Information on point-defect cluster stability and mobility is needed in order to parameterise models describing the nanostructure evolution of reactor pressure vessel steels under irradiation, but cannot be obtained directly from experimental measurements. The stability and mobility of vacancy-containing copper clusters and precipitates in iron has been quantitatively studied using advanced atomistic kinetic Monte Carlo methods, whose physical reliability has been improved by employing artificial intelligence techniques for the regression of the activation energies required by the model as input. Thus, the model validation was based, on the one hand, on comparison with available ab-initio (quantum mechanical) calculation for the verification of the used cohesive model, and, on the other, with other models and theories. The direct validation on experimental studies of copper precipitation appears to be computationally challenging, due to space- and time-scale limitations of the model.
|Number of pages
|Published - May 2009
|Studiecentrum voor Kernenergie