Abstract of PARAMAR Project

During the operation of future fusion reactors (ITER, DEMO), it is expected that materials containing the fusion plasma will be exposed to high fluxes of energetic neutrons. The effects of neutrons on materials involve i) the displacement of atoms from their equilibrium positions in the lattice creating point defects and cluster thereof, and ii) nuclear transmutation reactions leading to the formation of impurities such as He and H in the material. Predicting the evolution of these defects and impurities and their interactions is thus crucial to understand the microstructure change of materials and the degradation of their properties under fusion conditions. It is therefore essential to have an accurate knowledge of the defects, namely interstitials and vacancies, which form during collision cascades, i.e. at times as short as picoseconds. The MARLOWE program simulates atomic collisions in crystalline targets using the binary collision approximation. In order to obtain reliable results, it is imperative to accumulate statistics since the spatial extension of cascades can exhibit a huge variability from one cascade to the other. The version of MARLOWE that is currently available at the Nuclear Energy Agency is not parallelized. The project that we propose is to parallelize MARLOWE in order to decrease the CPU time required per run and to obtain a better damage statistics.