; A minimal introduction to C++ and to OOP
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

A minimal introduction to C++ and to OOP

VIEWS: 0 PAGES: 66

  • pg 1
									       Electromagnetic Physics in
       Geant4

                          Peter Gumplinger
                              TRIUMF

Partially based on presentations by A. Lechner, M.G. Pia, V. Ivanchenko, S. Incerti,
M. Maire, A. Howard, and adopted from Luciano Pandola (INFN-LNGS)
      EM Packages overview
   Models and processes for the description of EM interactions
    in Geant4 have been grouped into several packages
        Package                             Description
                        g-rays, e± up to 100 TeV, Hadrons, ions up to 100 TeV
       Standard
        Muons           Muons up to 1 PeV

        X-rays          X-rays and optical photon production

        Optical         Optical photons interactions

     High-Energy        Processes at high energy (> 10 GeV). Physics for exotic
                        particles

      Low-Energy        Specialized processes for low-energy (down to 250 eV),
                        including atomic effects

      Polarization      Simulation of polarized beams
   EM processes for g-rays, e±
Particle            Process                 G4Process

Photons    Gamma Conversion in e±   G4GammaConversion

           Compton scattering       G4ComptonScattering

           Photoelectric effect     G4PhotoElectricEffect

           Rayleigh scattering      G4RayleighScattering

  e±       Ionisation               G4eIonisation


           Bremsstrahlung           G4eBremsstrahlung


           Multiple scattering      G4eMultipleScattering

  e+       Annihilation             G4eplusAnnihilation
   EM processes muons
Particle           Process               G4Process

  m±       Ionisation            G4MuIonisation


           Bremsstrahlung        G4MuBremsstrahlung


           Multiple scattering   G4MuMultipleScattering

           e± pair production    G4MuPairProduction
        Process and Model concept
   The idea is that the same physics processes (e.g. Compton
    scattering) can be described by different models, which can be
    alternative or complementary in a given energy range
   For instance: Compton scattering can be described by
           G4KleinNishinaCompton
           G4LivermoreComptonModel (specialized low-energy, based on the
            Livermore database)
           G4PenelopeComptonModel (specialized low-energy, based on the
            Penelope analytical model)
           G4LivermorePolarizedComptonModel (specialized low-energy,
            Livermore database with polarization)
           G4PolarizedComptonModel (Klein-Nishina with polarization)
   Different models can be combined, so that the appropriate one is
    used in each given energy range ( performance optimization)
Part I: A short overview of the
“Standard” EM models
        “Standard” EM physics
   Release with the 1st version of Geant4, mainly based
    on the Geant3 experience
   Significant continued development in many aspects of
    EM processes simulation since then
   Used for many years in production in HEP
    experiments
       BaBar, SLAC (since 2000)
       LHC experiments ATLAS, CMS and LHCb (since 2004)
   Many common requirements for HEP, space, medical
    and other applications
       “Standard” EM physics
   The projectile is assumed to have energy > 1keV
   The material is described as homogeneous,
    isotropic, and amorphous
   The atomic electrons are quasi free - e.g. their
    binding energy is neglected (except for photo-
    electric effect)
   The atomic nucleus is fixed - e.g. the recoil
    momentum is neglected
        Standard EM packages
   Standard                                Validation, two configurations:
       g, e up to 100 TeV                       5 mm Pb/5 mm Scintillator
       hadrons up to 100 TeV
       ions up to 100 TeV                       10 mm Pb/2.5 mm Scintillator
   Muons                                   Data from NIM A262 (1987)
       up to 1 PeV                          229; NIM A274 (1989) 134
       Energy loss propagator
   Xrays
       X-ray and optical photons
   Optical
       Optical photon interactions
   High-energy
       Processes at high energy (E>10GeV)
       Physics for exotic particles
   Polarisation
       New package for simulation of
        polarized beams
           Particle Transport
   Common to all charged                Optical Photons:
                                         
    particles:                               Refraction and Reflection
                                             Absorption
       Ionization
                                             Rayleigh scattering
       Coulomb scattering from nuclei
                                             Wavelength shifting (wls)
       Cerenkov effect
                                             Mie scattering (9.4)
       Scintillation
       Transition Radiation
   Muons:
       Bremsstrahlung
       Nuclear Interaction
       (e+/e-) pair production
           Gamma and Electron Transport
                                               HEP calorimeter
   Photon processes:
       Conversion into e+e- pair
       Compton scattering
       Photoelectric effect
       Rayleigh scattering in LowE package
       Gamma-nuclear interaction in
        hadronic sub-package CHIPS
   e± processes:
       Bremsstrahlung
       Nuclear interaction in hadronic sub-      Medical linac
        package CHIPS
   e+ annihilation
    e-

