An experiment to compare the combined array and the product array for robust parameter design

The paper reports a robust parameter design experiment, where we have used Taguchi's product array and a combined array simultaneously. This was done as part of a research project dealing with experimental design to optimise the process of sheet metal spinning. We found that the classical analysis of the product array identified a control factor, which has an influence on the variance of the response. There were no indications of this effect in the combined array. A confirmation experiment supports the finding that this factor in fact is a control factor. It seems that its effect on the variance is due to several two- and three-factor interactions that were not large enough to be found in the combined array. In the authors' opinion this shows that the use of a product array and the classical analysis provides robustness against imprecise model assumptions.