先进中子学栅格程序KYLIN-Ⅱ输运模块并行优化开发

Parallel Optimization Development of Transport Calculation Module-in Neutronics Lattice Code KYLIN-Ⅱ

  • 摘要: 先进中子学栅格程序KYLIN-Ⅱ的输运计算模块采用了特征线中子输运计算方法,其计算精度较高,可适用于反应堆中复杂的燃料组件,然而当网格规模、能群数、特征线数较大时,计算时间较长,计算效率较低,因此需对其进行并行优化,以提高计算效率。通过性能分析,发现特征线扫描和高阶散射源计算较为耗时。本文通过基于MPI的能群并行、E指数优化、角通量球谐函数展开等方法实现了并行优化。基准题验证表明,并行优化计算精度较高,E指数优化对特征线扫描效率提升较好,角通量球谐函数展开对高阶散射源计算效率提升较好。经过优化后的KYLIN-Ⅱ的输运计算模块加速效果显著,可满足工程使用需求。

     

    Abstract: The transport calculation module of neutronics lattice code KYLIN-Ⅱ adopts the method of characteristics, which has higher calculation precision and can be applied to the complex fuel assemblies in the reactor. However, when the number of grids, energy groups and feature lines is large, the calculation time is longer and the calculation efficiency is lower. Therefore, parallel optimization is needed to improve the calculation accuracy. It is found that the calculation of characteristic line scanning and high-order scattering source updating is time-consuming through performance analysis. In this paper, parallel optimization was realized by parallel energy group based on MPI, E-exponential optimization, angular flux spherical-harmonic expansion and so on. The results show that the parallel optimization calculation precision is high, E-exponential is effective in calculation of characteristic line scanning, and angular flux spherical-harmonic function expansion is very effective in high-order scattering source updating. The optimized transport calculation module of KYLIN-Ⅱ has a significant acceleration effect and can meet project needs.

     

/

返回文章
返回