AI-assisted CFD optimisation of multi-element wing angle of attack for enhanced formula SAE aerodynamic performance

• A CFD–GA–ANN framework was developed to optimize Formula SAE wing aerodynamics. • The Menter k-ω SST model was shown to outperform k-ε with <15% prediction error. • MSHD and Selig 1223 airfoils were identified as achieving the most efficient results. • Downforce was increased by 14.8% on the front wing and 28.4% on the rear wing. • Aerodynamic drag was reduced, enhancing overall vehicle energy efficiency. This study presents a multi-level aerodynamic optimisation framework that integrates Computational Fluid Dynamics (CFD) with Artificial Intelligence (AI) techniques to enhance the performance of Formula SAE vehicles. Unlike conventional CFD or empirical approaches, the proposed framework employs a hybrid Genetic Algorithm–Artificial Neural Network (GA–ANN) model to optimise design parameters with greater accuracy and efficiency. Initial investigations refined the angles of attack (AoA) and airfoil geometries to maximise downforce while reducing drag and energy consumption. Two-dimensional CFD simulations established the baseline aerodynamic characteristics, which were subsequently extended to full three-dimensional analyses. Among turbulence models, the Menter k–ω SST demonstrated superior predictive capability compared to the standard k–ε, with discrepancies in drag and downforce below 15%. The MSHD and Selig 1223 airfoils emerged as the most aerodynamically efficient candidates. Systematic AoA adjustments from 4.5°, 28.0°, and 56.0° to 5.5°, 33.0°, and 59.5° for the front wing, and from 9.5° and 40.0° to 12.2° and 41.9° for the rear wing produced CFD predictions validated within a 5% error margin. The GA–ANN optimisation further enhanced aerodynamic performance, yielding a 14.8% increase in front-wing downforce and a 28.4% increase in rear-wing downforce, alongside reduced drag and improved energy efficiency. These findings demonstrate the potential of AI-assisted CFD optimisation to bridge high-fidelity simulations with tangible performance gains, offering a promising pathway for hybrid airfoil design and active control strategies in high-performance automotive applications. Formula SAE Car & Wing geometry & Cross-section prism layers (Meshes)