Research Topic

Direct numerical simulation of particle laden flows using extended Discontinuous Galerkin methods

Particulate flows are of high scientific and technological interest with lots of applications. The understanding of such flows is important in chemical engineering (separation), life science (blood flow) and basic understanding of nature (sedimentation in ocean or river beds). On both, the macro- and microscopic scale, the phenomena of particulate flows are not fully understood. There is still a lack of accurate and efficient solvers. Even with the computer power nowadays on big research and industry clusters the efficiency of a solver is still crucial especially for industrial applications. Methods for particulate flows can be separated in two general approaches: The first one is the so called Lagrangian approach which uses a mesh fitted to the surfaces of the particles for imposing the boundary conditions. The second group are immersed boundary methods.

The first immersed boundary method was proposed by (Peskin 1972) in the field of fluid-structure interaction for simulating flow patterns around heart valves. The new feature of this method was, that all calculations were done on a fixed cartesian grid. It was not needed to remesh and project the solution on the new grid in every timestep in order to be conform with the geometry. The key of his method was to impose the influence of the immersed boundary on the flow without remeshing.

In our method a sharp-interface cut-cell immersed boundary representation by using a level-set function is used to model the particle boundaries. The method is implemented into the higher-order DG framework of BoSSS at the department of fluid dynamics (FDY).

Recent Results

The immersed boundary methods was successfully implemented into the BoSSS framework and first test calculations in terms of one-way coupling between the particle and the fluid were proceeded. Results show good agreement with literature e.g. in comparison with body-fitted calculations. Nevertheless because if our splitting-type approach the current solver renders to be first order in time only. Possible improvements to tackle this will be investigated in future work.

Key Research Area

Particulate Flows, Discontinuous Galerkin Methods, Immersed Boundary Methods


Dennis Krause


Dolivostraße 15

D-64293 Darmstadt



+49 6151 16 - 24381


+49 6151 16 - 24404




krause (at) gsc.tu...

 Print |  Impressum |  Sitemap |  Search |  Contact |  Privacy Policy
zum Seitenanfangzum Seitenanfang