No description available.

Graph and network theory have become commonplace in modern life. So widespread in fact that most people not only understand the basics of what a network is, but are adept at using them and do so daily. This has not long been the case however and the relatively quick growth and uptake of network technology has sparked the interest of many scientists and researchers. The Science of Networks has sprung up, showing how networks are useful in connecting molecules and particles, computers and web pages, as well as people. Despite being shown to be effective in many areas, network theory has yet to be applied to mechanical engineering design. This work makes use of network science advances and explores how they can impact Computer Aided Design (CAD) data. CAD data is considered the most valuable design data within mechanical engineering and two places large collections are found are educational institutes and industry. This work begins by exploring 5 novel networks of different sized CAD collections, where metrics and network developments are assessed. From there collections from educational and industrial settings are explored in depth, with novel methods and visualisations being presented. The results of this investigation show that network science provides interesting analysis of CAD collections and two key discoveries are presented: network metrics and visualisations are shown to be effective at highlighting plagiarism in collections of students' CAD submissions. Also when used to assess collections of real world company data, network theory is shown to provide unique metrics for analysis and characterising collections of CAD and associated data.

Multi-Attribute Optimisation (MAO) is proposed as a tool for delivering high value products within the systems engineering approach taken in the automotive industry. This work focuses on MAO methods that use Computer Aided Engineering (CAE) analyses to build a metamodel of system behaviour. A review of the literature and current Jaguar Land Rover optimisation methods showed that the number of samples required to build a metamodel could be estimated using the number of input variables. The application of these estimation methods to a concept airbox design showed that this guidance may not be sufficient to fully capture the complexity of system behaviour in the metamodelling method. The use of the number of input variables and their ranges are proposed as a new approach to the scaling of sample sizes. As a corollary to the issue of the sample size required for accurate metamodelling, the sample required to estimate the error was also examined. This found that the estimation of the global error by additional samples may be impractical in the industrial context. CAE is an important input to the MAO process and must balance the efficiency and accuracy of the model to be suitable for application in the optimisation process. Accurate prediction of automotive attributes may require the use of new CAE techniques such as multi-physics methods. For this, the fluid structure interaction assessment of the durability of internal components in the fuel tank due to slosh was examined. However, application of the StarCD-Abaqus Direct couple and Abaqus Combined Eularian Lagrangian was unsuitable for this fuel slosh application. Further work would be required to assess the suitability of other multi-physics methods in an MAO architecture. Application of the MAO method to an automotive airbox shows the potential for improving both product design and lead time.

Creativity is a useful attribute for people to have. It allows them to solve unfamiliar problems, introduce novelty to established domains, and to understand and assimilate new information and situations - all things we would like computers to be able to do too. However, these creative attributes do not exist in isolation: they occur in a context in which people tend to solve problems routinely where possible rather than consider non-standard ideas. These more mundane attributes might also be useful for problem solving computers, for the same reasons they are useful for us. However, they are often ignored in attempts to implement systems capable of producing remarkable outputs. We explore how the study of both human and computational creativity can inform an approach to help computers to display useful, complete problem-solving behaviour similar to our own: that is, robust, exible and, where possible and appropriate, surprising. We describe a knowledge-based model that incorporates a genetic algorithm with some characteristics of our own approach to knowledge reuse. The model is driven by direct interactions with problem scenarios. Descriptions of the role or appearance of key themes and concepts in literature in functioning problem-solving systems is lacking; we suggest that they appear as artefacts of the operation of our model. We demonstrate that it is capable of solving routine problems flexibly and effectively. We also demonstrate that it can solve problems that would be effectively impossible for a genetic algorithm operating without the benefit of knowledge-driven biasing. Artefacts of the behaviour of the model could, in certain scenarios, lead to the appearance of non-routine or surprising solutions.