Calibrated POD Reduced-Order Models of Massively Separated Flows in the Perspective of Their Control Available to Purchase

Lots of studies have been performed to control boundary layer separation. In order to develop a control procedure that is energetically efficient, it is interesting to formulate active flow control problems like optimization problems [1]. To reduce the numerical costs associated to this approach, a POD Reduced-Order Model is used instead of the high-fidelity model formed by the Navier-Stokes equations. Resolving this low-order dynamical system provides the dynamics of the flow on the POD subspace and using the optimal control theory, it is possible to find the optimal control law determined on the low-fidelity model [2]. In the present paper, DNS velocity fields of the wake flow behind a circular cylinder and PIV measurements of a separated flow around an airfoil are used to build and calibrate low order models.