Ion Transport Simulation using Geant4 Hadronic Physics

Monte Carlo 2005, April 20, 2005 at Chattanooga

Ion Transport Simulation using Ion Transport Simulation using Geant4 Hadronic PhysicsGeant4 Hadronic Physics

Ion Transport Simulation using Ion Transport Simulation using Geant4 Hadronic PhysicsGeant4 Hadronic Physics

Koi, TatsumiKoi, TatsumiSLACSLAC

And Geant4 Hadronic Working GroupAnd Geant4 Hadronic Working Group


Contents• Cross sections

– NN total reaction formulae• Reactions

– Binary Cascade Light Ion– QGS Glauber

• Validation– Neutron Productions– Pion Productions– Neutron Yields– etc

• Conclusions


G4HadronicProcessGetMicrocopicCrossSection()

PostStepDoIt()

ModelsApplyYoursel()

Cross SectionsGetCrossSection()

When and where

an interaction will occur?

What will be generated

by this interaction?


Cross Sections• Total reaction cross section is defined by

• Many cross section formulae for NN collisions are included in Geant4– Tripathi, Shen, Kox and Sihver

• These are empirical and parameterized formulae with theoretical insights.

• G4GeneralSpaceNNCrossSection was prepared to assist users in selecting the appropriate cross section formula.

EMdisElTotR


References to NN Cross Section Formulae

implemented in Geant4• Tripathi Formula

– NASA Technical Paper TP-3621 (1997)• Tripathi Light System 　 (p, n ~ alpha)

– NASA Technical Paper TP-209726 (1999) • Kox Formula

– Phys. Rev. C 35 1678 (1987)• Shen Formula

– Nuclear Physics. A 49 1130 (1989)• Sihver Formula

– Phys. Rev. C 47 1225 (1993)


Inelastic Cross SectionC12 on C12


Models• Binary Cascade Light Ion

related talk “The Binary Cascade” by H. P. Wellisch

• QGS Glauberrelated talk “Parton String Models In GEANT4” by G. Folger


Binary Cascade ~Model Principals~

~related talk “The Binary Cascade” by H. P. Wellisch~

• In Binary Cascade, each participating nucleon is seen as a Gaussian wave packet, (like QMD)

• Total wave function is assumed to be direct product of these. (no anti-symmetrization)

• Participating means that they are either primary particles, or have been generated or scattered in the process of the cascade.

• This wave form have same structure as the classical Hamilton equations and can be solved numerically.

• The Hamiltonian is calculated using simple time independent optical potential. (unlike QMD)

• Collisions between participants are not considered. (unlike QMD)

xtip

tqxLLtpqx ii

ii 2

43

2exp2,,,


Binary Cascade ~nuclear model ~• 3 dimensional model of the nucleus is construc

ted from A and Z.• Nucleon distribution follows

– A>16 Woods-Saxon model– Light nuclei harmonic-oscillator shell model

• Nucleon momenta are sampled from 0 to Fermi momentum and sum of these momenta is set to 0.

• time-invariant scalar optical potential is used.


Binary Cascade ~Light Ion Reactions~

• Two nuclei are prepared according to this model (previous page).

• The lighter nucleus is selected to be projectile.• Nucleons in the projectile are entered with position a

nd momenta into the initial collision state.• Until first collision of each nucleon, its Fermi motion i

s neglected in tracking.• Fermi motion and the nuclear field are taken into acc

ount in collision probabilities and final states of the collisions.


Validation resultsNeutrons from 290MeV/n C12 on

Carbon

Iwata et al.,Phys. Rev. C64 pp. 05460901(2001)



Copper



Dual parton or quark gluon string model

– hadron hadron scattering-related talk “Parton String Models In GEANT4” by G. Folger

• In the approach based on the topological expansion, the Pomeranchuk pole is described by graphs of the cylindrical type, while the secondary Reggeons are described by planar graphs

• The planar case involves annihilation of valence quarks of the colliding hadrons, and a qq-bar string.


• In the cylindrical (Pomeron) case, the colliding hadrons simply exchange one or several gluons, resulting in color coupling between the valence quarks of the hadrons. They are connected by quark gluon strings.

• Breaking the strings leads to the 　　　　　　　　　　　 appearance of white hadrons.


Multiple Pomeron exchange

• The parameters of the Pomeron trajectory cannot at present be calculated, but are taken from fits to experimental data.

• (Ter-Martyrosian, Phys.Lett.44B,1973)


Hadron nucleus collisions• With respect to hadron hadron collisions, hadron

nuclear collisions offer the additional twist of multiple participating target nucleons.


Ion-ion reaction cross-sections.

• Ion-ion reactions simply add additional primary nucleon lines to the diagrams.

• The amplitudes calculated can be integrated to obtain reaction cross-sections for ion-ion collisions at high energies– From O(5A GeV) to O(10A TeV)– Predictions within about experimental errors.


