Continuous energy, multi-dimensional discrete ordinates transport calculations for problem dependent resonance treatment
Abstract
In the past twenty 20 years considerable progress has been made in developing new methods for solving the multi-dimensional transport problem. However the effort devoted to the resonance self-shielding calculation has lagged, and much less progress has been made in enhancing resonance-shielding techniques for generating problem-dependent multi-group cross sections (XS) for the multi-dimensional transport calculations. In several applications, the error introduced by self-shielding methods exceeds that due to uncertainties in the basic nuclear data, and often they can be the limiting factor on the accuracy of the final results. This work is to improve the accuracy of the resonance self-shielding calculation by developing continuous energy multi-dimensional transport calculations for problem dependent self-shielding calculations. A new method has been developed, it can calculate the continuous-energy neutron fluxes for the whole two-dimensional domain, which can be utilized as weighting function to process the self-shielded multi-group cross sections for reactor analysis and criticality calculations, and during this process, the two-dimensional heterogeneous effect in the resonance self-shielding calculation can be fully included. A new code, GEMINEWTRN (Group and Energy-Pointwise Methodology Implemented in NEWT for Resonance Neutronics) has been developed in the developing version of SCALE [1], it combines the energy pointwise (PW) capability of the CENTRM [2] with the two-dimensional discrete ordinates transport capability of lattice physics code NEWT [14]. Considering the large number of energy points in the resonance region (typically more than 30,000), the computational burden and memory requirement for GEMINEWTRN is tremendously large, some efforts have been performed to improve the computational efficiency, parallel computation has been implemented into GEMINEWTRN, which can save the computation and memory requirement a lot; some energy points reducing techniques have also been developed, improving the computational efficiency at the meanwhile preserving the accuracy. These efforts make the new method much more feasible for practical use.
Degree
Ph.D.
Advisors
Downar, Purdue University.
Subject Area
Nuclear physics|Computer science|Mathematics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.