proton

alpha
         Geant4 Urban MSC model
   Starting point: Lewis theory
    on transport equation of
    charged particles
   The Urban MSC model uses
    phenomenological functions to
    sample angular and spatial
    distributions after the simulation
    step
   The functions parameters are
    chosen to provide the same
    value of moments of the
    distribution as in Lewis theory
   See details in the Geant4
    Physics Reference Manual
(MSC)
        Muon EM Physics Simulation
   Main processes:                     Muon stopping power
       Ionization and bremsstrahlung    Precision about 2%
       e+e- pair production
       Muon-nuclear interactions in
        hadronic packages
        Total muon energy loss
           Hadron and ion EM physics
   Coulomb scattering
   Ionization
    Bethe-Bloch formula with corrections used for E>2 MeV
  dE             z 2  2me c 2  2 g 2  2 
                      ln
                                                T  C G   F              
     4N e r0 2 
               2
                                         1  c   
                                                   Z           zL1  z L2 
                                                                        2
                                                                            
  dx                       I          2  Tmax        2                  
                  C – shell correction
                  G – Mott correction
                  δ – density correction
                  F – finite size correction
                  L1- Barkas correction
                  L2- Bloch correction
                  Nuclear stopping
                  Ion effective charge
   Bragg peak parameterizations E<2MeV
        ICRU’49 and NIST databases
       Simulation of x-rays and
       optical photons
   Optical photons are created by              xray package
          Cherenkov effect (G4Cherenkov)
          Transition radiation (G4TransitionRadiation)
          Scintillation (G4Scintillation)
   Optical photons are hence managed by
          Reflection and reflection at boundary
           (G4OpBoundaryProcess)                 optical package
          Absorption (G4OpAbsorption)
          Rayleigh scattering (G4OpRayleigh)
          Wavelength shifting (G4OpWLS)
Standard EM Physics Summary

   Standard EM physics processes are available for
    gammas and charged particles from 1 keV and up
   EM processes are based on theoretical cross sections
    with corrections. During simulation, quantities are
    taken from tables calculated at initialization time
   Multiple scattering is handled by model functions which
    represent fits to Lewis transport theory results (not
    Moliere)
   Energy-range relation is used to compute energy loss,
    and to control step lengths and secondary production
Part II: A short overview of
the models. Low Energy
          Purpose
   Extend the coverage of Geant4 electromagnetic interactions with matter
        photons, electrons, hadrons and ions
        down to very low energies (sub-keV scale)

   Possible domains of applications
        space science
        medical physics
        Microdosimetry

   Choices of Physics models include
        Livermore library: electrons and photons [250 eV – 1 GeV]
        Penelope (Monte Carlo): electrons, positrons and photons [250 eV – 1 GeV]
        Microdosimetry models (Geant4-DNA project): [7 eV – 10 MeV]


                                                                                 37
         When/why to use Low Energy
         Models
   Use Low-Energy models (Livermore or Penlope), as an alternative
    to Standard models, when you:

       need precise treatment of EM showers and interactions at low-
        energy (keV scale)
       are interested in atomic effects, as fluorescence x-rays, Doppler
        broadening, etc.
       can afford a more CPU-intensive simulation
       want to cross-check other simulation (e.g. with a different model)

   Do not use when you are interested in EM physics
    > MeV
       same results as Standard EM models, performance penalty
 Livermore Models
Full set of models in Geant4 for electrons, g-rays
and ions based on the Livermore data libraries
          (cross sections and final states)
Energy range down to 250 eV. They include atomic
  effects, like fluorescence, Auger emission, etc.




                                                     39
        Livermore Models
   Validity Range: 250eV - 100GeV
   Included elements from Z=1 to Z=100
       Atomic relaxation: Z > 5 (EADL transition data)

   Database
       Mixture: experiments and theories
       Extracted from public libraries: EEDL, EPDL …

   Implementation in Geant4
       Total cross sections: photo-electric effect, diffusions
       Compton and Rayleigh, pair production and Bremsstrahlung
       Sub-levels integrated cross sections: photo-electric and ionization
       Energy spectra: secondary processes in e-
                 Available Livermore models
Physics                   Process                        Model                Low Energy    High Energy
Process                    Class                         Class                  Limit          Limit
Gammas

Compton           G4ComptonScattering     G4LivermoreComptonModel             250 eV       100 GeV

