Multidisciplinary Design Optimization for Space Applications

Abstract

Multidisciplinary Design Optimization (MDO) has been increasingly studied in aerospace engineering with the main purpose of reducing monetary and schedule costs. The traditional design approach of optimizing each discipline separately and manually iterating to achieve good solutions is substituted by exploiting the interactions between the disciplines and concurrently optimizing every subsystem. The target of the research was the development of a flexible software suite capable of concurrently optimizing the design of a rocket propellant launch vehicle for multiple objectives. The possibility of combining the advantages of global and local searches have been exploited in both the MDO architecture and in the selected and self developed optimization methodologies. Those have been compared according to computational efficiency and performance criteria. Results have been critically analyzed to identify the most suitable optimization approach for the targeted MDO problem.