Tony H. Shin

Neutron cross-talk characterization of liquid organic scintillators

Correlated neutrons from a Cf-252 spontaneous fission source were measured with a coincidence detection system. The system consisted of three EJ-309 cylindrical liquid scintillators (length and diameter of 7.62 cm) configured to quantify the effects of neutron cross-talk on the total observed correlated counts. Because scatter-based detectors are susceptible to neutron cross-talk, the discrimination between true correlated counts and cross-talk is significant for neutron coincidence measurements. This measurement aimed to mitigate the neutron cross-talk between detectors with adequate polyethylene shielding, and was compared to simulations. The fractional influence of cross-talk on the total observed correlated counts was 30% for neutrons above ~0.65 MeV at a detector-source-detector angle of 30°. Furthermore, the fractional influence of neutron cross-talk increases for decreasing detector-source-detector angles at a constant detector standoff. The angular distribution of the scattered neutron was also simulated, and shows an anisotropic distribution in the scatter detector frame of reference primarily dependent on kinematics of the neutron-hydrogen scatter and the detector volume.

Neutron Multiplicity Counting Moments for Fissile Mass Estimation in Scatter-Based Neutron Detection Systems

Abstract Neutron multiplicity counting (NMC) techniques are widely used for nuclear materials accountability and international safeguards applications to quantitatively evaluate characteristic properties pertaining to fissile material. Mathematical models for NMC moments have been previously derived for systems that use capture-based detectors; however, these models are not applicable when scatter-based detectors are used because of “neutron cross talk.” Neutron cross talk occurs when a single neutron scatters and deposits energy above threshold into multiple detectors causing spurious increase in multiplicity counts; this, in turn, has caused fissile mass to be overestimated when not treated. In this paper, we propose new mathematical models derived from point kinetics to correct for neutron cross-talk effects up to any arbitrary order N, where N denotes the maximum number of counts a single neutron can cause. The new models were used to estimate the fissile mass of plutonium metal and oxide samples with effective 240Pu mass ranging from 2.5 to 250 g. The adequacy of the models was confirmed using simulations of a conceptual scatter-based neutron multiplicity counter (e.g., organic scintillators) using MCNPX v2.7e with the PoliMi fission event generating extension. The fissile mass estimates with no correction for neutron cross-talk events yielded an average relative deviation from the true 240Pueff mass of 55.94% and 84.56% for metal and oxide samples, respectively. When neutron cross-talk events of order N = 2 are included in the model, the fissile mass estimates yielded an average relative deviation of 11.89% for metal and 13.21% for oxide samples. Accounting for neutron cross-talk events of order N = 3 resulted in fissile mass estimates with an average relative deviation of 9.58% and 10.51% for metal and oxide samples, respectively. These mass estimates were compared to a reference case (i.e., no neutron cross-talk effects) that yielded an average relative deviation of 6.81% and 4.77% for metal and oxide samples, respectively. The discrepancy between the estimates from the proposed model and the reference case is attributed to the assumed value of N, which sets a finite upper bound on the order of cross-talk events the model treats (i.e., the model for N = 3 assumes that a neutron will never cause more than three counts).

Fast multiplicity counter featuring stilbene detectors for special nuclear material assay

A fast neutron multiplicity counting system based on organic scintillators, i.e. EJ-309 and stilbene, has been developed and experimentally tested at the University of Michigan. The system is able to detect correlated photon and neutron multiplets emitted by the fission reaction, within a gate time of tens of nanoseconds. This counting strategy is exploited to quantify fissile mass, without the need of complex electronic circuitry and unfolding procedures, otherwise required in moderated systems. Moderated systems are traditionally based on helium-3 detectors and feature a gate time of hundreds of microseconds. Measurement precision is thus negatively affected by accidental coincidences. A prototypal version of the proposed well-shaped counter was assembled and tested in the laboratory, using a spontaneous fission and an (α, n) neutron source, i.e. 252Cf and PuBe respectively. Measured results show excellent agreement with the simulated model of the system. The viability of the system to discriminate time-correlated fission neutrons from random, uncorrelated neutrons was proved. Preliminary results of an experimental campaign, carried out at INL (Idaho National Laboratory), to characterize plutonium metal samples are also shown. Results show a monotonic increasing trend for the range of measured 240-Pu effective masses, i.e. 0.024-0.5 kg, enabling the measurement of the mass of an unknown sample.

Prompt neutron and gamma-ray correlations from Cf-252 spontaneous fission

New event-by-event fission models have prompt neutron and gamma-rays that are correlated in time, energy, and multiplicity, however there is limited measurement data available to validate these models. Measurement of high-order fission neutron and gamma-ray coincidences is difficult and there has previously been little motivation to measure properties of both particle types simultaneously. High-order Cf-252 spontaneous fission neutron and gamma-ray coincidences were measured with an array of 24 liquid organic and eight sodium iodide scintillation detectors. Measured coincidence data including neutron time-of-flight energy and measured gamma-ray pulse height distributions are compared with MCNPX-PoliMi simulation results from built-in and event-by-event fission models.

Plutonium metal spontaneous fission neutron cross-correlation measurements

A plutonium metal sample was measured by a fast-neutron multiplicity counter for characterization of spontaneous fission neutron anisotropy and for verification of MCNPX-PoliMi calculations. Accurate neutron angular distribution models are important to properly simulating fast neutron coincidence measurements for nuclear nonproliferation and safeguards applications. A majority of prompt neutrons are emitted from fully accelerated fission fragments; those neutrons carry momentum from the fission fragments, and thus an anisotropic neutron angular distribution is observed in the laboratory reference frame. The fast-neutron multiplicity counter was used with pulse shape discrimination techniques to produce neutron-neutron cross-correlation time distributions from spontaneous fission in a lead-shielded 0.84 g 240Pueff metal sample. Due to neutron anisotropy, the number of observed neutron cross-correlations varied as a function of angle between a detector pair and fission source. Fewer neutron correlations were observed at detector angles near 90 degrees, relative to higher and lower detector angles. Both the neutron coincident detections as a function of time difference and detector pair angle are compared with MCNPX-PoliMi calculations and show good agreement.