A Low-Fidelity Method for Aerofoil Shape Optimisation for Curvilinear Blade Kinematics

Abstract

This study develops a low-fidelity framework for aerofoil shape optimisation under curvilinear blade kinematics, using a hovering cyclorotor as a representative case. Aerofoil optimisation can improve cyclorotor efficiency by suppressing leading-edge vortex separation during dynamic stall, but conventional approaches rely on computationally expensive CFD-based optimisation. The proposed method uses a single-streamtube model to estimate the rotor throughflow and optimises the aerofoil camberline using separate leading- and trailing-edge criteria. Assessed across configurations with varying blade counts and chord lengths, the framework consistently identifies aerofoils that improve hover efficiency, quantified by Figure of Merit. For the baseline four-bladed configuration, the low-fidelity optimum achieves 77% of the Figure of Merit improvement obtained using high-fidelity optimisation at a fraction of the computational cost. The analysis also reveals an additional torque-minimising design family at increased chord lengths, highlighting the influence of trailing-edge loading. Aerofoil optimisation is also compared to blade-pitch kinematics optimisation, which improves the efficiency through similar control of the leading-edge vortex separation. While both approaches produce comparable improvements in efficiency, the optimised kinematics substantially reduces thrust. Aerofoil optimisation may therefore be more practical, as maintaining a target thrust with optimised pitch kinematics would require higher rotational speeds, potentially introducing structural and noise issues.