Page 119 - Design for Six Sigma a Roadmap for Product Development

94 Chapter Three

Step 3: Evaluate design alternatives. Several design alternatives might
be generated in the last step. We need to evaluate them and make a
final determination on which concept will be used. Many methods can
be used in design evaluation, including the Pugh concept selection tech-
nique, design reviews, design vulnerability analysis (El-Haik 1996,
Yang and Trewn 1999), and FMEA. After design evaluation, a winning
concept will be selected. During the evaluation, many weaknesses of
the initial set of design concepts will be exposed and the concepts will
be revised and improved. If we are designing a process, process man-
agement techniques will also be used as an evaluation tool.
The following DFSS tools are used in this phase:
■ TRIZ
■ QFD
■ Axiomatic design
■ Robust design
■ Design for X
■ DFMEA and PFMEA (design and performance failure mode–effect
analysis)
■ Design review
■ CAD/CAE (computer-aided design/engineering)
■ Simulation
■ Process management

3.7.3 Phase 3: Optimize the design (O)
The result of this phase is an optimized design entity with all functional
requirements released at the Six Sigma performance level. As the con-
cept design is finalized, there are still a lot of design parameters that
can be adjusted and changed. With the help of computer simulation
and/or hardware testing, DOE modeling, Taguchi’s robust design meth-
ods, and response surface methodology, the optimal parameter settings
will be determined. Usually this parameter optimization phase, in prod-
uct DFSS projects, will be followed by a tolerance optimization step. The
objective is to provide a logical and objective basis for setting manufac-
turing tolerances. If the design parameters are not controllable, which is
usually the case on the DFSS product projects, we may need to repeat
phases 1 to 3 of DFSS for manufacturing process design.
The following DFSS tools are used in this phase:

■ Design/simulation tools
■ Design of experiment