Ray effects mitigation through Monte Carlo coupling for detector problems

This work seeks to expand upon an existing method of mitigating ray effects through coupling deterministic transport methods with Monte Carlo results for low collision-count flux. Previous work relied on coupling the uncollided solution from Monte Carlo with deterministic methods in order to reduce ray effects in problems suffering this undesirable phenomena. It is shown in this work that this low level of coupling is not always sufficient to mitigate ray effects to an acceptable level. This work builds on that concept by introducing a formalism for a general collision count coupling method. Ray effects prove to be a major deficiency of deterministic codes when dealing with relatively low scattering materials. These materials include, among other things, near void air regions prevalent in detector problems. These regions are often large in extent and important in nonproliferation and detector problems compromising the primary buffer separating the source from the detector. Proper treatment of these regions is imperative to the goal of accurately solving transport problems of this type. In this work we show that the ray effects in these regions are severe in nonproliferation and detector problems, and that those effects can be effectively mitigated through collisional source coupling with Monte Carlo techniques. It demonstrates substantial improvement for a base test problems to show the quantitative impact on values of concern, such as detector response. Specifically the BeRP ball measured by the SNAP detector at a distance of 100-cm and the count rate measured for that problem approaching its asymptotic value faster with our new formalism than with straight S_N calculations.