| Standard Kinetic Monte Carlo (KMC) simulations with a catalog constructed a priori are constrained to a lattice and often ignore long-range elastic interactions. This strongly limits the application of the method to a relatively small subset of relevant problems. By combining the Activation-Relaxation technique (ART), a very efficient open-ended method to find saddle points, with an off-lattice topological classification scheme for local environments, kinetic ART (k-ART) overcomes these limitations. First, events consisting of initial state, saddle point and final state are generated for each local environment present in the system and stored for later re-use. The most important events are then reconstructed and refined for each instance that atomic environment appears, and one of those events is selected according to KMC rules. K-ART has successfully been applied to a wide range of complex materials, from point defects in metals to amorphous silicon. In this talk, I will describe the main features of k-ART, and present results on vacancy agglomeration in iron and defect relaxation in amorphous silicon. This work is performed in collaboration with L.K. Béland, F. El-Mellouhi, J.-F. Joly and N. Mousseau. |
|