Variance Reduction Techniques for Implicit Monte Carlo Simulations
Abstract
The Implicit Monte Carlo (IMC) method is widely used for simulating thermal radiative transfer and solving the radiation transport equation. During an IMC run a grid network is constructed and particles are sourced into the problem to simulate the transport of energy in the form of radiation. Particles that reach the end of a time step without losing their entire energy are moved to "census" and will continue in the next time step. In order to reduce the variance of the solution, a neutronics technique called "weight windows" is applied, which allows for a greater number of particles to reach the census. With a larger number of particles in each grid, obtaining good statistics becomes much more probable. Along with weight windows, two other techniques called "Roulette" and "Analog" are implemented to remove excess particles from the problem and reduce computational time. The implementation of these techniques should reduce the total variance of the solutions obtained using IMC and increase the efficiency of Monte Carlo particle simulations.
Citation
Landman, Jacob Taylor (2014). Variance Reduction Techniques for Implicit Monte Carlo Simulations. Honors and Undergraduate Research. Available electronically from https : / /hdl .handle .net /1969 .1 /152027.