CHANDLER Detector�Neutronics Modeling�

Alireza Haghighat

William Walters

​

Nuclear Science and Engineering Lab (NSEL)

Nuclear Engineering Program, Mechanical Engineering Dept.

Virginia Tech Research Center

Arlington, VA

​

Applied Antineutrino Physics 2015, Dec 7-8, 2015

1

Outline

• CHANDLER detector overview
• Inverse beta-decay neutron modeling
• Cosmic ray neutron modeling
• Spatial coincidence of neutron signals
• Cosmic neutron shielding

2

CHANDLER Detector

• 1x1x1 m³ plastic scintillator
• 16 layers of 16x16 scintillator cubes
• 17 layers of LiF-ZnS(Ag) (absorber & scintillator) neutron detector (ND) sheets between each layer
• Charge particles absorb in scintillator cubes
• Neutrons absorb in LiF-ZnS(Ag) sheets
• Each event can be localized to a single cube

3

3×3×3 microCHANDLER Prototype The open face clearly shows the optics of total internal reflection.

Detector Geometry

4

EJ-260 Scintillator

EJ-426 ND Scintillator

Polyethylene backing

Materials

5

Nlib=.80c ENDF-VII.1 pointwise cross sections, room temperature

Modeling IBD Neutron Source

6

Neutron angular dependency

Neutron Spectrum

IBD Neutron Modeling

7

Cosmic ray-generated neutrons

• Fast neutrons from cosmic rays can be a problem
• Fast neutron scatters in hydrogen
• Recoil proton deposits energy in scintillator, “looks” like a positron
• Neutron absorbs in Li-6 detector sheet

​

• These neutrons are much higher energy
• Large distance between proton absorption and neutron absorption
• How well is this distinguished from an IBD neutron-positron pair event

​

• Need to achieve a good signal-to-noise ratio (SNR)

8

Analyzing and Shielding of Cosmic ray

9

MCNP6 Atmospheric Model

• 65 km of atmosphere
• NO soil modeled
• 150 air cells with varying density and humidity (from USGS data)
• MODE n h p q g d t s a | z / k (13 particles)
• Manual cell importance
• Source
• Protons and Alphas
• Defined by the par=cg cosmic ray option
• Considering three parameters including lattitude, longitude, and date
• Tallies:
• Neutron flux as function of altitude
• Neutron current at sea level

​

​

10

Atmospheric Neutron Flux

11

Air Density

Neutron Flux

Ground level Neutron spectrum

12

Energy Spectrum

Integrated Current (bin values)

CHANDLER Coincidence Modeling

• Goal - model the coincidence from a single fast neutron:
• Recoil proton absorbed in scintillator
• Neutron absorbed in neutron detector
• Find the time and spatial correlation of these absorptions

​

• Source
• Neutron energy from previous step (average 107 MeV)
• Angular distribution from previous step (peaked in –z direction)

​

• PTRAC output
• Outputs individual particle events
• Processed using an in-house code

​

13

Neutron-proton coincidence

• Filter ptrac for events with:
• Single proton recoiled
• <6 MeV total energy
• Single neutron absorbed in detector material

​

• 0.54% of cosmic neutrons result in this coincidence
• Cosmic neutrons cause ~17,000 events/day; compared to ~500 antineutrino events/day

​

• Spatial and time correlation are shown
• Need to use this to reduce noise further

14

IBD vs. cosmic fast neutron coincidence

15

Fast neutron coincidence are more spread out, but still significant overlap

IBD vs. Cosmic fast neutron coincidence

16

Cosmic Fast Neutrons

IBD Neutrons

Time cut will not be very effective, while space cut can be

Spatial Correlation Anisotropy

• Cosmic neutrons – z directionality (from space), IBD neutrons – x directionality (from reactor)

​

• Z direction: +1 means the neutron is absorbed in the layer above the cube where the proton is absorbed, -1 is below
• i.e., there is no 0 for the z difference

​

• X/Y directions: 0 means the neutron is absorbed in a layer at the same x/y cube as the proton

​

​

17

IBD Neutron X-Bias

Cosmic fast neutron Z-Bias

Layers of ND

Cosmic ray

Spatial Cut

• Look at individual segments of the neutron detector (x, y, z)
• Order segments by their SNR
• Only accept segments with a high enough SNR

​

• Highest SNR
• +X (IBD neutrons peaked in this direction)
• +Z (Cosmic neutrons peaked away from this direction)

​

18

(0,0,1) means the neutron is absorbed in the layer immediately above the cube in which the proton is absorbed

​

Signal vs. noise tradeoff

• Add more segments inside the cutoff -> higher total signal, lower SNR (noise increases faster than signal)

19

Shielding of fast neutrons

• Previous results without shielding for fast neutrons gives a very low SNR for neutrino events (<0.5)
• Add layers of high density polyethylene (HDPE)
• How much shielding is required to cut the fast neutron signal to an acceptable level?
• 1-D model in MCNP6, tallying neutron current with different thickness of HDPE

20

Shielding of fast neutrons

• Decreases average energy slightly
• Decreases total current significantly (~2% at 1m of HDPE)

21

Conclusions & Future Work

• Cosmogenic neutrons account for a significant source of noise in the CHANDLER detector, unless shielded
• Depending on spatial cutoffs, an SNR (including cosmic neutrons as the only source of noise) of 0.2-0.5 can be obtained for the unshielded case
• With 1m of HDPE shielding, the cosmic fast neutron current can be reduced by ~98%; i.e., SNR increases by a factor ~50; therefore, a max. SNR can be ~25
• Further Analysis
• No account for any additional neutrons created in the shield from cosmic muons (ideally, the muon detectors would detect and veto these events)
• No account for the change in fast neutron coincidence with the changing neutron spectrum from shielding
• No account of gamma rays
• An optimum multi-layered (e.g., poly, Boron, lead) should be designed
• Experimental benchmarking against reactors (BR2 & North Anna)

22

Thanks!

Questions?

23

Appendix - PTRAC output file

24

Source particle #

Event type:

1000 src

20XX bank

5000 termination

9000 done

​

x,y,z

particle type

(1 =n, 9=p)