Page 610 - Design for Six Sigma a Roadmap for Product Development

564 Chapter Fifteen

M

Output y
Process
h(M,M*)
Combined
control
signal
M*
Noise factors
Figure 15.19 Two signals.

two signals work together to control the magnitude of transformation,
as illustrated by Fig. 15.19.
The combined signal is a function of two separate signals:
Combined signal h (M,M*) (15.9)

The two most common types of h(M,M*) are
1. h(M,M*) MM*
2. h(M,M*) M/M* or h(M,M*) M*/M

We will use the following example to illustrate how to find the correct
h(M,M*) and compute sensitivity and S/N.
Example 15.5 Chemical Formula for Body Warmers (Taguchi et al. 2000)
Disposable body warmers are chemical pockets that can be fit into sports or
outdoor outfits. The chemicals inside the body warmer will slowly react and
generate heat. The heat generated will keep the wearer warm in outdoor
conditions. The functional requirement (FR) of body warmer y is the heat
generated. There are two signals:
1. M heat-generating time. It has four levels: M 1 6, M 2 12, M 3 18,
M 4 24 (hours).
2. M* amount of ingredients. It has three levels: M* 70 percent, M*
2
1
100 percent, M* 130 percent standard amount.
3
Clearly, more time will generate more heat, and more chemicals will also gen-
erate more heat; in this example h(M,M*) MM* is a good choice. Table 15.7
gives a complete output data set for an inner-array run.
Using h(m,M*) MM*, we can transform Table 15.7 into a data set with
combined signal factor MM* as shown in Table 15.8.
Using Eqs. (15.6) to (15.8) and MINITAB, we find that the regression
equation is
Amount of heat 15.6 MM*
MSE 3738