Functional combination-oriented module identification for adaptable-function mechanical product design

The adaptable-function mechanical product is not only of multiple functions, but also of more complex physical structure than products of fixed functions; hence it is feasible to apply modular design methodology in its design and development, where module identification is critical to the success of the design. For the modular design of this kind of product, it is often necessary to formulate a module by combining components from multiple product functions, rather than from one function alone, let alone by decomposing a function first and then making multiple modules from that function. To this end, this paper attempts to study the strategies and methodologies for functional combination-oriented module identification suitable for the adaptable-function product design and other similar design problems. By investigation on the similarities of function, executing behavior, and physical structure, respectively, a similarity matrix is established first, from which the fuzzy clustering method is exploited to derive various module schemes. After that, an evaluation model is proposed to calculate the module scheme complexity and its redesign complexity, with which the schemes are evaluated and the optimal one is obtained. The identified modules can then be used for further modular design. Finally, the feasibility of the proposed methods is demonstrated by study of two cases, including a household toolkit consisting of a variety of tools, as well as a Francis turbine that is structurally more complex.