Computational Fluid Dynamics (CFD)
Focus of research
- dynamics of (near-wall) turbulence
- fluid mechanics of multi-phase flows
- direct numerical simulation and large eddy simulation
- specifically designed numerical methods
- data analysis and physical modeling
Because of the wide range of length and time scales in turbulent flows, direct numerical simulation (DNS) is very demanding in terms of computational resources. This is true for actual simulations as well as for the subsequent data analysis. Therefore, the use of efficient numerical algorithms for massively parallel computing systems is of particular importance.
DNS of fluid-particle systems
The intercation between turbulent flows and solid particles is a question of technological relevance, with a wide range of applications in hydraulics, meteorology, process engineering, energy technologies, bio-medical flows etc. We study fluid-particle systems by means of direct numerical simulation with fully resolved phase boundaries. The aim is to analyze processes such as the formation of particle agglomerations and the enhancement /attenuation of turbulence intensity. In the long run, our object is to contribute to an improvement of commonly used engineering models for multiphase flow.
Photo: M. Tischmacher
Dynamics of coherent structures
An understanding of fundamental dynamic processes in turbulent flows is essential in determining scaling laws and can be understood as a prerequesite for technological applications. As an example, detailed knowledge of the regeneration mechanism of turbulent near-wall structures can inspire strategies for boundary layer control.
One of the phenomena studied in our group is the generation of secondary flow in a square duct due to coherent structures. Another topic of interest is the analysis of turbulent flows with stratification, which are of great importance to geo-physical applications (e.g. ocean currents). For stable stratification there is a competition between damping buoyant forces and the intrinsic instability of the shear flow.
LES of complex flows
Large Eddy Simulation (LES) is very suitable for the calculation of turbulent flows in cases where no direct simulation is possible because of geometrical complexity and/or high Reynolds numbers. This applies in particular to cases with unsteady flow separation where purely statistical models typically do not provide accurate predictions.
Data of the numerical simulations is available here after peer-review.