An automatic adaptive mesh generation method for weight window in Monte Carlo particle transport. (May 2016)

• Record Type: Journal Article
• Authors: Zhao, Jinbo; Li, Xinmei; Wu, Bin; Zheng, Huaqing; Hao, Lijuan; Song, Jing; Long, Pengcheng
• Is Part Of: Annals of nuclear energy. Volume 91(2016:May)
• Page Start: 105
• Page End: 110
• Publication Date: 2016-05
• Subjects: Variance reduction technique -- Mesh-based weight window -- Shielding calculation -- SuperMC; Nuclear energy -- Periodicals; Nuclear engineering -- Periodicals; 621.4805
• DOI: 10.1016/j.anucene.2015.12.011
• Languages: English
• ISSNs: 0306-4549

Abstract

Highlights: We develop an automatic adaptive mesh generation method. Coarse meshing method based on user's interested area is developed. Fine meshing method based on mean free path has been further improved. The mesh generation method is proved to be convenient and effective. Abstract: The variance reduction technique of mesh-based weight window was widely used in reactor shielding calculation. However, there are disadvantages in the traditional calculation procedure with weight window method based manual operation. It is time consuming to perform iterative calculations to get an effective weight window. Even for experts, they can only give an approximate division range of mesh parameters. Iteratively mesh plotting is needed for the under-bias or over-bias of particles. To solve the problem, an automatic adaptive mesh generation method was proposed for weight window based on the characteristics of particle's transport. The method involved coarse and fine meshing. Coarse mesh boundary parameters can be automatically generated based on the division of the interested area. Fine mesh boundary parameters can be produced by the automatic process of the mean free paths at the current energy of the particle. Coarse meshing and fine meshing were combined to generate the final meshing scheme. The fusion shielding benchmark and the ITER T426 benchmark were selected as the benchmark cases. The accuracy and efficiency of the proposed method were demonstrated.