Stochastic geometry capability in MCNP5 for the analysis of particle fuel

A stochastic geometry capability has been implemented into the MCNP5 Monte Carlo code. This capability allows the analysis of TRISO particle fuel, allowing for the random locations of the fuel kernels within the graphite matrix. The method has been compared to a MCNP5 benchmark calculation of randomly placed fuel kernels within a box with reflecting boundaries as well as multiple realizations of lattices where the microspheres are randomly located within the equivalent cubical cells. Comparisons are also made to MCNP5 calculations where the fuel kernels are fixed on a cubical lattice. Our preliminary results for infinite medium configurations indicate that the new stochastic geometry capability in MCNP5 is an accurate and efficient approach to analyze TRISO particle fuel configurations.