Objectives

To assess the feasibility of bridging the gap between atomistic scale and microstructural scale modelling.

The potential importance of this work is apparent. If we can understand how variations on the atomistic level modify the microstructure, we will be in a position to engineer microstructures by making modifications at the atomic level.

Approach

Generation of parsimonious rules for the CA engine from atomistic scale simulations.

Possible approaches were reviewed in the preliminary report.

Here we demonstrate the feasibility by addressing a specific problem.

Conclusions

1. CA can be used to simulate microstructural evolution using rules generated by atomistic simulations.

2. This was demonstrated using the specific example of the adsorption / desorption of inert gas atoms on a Ca (111) surface.

3. The computational benefits if using atomistic based rules cellular automata (AR-CA) are considerable. We have shown a computational gain over MD simulations of the same size of 10⁵.

4. The computational gain can be increased significantly by using multiple atom CA cells of a hierarchical approach.

5. Specific conclusions from the first implementation of an AR-CA model are:
   1. island growth is observed.
   2. island growth is a strong function of temperature and gas atom species.
   3. island growth rates depend on the radius of the boundaries. Consequently, a ripening process of island growth is observed over a specific temperature range.

back to contents