Improved Convergence Rate of Multi-Group Scattering Moment Tallies for Monte Carlo Neutron Transport Codes.

Monte Carlo Neutron Transport;Hybrid Methods;Tallying Multi-Group Cross Sections;Scattering Moments;Nuclear Engineering and Radiological Sciences;Engineering;Nuclear Engineering and Radiological Sciences

Nelson, AdamCollins, Benjamin Steven ;

University of Michigan

关键词: Monte Carlo Neutron Transport;    Hybrid Methods;    Tallying Multi-Group Cross Sections;    Scattering Moments;    Nuclear Engineering and Radiological Sciences;    Engineering;    Nuclear Engineering and Radiological Sciences;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/110342/nelsonag_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Multi-group scattering moment matrices are critical to the solution of the multi-group form of the neutron transport equation, as they are responsible for describing the change in direction and energy of neutrons.These matrices, however, are difficult to correctly calculate from the measured nuclear data with both deterministic and stochastic methods.Calculating these parameters when using deterministic methods requires a set of assumptions which do not hold true in all conditions.These quantities can be calculated accurately with stochastic methods, however doing so is computationally expensive due to the poor efficiency of tallying scattering moment matrices.This work presents an improved method of obtaining multi-group scattering moment matrices from a Monte Carlo neutron transport code.This improved method of tallying the scattering moment matrices is based on recognizing that all of the outgoing particle information is known a priori and can be taken advantage of to increase the tallying efficiency (therefore reducing the uncertainty) of the stochastically integrated tallies.In this scheme, the complete outgoing probability distribution is tallied, supplying every one of the scattering moment matrices elements with its share of data.In addition to reducing the uncertainty, this method allows for the use of a track-length estimation process potentially offering even further improvement to the tallying efficiency.Unfortunately, to produce the needed distributions, the probability functions themselves must undergo an integration over the outgoing energy and scattering angle dimensions.This integration is too costly to perform during the Monte Carlo simulation itself and therefore must be performed in advance by way of a pre-processing code.The new method increases the information obtained from tally events and therefore has a significantly higher efficiency than the currently used techniques.The improved method has been implemented in a code system containing a new pre-processor code, NDPP, and a Monte Carlo neutron transport code, OpenMC.This method is then tested in a pin cell problem and a larger problem designed to accentuate the importance of scattering moment matrices.These tests show that accuracy was retained while the figure-of-merit for generating scattering moment matrices and fission energy spectra was significantly improved.