A comparison of Monte Carlo methods for neutron leakage at assembly level

Overview of methodology for spatial homogenization in the Serpent 2 Monte Carlo code

“…The currently distributed version of Serpent 2 (version 2.1.26) does not yet support the calculation of directional diffusion coefficients, but recent work on the topic has been carried out in a related doctoral thesis (see Dorval and Leppänen, 2015;Dorval, 2016b). The methodology, together with the new leakage model developed as part of the same thesis work (Dorval, 2016a), will be included in the official version of Serpent 2 within the future.…”

“…There exist also Monte Carlo specific methods in which the correction is applied during the transport simulation, for example, based on the use of complex weights (Yamamoto, 2012) or albedo iteration (Yun and Cho, 2010). One of such methods was developed for Serpent by Dorval (2016a), but at the time of this writing the capability was not yet included in the distributed version (update 2.1.26). Smith (2016) has recently pointed out that there are several fine details related to the implementation of nodal diffusion methods, which have not been published until recently or discussed in the open literature.…”

Monte Carlo method for solving a B1 equation with complex-valued buckling in asymmetric geometries and generation of directional diffusion coefficients

“…The capability of calculating direction-dependent diffusion coefficients is desirable because the direction dependence of diffusion coefficients is notable in a fertile fuel assembly (Faure and Marleau, 2017) and a void containing fuel pin cell (Yamamoto, 2012). Previous works on the directional diffusion coefficients by the Monte Carlo method are based on various techniques, and they are published in Milgram (1997), Gelbard and Pego, (1979), Yamamoto (2012), Dorval and Leppänen (2015), Dorval (2016a), and Dorval (2016b). In Yamamoto (2012), the neutron transport equation for the heterogeneous B 1 method is explicitly solved by introducing complex-valued weights in the random walk process of a Monte Carlo calculation for leakage-corrected calculations.…”