基于抽样方法的特征值不确定度分析

Eigenvalue Uncertainty Analysis Based on Statistical Sampling Method

  • 摘要: 核数据是反应堆物理计算的基础数据,研究其不确定度对反应堆物理计算引入的不确定度,对提高反应堆的安全性和经济性具有重要意义。本文基于抽样理论研究了反应堆物理计算不确定度分析的方法,研发了不确定度分析程序UNICORN。基于ENDF/B-Ⅶ.1评价数据库,使用NJOY程序开发了多群协方差数据库。采用UNICORN程序和多群协方差数据库对三哩岛燃料棒和基准题RB31的k进行了不确定度分析,得到核数据库中各分反应道截面的不确定度对k造成的不确定度。结果表明:238U(n,γ)截面对三哩岛燃料棒k造成的不确定度最大,相对不确定度达0.4%左右;协方差数据库的不同来源会对不确定度分析结果造成一定影响。

     

    Abstract: The nuclear data are the basic data for reactor physics calculation. It is significant to study the contribution of their uncertainty to the uncertainty of reactor physics calculation for improving the safety and economy of reactor. Based on the ENDF/B-Ⅶ.1 library, the covariance library including the variance and covariance for all the basic cross sections and all the groups was generated using the NJOY code. The uncertainty analysis code UNICORN for reactor physics calculation was developed based on the statistical sampling method. Based on the UNICORN and covariance library, the uncertainties of k for Three Mile Island (TMI) lattice and RB31 benchmark, due to the uncertainty of basic cross sections, were analyzed. The results indicate that the uncertainty of 238U(n, γ) cross section is the largest uncertainty to the kof TMI lattice and the relative uncertainty to the k is up to about 0.4%, and different sources of covariance library effect the uncertainty of analysis result.

     

/

返回文章
返回