Polarized         G4ComptonScattering     G4LivermorePolarizedComptonModel    250 eV       100 GeV
Compton
Rayleigh          G4RayleighScattering    G4LivermoreRayleighModel            250 eV       100 GeV

Polarized         G4RayleighScattering    G4LivermorePolarizedRayleighModel   250 eV       100 GeV
Rayleigh
Conversion        G4GammaConversion       G4LivermoreGammaConversionModel     1.022 MeV    100 GeV

Photo-electric    G4PhotoElectricEffect   G4LivermorePhotoElectricModel       250 eV       100 GeV

Electrons

Ionization        G4eIonisation           G4LivermoreIonisationModel          250 eV       100 GeV

Bremsstrahlung    G4eBremsstrahlung       G4LivermoreBremsstrahlungModel      250 eV       100 GeV

                                                                                                     41
Photon models (1)
   Compton scattering (incoherent)
       Scattered photon energy: from Klein Nishina formula
        Modified by the Hubbel form factor obtained from EPDL97
       Angular distributions of scattered photon and recoil electron
        from EPDL97

   Rayleigh scattering (coherent : no energy loss)
       Angular distribution from Rayleigh formula
       Include the Hubbel form factor from EPDL97



                                                                    42
Photon models (2)
   Photoelectric effect
        Cross section integrated over shells and cross section by shell from EPDL
        Several angular distribution generators available (naive, Sauter-Gravila,
         Gravila)
        De-excitation : managed by the atomic relaxation process
             Initial vacancy and cascade of resulting vacancies

   Pair conversion
        e+ and e- energies computed from Bethe-Heitler formula
         Include Coulomb correction
        Tsai differential cross section for energy and polar angle computation
        Polar angular distribtuion: symmetric
        Azimuthal angle distribution: isotropic


                                                                                     43
Electron models
   Bremsstrahlung
       Parametrisation from EEDL
       16 parameters


   Ionisation
       Parametrisation using 5 parameters by
        shell


                                                44
Livermore Models – cross
sections, ranges

  Photo-electric
    Hydrogen       Photo-electric
                                       Gamma
                       Neon
                                      Conversion
                                         Lead




                     Electron Range
                       (tag: 9.2-4)
         Livermore Polarized Models
   Describe in detail the kinematics of polarized photon
    interactions
   Based on the Livermore database
   Applications of such developments include the design
    of new space mission for the detection of polarized
    photons
   Documentation on the MC methods could be found:
       Nucl. Instrum.Meth. A566: 590-597, 2006 (Photoelectric)
       Nucl. Instrum.Meth.A512: 619-630, 2003 (Compton and Rayleigh)
       Nucl.Instrum.Meth.A452:298-305,2000. (Pair production)
   Currently available: Compton and Rayleigh
      Eg. polarized Compton cross section

d      
            1 h  h          2 
                              h
            The Klein Nishina cross section:
                   2
    r0 2 
      2       0
                    2  4 cos 
d 4 h 0  h h 0              
       where
       h0 : energy of the incident photon
       h : energy of the scattered photon
        : angle between the two polarization vectors

The code properly reproduces polarized photon interactions
and also the secondary polarization acquired after a Compton interaction
         Penelope Models
   Geant4 includes the low-energy models for e± and g-rays from the
    Monte Carlo code PENELOPE (PENetration and Energy LOss of
    Positrons and Electrons)
      Nucl. Instrum. Meth. B 207 (2003) 107



   The physics models have been specifically developed by the
    Barcelona group (F. Salvat et al.) and a great care was dedicated to
    the low-energy description (atomic effects, fluorescence, Doppler
    broadening, etc.)

   Mixed approach: analytical, parametrized & database-driven
      applicability energy range: 250 eV  1 GeV



   Includes also positrons (not described by Livermore models)
                                                                           48
Valid from Geant4 9.3 BETA


          Penelope in Geant4
    Reliability of the physics models
         Extensively tested by the Penelope group itself (several papers)
    Penelope coding
         Original in FORTRAN77  Version 2001 re-engineered in Geant4 (C++)
    Corresponding physics models in Geant4:
G4PenelopeComptonModel                             G4PenelopeAnnihilatio
G4PenelopeRayleighModel                            nModel

G4PenelopeGammaConversi g-rays G4PenelopeBremsstrahl
onModel                        ungModel

