Upload
amber-conley
View
218
Download
1
Tags:
Embed Size (px)
Citation preview
Maria Grazia Pia, INFN Genova and CERN 1
An OO model for intra-nuclear transport
Maria Grazia PiaL. Bellagamba, A. Brunengo, E. Di Salvo
for the Geant4 Collaboration
Maria Grazia Pia, INFN Genova and CERN 2
Hadronic physics in Geant4 The most complete hadronic simulation kit
on the market data-driven, parameterisation-driven and
theory- driven models complementary and alternative models
Common basic approach: expose the physics through OO design, to provide the transparency required for the validation of physics results
Maria Grazia Pia, INFN Genova and CERN 3
A look at the past...
Hadronic simulation was handled through “packages” monolithic: either take all of a package or nothing difficult to understand the physics approach hard to disentangle the data, their use and the
physics modeling
...keeping in mind that this is complex physics, in some cases with poor support of experimental data
Maria Grazia Pia, INFN Genova and CERN 4
Gheisha Fluka Calor
Crystal Barrel and BaBar IFR
An example:
(Geant3)
Maria Grazia Pia, INFN Genova and CERN 5
Energy range of hadronic models
Evaporation phase Pre-equilibrium: O(100 MeV) Intra-nuclear transport: ~100 MeV - 5 GeV
(the “resonance” region) The high energy régime
Maria Grazia Pia, INFN Genova and CERN 6
The context
The Hadron Kinetic Model is situated in the context of Geant4 theory-driven hadronic models
It covers the energy range of intra-nuclear transport in Geant4
It must satisfy the requirements as a model for intra-nuclear transport to be
used directly by the processes as a back-end to higher energy models
Maria Grazia Pia, INFN Genova and CERN 7
Maria Grazia Pia, INFN Genova and CERN 8
The physics idea
Introduce the concept of “time development”
into the traditional approach of intra-nuclear cascade
Provides means to address interactions and transport with more sophisticated modeling, i.e. with greater precision
Maria Grazia Pia, INFN Genova and CERN 9
Problem domain analysis
Identify what the model should do (not how it should do it!)
Decompose the domain into the basic physics concepts
Further iteration of the decomposition within each sub-domain
Close collaboration of OO “experts” with theoreticians
Maria Grazia Pia, INFN Genova and CERN 10
Model domains
Model-specific The time-development management The scattering of interacting particles The intra-nuclear transport
Common to other theoretical models The 3D modeling of the nucleus The description of the interacting particles
Maria Grazia Pia, INFN Genova and CERN 11
Maria Grazia Pia, INFN Genova and CERN 12
Features of the OO design Mapping of physics onto OO design Openness to extension and evolution
new channels, new experimental data, new parameterisations, new theoretical models, new algorithms...
Flexibility for reuse of components in other contexts e.g. scattering component common to other models
Extensive use of patterns to cope with the physics complexity through common
archetypes (Composite pattern) to handle alternative algorithms or physics options (Strategy
pattern)
Maria Grazia Pia, INFN Genova and CERN 13
Maria Grazia Pia, INFN Genova and CERN 14
Maria Grazia Pia, INFN Genova and CERN 15
The Kinetic AlgorithmA step by step updating of a particle vector: Create a vector of particles, assign initial particle types,
coordinates and momenta etc., assign initial value for the time evolution parameter
For a given step of the time evolution parameter find pairs of particles, according to a collision criterion, which are assumed to collide and particles which, according to their lifetimes, are assumed to decay
Perform particle collisions and particle decays, determining the generation of outgoing particles; during this step particle coordinates and momenta are updated (particle propagation)
Starting from , perform the next step(all this taking into account Pauli blocking)
Maria Grazia Pia, INFN Genova and CERN 16
The Scatterer Two-body hadron elastic and inelastic scattering
including resonance excitation and deexcitation, particle absorption etc.
Physics processes implemented: baryon-baryon interactions (including interactions of baryon resonances) baryon-antibaryon annihilation meson-baryon interactions meson-meson interactions
Cross sections are calculated from: tabulations of experimental data (by means of fits or interpolations) parameterisations according to algebraic functions from other cross sections via general principles (detailed balance, AQM...)
Yes, it is very complex indeed...
Maria Grazia Pia, INFN Genova and CERN 17
The field transport At every step coordinates and momenta of all particles travelling
through the nucleus are updated from the values at the time of previous interaction to the current time
For transportation purpose particle-nucleus interaction is described through phenomenological potentials (optical potentials)
The particle propagation can be operated via a cascade approach, i.e. along a straight line trajectory
– using potentials only to evaluate the final particle energy at the end of the step
via a numerical integration of the equations of motion– using the classical Runge-Kutta method (code reuse from Geant4 Geometry-
Transportation)
Maria Grazia Pia, INFN Genova and CERN 18
What is new in this model? In terms of physics... The kinetic algorithm The detailed and extensive scattering module The precise field transport The wealth of advanced theoretical modeling in each
of its details and the thorough exploitation of existing experimental data
The combination of all the above
The transparency of the physics
Maria Grazia Pia, INFN Genova and CERN 19
What is new in this model?In terms of software... It fits into a general framework
easily interchangeable with other models at a fine granularity
It is based on a solid OOAD open to extension open to evolution capable to host alternative models/algorithms/data
Ample code reuse from the G4 Geometry/Transportation
– fully exploits the knowledge of mathematical experts by other models
– powerful general components like the Scatterer
OOAD contributes to make the physics transparent
Maria Grazia Pia, INFN Genova and CERN 20
Conclusions
Geant4 Hadron Kinetic Model represents a new approach to intra-nuclear transport
OOAD has been the key to handle the underlying complex physics domain
OOAD has allowed to clearly expose the physics, thus contributing to the validation of physics results
The sound OOAD makes the model open to evolution and extension