Open- and closed loop control on a D-shaped bluff body equipped with Coanda actuation

Abstract

The present study investigates drag reductions and efficiency increases by open-and closed-loop Coanda actuation on a D-shaped bluff body. Open-loop measurements are performed by scanning the actuation parameter space spanned by the actuation frequency and the moment coefficient. At a maximum power coefficient of 0.22 a drag reduction of 33 % relative to the unactuated case is observed. Closed loop control is conducted using machine learning control (MLC). MLC is a model-free control methodology that seeks to optimize a predefined cost function. Two cost functions are optimized , the drag coefficient and the sum of drag & momentum coefficients. MLC yields a drag reduction of up to 27 % and a maximum power ratio of 0.154, which are comparable to the open-loop results.