Page 213 - Design for Six Sigma a Roadmap for Product Development

186 Chapter Six

6.2 Design Analysis
Axiomatic design provides the zigzagging method as the means to define
the process and physical structures. A structure can be defined as an
input-output or cause-and-effect relationship of functional elements.
Graphically, it is depicted in a “block” diagram that is composed from
nodes connected by arrows depicting the relationships. A structure
should capture all design elements within the scope and ensure correct
flowdown to critical parameters. A structure is captured mathematically
using mapping matrices with matrices belonging to the same hierarchi-
cal level clustered together. Hierarchy is built by the decomposing
design into a number of simpler functional design matrices that collec-
tively meet the high-level functional requirements identified in step 2 of
the DFSS algorithm. The collection of design matrices forms the con-
ceptual physical or process structure. A structure provides a means to
track the chain of effects for design changes as they propagate across the
design. The decomposition starts by the identification of a minimum set
of functional requirements that deliver the design tasks as defined by
the customer and obtained from phase 2 QFD. The decomposition is
guided by both the design axioms and the creative and heuristic process
of function definition through logical questions as offered by the zigzag-
ging method (Chap. 7).The efficient use of design principles can gear the
synthesis and analysis activities to vulnerability-free solutions. The
deployment of design axioms seems to be promising for two reasons:
(1) history tells us that knowledge that is based on axioms will con-
tinue to evolve as long as the axioms are maintained, and (2) it
strengthens the emphasis placed by robust design on the functional
structure of the design. The deployment of design axioms achieves
two main objectives: to uncouple or decouple design and to reduce
design complexity (see Chaps. 5 and 7). Both objectives will eliminate
or reduce conceptual design vulnerabilities.

6.3 DFSS Design Synthesis
The design mappings (matrices) conducted in the DFSS algorithm
(Chap. 7) are conceptual representations at different hierarchical levels.
These matrices don’t stand alone, and a complete solution entity for the
design project needs to be synthesized in the physical structure and
later in the process structure. The ultimate goal of design mapping is to
facilitate design detailing when the mathematical relationships are iden-
tified in the form of transfer functions. Design mapping, a design analy-
sis step, should be conducted first prior to design synthesis activities.
A detailed transfer function is useful not only for design optimization
but also for further design analysis and synthesis activities, in particular,