PANDA Code Application to the OECD/NEA 2D/3D MOX Assembly Benchmark Calculations

Panda is an all purpose dicrete ordinates transport code developed at CEA/Bruyeres-le-Châtel. It is applied to stochastic neutronics applications like backward time dependent neutron number and count fluctuations or extinction probability calculations. The code can also be used for more conventional applications like the effective multiplication factor. It is based on a classic discrete ordinate (SN) transport solver which can be used for 1D, 2D and 3D computations. In order to validate PANDA code in two and three spatial dimensions, benchmark computations are performed. In this paper we present PANDA results on the international OECD/NEA C5G7 MOX benchmark. This test problem was designed by the expert group on 3D radiation transport benchmark of the OECD/Nuclear Agency in order to assess the ability of current deterministic transport codes to treat two and three dimensional core problems without homogenisation. The obtained solutions seem to be not fully converged with regard to the MCNP reference solutions. Accurate solutions in 3D will require a high level of spatial and angular discretization leading to very expensive calculations in term of CPU time and memory usage