Validation of EM Part of Geant4 February 22, 2002 @ Geant4 Work Shop - - PowerPoint PPT Presentation

1

Validation of EM Part of Geant4

Tsuneyoshi Kamae/Tsunefumi Mizuno February 22, 2002 @ Geant4 Work Shop

2

Purpose and Plan of this Talk

• Particle Ionization Energy Loss in Matters
• Bethe-Bloch Formula (pp. 3-5)
• Landau Distribution (pp. 3-5)
• Range (p. 6)
• Processed related to electromagnetic shower
• Pair creation (pp. 8-9)
• Bremsstrahlung (p. 10)
• Moller scattering/BhaBha scattering (pp. 11-12)
• Comparison of shower profile with experimental data

(pp. 13-14)

• Comparison of shower profile with EGS4 (pp. 15-17)

We have validated EM processes in Geant4 important for gamma-ray satellite GLAST. Energy range of GLAST is 20 MeV - 300 GeV. So we need to validate physics processed down to ~10 MeV and lower.

• We use Geant4 2.0

Tracker (Si, Pb) Calorimeter (CsI)

3

Ionization Energy loss in matters

• Shoot proton/electron in Pb and Si.
• Cutoff Length:

0.4mm(e-), 0.1mm(others) Thick absorber: Bethe-Bloch formula Thin absorber: Landau distribution Check mean energy loss (thick absorber), shape and most probable energy loss (thin absorber).

4

Bethe-Bloch formula for protons

Geant4 well reproduces ionization energy loss in thick Si and Pb of proton down to 10 MeV

5

Landau Distribution for protons and electrons

Fluctuation of energy loss and the most probable value in thin Si and Pb are appropriate. protons electrons

6

Range of protons

Shoot protons of 200 MeV and 1 GeV in CsI, Pb and W Show good agreement within 0.6%. Compare the results with NIST data:

(http://physics.nist.gov/Ph ysRefData/Star/Text/conte nts.html)

7

Processes related to EM shower

• Pair creation (pp. 8-9)

(and compton scattering and photo-electric effect)

• Bremsstrahlung (p. 10)
• Moller scattering (pp. 11-12)
• Bhabha scattering (pp. 11-12)

We need to validate these processes, especially in low-energy, since

• Pair creation and EM shower are key

processes to determine gamma-ray energy and direction in the GLAST.

• Low energy electrons may suffer large-angle

scattering and cause trigger in the GLAST.

e- Tracker (Si-Strip detectors) Calorimeter (CsI scintillator)

• Cutoff

0.04mm(e-) 0.01mm(others)

8

Pair creation – cross section

Geant4 correctly calculates cross section down to ~100 keV Compare the cross section in Pb calculated by Geant4 with that of a reference

(http://physics.nist.gov/Phy sRefData/Xcom/Text)

9

Pair creation-angular distribution of emitted electrons

• We found discrepancies for low

energy electrons and fixed the code e- 4 MeV

0.125 0.25 0.5 0.375 Theta [rad] Events/Bin Events/Bin 0.125 0.25 0.5 0.375 Theta [rad]

e- 18 MeV

• Shoot 20 MeV gammas into Pb

absorber (10% RL)

• Compare the angular distribution
• f generated e- with that of

theoretical formula

10

Bremsstrahlung -- angular distribution of generated photons

Geant4 well reproduces the angular distribution

• f photons generated

via bremsstrahlung down to ~1 MeV

• Shoot 20 MeV e- into Pb absorber

(10% RL)

• Compare the angular distribution
• f generated photons with that of

theoretical formula gamma 2 MeV gamma 6 MeV gamma 10 MeV

11

Moller/Bhabha scattering (scattering with e-/e+)

• Shoot e- of 20 MeV and 100 MeV

into Pb absorber (1% RL)

• Compare the angular distribution
• f scattered electrons with that of

theoretical formula

12

Moller scattering (with e-) G4 well reproduces the theoretical formula down to 20 MeV 100 MeV Bhabha scattering (with e+) 20 MeV 20 MeV 100 MeV Geant4 Theoretical furmula

13

Vacuum

Layer 0 Layer 3 Layer 11 Material : Water

28 cm

top view side view 10 rings in each layer 11 Water Layers electron

20 cm 400 cm

Layer 4

40 cm Ring ID width

0 ~ 3 1 cm 4, 5 2 cm 6 ~ 10 4 cm

Ring 10 Ring 0 Layer ID thickness 0 ~ 3 20 cm 4 ~ 11 40 cm

EM shower profile – Crannell experiment (1)

• Cranell experimtns:

Carol Jo Crannell, et al., Phys. Rev. 184 (1969) 426

• Compare the G4 results

with that of experimental data and EGS4 (Professor Nelthon and Liu).

• shoot 1GeV e- into water tank

(10 rings and 11 layers)

14

EM shower profile – Crannell experiment (2)

Shower profile of G4 is narrower than that of experiment and EGS4 longitudinal profile lateral profiles

15

EM shower in Pb layers (1)

To investigate the EM shower profile in the GLAST Tracker, we constructed 10 Pb layers(0.1RL each) with air (3cm). The radius is 0.2 Moliere radius (core of shower)

20 MeV, longitudinal profile

Geant4 EGS4

lateral profiles

Pb air

16

EM shower in Pb and air (2)

500 MeV, longitudinal profile lateral profiles 5 GeV, longitudinal profile lateral profiles

Geant4 EGS4 Geant4 EGS4

• Energy deposition is higher in Geant4 (low energy region)
• Geant4 shows narrower shower profile (and could not be

solved by fixing the angular distribution of pair creation)

17

Effect of cutoff energy (length)

0.4mm, 0.04mm (e-), 0.1mm, 0.01mm(others)

• The discrepancy can

not be attributed to cutoff length (energy)

Geant4 EGS4

The reason of the narrower shower profile is unrevealed.

18

Summary

• We have validated EM part of Geant4 (2.0).
• Energy loss in matters is well simulated down to

~10 MeV.

• Angular distribution of pair-created e- was found to

be inappropriate and fixed.

• Cross section of pair creation, angular distribution
• f bremsstrahlung/Moller scattering/Bhabha

scattering were validated down to ~10 MeV.

• Shower profile of G4 is slightly narrower than that
• f experimental data and EGS4. The reason of this is

unknown.