Achieving Design Target in the Presence of Functional Coupling 2007-01-1208

The primary objective of design is to achieve the target value of its function. While principles and techniques of Robust Design address the issue of achieving target values in the presence of different types of variations and disturbances, there exists a unique challenge in achieving design targets when multiple response functions are interrelated. In order to overcome the challenge, we must avoid functional couplings and obtain the interrelationship structure as flexible as possible. In the Axiomatic Design process, such interrelationships are represented by coupling terms in a design matrix. From the targeting aspect of design, it is important to achieve a desirable design matrix structure to, first, avoid any functional coupling in a design matrix and, secondly, maximize allowable sequences of adjusting DPs. This can be achieved by examining the structure of a design matrix to identify off-diagonal terms causing coupled and decoupled sub-structure and then eliminating an optimum set of off-diagonal terms. This paper presented a structured method based on the graph representation to effectively determine the optimum set of off-diagonal terms to eliminate from a large design matrix.