Non-overlapping Local/Global (NLG) Iteration Scheme for 2D Transient Method of Characteristics (MOC)

To lessen these issues, non-overlapping local/global (NLG) iteration has been proposed. The advantages of NLG iteration are that it is natural for parallel computing and competitive to the partial current-based Coarse Mesh Finite Difference (p-CMFD) acceleration in terms of computing performance. Up to now, NLG iteration has been developed only for steady-state transport calculation. In this paper, NLG iteration is extended to have the capability of transient transport calculation. As a transient transport kernel, the method of characteristics (MOC) is chosen, and applied to several test problems. NLG iteration with p-CMFD has been successfully extended to have transient capability. As in the steady-state case, NLG iteration and p-CMFD acceleration give identical results if the solutions are converged. In addition, due to the good initial conditions for each time step, the computing performance is very promising in transient calculations compared to p-CMFD acceleration. Since NLG iteration is natural for parallel computing, the computing performance could be enhanced further if local problems are solved independently with parallel computing nodes. Parallel computing is planned for future work