Subdivision Surface-Based Geometry Modeling of Fighter Type Aerodynamic Shapes

Abstract

Classical CAD-based modeling approaches are often time-consuming and difficult to parameterize robustly for complex aerodynamic shapes, especially when large design spaces are involved. They tend to have numerical instabilities during deformation, making robust and flexible surface modeling challenging. To integrate such models into optimization processes, CAD-based Reduced-Order Models are commonly used. However, for high-dimensional problems, their interpolation errors dominate the overall modeling accuracy. This Limitation strongly affects configurations with high geometric integration, such as fighter type aircraft, which feature numerous topologically complex geometry details. This work introduces an alternative approach for generating CAD-compatible surfaces based on subdivision surfaces originating from computer graphics. The method provides analytically smooth and topologically flexible surface representations that enable robust and efficient modeling. A process concept for global aerodynamic optimization using subdivision-based Surface models is presented, and the successful application of the individual process components to a fighter type configuration is demonstrated. Furthermore, the capability to perform RANS-based CFD on these surfaces is demonstrated with two different solvers for a complex fighter type shape.