Page 285 - Design for Six Sigma a Roadmap for Product Development

Axiomatic Design 255

as having an overall uncoupled or decoupled design by conducting the
process mapping and physical mapping concurrently.
As expressed by Table 8.2, the following scenarios are observed:
■ An overall uncoupled design is achieved only when both mappings
are uncoupled.
■ An overall decoupled design is achieved when
Both mappings are decoupled, having similar triangular orientations.
Either mapping is uncoupled while the other is decoupled.
■ An overall coupled design is achieved when
At least one mapping is coupled.
Both mappings are decoupled with different triangular orientations.

With everything equal and left to chance, the odds are given by the
probability distribution shown in Fig. 8.11. Design should not be left to
chance, and the design community should endorse and adopt the DFSS
approach.
In addition, Table 8.2 indicates where it is easier to implement a
change for a problem solving, before or after release, without caus-
ing new problems or amplifying existing symptoms of the FRs. A
design change in the form of a DFSS project can be more easily
implemented and controlled in the case of uncoupled and decoupled
designs than in the case of a coupled design. A DFSS project may be
launched to target soft or hard design changes. Whether soft
changes are solution-effective depends on the sensitivities, nominal
and tolerance settings. Hard changes require alterations of the PV
array, the DP array, or both, namely, a new design. Hard changes are
usually followed by a soft-change phase for tuning and adjustment.
In either case, the major cost is the expense of controlling the solu-
tion implementation.
Altering (hard changes) or adjusting (soft changes) the x values can
be used to uncouple or decouple a system. Unfortunately, hard changes

1/16

Uncoupled design
Decoupled design
Coupled design
6/16
9/16
Figure 8.11 Probability of design.