Exploring the use of a deterministic adjoint flux calculation in criticality Monte Carlo simulations

A. Jinaphanh, J. Miss, Y. Richet, N. Martin, A. Hébert

International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering (M&C 2011), Rio de Janeiro, RJ, Brazil,
May 8-12, 2011

Abstract

The paper presents a preliminary study on the use of a deterministic adjoint flux calculation to improve source convergence issues by reducing the number of iterations needed to reach the converged distribution in criticality Monte Carlo calculations. Slow source convergence in Monte Carlo eigenvalue calculations may lead to underestimate the effective multiplication factor or reaction rates. The convergence speed depends on the initial distribution and the dominance ratio. We propose using an adjoint flux estimation to modify the transition kernel according to the Importance Sampling technique. This adjoint flux is also used as the initial guess of the first generation distribution for the Monte Carlo simulation. Calculated Variance of a local estimator of current is being checked.