Generative Shape Optimization of a Rigid Wingsail Aerofoil using Bézier Parameterization and CFD Validation

With regulatory and market pressure to decarbonize merchant shipping, Wind-Assisted Propulsion Systems (WAPS) have regained attention as a fuel-saving technology. This study presents a generative design workflow for a two-dimensional wingsail aerofoil tailored to low-speed maritime operation. Geometry was parameterized using 9th-order Bézier curves and optimized in MATLAB using a constrained formulation of the fmincon solver with an objective equivalent to maximizing lift-to-drag ratio (CL/CD) under a thin-aerofoil, inviscid formulation. The optimization returned a candidate profile with a theoretical peak CL/CD of 21.51 at 3.23° angle of attack. To account for viscous effects and validate the design, RANS CFD simulations were performed using the SST k–ω turbulence model on a C-grid mesh with y+<1. CFD predicted a peak CL/CD of 40.91 at 20 m s⁻¹ and 3.23° AoA — an improvement of ≈90% relative to the inviscid prediction — and revealed a sensitivity to leading-edge radius that causes stall onset between 10° and 11°. A techno-economic case study for a 35 m × 5-sail retrofit on M.V. BANGLAR ARJAN (transatlantic route) estimates an approximate voyage energy saving of 1.86% under the assumptions stated. We discuss the limitations of inviscid optimization for marine wing sails and recommend integrating viscous models and geometric constraints (minimum leading-edge radius and thickness distribution) in future generative loops. The results demonstrate the practicality of generative Bézier-based optimization for wing sail aerofoil design while identifying specific improvements required for robust operational performance.