Efficient Information Exchange/Storage between two Models in Multiscale Simulations of Complex Fluids and Cell Tissue

Using a Heterogeneous Multiscale Modeling (HMM) approach we couple two
different formulations of a system: one for the macroscopic behaviour
in the whole domain but where we neglect the dynamics of microscopic
entities, and a microscopic model for the actual microscopic entities
which, however, does only have a very limited extend in space and
time. These two systems have to exchange information, but there is a
lot of redundancy. To minimize this redundancy, symmetries can be
exploited by appropriate transformations and by the intermediate
storage of information. The development of a database is therefore
closely related to the study of relevant information, symmetries and
correlations in the natural systems we want to model. Another
challenge is to develop an efficient implementation of such a
framework for models we run in parallel, i.e. where data is
distributed.
We have three driving applications for which ideas can be tested: two
multiscale models for complex fluids (hard-spheres suspensions & fully
resolved blood) and a potential multiscale model for cell tissue.