This work looks at two different "Design of Experiments" (DoE) methods for
defining an operating window in the sheet metal stamping process. The first
involves the use of replicates at the different experimental points, while
the second is a nonreplicated method. The two methods are compared by look
ing at the relationship results produced and the indication of variation in
the process. It is found that the results from both the methods are very s
imilar. However, the replicated method provides a greater level of confiden
ce in the results. In the stamping process, where performing large numbers
of replicates is expensive in both time and money, the nonreplicated method
provides a cost effective way of understanding the process. (C) 2001 Elsev
ier Science B.V. All rights reserved.