Enriched Functional Modelling Software to Assess Aerospace System Architectures

Abstract

The aerospace manufacturing industry is transforming, bringing in electrical and autonomous technologies into both products and production processes at a faster pace than the past. Yet manufacturers are not exploring radically different variants or product architectures early in the product development process, which is when changes in design are most economical. A link is needed in the design process between the problem at hand and the design phase, where not just radically different architectures can be analysed and thus represented together, but also where properties unrepresentable by current CAD systems are possible to analyse. To tackle these challenges and to strengthen the available set of tools and methods used before a geometry is available, this thesis has focused on the enhancement of an existing functional modelling software developed at the department of Industrial and Materials Sciences at Chalmers University of Technology. The tool uses the enhanced function-means (EF-M) method to represent product architectures side-by-side with alternatives. The problem was divided into four functionality levels, which were to be implemented sequentially into the tool. The four functionalities consisted among others of implementing interface visualisation and characterisation, extraction of the interfaces in design structure matrices (DSMs), the addition of a function-component library and the connection of the tool to an external, third-party simulation software. The benefits of this development are the utilisation of historic data of the organisation in the EF-M models and the connection of the tool to established methods. The DSM models are well-defined in the literature and their extraction from the tool means that the modelled products can be analysed further in external software. Furthermore, the enhanced tool paves the way for architecture discovery by simulation, since many steps of the process can with relative ease be automated.