Abstract: This paper presents a multi-objective shape optimization method for the conceptual design of reentry vehicles. The class function/shape function transformation technique is used to represent briefly the various shapes of the reentry vehicles with relatively few variables. The aerodynamic characteristics of the vehicles are predicted by an engineering method which accounts for the effects of real-gas flow across the oblique shock wave. The multi-objective evolutionary algorithm non-dominated sorting genetic algorithm (NSGA-Ⅱ) which produces a set of Pareto solutions is employed as the optimization strategy to trade off the maximization of lift-drag ratio, volumetric efficiency, and the minimization of overall heat rate. Representative configurations are then obtained using a subtractive clustering algorithm and an optimum solution is visually identified with level diagrams which is a graphical rephical representation of the Pareto front and set. Design optimization of a bicone baseline configuration demonstrates that the trimmed lift-drag ratio is improved from 0.65 through 0.94 without reducing volumetric efficiency. This method also contributes a great deal to the reduction of the design space of a conceptual design and the determination of aerodynamic shape parameters.

Key words: reentry vehicle, conceptual design, class function/shape function, aerodynamic shape optimization, multi-objective optimization, level diagram.