Research On The Method Of Coupled 2D MOC And 1D NEM Whole Core Calculation

The current generation core physics analysis theory based on the ’two-step’ method employs a lot of approximations, which makes it hard to handle the complex core design.However, considering the state of the art computational capabilities, the direct three dimensional whole core transport calculation is too time and resources consuming to be competent in engineering applications. In order to shorten computing time while still retain high accuracy of the whole core transport calculation, the coupled 2D/1D iteration calculation strategy is under extensive research worldwide. In this strategy, the method of characteristics(MOC) is chosen for 2D calculation because of its good adaptation to geometries and accuracy comparable to the Monte Carlo method, while a variety of options are available for the axial solver.The coupled 2D MOC and 1D nodal expansion method (NEM) whole core calculation code was developed based on the two dimensional MOC calculation module in the open source lattice calculation code DRAGON. The practicability of the coarse mesh finite difference (CMFD) non-linear iteration method applied to the cell based 2D MOC calculation was studied. And the CMFD acceleration algorithm module was programmed and the iteration strategy was implemented in the DRAGON code, then the stability and the acceleration method of CMFD equations were also studied. Furthermore, according to the feature that MOC transport sweep calculation is carried out for each ray on each spatial angle independently, the 2D MOC calculation module in the original DRAGON code was parallelized using shared memory multi-thread OpenMP parallel environment. Finally, the coupled 2D MOC and 1D NEM whole core calculation method was studied based on the previous two researches. Under the 3D global CMFD framework, the coupled 2D MOC and 1D NEM whole core calculation code was developed making best use of application programming interfaces (API) of DRAGON code and the CLE2000 control language. The SUB-PLANE scheme was also introduced to improve the calculation accuracy when dealing with intensive axial heterogeneous problems.The verification results show that, the CMFD acceleration algorithm developed in this thesis has better acceleration effect compared with 2D MOC acceleration methods implemented in the original DRAGON code, and better stability compared with the similar code OpenMOC. The progressive source extrapolation acceleration convergence technology reduces the source iteration times when solving the CMFD equations. The parallelized 2D MOC module using OpenMP could obtain high parallel efficiency with multi-threads. The code developed in this thesis based on the coupled 2D MOC and 1D NEM whole core calculation method under the 3D global CMFD framework is capable of solving the 3D whole core Pin-by-Pin issues effectively, with an accuracy close to that of the direct 3D whole core transport calculation while saving a lot of computing time and resources. The introduced SUB-PLANE scheme is capable of improving the accuracy with only a small increase in computation effort when solving problems with strong axial heterogeneity.