Multidisciplinary design optimization of part geometry in CAD

Abstract

Multidisciplinary design optimization (MDO) is the process of searching for designs which best satisfy a set of objectives, while respecting that the appropriateness of a design is dependent on more than just a single discipline. Traditionally, MDO has been conducted on a higher system level where there is less granular technical detail and more general abstractions applied to the design. With the advances of modern computing and commercially available software tools, it is now possible to conduct MDO at a more granular level such that the part geometry in CAD can be guided by the process. In this thesis, an MDO workflow is created for purposes of improving the design of a manufactured baseplate. Off-the-shelf commercial software tools (ANSYS and OptiSLang) are used to develop the modular workflow, through an iterative improvement process. Final designs are successfully generated from the workflow and compared to baseline and reference designs, with respect to the temperature, cost and frequency properties related to the design. The results obtained in this project demonstrate the potential for MDO to improve the way in which product design and development is conducted in the future.