Enhancing the performance of a tensioned metastable fluid detector based active interrogation system for the detection of SNM in <1 m 3 containers using a D–D neutron interrogation source in moderated/reflected geometries

This paper describes the development of a SNM detection system for interrogating 1m³ cargos via the combination of a D–D neutron interrogation source (with and without reflectors) and tensioned metastable fluid detectors (TMFDs). TMFDs have been previously shown (Taleyarkhan et al., 2008; Grimes et al., 2015; Grimes and Taleyarkhan, 2016; Archambault et al., 2017; Hagen et al., 2016) to be capable of using Threshold Energy Neutron Analysis (TENA) techniques to reject the $\sim$2.45 MeV D–D interrogating neutrons while still remaining sensitive to >2.45 MeV neutrons resulting from fission in the target (HEU) material. In order to enhance the performance, a paraffin reflector was included around the accelerator head. This reflector was used to direct neutrons into the package to increase the fission signal, lower the energy of the interrogating neutrons to increase the fission cross-section with HEU, and, also to direct interrogating neutrons away from the detectors in order to enhance the required discrimination between interrogating and fission neutrons. Experiments performed with a ²³⁹ Pu–Be neutron source and MnO₂ indicated that impressive performance gains could be made by placing a parabolic paraffin moderator between the interrogation source and an air-filled cargo container with HEU placed at the center. However, experiments with other cargo fillers (as specified in the well-known ANSI N42.41-2007 report), and with HEU placed in locations other than the center of the package indicated that other reflector geometries might be superior due to over-"focusing" and the increased solid angle effects due to the accommodation of the moderator geometry. The best performance for the worst case of source location and box fill was obtained by placing the reflector only behind the D–D neutron source rather than in front of it. Finally, it was shown that there could be significant gains in the ability to detect concealed SNM by operating the system in multiple geometric configurations. Worst case scenarios were created by filling the box with hydrogenous material and placing the HEU as far away as possible from the neutron source. The performance of the system in the worst-case scenarios were greatly improved by exchanging the location of the accelerator and the opposite TMFD panel half way through interrogation. Using this operation, scenarios with positions of the concealed SNM that were once the most challenging to successfully detect became readily detectable.