Modeling of boundary conditions in embedded lattice calculations

Abstract

The coupling of the fine-mesh high-order transport operator with a coarse-mesh low-order approximation, in order to reduce the computational requirements when simulating whole reactor cores, is studied here through the boundary conditions for the fine-mesh solver. A recently developed software, solving the neutron transport equation though a discontinuous Galerkin Finite Element discrete-ordinates method, is used for the fine-mesh high-order solver, and the coarse-mesh low-order solver is simulated by a coarsening process. As a first step, a verification and validation process is necessary to be performed. This is carried out using the softwares DRAGON and MCNP, respectively, as the references in the verification and validation processes. A very good agreement is obtained during the verification process, while for the validation, results show that different quadratures should be considered in order to improve the accuracy. In addition, a parametric study is presented, where the different parameters of the spatial and angular discretizations are analyzed, in order to understand the behavior of the solver for different configurations. A second part of the work consists in studying the effect of coarsening the boundary conditions for a particular problem (C5G7 benchmark). This coarsening is performed to simulate the low-order approximation of the boundary conditions obtained with the coarse-mesh solver. Understanding the loss of accuracy for the fine-mesh calculations is necessary in order to improve the fine-mesh/coarse-mesh coupling for the neutronic solvers.