Multidisciplinary design techniques applied to conceptual aerospace vehicle design

Multidisciplinary design optimization (MDO) is an emerging discipline within aerospace engineering. Its goal is to bring structure and efficiency to the complex design process associated with advanced aerospace launch vehicles. Aerospace vehicles generally require input from a variety of traditional aerospace disciplines--aerodynamics, structures, performance, etc. As such, traditional optimization methods cannot always be applied. Several multidisciplinary techniques and methods have been proposed as potentially applicable to this class of design problem. Among the candidate options are calculus-based (or gradient-based) optimization schemes and parametric schemes based on design of experiments theory.; A brief overview of several applicable multidisciplinary design optimization methods is included in this dissertation. Methods from the calculus-based class and the parametric class are reviewed, but the research application reported in this work focuses on methods from the parametric class.; A vehicle of current interest was chosen as a test application for this research. The rocket-based combined-cycle (RBCC) single-stage-to-orbit (SSTO) launch vehicle combines elements of rocket and airbreathing propulsion in an attempt to produce an attractive option for launching medium sized payloads into low earth orbit. The RBCC SSTO presents a particularly difficult problem for traditional one-variable-at-a-time optimization methods because of the lack of an adequate experience base and the highly coupled nature of the design variables. MDO, however, with its structured approach to design, is well suited to this problem.; This dissertation presents the results of the application of Taguchi methods, central composite designs, and response surface methods to the design optimization of the RBCC SSTO. Attention is given to the aspect of Taguchi methods that attempts to locate a "robust" design--that is, a design that is least sensitive to uncontrollable influences on the design. Near-optimum minimum dry weight solutions are determined for the vehicle.; This dissertation concludes with a summary and evaluation of the various parametric MDO methods employed in this research. Recommendations for additional research are provided.