Type:

Year:

Author:

Supervisor:

Institution:

Page(s):

Abstract:

This work investigates the development of an automated design method aimed at facilitating synthesis through the generation of a range of feasible and optimally directed design alternatives. The method is aimed at assisting designers in the exploration of performance limits and trade-offs for synthesis tasks as well as reducing design time. Due to their complexity and their multidomain nature, microelectromechanical systems (MEMS) are the design domain of application chosen for this exploration. MEMS design is still performed by hand, despite a growing interest in facilitating their design and shortening their time to market. The novel method combines a multicriteria generate-and-test search algorithm, called Burst, with a Connected Node System (CNS) design representation. The methodprovides automatic links to multiphysics simulation for quantitative evaluation of designs performance, as well as the possibility for designers to choose preferred solutions in archives of Pareto optimal designs. Different case studies are used in order to test the potential of the method, and extend its capabilities to solve a variety of design tasks. Characteristics of flexibility and scalability of the method are validated through complex MEMS design tasks. The appraisal is provided by practical applications such as microresonators developed for the mobile and satellite industry. The tasks examined range from size optimisation of longitudinal free-free-beam resonators to more complex topology optimisation of sandwich resonators. The design objectives examined are bulk resonant frequencies, motional resistance and quality factor. The results not only successfully proved the effectiveness, flexibility and extendibility of the novel method, but also offered a better understanding of potential and limits of synthesis methods, as well as the basis for their future improvement.