Assessment of the Random Ray Method for Neutron Transport Problems with Strong Angular Anisotropy

The random ray method (TRRM), a variant of the method of characteristics (MOC), has shown promise for three-dimensional transport calculations due to its use of stochastic quadrature. Unlike the traditional MOC, which relies on a fixed set of characteristic rays, TRRM generates a new set of random rays for each iteration, effectively performing a Monte Carlo integration over the angular domain. This gives it the potential to accurately handle transport problems with strong angular anisotropy, a significant challenge for deterministic methods, which often suffer from ray effects.This study investigates TRRM’s performance in such problems, hypothesizing that it can resolve strongly anisotropic flux provided the statistical uncertainty is sufficiently low. To validate this, two benchmark problems known for causing strong flux anisotropy are analyzed: a lattice geometry with strong absorbers and a modified KAPL (Knolls Atomic Power Laboratory) benchmark with void ducts that induce neutron streaming. Calculations are performed using both TRRM and the traditional MOC (via the OpenMOC code).The results demonstrate that the traditional MOC produces severe ray effects in both test cases, even with fine angular discretization. In contrast, TRRM successfully mitigates these ray effects, yielding smooth and physically accurate flux distributions that converge as the number of rays and tracking length increase. Furthermore, a comparative analysis shows that TRRM achieves a higher level of accuracy with significantly fewer computational integrations than MOC, highlighting its superior computational efficiency for these challenging problems. The study concludes that TRRM is a robust and efficient alternative to deterministic methods for neutron transport problems characterized by strong angular anisotropy.