A POD-based Reduced Order Design Scheme for Shape Optimization of Air Vehicles

Abstract

This paper describes the development of an innovative ’Proper Orthogonal Decompositionbased Reduced Order Design Scheme’ method to reduce the number of shape design variables needed for an e cient geometry optimization of air vehicles, while maintaining the generality of the design space. This POD-RODS method is based on the Proper Orthogonal Decomposition technique of extracting POD modes from a family of candidate con gurations and constructing a reduced order design space, in which the POD coe cients act as the new shape design variables. The candidate con gurations can be selected based on a designer’s past experience and entailing intuition. Established with the goal of developing a prototypical design process for unconventional air vehicle shape design, the POD-RODS method, when linked with a streamlined multi-disciplinary optimization framework, o ers a computationally e cient and cost e ective methodology for the conceptual design and con guration development of air platforms. The geometry optimization scheme developed herein is demonstrated for its robustness on a supersonic tailless air vehicle, which o ers a medium level of complexity and a much challenging design space to fully utilize the POD-RODS method.