Towards Adaptive Closed-Loop Control of Transonic Cavity Flows

We present results from the development of an active/adaptive controller for the suppression of cavity pressure loads. A computational test bed for adaptive feedback flow control has been developed comprised of the CRAFT CFD flow solver, a recursive ARMARKOV algorithm for system identification, and an adaptive disturbance rejection algorithm. Control is implemented using unsteady surface pressure sensors on the cavity walls and an array of zero-net mass-flux actuators at the leading edge. These novel simulations complement companion experiments conducted at the University of Florida and help screen potential control algorithms and actuators. Preliminary experimental results available so far show suppression of the tones as well as broadband levels at low Mach numbers. Computationally, the results suggest that modeling the actuator dynamics is potentially important. The overall concept of examining control strategies through the use of a virtual test bed is promising.