Large-Eddy Simulation of Separated Flows Using a New Integral Wall Model
Speaker: Francois Cadieux, Flow Physics and Computation Lab, Johns Hopkins University
Presentation
Seminar Slide Deck (PDF-2.2MB)
Abstract
Wall-resolving large eddy simulation (LES) can accurately recover experimental and time-averaged direct numerical simulation (DNS) results for many separated flows of interest for engineering applications. However, wall-resolving LES is limited to moderate Reynolds numbers due to the strong requirement of resolving the viscous sub-layer. To perform turbulence resolving simulations of high Reynolds number flows, wall-modeled LES (WMLES) and hybrid RANS-LES methods must be employed. However, typical equilibrium wall-models for LES do not take pressure gradients into account, and many hybrid RANS-LES methods require very fine mesh spacing in the wall-normal direction to reliably capture separation and reattachment. The newly proposed integral wall model for LES (iWMLES) addresses these issues through the addition of non-equilibrium terms in an integral version of the momentum equation. First, tests of the iWMLES approach on canonical boundary layers and flow over wall attached cubes are reviewed. Then a laminar separation bubble over a flat plate is simulated to demonstrate the capability of iWMLES to capture separation, transition to turbulence, and reattachment. iWMLES predictions for a separating and reattaching fully turbulent boundary layer flow are compared against wall-resolved LES. The impact of including non-equilibrium terms is also investigated.
Biography
Francois Cadieux is currently a postdoctoral fellow at the Flow Physics and Computation Lab under the supervision of Prof. Rajat Mittal at Johns Hopkins University. His research focuses on developing higher-fidelity computational fluid dynamics simulation tools that are fast enough to inform the design of engineering applications like wall-modeled large-eddy simulation. Cadieux is also investigating active flow control technologies to mitigate flow separation and delay stall. During his graduate work with Prof. Julian A. Domaradzki, he probed the limits of under-resolved large-eddy simulation for separated flows applicable to small unmanned aerial vehicles and low pressure turbines, and quantified the importance of numerical dissipation effects at those resolutions. He obtained his Ph.D. in aerospace engineering from the University of Southern California in 2015.