucial is not
complete.
The Italian software
group is addressing the validation of physics processes directly involved in
the detection of high energy photons. The first process tested is
G4GammaConversion, designed to simulate the Pair Production process. As a
consequence of the tests, a bug in the energy distribution of the secondary electron-positron
pair was found; we have corrected the implementation of the present release of
G4 (this correction will be inserted in future releases of GEANT4) and we have
obtained an energy distribution consistent with the literatuture [12]. We plan
to proceed on testing the G4 simulation toolkit, both addressing the concrete
implementation of its classes and comparing its results with available
experimental data.
The Italian software
group plans also to extend the G4 physics capabilities, inserting for example a
description of the gamma conversion that depends also on the polarisation of
the incoming gamma.
For GLAST a major
validation will come from the Balloon flight that is using GEANT4 for the Monte
Carlo simulations; comparison with both test beam data and flight data will be
of invaluable use in this phase. Some preliminary results coming from
comparisons between the simulations and real data are already available thanks
to the work of the Japanese software group (see next section).
For the GEANT4
activities related directly to GLAST, two groups (Italy and Japan) have till
now worked actively on simulations with the following main results:
Some discrepancies with test beam data have been found for low energy energy
loss of protons; this has to be addressed more carefully, probably shifting
from the standard electromagnetic package of GEANT4 to the low energy one.
The geometry for the balloon simulation is still hard coded in the
simulation; going to the XML representation [6] should not require a large work.
To conclude, we think
that GEANT4 efforts in GLAST have demonstrated the reliability of the toolkit
and the concrete possibility to use it as a production software in the GLAST
experiment.
For a full prototype
of GEANT4 useful for GLAST we need
The final word will be
possible after the completion of the previous steps. Further analysis are
needed, especially on the physical validation domains; but we are quite
confident, also because of the involvement on GEANT4 of other big experiments,
like ATLAS, LHCb, ALICE, BaBar.
[1] Geant4: LCB Status Report/RD44,
CERN/LHCC-98-44, 1988
[2] http://www.space.dera.gov.uk/space_env/geant_mn.html
[3] S. Chauvie et al., Electromagnetic Physics,
Proceedings of MC2000
[4] S. Chauvie et al., Geant4 Low Energy
Electromagnetic Physics, These Proceedings
[5] J.P. Wellisch, Hadronic Shower Models in
Geant4, CHEP2000 Proc.
[6] http://www.fisica.uniud.it/~riccardo/research/glast/GDD
[7] http://www.fisica.uniud.it/~riccardo/research/glast/gallery
[8] http://www-heaf.hepl.hiroshima-u.ac.jp/glast/glast-hp/Geant4/geant4_research.html
[9] http://www-glast.slac.stanford.edu/LAT/balloon/
[10] http://www.ge.infn.it/geant4/lowE/examples/gammaray_telescope/index.html
[11] Atlas GEANT4
validation page: http://atlas.web.cern.ch/Atlas/GROUPS/SOFTWARE/OO/domains/simulation/G4PhysicsStudies/
[12] B. Rossi, High-Energy
Particles, Prentice-Hall
_______________________________________________________________________________________