Acceleration of some sort is needed for within-group, multigroup, and eigenvalue problems. The most straightforward approach would be to define an acceleration (coarse) mesh on the boundary of which the angular flux (or functions of it) are computed. The materials within the mesh could be homogenized as well. This facilitates coarse mesh rebalance and coarse mesh finite difference. Moreover, one could, in theory, implement a variety of potential schemes based on the boundary information.
Acceleration of some sort is needed for within-group, multigroup, and eigenvalue problems. The most straightforward approach would be to define an acceleration (coarse) mesh on the boundary of which the angular flux (or functions of it) are computed. The materials within the mesh could be homogenized as well. This facilitates coarse mesh rebalance and coarse mesh finite difference. Moreover, one could, in theory, implement a variety of potential schemes based on the boundary information.