| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 1728822 | Annals of Nuclear Energy | 2012 | 8 Pages |
In this paper a computational scheme is presented to calculate a small number of the largest alpha eigenvalues and the corresponding eigenvectors of transport problems, that is moderate dimension dominant alpha eigenspaces of discretized transport matrices. The aim of the scheme is to provide a computational tool for practical reactor applications with real-life size and complexity, e.g. metal cooled fast systems with strong spatial heterogeneity. The scheme is based on the 3D multigroup Sn equations that are spatially discretized with the discontinuous finite element method. Both implicitly restarted Arnoldi method (IRAM) and Sub-space Iteration (SSI) were applied to the equations, in order to compare their efficiency. The results show that first is superior in terms of computation time requirement, but the second is a viable option for test and development due to its stability and reduced memory requirements. Finally, calculation results are presented for the Venus experimental Accelerator Driven System (ADS) core.
► A computational scheme was developed to calculate moderate dimension fundamental alpha eigenspaces of 3D transport problems. ► The scheme is based on the multigroup Sn method with discontinuous finite element discretization, the equations are solved using Arnoldi method and sub-space iteration. ► Calculation results are presented for the Venus experimental ADS core.
