Multi-Fidelity Gradient-based Approach to the Aerostructural Optimization Benchmark Problem

This paper describes our approach and results for the benchmark model and optimization problem proposed by working group 1 for the special session on High-Fidelity Aeroelastic Design Optimization Applications and Benchmarks at the 2026 AIAA SciTech Forum. A gradient-based multi-fidelity constrained optimization approach is applied to this aeroelastic fuel burn minimization of a Boeing 717-like wing with a lift and a buffet constraint and utilizing three planform and one flow design variable (aspect ratio, sweep, taper and angle of attack). The highest and lowest fidelity levels considered are Euler and panel solutions, respectively, all combined with a modal structural solver. The coupling of the two solvers and adjoint gradient computations are accomplished with FUNtoFEM. The optimization results show that a 20% reduction in fuel burn is possible while producing enough lift for cruise without buffeting. The multi-fidelity approach outperforms the high-fidelity one in terms of both robustness and computational cost for most starting points considered.