Resonance self-shielding calculation using subgroup method and ABC algorithm

In this study, we propose a new method to optimize subgroup parameters for resonance self-shielding calculation. Our approach integrates the merits of both the subgroup method and ABC optimization technique to effectively evaluate self-shielded resonance cross-sections. The ABC algorithm is used to obtain subgroup level in a way that guarantees reproduction of shielded effective cross sections in the subgroup formulation. The temperature dependency of the cross-section is included in both subgroup level and subgroup weight. We used the conjugate gradient method based on the normal equations (CGNR) to evaluate the subgroup weights. An iteration technique is also used to consider the resonance interference. The proposed method is verified by analyzing Rowlands benchmark problems and Mosteller benchmark problems and comparing the obtained results with corresponding Monte Carlo solutions. The multiplication factor results show small errors and also good agreement. (C) 2014 Elsevier Ltd. All rights reserved.