应用CMFD加速区域分解的并行MOC

Spatial Domain-Decomposed Parallel MOC Accelerated by CMFD

1. Department of Engineering Physics, Tsinghua University, Beijing, 100084, China;
2. Science and Technology on Reactor System Design Technology Laboratory, Nuclear Power Institute of China, Chengdu, 610041, China

摘要: 空间区域分解适合于大规模并行求解中子输运方程,但是子区域的增多会导致收敛变慢。为了克服这一缺点,采用粗网有限差分(CMFD)技术对空间区域分解的并行特征线方法(MOC)进行加速。使用ScaLAPACK求解CMFD粗网扩散方程;CMFD的粗网解既用来修正细网标通量,又用于修正内界面角通量。一维MOC数值结果表明,对于区域分解并行的MOC,CMFD技术是一种十分高效的加速方法,可以显著提高收敛速度。
- 区域分解 /
- CMFD加速 /
- 特征线方法 /
- 内界面修正
Abstract: Spatial domain-decomposing is applicable for solving the neutron transport equation on massively parallel architectures. However, the more sub-domains there are, the slower the convergence is. Then CMFD is used to accelerate the spatial domain-decomposed parallel MOC in order to overcome the problem. The CMFD coarse diffusion equation is solved by ScaLAPACK, and the CMFD results are used to update both fine-mesh scalar fluxes and inner-interface angular fluxes. Numerical results on 1D slab example illustrate that CMFD is an efficient acceleration method for the spatial domain-decomposed parallel MOC, which improves the convergence remarkably.
- Spatial domain-decomposing /
- MOC /
- CMFD /
- Interface flux correction