The GEANT SimulationPackage and its use inCompton Telescope Design
The GEANT SimulationPackage and its use inCompton Telescope Design
R. Marc KippenSpace and Remote Sensing Sciences GroupLos Alamos National Laboratory
Astronomy with Radioactivities IV May 26–30, 2003 – 2 –R. Marc Kippen
The Role of Simulation in DesignThe Role of Simulation in Design
+ =
: Realistic componentperformance
6 Expensive and timeconsuming
6 Inflexibleconfiguration
6 Unrealisticenvironment
: Inexpensive andcomparatively rapid
: Flexibleconfiguration
: Flexible environment6 Must model the
componentperformance
MEGA Prototype
Prototypes, Balloons, etc.
MEGA Prototype Simulation Model
Simulations, Models, etc.
MEGA Flight Concept
Scientific Mission
n Successful flightexperiment, where:® Realistic estimates
of performancehelp “sell” themission
® Instrument designis optimized forscientific missionand environment
Astronomy with Radioactivities IV May 26–30, 2003 – 3 –R. Marc Kippen
Simulating Compton TelescopesSimulating Compton Telescopes
n Analytical modeling ofCompton imager physicalresponse is impractical due tocomplexities of geometry,scattering, and secondaryproduction
n The most viable approach isMonte Carlo radiation transportsimulation — probabilistictracking individual “testparticles”
n Other simulations important toinstrument design: mechanical,thermal, electronics, etc.
PhotonsElectrons
Astronomy with Radioactivities IV May 26–30, 2003 – 4 –R. Marc Kippen
Instrument Simulation FrameworkInstrument Simulation Framework
ScienceGoal
Inputs
BackgroundInputs
OrbitalEnvironment
Model
PhysicalSimulation
Engine
InstrumentMass Model
SpacecraftMass Model
Test Dataand/orModels
InstrumentalEffects Engine
Data Processingand Analysis
Mechanical Model(s)
Physics Data/Models
SciencePerformanceEvaluation
AuxiliaryData/Models
Iterate at All Levels
Credible Simulation Requires Credible Inputs at All Levels
Astronomy with Radioactivities IV May 26–30, 2003 – 5 –R. Marc Kippen
n Requirements for ComptonTelescope simulations:
® Detailed electromagnetic physicsfor direct telescope response(~1 keV – 100 MeV)
® Competent hadronic cascadephysics for simulation of promptcosmic-ray–induced background
® Isotope excitation and radioactivedecay for simulations of delayedactivation-induced background
® Convenient and flexible handlingof complex geometry andmaterials for rapid design studies
® Modern, modular architecturethat allows customization
Monte Carlo Radiation Transport PackagesMonte Carlo Radiation Transport Packages
n The particle and nuclearphysics communities havedeveloped several “general-purpose” Monte Carlo transportpackages, including:
® EGS® FLUKA® HETC/MORSE/MICAP® CALOR® MCNP/MCNPX® GEANT/GEANT4
Astronomy with Radioactivities IV May 26–30, 2003 – 6 –R. Marc Kippen
Capabilities of GEANT4Capabilities of GEANT4
n GEANT := GEometry And Tracking
n Complex 3D geometry, materials,MC transport, and visualization inone package
n Developed & maintained by CERN+ large collaboration
n Modern, object-oriented (C++)“toolkit” architecture
n Comprehensive (nearly) suite of EMand hadronic physics
n Straightforward installation and useon many platforms® Wintel, Sun, HP, Linux, Darwin
n ESA Space Specific Modules
n General Source Particle Module® Tookit for input spatial/spectral sampling
n Radioactive Decay Module® Provides the capability to model
activation-induced background inorbit
® Uses detailed Evaluated NuclearStructure Data Files
n Low-energy EM physics® Uses detailed cross sections from LLNL
Evaluated Photon/Electron/Atomic DataLibraries
® Applicable above ~250 eV® Ties X-ray and Gamma-ray applications
´ Important omission: electronbinding effects in Compton
geant4.web.cern.chgeant4.web.cern.ch www.space.qinetiq.comwww.space.qinetiq.com
Astronomy with Radioactivities IV May 26–30, 2003 – 7 –R. Marc Kippen
