Numerical Simulations of Turbulent Boundary Layers

by Dr Philipp Schlatter (KTH Mechanics)

Friday 5 Jun 2009, 13:15 → 14:15 Europe/Stockholm

122:026

In this presentation, an overview of the research performed at KTH Mechanics involving simulations of fluid flows is given. This research is organised as a part of the Linné Flow Centre, a centre of excellence founded in 2007 and consisting of both experimental, theoretical and numerical groups of three departments at KTH Stockholm including Mechanics, Numerical Analysis and Aerodynamics. The focus of the research is on basic studies and method development for various types of flows ranging from micro-fluidics up to large-scale computations of high-Reynolds number turbulence and geophysical flows. The showcases presented here are examples from three areas which are the traditional strengths at KTH: i) Transition from laminar to turbulent flow, ii) fully-developed turbulent wall-bounded flows and ii) flow control. The first part is concerned with the changeover of a flow from the ordered laminar state into the chaotic turbulent flow behaviour, a process termed laminar-turbulent transition. This is illustrated via a stability analysis of the jet-in-crossflow configuration. To this end, both linear and non-linear direct numerical simulations (DNS) are employed to determine the (undisturbed) flow behaviour, the steady-state solution, and an eigenvalue analysis using the time-stepper approach to determine the three-dimensional instability modes. In a second part, the research on high-Reynolds number turbulence is documented with a large-scale DNS of the generic turbulent flow over a flat surface. This simulations is one of the largest to be performed to date, employing on the order of one billion grid points. The obtained results allow for the first time a direct comparison of such a flow to a laboratory experiment, showing excellent agreement in both mean and fluctuating quantities. This allows us to gain new insights into the turbulent processes close to solid walls, cross-validating both experimental techniques and numerical methods. In a last part, the possibilities of actively affecting a flow via small changes in the setup is dicussed, i.e. flow control. In the particular example we will look at a passive control mechanism designed to reduce frictional losses due to the appearance of turbulence on a flat surface. Streamwise flow disturbances are introduced via small roughness elements; this changed mean flow is then stabilising the growth of certain instability waves which are known to be among the most dangerous instabilities for such a flow case. Visualisations are presented which illustrate the effectiveness of this passive control strategy.