A graph-based data structure for assembly dimensional variation control at a preliminary phase of product design

Dimensional quality control remains a challenge in assembly of complex products. Currently dimensional variation control research efforts include tolerance analysis and allocation, fixture layout design, assembly sequence planning, and others, without systematically viewing assembly as a proxy for a wide range of design decisions to produce final products with least cost, most productivity and best quality. A literature review in current assembly modelling methods shows that a unified data structure has yet to be developed at a preliminary design phase in order to design product and plan assembly process automatically. This paper serves to develop such a data structure, which captures heterogeneous product and assembly process information available at a preliminary design phase and unifies them in a data structure represented as a hierarchical graph. The graph-based data structure, on one side, facilitates automatic product design and assembly process planning at the preliminary design phase by utilising research results developed in graph theory, and on the other side, allows easier embedding the assembly model into computer aided design (CAD) software since they share the similar data structure. The presented graph-based data structure is illustrated by a sport utility vehicle (SUV) side frame assembly.