Effects of Atomic Electron BindingEffects of Atomic Electron Binding
n Suppresses forward scattering,particularly at low energies
n Suppresses total scatteringprobability at low energies
0.4
0.3
0.2
0.1
0.0
sto
t (cm
–1)
10 100 1000 10000
Photon Energy (keV)
Atomic Silicon
Free e- Atomic e-
0.08
0.06
0.04
0.02
0.00
ds/d
W (b
arn
· sr–1
)
1801501209060300
Photon Scatter Angle (deg)
Atomic Silicon
100 keV
Free e- Atomic e-
n GEANT4 Low-energy Compton process includes these effects
†
d2sdWdk
Ê
Ë Á
ˆ
¯ ˜
i
=ro
2
4k f kko
2
Ê
Ë Á
ˆ
¯ ˜
k f
ko+
kok f
- sin2 jÊ
Ë Á Á
ˆ
¯ ˜ ˜
dpz
dkJi pz( )
Astronomy with Radioactivities IV May 26–30, 2003 – 8 –R. Marc Kippen
Doppler Broadening Physics & EffectsDoppler Broadening Physics & Effects
1.0
0.8
0.6
0.4
0.2
0.0
Norm
alize
d Av
erag
e J(
Q)
-30 -20 -10 0 10 20 30Bound Momentum Q (keV/c)
Compton ProfilesJ(Q)
CZTHWHM = 4.5 keV/c
SiliconHWHM = 2.1 keV/c
k = ko -kokEo
1 - cosj( ) - pz ko - k
For bound atomic electron:
kfree = ko -kok
moc2 1 - cosj( )
For free electron: pz = 0 ; Eo = moc2
†
fi Dk = k - kfree; Dj = j -j free
Doppler broadening error:
jko
k
Eo, pz
E, p
Biggs,Mendelsohn,& Mann (1975)
14
12
10
8
6
4
2
0
Scat
ter A
ngle
Erro
r (de
g)
1801501209060300Scatter Angle (deg)
CZT Silicon
100 keV
1 MeV
Astronomy with Radioactivities IV May 26–30, 2003 – 9 –R. Marc Kippen
GLECS & G4LECSGLECS & G4LECS
n GLECS = GEANT Low-Energy Compton Scattering® Thanks to Doug Swartz (USRA, Huntsville) for early help
n Incorporates Doppler broadening into GEANT & GEANT4
n Algorithm based closely on EGS Implementation® Namito, Ban, & Hirayama, NIM A349, 489 (1994)® Relativistic impulse approximation (ignore atomic electron interactions)® Uses EPDL for total cross sections® Uses EPDL differential cross sections (scattering form factors)® Uses shellwise Compton profiles (Biggs, Mendlesohn, & Mann 1975) to
sample Doppler broadened scattered photon energies® Also fixes Rayleigh (coherent) scattering physics with EPDL data® Computing performance within 5% of G4LowEnergy classes
n Soon to come: combined polarization and Doppler broadening
Astronomy with Radioactivities IV May 26–30, 2003 – 10 –R. Marc Kippen
Verification of G4LECSVerification of G4LECS
n G4LECS compared to synchrotron beam experiment® Namito, Ban, Hirayama, et al. (1994, 1995)
Experiment(Polarized Beam)
Simulation(Unpolarized Beam)
Astronomy with Radioactivities IV May 26–30, 2003 – 11 –R. Marc Kippen
Test ResultsTest Results
10-6
10-5
10-4
10-3
10-2
Coun
ts (p
eak
norm
alize
d)
403836343230
Scattered Energy Deposition [keV]
Data (Namito et al.) G4LECS Simulation G4LowEnergy Simulation
v4.4.1
Carbon 40 keV
10-6
10-5
10-4
10-3
10-2
Coun
ts (p
eak
norm
alize
d)
403836343230
Scattered Energy Deposition [keV]
Data (Namito et al.) G4LECS Simulation G4LowEnergy Simulation
v4.4.1
Copper 40 keV
10-6
10-5
10-4
10-3
10-2
Coun
ts (p
eak
norm
alize
d)
403836343230
Scattered Energy Deposition [keV]
Data (Namito et al.) G4LECS Simulation G4LowEnergy Simulation
v4.4.1
Lead 40 keV
n Good agreement in Compton andRayleigh peaks (and Ge-K escape)
n Some differences in multi-Compton continuum probably dueto approximated geometry
Astronomy with Radioactivities IV May 26–30, 2003 – 12 –R. Marc Kippen
Application to Compton Telescope DesignApplication to Compton Telescope Design
Doppler Limit Angular ResolutionZoglauer & Kanbach, Proc. SPIE 4851, 1302 (2003)
Astronomy with Radioactivities IV May 26–30, 2003 – 13 –R. Marc Kippen
Telescope Design Study ExampleTelescope Design Study Example
Si + CZTmulti-scatterDesign
14121086420An
gula
r Res
. (FW
HM, d
eg)
10 100 1000 10000Energy (keV)
100¥100 cm2
NoDoppler
10-3
10-2
10-1
100 E
2 • Se
nsitiv
ity (k
eV ·
cm-2 s
-1)
10 100 1000 10000Energy (keV)
1 Msec
BATSECOMPTEL
New Design
Astronomy with Radioactivities IV May 26–30, 2003 – 14 –R. Marc Kippen
http://nis-www.lanl.gov/~mkippen/actsim/
http://nis-www.lanl.gov/~mkippen/actsim/