G4PenelopePhotoElectric e±     G4PenelopeIonisationM
Model                          odel
                                                                               49
                 When/how to use Penelope models
        Use Penelope models (as an alternative to Livermore or
         Standard models) when you:
                need precise treatment of EM showers and interactions at low-
                 energy (keV scale)
                are interested in atomic effects, as fluorescence x-rays, Doppler
                 broadening, etc.
                can afford a more CPU-intensive simulation
                want to cross-check an other simulation (with a different model)
                are interested in low-energy positrons (only choice in Geant4)

       Do not use when you are interested in EM physics > MeV
               same results as Standard EM models, performance penalty


                                                                                     50
                                Penelope verification and validation
              If G4Penelope gives the same results as Penelope-Fortran  take for granted
              the (large) validation work performed by the Penelope group
     Additional validation within Geant4 for e± and g-rays (all EM models)

                                         g-ray                  Rayleigh scattering
Attenuation coeff. (cm2/g)




                                      attenuation
                                      coeff in Al                50 keV g-ray in Au

                                                2=15.9       Penelope FORTRAN
                                                =19
                                                              G4Penelope
                                                p=0.66
                              NIST data
                              Penelope

                                               Energy (MeV)                           cos   51
       Example: Doppler broadening
       in Compton scattering
Compton scattering: electrons bound and not at rest (as assumed
for Klein-Nishina)  change of angular distribution, reduction of XS

                                     Penelope model includes it (via
    Au, 50
                                         analytical approach)
    keV g-ray
                                     Livermore model also deals with
                                   Doppler broadening (EGS database
                                               approach)
                                     Good agreement Penelope-
                                            Livermore
                                      Standard model includes cross
                                     section suppression, but samples
                                   final state according to Klein-Nishina
                                                                       52
        Ion energy loss model
   Describes the energy loss of ions heavier than Helium
    due to interaction with the atomic shells of target
    atoms

   The model computes
       Restricted stopping powers, which determine the continuous energy
        loss of ions as they slow down in an absorber (more details on next
        slides)
       Cross sections for the production of δ-rays (Note: δ-rays are only
        produced above a given threshold), which inherently also govern the
        discrete energy loss of ions

   Primarily of interest for
       Medical applications         G4IonParametrizedLossModel
       Space applications
                                                                              53
         Using ICRU73 Tables
   ICRU 73 stopping powers are available for a range of
    elemental and compound materials:

       In order that the ICRU 73 tables are used by the ion model,
        materials must have names of Geant4 NIST materials
       Either Geant4 NIST materials are used, or user-specific materials
        are created having the same name as materials in Geant4 NIST
        data base.
       Note: ICRU 73 stopping powers are not available for all NIST
        materials.
       Available stopping powers can be looked up in the following classes
        of the Geant4 material sub-package
        ($G4INSTALL/source/materials):
            G4SimpleMaterialStoppingICRU73 (ions up to Ar)
            G4MaterialStoppingICRU73 (ions up to Ar)
            G4IronStoppingICRU73 (Fe ions & ions scaled from Fe)
Ion stopping powers (1/2)
   Electronic stopping powers: important ingredient to determine the mean
    energy loss of ions along simulation steps
        Impacts the ion range (for example)


   Restricted stopping powers: account for the fact that the continuous energy
    loss     description     is    restricted    to     energies  below     Tcut
    (where Tcut denotes the lower production threshold of δ-rays)

   Restricted     stopping     powers         are   calculated     according   to
    (T = kinetic energy per nucleon)
        T < TL: Free electron gas model
        TL ≤ T ≤ TH: Interpolation of tables or parameterization approach
        T > TH: Bethe formula (using an effect. charge) + high order corr.


                                                                                55
Ion stopping powers (2/2)
   Parameterization approach
    Model incorporates ICRU 73 stopping powers into Geant4

   ICRU73 model
        Covers a large range of ion-material combinations: Li to Ar, and Fe
        Stopping powers: based on binary theory
        Special case: water
             Revised ICRU 73 tables of P. Sigmund are used (since Geant4 9.3.b01)
             Mean ionization potential of water of 78 eV
        Current model parameters (Geant4 9.3.b01):
             THigh = 10 MeV/nucleon (except Fe ions: TH = 1 GeV/nucleon)
             TLow = 0.025 MeV/nucleon (lower boundary of ICRU 73 tables)
        For ions heavier than Ar
             Scaling of Fe ions based on effective charge approach

                                                                                 56
         How to use the new model ?
   Model name: G4IonParametrisedLossModel

   Designed to be used with G4ionIonisation process (of standard EM package)
        Not activated by default when using G4ionIonisation
        Users can employ model by utilizing SetEmModel function of
         G4ionIonisation process


