A hybrid stochastic/deterministic method for transient, three-dimensional neutron transport

The objective of this research is to develop and evaluate a method for solving the time-dependent, three-dimensional Boltzmann transport equation with explicit representation of delayed neutrons. A hybrid stochastic/deterministic technique is utilized with a Monte Carlo code embedded inside of a quasi-static kinetics framework. The time-dependent flux amplitude, which is usually rapidly varying, is computed deterministically by a conventional point kinetics algorithm. The point kinetics parameters, reactivity and generation time as well as the flux shape, which is usually slowly varying in time, are computed stochastically during the random walk of the Monte Carlo calculation.

To verify the accuracy of this new method, several computational benchmark problems from the Argonne National Laboratory benchmark book, ANL-7416, were calculated.

The results are shown to be in reasonably good agreement with other independently obtained solutions. The results obtained in this work indicate that the method/code is working properly and that it is economically feasible for many practical applications provided that a dedicated high performance workstation is available.