论文部分内容阅读
在具有全局特性的蒙特卡罗输运精细计算的问题中,传统的MCNP(Monte Carlo N Particle Transport Code)局部减方差方法很难得到理想的计算结果,全局减方差方法(Global Variance Reduction,GVR)则是一种有效的解决方法。针对中国聚变工程试验反应堆(Chinese Fusion Engineering Testing Reactor,CFETR)的中子输运过程中减小全局方差的问题,将多种形式的GVR方法应用到柱状CFETR中子学模型的计算中。依据不同的中子分布信息,在算例中应用和对比了6种不同形式的GVR权窗,并对不同GVR方法的品质因子(FOMG)、标准差(σ)和有效计数率(Scoring)进行了分析。与AN(MCNPanalog method)相比,GVR方法的FOMG有很大的增长,误差在空间的分布也更加平缓,且具有更高的Scoring。在前人提出的全局减方差的基础上,在计算中应用一些新的GVR形式(能量、径迹数等),计算结果表明,基于中子通量的GVR方法的全局计算效率较AN提高了6.43倍。此外,基于中子能量的全局减方差方法也是一种可行的GVR应用形式,其与AN比较,计算效率提高了5.11倍。综上,基于中子通量的GVR方法具有最佳的全局减方差效果。
In the Monte Carlo N Particle Transport Code (MCNP) local subtraction method, it is difficult to obtain the ideal calculation results in the global Monte Carlo N Particle Transport Code (CFN) refinement calculation. The Global Variance Reduction (GVR) It is an effective solution. In order to reduce the global variance in the neutron transport of the Chinese Fusion Engineering Testing Reactor (CFETR), various forms of GVR methods are applied to the calculation of the cylindrical CFETR neutron model. According to the different neutron distribution information, six different forms of GVR weighting windows are applied and compared in the example, and the quality factor (FOMG), standard deviation (σ) and effective counting rate (Scoring) of different GVR methods are compared Analysis. Compared with the AN (MCNPanalog method), the FOMG of the GVR method is greatly increased, the distribution of the error in the space is more flat, and the Scoring of the GVR method is higher. Based on the global variance reduction proposed by the predecessors, some new GVR forms (energy, track number, etc.) are applied to the calculation. The calculation results show that the global calculation efficiency of GVR based on neutron flux is higher than that of AN 6.43 times. In addition, the global subtraction method based on neutron energy is also a feasible form of GVR application, which increases the computational efficiency by 5.11 times compared with AN. In summary, the GVR method based on neutron flux has the best effect of global minus variance.