Efficient methods for uncertainty propagation in Monte Carlo are essential to characterise the impact of model uncertainties in geometry, material densities, nuclear data, etc. on the results of interest. In the past, a Polynomial Chaos framework was proposed and applied to multi-group (MG) Monte Carlo problems, and showed significant promise. As part of this project, the Polynomial Chaos framework for nuclear data uncertainty quantification was implemented to power iteration in MG SCONE. The goal is to investigate how it fares in a real MG benchmark, with the final aim of extending it to continuous energy (CE) models.

Project lead: Theophile Bonnet