Page 233 - Design for Six Sigma a Roadmap for Product Development
P. 233
204 Chapter Six
USL
6
Transfer Function
T
LSL 6
1 2
DP
DP' DP''
Figure 6.12 Transfer function and optimization.
Consider two levels or means of a design parameter (DP), level 1
(DP') and level 2 (DP''), having the same variance and distribution. It
is obvious that level 2 produces less variation in the FR than level 1.
Level 2 will also produce a lower quality loss (see Chaps. 13 to 15) sim-
ilar to the scenario at the bottom of Fig. 6.13. The design produced by
level 2 is more robust than that produced by level 1. When the distance
between the specification limits is 6 times the standard deviation, a
Six Sigma optimized FR is achieved. When all design FRs are released
at this level, a Six Sigma design is obtained.
The black belt and the rest of the DFSS team needs to detail the
transfer functions in context of the DFSS algorithm if they want to opti-
mize, validate, and predict the performance of their project scope in the
use environment.
However, the team should be cautious about the predictability of
some of the transfer functions due to the stochastic effects of noise fac-
tors in the use environment that are particularly hard to predict. The
team should explore all knowledge to obtain the transfer functions
desired, including drawings. For example, stackup based on tolerances
may contain descriptions of functionality that is based on lumped-mass