Title: Design-for-six-sigma for multiple response systems

Authors: Gordon J. Savage, Young Kap Son

Addresses: Systems Design Engineering, University of Waterloo, Waterloo, ON N2L 3G1, Canada. ' Systems Design Engineering, University of Waterloo, Waterloo, ON N2L 3G1, Canada

Abstract: An important upstream activity in the overall design of a system is the allocation of the means and tolerances. This is a daunting task because of the need to satisfy multiple competing demands that arise from performance, cost and quality policies. Herein, the so-called design-for-six-sigma is adopted for the allocation process whereby the philosophy of zero defects in Six Sigma is now applied to the system performances. Probability-constrained optimisation is invoked. Robustness and cost measures are required. The production cost provides the objective function to be minimised and a maximum allowable system probability of non-conformance (e.g. a system defect rate) provides the primary design constraint. The design of an electro-mechanical servo system with three responses, three control variables and two noise variables is given. Means and tolerances for the three control variables plus system costs for progressively lower product defect rates show the practicality and potential of the approach.

Keywords: mean design; tolerance design; probability; design for six sigma; DFSS; defect rates; production costs; multiple responses; optimisation; electromechanical servo systems.

DOI: 10.1504/IJPD.2008.016369

International Journal of Product Development, 2008 Vol.5 No.1/2, pp.39 - 53

Published online: 22 Dec 2007 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article