Abstract: Monte Carlo transport methods are the most accurate techniques for solving neutron transport problems in nuclear technology applications. The accuracy results from the ability to sample continuous distributions in space, angle, and particularly, energy. While Monte Carlo methods, in general, do not suffer from discretization errors, they are limited by statistical convergence that goes as the one over N-squared. Thus, the added accuracy is offset somewhat by slow statistical convergence. Regardless, due to advanced computing power and advances in variance reduction techniques, Monte Carlo methods have become increasingly attractive for a wide range of problems in nuclear modeling and simulation. This talk will present a detailed introduction to Monte Carlo methods for three-dimensional, continuous-energy, neutral particle transport. In addition to presenting a comprehensive introduction to the particle random walk process, we will discuss advanced hybrid-deterministic methods that have enabled the use of Monte Carlo transport across a range of problem domains that used to be impractical due to slow convergence.
Speaker’s Bio: Thomas M. Evans is a distinguished Research and Development staff member and the group leader of the High Performance Computing (HPC) Methods for Nuclear Applications group in the Nuclear Energy and Fuel Cycles Division at the Oak Ridge National Laboratory. He has a Ph.D. in nuclear engineering (1997) and MS in health physics from the Georgia Institute of Technology and a BS in physics and astronomy (1992) from Haverford College. His research interests are in the areas of radiation and particle transport, linear and nonlinear solvers, algorithms and methods for HPC, and computational coupled physics simulation. He is currently the Exascale Computing Project Applications Development focus area deputy in charge of the energy applications portfolio.
Last Updated: November 4, 2021 - 1:09 pm