SINGLE AND MULTI-OBJECTIVE OPTIMIZATION OF THE AERODYNAMICS OF PROJECTILES USING GENETIC ALGORITHMS

Abstract

This paper is devoted to the use of genetic algorithms (GA) in order to optimize the aerodynamic of projectiles. The single and multi-objective optimization framework PYMOO [1] is used in conjunction with PRODAS software [2], a semi-empirical software assessing aerodynamic coefficients. First, single (SOO) and multi-objective optimizations (MOO) using GA are assessed on two test functions: the Gomez-Levy function [3] for a SOO under constraints, and the Binh & Korn [4] problem for a MOO. In a second time, PRODAS software is coupled with PYMOO. The first considered case is the SOO of the axial force coefficient of a 155 mm spin-stabilized projectile. The optimal configuration found by PRODAS/PYMOO is compared to additional Computational Fluid Dynamics computations in order to emphasize the dependence of the results to the aerodynamic prediction tool. Finally, both SOO and MOO are performed on a guided 155 mm fin-stabilized artillery projectile equipped with a Canard Actuation System (CAS). The single-objective optimization is based on the maximization of the normal force coefficient CNα whereas the MOO adds the drag coefficient CXO as a second objective.