Nuclear data uncertainty propagation with the stochastic sampling module in RMC

In this paper, an integrated, built-in stochastic sampling module, RMC-SS, is developed in the Reactor Monte Carlo code RMC. Comparing to traditional stochastic sampling tools which take the calculation code as a black box, RMC-SS is much more user-friendly and memory-efficient. Instead of storing perturbed nuclear data library, perturbation factors are prepared and stored, which can reduce the memory consumption of storing library. After reading the nominal nuclear data library and perturbation factors, RMC automatically repeats transport or burnup calculations. Each calculation will be assigned a unique random number seed to consider the statistical uncertainty. Finally, uncertainties of results are calculated by RMC. No post processing is required. Uncertainty analyses are performed in two in two problems, a bare sphere benchmark which represents a standard Monte Carlo transport calculation, and a PWR pin cell burnup benchmark, which represents a Monte Carlo burnup calculation, with RMC-SS. (authors)