Visual dependency structure matrix for multidisciplinary design optimization tradeoff studies

The design of complex products, such as automobiles and aircraft, necessitates modeling of the physics within numerous participating disciplines that are inherently linked through the transference of critical data. The resulting complexity makes it essentially impossible for designers or design teams, regardless of the number of participants or expertise of the participants, to intuitively understand the interdependencies of the disciplinary analyses and tasks. Recent methods and tools being developed in the field of multidisciplinary design optimization provide a means of representing these inherent couplings, as well as a way in which tradeoffs between accuracy and efficiency may be explored. It is with this in mind that the visual dependency structure matrix has been developed in this work. The visual dependency structure matrix is a web-based framework, coupled with underlying cost and error models, that enables designers to explore the possibility of eliminating or suspending couplings interactively, beforeor duringacomplex analysis run.The paper presents the visual dependency structure matrix framework and shows how the visual dependency structure matrix can be used as a mechanism for designers to cut computational costs during the design process without sacrificing the desired accuracy of the system level process.