Modelling and analysis of system robustness for mechanical product based on axiomatic design and fuzzy clustering algorithm

Robust performance is the most important concern in the design of any product, especially in system design stage that precedes parameter design, because it actually determines the attainable level of product robustness in the parameter design phase. In this article, a framework of modelling and analysis of system robustness is proposed, which includes system modelling, cluster analysis and design of experiments. In the process of system modelling, the metamodel of general design theory was utilized to describe the function–structure model of product design, and the customer needs are transformed into functional requirements. Based on the independent axiom and zigzag mapping mode of axiomatic design, the functional requirements are mapping to design parameters, and the design matrix is created, which is then converted into design structure matrix by identifying the relationship between functional requirements and the sensitivity of functional requirements to design parameters. The fuzzy clustering algorithm is utilized to cluster the design parameters and to group the system components into modules in design structure matrix, and the interface among modules can be identified and system robustness incidence matrix is developed. Then the incidence parameters are considered as controllable factors, and experimental design techniques are utilized to analyse the influence of incidence parameters on the design objectives, if any, that may result in a robust system. The proposed framework is illustrated with the trolley design of overhead travelling crane