Computational Physics

We study and apply virtual particle models, with strong emphasis on fully discrete particle models which belong to the general class of Cellular Automata. Furthermore, as a generalization of the strict Cellular Automata class, continuous particle models, possibly with long range interactions, are addressed.  We continue to investigate specific instances of Cellular Automata, the Lattice Gas Automata (LGA) and Lattice Boltzmann Models (LBM).

The LBM was applied to study among other flow in porous media, advection-diffusion processes, and hemodynamics. For details we refer to the LBM pages.

Another important test-bed is the continuous virtual particle model with long range interactions, where the aim is to investigate the computational structure and load-balancing performance of hierarchical algorithms for very large systems of non-uniformly interacting agents. Here, we plan to study in detail the mapping of hierarchical particle models onto special hybrid computer architectures.

Finally, we study virtual particle models that mimic wave phenomena. Both CA's and continuous particle models capable to model wave phenomena are developed. Here, the aim of the research is to further investigate the potency of virtual particle models to mimic a large variety of natural phenomena. It is anticipated that the computational structure of these new virtual particle models have a strong resemblance to the other models that we study, thus allowing direct mapping of such new application domains to parallel computers.