Method for Functional Design of Axiomatic Design Based on Connections

Axiomatic design（AD） is a popular design method,and satisfying the independence axiom is the basis of AD.However,AD doesn＇t provide methods to decompose functions then keep them independent and to handle coupled design.A few of ways of handling coupled design are mainly passive resolutions when coupled design exists,but not efficient to each product design.Hence,this paper presents an innovative approach to design and decompose functions of complex products based on functional connections,aiming at actively avoiding functional coupling.By contrasting with component networks,four kinds of relations among functions are identified,including spatial,energy,material,and information connection.Then the definitions of these relations and the dominant connection are given.Based on the definitions,the principles of functional decomposition and design are developed,in which each non-leaf function is broken into sub functions centered on its dominant connection with avoidance of functional cross and coupling,and sequentially satisfies the independence axiom.Then the operational flow of the proposed approach is constructed.Determining the dominant connection of a function,decomposing the function into sub functions in terms of the dominant connection and reverse examination and optimization are planed as the core steps in each zigzagging.Input process output（IPO） analysis is introduced to obtain the dominant connection of a function,some rules for examining and optimizing the decomposition results reversely according to oriented object theory are presented as well.An illustrative example about the pouring function of squeeze casting equipments presented demonstrates how to use the proposed approach,and indicates its effectiveness.The proposed approach expands the principles of AD,constructs a guidance policy for independent functional design of complex products based on AD,and can help decrease or actively avoid coupled design and improve design efficiency.