Neutron Detector Simulations for Fast Neutrons with GEANT4

1
Neutron Detector Simulations for Fast Neutrons
with GEANT4

• Brian Roeder
• LPC Caen

2
Motivation

• Want to design a next generation neutron
  detector array.
• Array to detect neutrons over a large energy
  range (1 MeV lt Enlt 150 MeV) with good energy and
  angular resolution.
• Would like the capability to detect multiple
  neutron events for breakup experiments and
  ?-multiple N decays.

3
Development Issues

• Response of detector modules
• Detection efficiency (materials)
• ?-N discrimination
• Minimization of Cross Talk
• Geometry of array setup
• Development of offline cross talk rejection
• Optimization of detector array for different
  types of experiments

4
Monte Carlo Sims. With GEANT4

• GEANT4 has the following advantages
• Built-in, 3D Visualization of detector and
  events!
• Tracking of all particles before and after
  detection (important for Cross Talk simulation).
• Realistic design of detectors with all materials
  easy to change materials without rewriting whole
  simulation.

5
Overview of Sim. Validation

• Create a simulation in GEANT4 of an existing
  detector module with measured detection
  efficiency DEMON
• Test neutron scattering models provided by GEANT4
  vs. DEMON efficiency and other simulations (e.g.
  MENATE).
• If necessary, modify existing scattering models
  in GEANT4 to obtain a more realistic simulation

DEMON I. Tilquin et al. NIM A365, (1995), Pg.
446
MENATE P. Désesquelles et al., NIM A307 (1991)
366
6
The MARGOT simulation

• Tube of BC-501A liquid scintillator with DEMON
  module dimensions
• Can measure energy deposited, position of neutron
  hit, and time of flight
• Currently converts energy loss of ions by hand
  into electron-equivalent (MeVee) energy using
  Cec79 equations.
• Tracks all particles produced in neutron
  interactions.
• Can produce different neutron beams
• Detection Threshold set at 0.5 MeVee.

R.A. Cecil et al., NIM 161 (1979) Pg. 439
7
Standard GEANT4 neutron elastic scattering

• LEElastic scattering model

Points Cyril Varignon these LPC Caen (1999)
8
Improvement with JENDL data set
9
Addition of Standard Inelastic Model

• Corrects for drop in efficiency after 20 MeV.

10
Evaluation

• Over-estimates efficiency between 2 and 5 MeV.
• Under-estimates efficiency after 25 MeV.
• Does not produce heavy-ion residuals, gammas from
  (n,n?) events. Seems to randomly produce d,t,?,
  etc.
• Would like a more-realistic simulation with the
  KNOWN organic scintillator reactions!

11
Inelastic Neutron Scattering for Engt4 MeV

• Mostly inelastic scattering, transfer, and
  breakup reactions on 12C.
• The GOOD 12C(n,p)12B and 12C(n,?)9Be
• The BAD 12C(n,3?)n and 12C(n,np)11B
• The UGLY 12C(n,n ?) and 12C(n,2n)11C
• Cross sections for these reactions are known,
  have been used in other simulations e.g. MENATE.

A. Del Guerra NIM 135 (1976) Pg. 337
M. Labiche these LPC Caen (1999)
12
GEANT4 NeutronHP model

• Includes inelastic reactions specifically with
  reference to total cross sections.
• Data and parameterizations available from thermal
  energies to 20 MeV with the G4NDL.
• With a few modifications, can be extended to
  100 MeV by adding cross section data from MENATE
  ? LEFastNeutron Model

13
Comparison MENATE vs. GEANT4 LEFastNeutron
14
Results of LEFastNeutron model
15
What about the 1 to 4 MeV efficiency?
16
Results

• Have developed a Monte Carlo simulation in GEANT4
  for BC-501A scintillator.
• Reasonable agreement between simulation and
  measured efficiency
• More realistic representation of inelastic
  reactions with LEFastNeutron model

17
Future Work

• Test how GEANT4 simulates angular distributions
  of scattered neutrons
• Study Cross-Talk for different neutron
  energies.
• Develop setup of future array and test cross
  section rejection schemes.

18
Special Thanks

• Marc Labiche for initial help with GEANT4.
• Franck Delaunay for providing the MENATE
  calculations presented.
• Other collaborators Nigel Orr, Miguel Marqués,
  Benoit Laurent, writers of GEANT4.
• We acknowledge the financial support of the
  European Community under the FP6 "Research
  Infrastructure Action-Structuring the European
  Research Area" EURISOL DS Project contract no
  515768 RIDS. The EC is not liable for the use
  that can be made of the information contained
  herein.

19
FIN
20
Discussion

• Is there some other way to improve result at low
  energy other than conic beam spot or adjustment
  of cross sections?
• Is it proper to leave out alpha reactions to
  improve simulation of efficiency?
• Suggestions from meeting participants on how this
  simulation can be improved.