Abstract:Effective delay of turbulent boundary layer separation could be achieved via closed-loop control. Constructing such a system requires that sensor data be processed, real-time, and fed into the controller to determine the output. Current methods for detection of turbulent boundary layer separation are lacking the capability of localized, fast and reliable identification of the boundary layer state. A method is proposed for short-time FFT processing of time series, measured by hot-film sensors, with the purpose of identifying the alternation of the balance between small and large scales as the boundary layer separates, favoring the large scales. The method has been validated by comparison to other criteria of separation detection and over a range of baseline and controlled flow conditions on a "simplified" high-lift system, incorporating active flow control.