Abstract:
A multifidelity, multipoint aerodynamic blade shape optimization was conducted to design a realistic, full-sized proprotor, representative of recent industry tiltrotor and lift+cruise UAM vehicle designs. The proprotor was designed to achieve a disk loading of 8 psf in hover at sea level standard day and 1.9 psf in cruise at an altitude of 4000 ft above ground level with a multipoint efficiency optimization target. A low-fidelity optimization was first conducted using a differential evolution algorithm with CAMRAD II's uniform inflow model, followed by a mid-fidelity trim using CAMRAD II's nonuniform inflow and free-wake models, a high-fidelity verification using a hybrid RANS/LES approach in FUN3D, and finally a high-fidelity optimization on the low-fidelity optimized blade shape with a gradient-based method using a uRANS approach in FUN3D. The low-fidelity optimization resulted in a proprotor that achieved a hover figure of merit of 0.830 and a propulsive efficiency in cruise of 0.904. Results from the low-fidelity optimization, mid-fidelity trim, and high-fidelity verification were compared to highlight differences in predicted blade span loading and aerodynamic efficiency between the different aerodynamic solvers. It was shown that the low-fidelity solver compared least favorably with the high-fidelity aerodynamic solver. Lastly, the high-fidelity optimization further reduced the torque in cruise by approximately 75 ft-lb, with negligible changes in hover figure of merit and propulsive efficiency in cruise. Negligible blade shape differences were observed between the low-fidelity optimized design and the results from the sequential high-fidelity optimization with the largest difference being collective pitch in both hover and cruise, respectively.