The development and application of an improved reactor analysis model for fast reactors

Accuracy in neutron cross sections calculation and consistency in reactor physics are fundamental requirements in advanced nuclear reactor design and analysis. The work presented in this dissertation focuses on the development and advanced application of a reactor analysis model with updated cross section libraries that is suitable for online cross section generation for fast reactors. Research has been performed in two areas of interest in reactor physics.;The first target of the research is to develop effcient modeling capacity of the 1- D lattice code MICROX-2 for its neutron spectrum calculation based on Collision Probability Method (CPM). Expanded master cross section libraries have been generated based on updated nuclear data and optimized fine-group energy structure to accommodate both thermal and fast reactor spectra as well as to comply with the need for advanced fuel cycle analysis. After verifying the new libraries, the solution methods have been reviewed and updated, including the update of interpolation scheme for resonance self-shielding factors and improvement of spatial self-shielding models for various fuel assembly geometries. The assessment of the updated lattice calculation models has shown that the prediction accuracy of lattice properties represented by the eigenvalue and reaction rate ratios is improved, especially for fast neutron spectrum lattices of which the importance of neutrons in the unresolved energy range is high.;The second target of the research is to improve the accuracy of few-group nuclear cross section generation for the reactor core calculation. A 2-D pin-by-pin lattice model has been developed based on embedded CPM within the framework of the Nodal Expansion Method (NEM), which is capable of modeling the heterogeneity of the fuel assembly. Then, an online cross section generation methodology along with discontinuity factors has been developed based on Iterative Diffusion- Diffusion Methodology (IDDM), which can minimize the inconsistency in physics parameters by feeding the actual core condition into the cross section generation by the 2-D lattice code. In order to facilitate the iterative scheme between the 2-D lattice and core calculation, appropriate interface routines are used for accurate and consistent data transfer. Finally the overall physics method, starting from the 1-D lattice calculation to the core calculation, has been validated against benchmark problems, and promising results have been observed in core eigenvalue and power distribution comparisons.