What’s New in 1.0.5

New Features

  1. 多群常数库读入后直接在GPU上分配内存,减少GPU-CPU数据传递;

  2. 多群常数库从RLIB-1.04c更新至RLIB-1.0.5;

  3. 增加了通量分布、核子密度分布等输出;

Bug Fixes

  1. a6c6e03: incorrect scatter start and end group index

  2. 2a94f05: type error if library is None

  3. ef3c9f3: incorrect reflector heights set

  4. 4a9f326: incorrect length of group_bins

  5. e912c9a: mesh_index not correct when get source slope

  6. b64220e: typo which cause address out of bounds

  7. 769a34a: heights may exceed plane boundary

  8. 944dca1: assembly track endpoints may not include in segments points

  9. 39e3e35: duplicated vertexes may generated when clipping cell

  10. 68151a2: incorrect reporting error about if track linked

  11. f3d84ba: use almost_equal to avoid equal lines not equal after rotation

  12. c5002b2: forget to add source slope term

  13. baa706a: line point overlap bug if nr = 0 in SectMeshing

  14. 480ea14: forget to reset flux_slope and a typo

  15. 3ef2374: forgot to push_back values into cell_mesh_index

  16. 5856478: cell_avg_power not zero in non-fuel pins

  17. 8c7faf7: incorrect logics in get_plane_index

  18. 0159119: miss cudaDeviceSynchronize

  19. 843e87a: use sector meshing instead of rectangle meshing for hexagonal cases to avoid bugs

  20. e94e577: incorrect pad_width used by map resizing

  21. 995e119: the number of nearest cells more than 16 some times

  22. 6f0f0e3: rotational theta can not kept by cloning of SectMeshing

  23. 22947d1: boundary exceed if the core has only one assembly

  24. 12106c8: insufficient length for nearest_index

  25. 0a8a912: cudaFree bug