Predictions of the effect of wing camber and thickness on airfoil self-noise

Two wing self-noise modeling methods were used to predict the effect of airfoil camber and thickness on wing self-noise and their relationship with lift. Both a semi-empirical airfoil self-noise prediction code called NAFNoise, and a noise metric which uses steady RANS models were used for the investigation. Development and previous validations of the NAFNoise code with experimental measurements are summarized to understand the limitations of the code. Predictions were made at low speed and moderate Reynolds number similar to the environment of a small unmanned aerial system. Self-noise predictions are plotted with lift coefficient for a series of camber and thickness changes. The boundary layer was tripped and left untripped to understand the effect boundary layer transition has on airfoil noise versus airfoil aerodynamic performance. Analysis indicated that increasing airfoil camber leads to higher overall sound level at lower angles of attack. An increase in airfoil camber increases lift at lower angles of attack. NAFnoise models with untripped boundary layers predict that increasing airfoil thickness leads to higher overall sound level, with largest increase in overall sound level at lower angles of attack. Results were inconclusive as to whether increasing camber is an effective way to increase lift coefficient and lift over drag without significantly increasing airfoil noise, because the conclusion was dependent on modeling methodology. NAFNoise models with untripped boundary layers indicated that increasing camber would result in a beneficial increase in section lift coefficient & lift over drag (up to about 8% camber) with minimal increase in noise production. RANS based noise metric and exclusion of the laminar boundary layer vortex shedding model from NAFNoise predictions indicated an increase in noise with section lift coefficient independent of airfoil camber.