An Automatic Aerodynamic Shape Optimisation Framework Based on DAKOTA

Abstract

Aerodynamic design is a must and fundamental aspect for better performance of an aircraft. This work describes a framework for aerodynamic shape optimisation (ASO) based on high-fidelity Computational Fluid Dynamics (CFD) analyses and an open source optimisation toolkit DAKOTA. The free form deformation (FFD) technique is adopted for parameterization of geometries, followed by a transfinite interpolation (TFI) method for grid deformation. CFD simulations are performed by full Navier-Stokes equations in conjunction with turbulence model. CFD, FFD and TFI modules are integrated with DAKOTA by a black-box interface, a pre-processing and a post-processing. The transonic RAE2822 airfoil is employed for demonstration of the effectiveness and capability of the developed framework. Several cases with different optimisation algorithms, design variables and turbulence models are performed. Optimum solutions with significant improvement of lift-drag ratio are presented to demonstrate the capability of the framework in this paper.