Journal article Open Access
Demazière C; Tatidis A; Vinai P
Capitalizing on some earlier work, this paper presents a novel Monte Carlo-based approach that allows estimating the neutron noise induced by stationary perturbations of macroscopic cross-sections in the frequency domain. This method relies on the prior computation using Monte Carlo of modified Green’s functions associated to the real part of the dynamic macroscopic cross-sections, mimicking equivalent subcritical problems driven by external neutron sources. Once such modified Green’s functions are estimated, the neutron noise induced by any type of perturbations can be recovered, by solving a linear algebra problem accounting for the interdependence between the real and imaginary parts of the governing balance equations. The newly derived method was demonstrated on a large homogeneous test system and on a small heterogeneous test system to provide results comparable to a diffusion-based solver specifically developed for neutron noise applications. The new method requires the specification by the user of the real part of the Fourier transform of the macroscopic cross-sections. This is accomplished using ACE-formatted cross-section files defined by the user. Beyond this input data preparation, no change to the Monte Carlo source code is necessary. This represents the main advantage of the proposed method as compared to similar efforts requiring extensive modifications to the Monte Carlo source code.