In the early stages of engineering design a considerable amount of experience and knowledge of past designs is used to build and evaluate new empirical models with known design relationships. However, conventional computer-based systems which aim to assist this stage have tended to concentrate on the analytical aspects of the process and have not been successful in accessing this expertise and benefiting from it during synthesis. The paper presents some of the results of a programme of research into methods of representing the knowledge to support modelling during these early stages of the design process. Key features of the modelling at this stage are the use of abstract representations, reuse of past design information, partitioning of designs, and synthesis of concept structures. The work is based on an experimental system, NODES, which was developed to model knowledge of design objects and their associated numerical relations. The utility of the system in creating, evolving and evaluating design solutions is illustrated through a worked example in a typical engineering design application.
