Adjoint gradient-based approach for aerodynamic optimization of transport aircraft

Kurzfassung

Aerodynamic design of transport aircraft has been steadily improved over past several decades, to the point where today highly-detailed shape control is needed to achieve further improvements. Aircraft manufacturers are therefore increasingly looking into formal optimization methods, driving high-fidelity CFD analysis of finely-parametrized candidate designs. We present an adjoint gradient-based approach for maximizing the aerodynamic performance index relevant to cruise-climb mission segment by modifying the aircraft wing shape. The test baseline geometry is a wing-body configuration based on the Do-728 regional jet, over which a hybrid structured-unstructured CFD grid is generated. For primal and adjoint aerodynamic analysis we employ the DLR TAU flow solver in RANS mode, with Spalart-Allmaras turbulence model. In each optimization cycle, multiple flight points are evaluated in order to make the final shape more robust to off-design flight conditions. Constraints are placed on pitching moment and wing root bending moment, to counter shape modifications unfavorable to trimming and structural mass. The wing shape is modified by free-form deformation (FFD), with positions of FFD control points being the design parameters. An SQP optimizer is used to drive the optimization.