Experimental Investigation of the Aerodynamic Interaction Between Ducts and Actuator Discs

The present study aims at providing experimental performance analysis of ducted wind turbines (DWTs) through momentum theory. To simplify the DWT model, a screen emulating the rotor is adopted by applying the actuator disc theory. Two duct models and five different screens are employed for different configurations to investigate the aerodynamic performance of the DWT in this experimental study. Duct 1 (C-T,C-duct = 0.91) is characterized by an aerofoil-shaped cross-section and duct 2 (C-T,C-duct = 1.17) has a cambered shape. Specifically, the five screens vary in porosity from low (C-T,C-screen = 0.46) to high thrust coefficient (C-T,C-screen = 1.12). Results show that the total thrust coefficient (CT, DWT) varies with screen thrusts and ducts geometry. The aerofoil-shaped duct 1, caused by the non-linear behaviour of the aerodynamic force on an aerofoil-shaped body at increasing angle of attack, is more sensitive to the screen loading compared to duct 2. Thrust on duct 1 is highly dependent on screen thrusts owning to the existence of inward oriented lift, while thrust on cambered plate duct 2 stays constant. The high thrust and solid blockage this duct creates, accelerates the flow through the lesser loaded centre where the screen is located. The data also show that the growth of duct thrust has a positive effect on the overall performance of DWT. Both high screen loading and inappropriate aerofoil shape duct may lead to flow separation which causes the momentum loss for DWT. Moreover, the optimal value of C-T,C-screen = 0.89 for a bare wind turbine doesn't seem apply to the tested DWTs. The optimal value of CT, DWT depends both on duct geometry and the screen loading. The velocity measurements further indicate that flow separation occurs at the duct 1 external (pressure) surface, and shows the presence of stalled flow around centre line inside the duct. Finally, a comparison of the pressure distribution of duct with different screen loadings reveals that adding a screen inside duct 1 leads to loss in lift and hence to mass reduction.