Conference Paper

Aerodynamic Shape Optimization with CAD-Based Geometric Parameterization


H. M. Hajdik, A. Yildirim, and J. R. R. A. Martins


AIAA SciTech Forum, 2023



Multidisciplinary design optimization (MDO) is a powerful tool for aircraft design. These design optimization methods often rely on the free-form deformation (FFD) method to parameterize geometry. However, the geometric parameterization with the FFD approach is not based on the original computer aided design (CAD)-based geometry definition. As a result, the optimized designs obtained from the FFD approach are not compatible with standard design and manufacturing workflows. CAD in the loop of a high-fidelity MDO framework can then bridge this gap between MDO methods and design or manufacturing needs. To effectively integrate CAD into optimization, the surface mesh points’ relationship to the CAD geometry needs to be tracked in order to calculate derivatives for gradient-based optimization and morph the mesh as the geometry changes. In this work, we present ongoing efforts on coupling MDO to CAD as well as a comparison of results obtained from traditional FFD methods. Results for aerodynamic shape optimization are discussed. We show that FFD-based optimizations and CAD-based optimizations for airfoil cases and a wing case starting from the same baseline designs can achieve the same optimized design, validating both the surface mesh tracking algorithm and the derivatives for the CAD method. The presented method is a critical prerequisite for integrating CAD-based geometric parameterizations in simulation based MDO frameworks.