   Restricted to one Geant4 particle type: G4GenericIon
        Note: The process G4ionIonisation is also applicable to alpha particles
         (G4Alpha) and He3 ions (G4He3), however the model must not be activated
         for these light ions



                                                                                57
Geant4 for microdosimetry
   History : initiated in 2001 by Petteri Nieminen (European Space Agency /
    ESTEC) in the framework of the « Geant4-DNA » project

   Objective : adapt the general purpose Geant4 Monte Carlo toolkit for the
    simulation of interactions of radiation with biological systems at the cellular
    and DNA level (« microdosimetry »)

   A full multidisciplinary activity of the Geant4 low energy electromagnetic
    Physics working group, involving physicists, theoreticians, biophysicsts…

   Applications :
        Radiobiology, radiotherapy and hadrontherapy
         (eg. prediction of DNA strand breaks from ionising radiation)
        Radioprotection for human exploration of Solar system
        Not limited to biological materials (ex. Silicon)



                                                                                      58
Geant4 for microdosimetry
   Several models are available for the description of physical processes
    involving e-, p, H, He, He+, He++

   Include elastic scattering, excitation, ionisation and charge change

   For now, these models are valid for liquid water medium only

   Models available in Geant4-DNA
        are published in the literature
        may be purely analytical or use interpolated cross section data

   They are all discrete processes



                                                                             59
          Valid from Geant4 9.3 BETA                          • Models in black are analytical
                                                              • Models in purple use interpolated data




                Physics models in Geant4 DNA
                             e                    p                       H                 , He+, He
   Elastic           > 7.4 eV – 10 MeV
 scattering          Screened Rutherford
                      > 7 eV – 10 MeV
                                                  -                        -                       -
                          Champion

 Excitation
                                            10 eV – 500 keV
A1B1, B1A1, Ryd      7.4 eV – 10 MeV          Miller Green
                                                                           -
A+B, Ryd C+D,          Emfietzoglou        500 keV – 10 MeV
 diffuse bands                                    Born
                                                                                           Effective charge
   Charge                                   1 keV – 10 MeV        1 keV – 10 MeV          scaling from same
   Change                     -                                                             models as for
                                               Dingfelder            Dingfelder
                                                                                                proton
 Ionisation                                100 eV – 500 keV
                     12.6 eV – 30 keV            Rudd            100 eV – 100 MeV
1b1, 3a1, 1b2, 2a1
      + 1a1
                           Born            500 keV – 10 MeV            Rudd
                                                 Born
                                                                                                         60
Part IV: How to use the EM
physics and to set-up the
physics list
      EM Physics Constructors for
      Geant4 9.3 - ready-for-the-use
G4EmStandardPhysics           – default
G4EmStandardPhysics_option1 – HEP fast but not precise
G4EmStandardPhysics_option2 – Experimental
G4EmStandardPhysics_option3 – medical, space
G4EmLivermorePhysics
                                    Combined Physics
G4EmLivermorePolarizedPhysics
                                     Standard > 1 GeV
G4EmPenelopePhysics                LowEnergy < 1 GeV
G4EmDNAPhysics

   $G4INSTALL/source/physics_list/builders
   Advantage of using of these classes – they are tested on
    regular basis and are used for regular validation
         Example: change default model
   Process class G4ComptonScattering
   Default model in G4 9.3 is G4KleinNishinaCompton
    (EM Standard)
   There are alternative Livermore and Penelope models
       Let’s try G4PenelopeComptonModel

G4double limit = 1.0*GeV;                           create process
if ( particleName == “gamma” ) {
   G4ComptonScattering* cs= new G4ComptonScattering();
   G4PenelopeComptonModel* aModel = new
        G4PenelopeComptonModel();               Create model and
    aModel->SetHighEnergyLimit(limit);           set energy limits
    cs->AddEmModel(0, aModel);    // 1st parameter - order
    pmanager->AddDiscreteProcess(pef);
                                            Set the model for
}
                                               the process
         How to extract Physics ?
   Possible to retrieve physics quantities via G4EmCalculator
    or directly from the physics models
       Physics List should be initialized
   Example for retrieving the total cross section (cm-1) of a
    process with name procName: for particle partName and
    material matName
G4EmCalculator emCalculator;
G4Material* material =
  G4NistManager::Instance()->FindOrBuildMaterial(“matName);
G4double massSigma = emCalculator.ComputeCrossSectionPerVolume
   (energy,particle,procName,material);
G4cout << G4BestUnit(massSigma, "Surface/Volume") << G4endl;

A good example:
   $G4INSTALL/examples/extended/electromagnetic/TestEm14

								
To top
;