Direct Numerical Simulations for Adjoint-based Optimal Flow and Noise Control of a Backward-Facing Step

The advances in the understanding compressible turbulent flows alongside with the im­provements in computing capacity, enable researchers to tackle more ambitious problems that surpass human intuition. Hence, it becomes necessary to rely on methods that give an insight on the underlying physical mechanisms that govern these turbulent flows to perform non-trivial tasks such as flow control. A recently developed adjoint-based opti­mal flow control framework for compressible flows with non-constant viscosity has been appended to an existing well validated in-house direct numerical simulation (DNS) code (HiPSTAR). A three-dimensional backward-facing step (BFS) flow is used as a simplified cabin noise problem. The turbulent flow over a target region is optimally controlled to minimize the potential flow-induced vibrations in a structure at the same location. A cost function defined in wavenumber space is used to guide the actuation towards a specific range of flow scales. A total cost reduction of 0.48% was achieved over the non-actuated case. Wavenumber-frequency decomposition is applied in both controlled and uncontrolled scenarios to achieve a further understanding of the optimal flow actuation. This analysis shows a change in the flow characteristics, shifting the wavenumber and directivity of some flow structures.