As nuclear industry technology advances, the advanced nuclear systems typically incorporate large shielding block and complex channels. Research on methods for these deep-penetration shielding calculations not only has theoretical significance, but also has broad application prospects, finding value in fields such as nuclear engineering, radiation shielding, and medical applications. Existing angle-informed methods approximate the angular flux using other information more accessible without explicitly calculating the angular flux, leading to some limitations of these methods. Given this scenario, a method for strongly anisotropic, deep-penetration radiation shielding problems is presented in this paper. Based on the theories of CADIS and FW-CADIS, this method generates weight windows which are not only dependent upon space and energy but also direction. Instead of deterministic adjoint calculation, this method uses cascading method and Monte Carlo pre-calculations to obtain importance function and generate weight windows. Then, a self-written Monte Carlo particle transport simulation program has been developed to validate this method. Results are presented for the FNG Benchmark Experiment on Silicon Carbide (SiC) and other examples. Results indicate that angle-dependent weight windows generated by this method can reduce variance effectively. This preliminary work indicates the angle-dependent variance reduction method may be highly effective for practical shielding problems with strong anisotropies. Further tests have been proposed to verify the method’s effectiveness in practical scenarios.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

An Angle-Dependent Variance Reduction Method for Monte Carlo Deep-Penetration Particle Transport Calculations

  • Lu Yiran,
  • Lyu Huanwen,
  • Wang Xueqing

摘要

As nuclear industry technology advances, the advanced nuclear systems typically incorporate large shielding block and complex channels. Research on methods for these deep-penetration shielding calculations not only has theoretical significance, but also has broad application prospects, finding value in fields such as nuclear engineering, radiation shielding, and medical applications. Existing angle-informed methods approximate the angular flux using other information more accessible without explicitly calculating the angular flux, leading to some limitations of these methods. Given this scenario, a method for strongly anisotropic, deep-penetration radiation shielding problems is presented in this paper. Based on the theories of CADIS and FW-CADIS, this method generates weight windows which are not only dependent upon space and energy but also direction. Instead of deterministic adjoint calculation, this method uses cascading method and Monte Carlo pre-calculations to obtain importance function and generate weight windows. Then, a self-written Monte Carlo particle transport simulation program has been developed to validate this method. Results are presented for the FNG Benchmark Experiment on Silicon Carbide (SiC) and other examples. Results indicate that angle-dependent weight windows generated by this method can reduce variance effectively. This preliminary work indicates the angle-dependent variance reduction method may be highly effective for practical shielding problems with strong anisotropies. Further tests have been proposed to verify the method’s effectiveness in practical scenarios.