Large Eddy Simulation (LES) is the most promising approach for the prediction of complex turbulent flows. By resolving fluctuating quantities, LES has the potential to become a widely used simulation tool for flows of practical relevance. Currently, the use of LES for high Reynolds number complex flows is limited by its extraordinary resolution requirements that are present due to the need to resolve the inner wall layer. Models bridging the wall layer are able to solve this resolution dilemma - but currently available wall models suffer from severe constraints. lt is the aim of this proposal to improve existing and to develop novel wall models for LES of complex, high Reynolds number flows. The research will focus on pressure induced separation and reattachment and on the effects of streamwise curvature, axial vorticity and compressibility. It will start from careful a priori testing of existing approaches and an approach, recently proposed by the applicants. The data used will be based on well resolved Direct Numerical Simulations (DNS). Based on the experiences made during these test, a posteriori testing will be made on test cases commonly defined within this French-German research group thus allowing for detailed comparisons of the developed approaches with complementary methods developed by the other members of the research group.
DFG Programme
Research Units
International Connection
France
