Describing and evaluating functionally integrated and manufacturing restricted product architectures

Artikel i vetenskaplig tidskrift, 2018

Within manufacturing companies, the architectural description of how a product is built is typically well defined while the architecture of the product from a functional view describing how the functional requirements are met in the product is often less articulated. For products composed of many components (modular products) such descriptions are clear, whereas few representation schemes are available that treat highly functionally integrated components, where all the functions are satisfied by one integral, homogenous physical structure. In this paper, an approach to describe the architecture for integrated components in the aerospace industry is described. Different regions of the component, termed as sections, are assumed to satisfy the functions required of the structure which are often manufactured by joining (welding) different segments. By assigning sectional and functional information to different manufactured segments of the structure, graphs are created that link the functional requirements and sections. Two different methods, one based on set compositions and other on creating an enhanced function-means (EF-M) tree are used to link the functions to the sections of the component, resulting in different graphs for different types of manufacturing splits for the same component. Comparison of the methods is then carried out using properties of the graphs produced. The method that utilises set compositions performs well for entire component descriptions while the method that uses an EF-M tree to create a graph describes sections (regions) of the component well so that functional relationships can be better described (integration of already existing design knowledge). The product descriptions created can help designers to identify how alternative manufacturing splits impact the functionally defined product architecture which in turn enables both improved manufacturing and improved design decisions.