Multipoint aerodynamic wing optimization in viscous flow

A methodology that allows for efficient aerodynamic optimization of wings with a full account of typical viscous effects is proposed. It extends earlier work by A. Jameson1-3 on wing optimization for inviscid flows. The optimization process is based on control theory, which is employed to derive the adjoint equations. Accurate and consistent modeling of viscous effects is essential in wing design and is implemented in the approach described here by viscous-inviscid interaction. The solution involves interaction between the Euler solver and the two-dimensional boundary layer. Although this technique is limited by the known assumptions of the boundary-layer approximation, it is very well suited for civil aircraft wing design in cruise conditions. For applications where large viscous-dominated regions of separation are present, substantially more expensive design methods based on the Navier - Stokes equations have to be used. The method developed in this work has resulted in a practical engineering tool because it combines the benefits of the fast adjoint equation-based technique and a very economical boundary-layer approach. The cost required for viscous calculations is similar to that for inviscid flows. Results for three-dimensional wing design in viscous flow are possible, even with the use of a PC. The method is demonstrated for a single, clean wing, and wing-body configuration.