Evaluation of memory-preserving models for particle transport in random media

Abstract

A shortcoming of the Levermore-Pomraning (LP) model is its inaccuracy to treat the transport of particles in a diffusive material alternating with a transparent one. Enhanced-memory LP models have been proposed by Zimermann et al. (1991) and Larmier (2018) to circumvent this issue. The main purpose of this paper is to evaluate the PBS2 model also known as Algorithm C in these geometries. A second objective is to generalize it to an arbitrary number of interfaces allowing the same code to give either an approximation of the solution with a few interfaces or a reference result by increasing their number. A third one is to give a mathematical formulation of these models.