Preliminary results of cross section predictions by QGS-Glauber

Prelim

inary Difference in Pb comes form mainlyEM dissociation effect

4.2 GeV/n C ions

156A GeV Pb ions

p C P C Pb


Summary of Cross Section and models for N-N Inelastic

Interaction in Geant4

Cross Sections

Models

Tripathi & TripathiLightSystem ~10 GeV/A

Kox & Shen ~10 GeV/A

Binary Cascade Light Ions 10 GeV/A

~100 MeV/A Sihver

~5 GeV/A QGS - Glauber

Energy 1 GeV 10 GeV100 MeV

QGS-Glauber is not yet included the release


Other Ion related processes already implemented in

Geant4 • Ionization Energy Loss which dedicated to Ions• Multiple Scattering

related talk “GEANT4 "Standard" Electromagnetic Physics Package” By M. Marie

• EM Dissociation• Abrasion-Ablation Model

– Macroscopic model for nuclear-nuclear interactionrelated talk “Implementation Of Nuclear-Nuclear Physics

In The GEANT4 Radiation Transport Toolkit For Interplanetary Space Missions” By P. Truscott

All these processes work together for Ion transportation in Geant4


Validations• Neutron Production

– Double Differential Cross Section– Angular Distribution

• Thick Target Neutron Yield• Pion Production• Fragment Production


Validation resultsNeutrons from 400MeV/n Ne20

on Carbon



on Copper


Validation resultsNeutrons from 560MeV/n

Ar40 on Lead


Quantitative comparison between

the measured and calculated cross sections

R = (σ calculate - σ measure ) /σ measure


Distribution of Rs Carbon Beams

C 400MeV/ n

- 100

0

100

200

0 20 40 60 80

Laboratory Angle [Degree]R

atio

%

C 290MeV/ n

- 100

0

100

200

0 20 40 60 80

Laboratory Angle [Degree]

Rat

io %

Target Materials


Overestimate

Underestimate

209


Distribution of Rs Neon Beams

Target Materials

Ne 400MeV/ n

- 100

0

100

200

0 20 40 60 80


Rat

io %

Ne 600MeV/ n

- 100

0

100

200

0 20 40 60 80


atio

%



Distribution of Rs Argon Beams

Ar 560MeV/ n

- 100

0

100

200

0 20 40 60 80


atio

%

Target Materials

Ar 400MeV/ n

- 100

0

100

200

0 20 40 60 80


Rat

io %



Distribution of Rsfor QMD and HIC Calculation

(done by original author)


100%


-100%Overestimate

Underestimate

R = 1/σ measure

x(σ measure -σcalculate )

QMD HIC


Validation results Pions from 1.05 A GeV/c C on Be, C, Cu and Pb

J. Papp, LBL-3633, (1975)


Thick Target Neutron Yield

• Thick target is a target which can stops incidence heavy ions completely.

• Not only a reaction model but also other ion related process of Geant4 are tested by this validation.


Neutron YieldArgon 400 MeV/n beams

Carbon Thick Target Aluminium Thick Target

T. Kurosawa et al., Phys. Rev. C62 pp. 04461501 (2000)


Neutron YieldArgon 400 MeV/n beams

Copper Thick Target Lead Thick Target



Neutron YieldFe 400 MeV/n beams

CarbonThick Target Aluminum Thick Target



Neutron YieldFe 400 MeV/n beams

Copper Thick Target Lead Thick Target



Fragment ProductionSi 490 MeV/ n on C

1

10

100

1000

Al Mg Na Ne F O N C

Particle Species

Cro

ss S

ectio

n [m

b]

DATAG4

Si 490 MeV/ n on H

1

10

100

1000

Al Mg Na Ne F O N C

Particle Species

Cro

ss S

ectio

n [m

b]

DATAG4

F. Flesch et al., J, RM, 34 237 2001


Fragment ProductionSi 453 MeV/ n on Al

1

10

100

1000

Al Mg Na Ne F O N C

Particle Species

Cro

ss S

ectio

n [m

b]

DATAG4

Si 490 MeV/ n on Cu

1

10

100

1000

Al Mg Na Ne F O N C

Particle Species

Cro

ss S

ectio

n [m

b]

DATAG4

F. Flesch et al., J, RM, 34 237 2001


The people involvedJ. P. Wellisch (CERN)G. Folger (CERN)B. Trieu (CERN)P. Truscott (ESA)I. Corneliu (INFN)


Conclusions• Now Geant4 has abundant processes for

Ion interactions with matter.• Without any extra modules, users may

simulate ion transportation in the complex and realistic geometries of Geant4

• Validation has begun and the first results show reasonable agreement with data. This work continues.



Carbon



Carbon



Copper



on Copper



Ne20 on Lead



on Carbon



Ne20 on Lead


Neutron YieldXe 400 MeV/n beams

CarbonThick Target Aluminum Thick Target


Neutron YieldXe 400 MeV/n beams

CopperThick Target Lead Thick Target


Neutron YieldSi 800 MeV/n beams

Carbon Thick Target Copper Thick Target

Documents

Ion Transport Simulation using Geant4 Hadronic Physics