A tool for holistic optimization of mechatronic design concepts

Sammanfattning: Designing mechatronic systems is challenging. Even so, mechatronicproducts are regarded as one of the most important means in order toinnovate in many large industries, e.g. automobile and aerospace. Theinherent difficulty lies in the multiple engineering domains involved and howthese are treated during the development process. A holistic approach, whichtreats these domains concurrently, is needed in order to find hidden synergiesbetween domains and in the end find an optimal product design based ongiven requirements.Traditional product development methodologies, even ones with amechatronic design focus, rely on treating the individual domains separatelyand only integrating them at a point in time rather far into the detailed designphase of the development process. In addition, in order to solve problems thatarise in later design stages, the design engineers need to backtrack to earlierstages and in the worst case repeat substantial parts of the work. This is bothtime and cost inefficient.This dissertation summarizes and extends previously published research bythe author about a novel mechatronic design methodology and a supportingsoftware tool. The goal of the design methodology is to enable design of betterproducts, in terms of e.g. development cost, size, and sustainability, by findingsynergies and reducing backtracking through better analysis of systemconcepts. The methodology relies on using an optimizer to efficientlydetermine the potential of a system concept, described by combining a numberof component models, e.g. motors, transmissions, and structural ones, from acomponent library and specifying their parameters.A number of design cases have been evaluated, some of which are presentedin this dissertation, including both one where a physical prototype is built, anda few virtual ones. These design cases have been instrumental in evaluating themethodology as well as the software tool.It is concluded that the design methodology is able to properly evaluate aconcept against competing concepts, and that it is a useful addition to existingmethodologies. However, as always, a number of improvements are possible,some of which are presented in a concluding section.