Airfoil selection, optimization and analysis for a solar-powered unmanned aerial vehicles

This paper presents a method for obtaining an optimal airfoil for a Solar-powered Unmanned Aerial Vehicle (UAV). This method uses B-spline shape functions along with control points to define the geometry of the airfoil and Pattern Search (PS) optimization algorithm to obtain the optimal airfoil for the UAV from a given initial vector. The wing of the UAV was considered to operate at Reynolds Number of 2×105. Hence, existing airfoils from a low Reynolds Number airfoil database [7] which performed the best at the considered Reynolds Number were selected and interpolated together in XFLR5 to obtain the initial vector. An XFOIL-MATLAB interface was built for the evaluation of the aerodynamic performance of each generated airfoil during the optimisation process. The conditions of evaluation were Reynolds Number = 2×105, Angle of Attack ranging from −5° to 12°. The objective function for the optimisation algorithm includes the lift-drag characteristics of an airfoil in order to optimise glide and range and also minimise power consumption in the operating angle of attack of the UAV. This operating range varies from angles −2° to 4° during different phases of the flight. Constraints for the program includes manufacturing feasibility of the wing with solar panels for which penalty is included in the objective function. Due to limited flexibility of the solar panels, the curvature limitation on the suction surface of the airfoil, towards its trailing edge, is taken into account by placing limited number of control points in that area [2] and having a restriction on the degrees of freedom of the same. The resulting airfoil has an average of 13.93% increase CL/CD and 15.21% increase in CL1.5/CD compared to the selected initial airfoil.