At the center of the activities is the simulation and visualization of fluid flows, heat and radiation transfer. Our favorite flows are the convective motions in stars. The vast majority of stars - including our sun - posesses extensive zones where the energy, released deep down in the stellar interior, is transported outwards by motions of hot stellar material rather than by radiation. These convection zones may drastically influence the structure of the whole star. In itself or in combination with stellar rotation and pulsation convection generates a host of phenomena in the various stellar types. Hence stars are fascinating physical entities in their own right. Even old questions are still unanswered. For example, is the usually accepted scenario for the generation of the solar magnetic field, together with its polar reversals, really the correct one, and what are, anyway, the processes in some detail? Stars are also essential building blocks of the galaxies, whence additional interest in chemical evolution etc. derives. More recently, due to the flurry of discoveries and explorations of planetary systems detailed knowledge about the central stars is increasingly requested in ever subtler investigation of planetary properties and evolution.

In the last few decades modelling of convection zones in three dimensions by large-scale computations has become increasingly feasible and is being performed by a number of groups. The problems which our group is working on include

3D models of compressible convection (including realistic microphysics, radiative transfer, and magnetic fields) in the solar and stellar case, in particular also high resolution studies

Pulsation-convection coupling (cepheids, solar-like stars)

investigation of momentum closure models of compressible convection

semiconvection

the role of subgrid modelling

and other items

In order to perform these studies major software developments, in particular of our software package ANTARES, have been done and are being pursued also presently in order to have codes available which are highly advanced from the standpoint of numerics and software design. Earlier on, we also had to develop software for visualization purposes. We now highly appreciate the existence of very extensive, freely available visualization and analysis software packages such as Paraview, largely based on the also freely available visualization library Visualization Toolkit (VTK).

The research of H.J. Muthsam has, in the course of the years, been supported by the following institutions to whom I express my gratitude:

• Alexander von Humboldt Stiftung, Germany

• Austrian Federal Ministry of Science and Research

• Austrian Science Foundation

• Max Planck Society, Germany

• Research Fonds of the Austrian National Bank

• Research Fonds of the City of Vienna