Simulation Study to Develop Spatial Multiplication Model in Neutron Multiplicity Counting

The point model (PM) is commonly used in neutron multiplicity counting to relate the correlated neutron detection rates (singles, doubles, triples) to item properties (mass, (?,n) reaction rate and neutron multiplication). The PM assumes the probability that a neutron will induce fission is a constant across the physical extent of the item. However, in reality, neutrons near the center of an item have a greater probability of inducing fission than neutrons from the edges. As a result, the neutron multiplication has a spatial distribution. The form of this distribution is assumed within the approach being developed and partially described in this paper, referred to as the Spatial Multiplication Model (SMM). While the singles rate is a function of the average neutron multiplication and is notaffected by the spatial distribution, the higher order detection rates (doubles, triples, etc.) are functions of the neutron multiplication raised to an integer power. In the PM, where no spatial distribution of neutron multiplication is assumed, these values are estimated by raising the average neutron multiplication to a power. However, since in reality the neutron multiplication exhibits significant spatial variation across an item, the average value of the neutron multiplication raised to a power is not equal to the average of the neutron multiplication distribution raised to a power. In order to correct this limitation of the PM, within the SMM the specified spatial multiplication distribution is first raised to integer powers and then averaged. The resulting values are used in the existing PM equations in place of the traditional neutron multiplication terms.The preliminary results have yielded significantimprovement in terms of accuracy of reconstructed assayed item mass.The focus of this paper is on high-fidelity neutron transport simulations that were developed to address one of the crucial elements of the SMM – evaluating the form ofthe spatial distribution of multiplication for several cases of Pu metal cylinders assayed by an Active Well Coincidence Counter (AWCC). The shape of this dependence serves as an input into the SMM. We also present preliminary quantitative simulated comparison of PM and SMM for select cylindrical shaped Pu-metal samples in AWCC.