An efficient Monte Carlo fission source convergence acceleration strategy adapted from the survival biasing technique

The current effective methods for accelerating fission source convergence in criticality calculation such as the super-history powering and Wielandt method are proven to reduce the dominance ratio. As a result, source convergence can be achieved with a smaller number of iteration cycles. However, none of these methods could improve the convergence time in most of the cases, despite the reduced number of iteration cycles. In particular, the conventional source acceleration mechanism adapted in the super-history powering and Wielandt method involve banking fission neutrons for later tracking. Here, reactive fission sites are only stored during tracking higher-generation fission neutrons. We postulated that the conventional acceleration strategy is computationally inefficient and will consequently result in a remarkable increase in the convergence time. To rectify the problem, we present an acceleration strategy which is adapted from the survival biasing technique. In this work, the new strategy is targeted to increase the chance of searching for reactive high-generation fission sites whilst eliminating most of the computational cost of simulating neutrons that are not able to survive until higher generation fission events. Numerical verification results show that the new method is feasible and effective, which can save up to 87% of the convergence time according to the selected acceleration parameter. (C) 2019 Elsevier Ltd. All rights reserved.