Richard B. Williams

Neutron and gamma discrimination by compensation of long components in liquid scintillators

A technique for eliminating the long component for gamma pulse response in liquid scintillators has been developed. The analytical relationships between the values of the electrical components and the light response, approximated with two and three exponential components, have been derived. A simple reset preamplifier was built for acquiring and digitizing the charge pulse information. The neutron charge pulses have been measured at a time scale of 50 mus. The existence of a very long time-scale component (tens of microseconds) has been demonstrated for neutron pulses. An analysis of experimental data for pulse-shape discrimination is presented.

Electronics and signal processing for prompt radiation

Active interrogation with pulsed bremsstrahlung beams can saturate detectors and produce high count rates of overlapping pulses in the prompt region (<1 mus) after the interrogating pulse. We describe a method to eliminate saturation by modifying the photomultiplier voltage divider and by gating intermediate pairs of dynodes and the anode. To process the high count rate of overlapping pulses, we convert the output current pulses to charge-pulse steps that can be digitally processed more easily and rapidly in real time. We discuss the application to LaBr3, liquid, and plastic scintillators and present some preliminary data.

Development of a Liquid Scintillator Neutron Multiplicity Counter (LSMC)

A new neutron multiplicity counter is being developed that utilizes the fast response of liquid scintillator detectors. The ability to detect fast (vs. moderated) fission neutrons makes possible a coincidence gate on the order of tens of nanoseconds (vs. tens of microseconds). A neutron counter with such a narrow gate will be much less sensitive to accidental coincidences making it possible to measure items with a high single neutron background to greater accuracy in less time. This includes impure Pu items with high (alpha,n) rates as well as items of low mass HEU where a strong active interrogation source is needed. Liquid scintillator detectors also allow for energy discrimination between interrogation source neutrons and fission neutrons, allowing for even greater assay sensitivity. Designing and building a liquid scintillator multiplicity counter (LSMC) requires a symbiotic effort of simulation and experiment to optimize performance and mitigate hardware costs in the final product. We present preliminary Monte Carlo studies using the GEANT toolkit along with analysis of experimental data used to benchmark and tune the simulation.

The IAEA Universal Nondestructive Assay Data Acquisition Platform (UNAP)

The Universal NDA Data Acquisition Platform (UNAP) will be the next generation data acquisition system for the International Atomic Energy Agency (IAEA) attended and unattended non-destructive assay measurement equipment. The system will also be the principal data acquisition module for the Japan Nuclear Fuel Limited (JNFL) MOX Fuel Fabrication Plant (J-MOX) safeguards project. The primary goal of the UNAP development is for the new module to become the IAEA standard and replace existing Non-Destructive Assay (NDA) modules such as the JSR-12, the MiniGrand, the Advanced Multiplicity Shift Register (AMSR), and the JSR-14. The inputs to the UNAP will be a superset of the existing inputs of all these previous modules with flexibility to anticipate future developments.

In-the-flash pulse shape discrimination with liquid scintillators

Single neutron and gamma charge pulses were captured using a fast ADC. These data were used to populate simulated 5-µs active interrogation flash events. Analysis of these simulations revealed that long pulse tail accumulation during the flash leads to an unpredictable charge background for several microseconds after the flash. The existence of this changing background will make retrieving prompt neutron information immediately after the flash problematic. The tail accumulation is accentuated for every pulse in the flash, making the time just after the last pulse in the flash the worst possible time to measure prompt neutron data. Obtaining prompt neutron data during the flash—particularly near the beginning of the flash, where background tail accumulation is minimal—is feasible if shielding is used to control the count rates, but requires fast gamma-neutron pulse shape discrimination. A means of accentuating the pulse shape difference between neutron and gamma data is given, along with a means of compensating for the long pulse components in the dominant gamma pulses.