An adaptive p-refinement strategy applied to nodal expansion method in 3D Cartesian geometry

Abstract

The aim of this work is to develop a coarse mesh treatment strategy using adaptive polynomial, p, refinement approach for average current nodal expansion method in order to solve the neutron diffusion equation. For performing the adaptive solution process, a posteriori error estimation scheme, i.e. flux gradient has been utilized for finding the probable numerical errors. The high net leakage in a node represents flux gradient existence between neighbor nodes and it may indicate the source of errors for the coarse mesh calculation. Therefore, the relative Cartesian directional net leakage of nodes is considered as an assessment criterion for mesh refinement in a sub-domain. In our proposed approach, the zeroth order nodal expansion solution is used along coarse meshes as large as fuel assemblies to treat neutron populations. Coarse nodes with high directional net leakage may be chosen for implementing higher order polynomial expansion in the corresponding direction, i.e. X and/or Y and/or Z Cartesian directions. Using this strategy, the computational cost and time are reduced relative to uniform high order polynomial solution. In order to demonstrate the efficiency of this approach, a computer program, APNEC, Adaptive P-refinement Nodal Expansion Code, has been developed for solving the neutron diffusion equation using various orders of average current nodal expansion method in 3D rectangular geometry. Some well-known benchmarks are investigated to compare the uniform and adaptive solutions. Results demonstrate the superiority of our proposed strategy in enhancing the accuracy of solution without using uniform high order solution throughout the domain and increasing the number of unknowns.