Page 178 - Design for Six Sigma a Roadmap for Product Development

Design for Six Sigma Project Algorithm 151

Solution Table Legend
Function 1 2 … … j m
FR : Function indexed i in
i
FR 1 S 11 S 12 … … S 1j … S 1m the functional structure
FR 2 S 21 S 22 S 2j S 2m S : The physical solution
j
of solution j
:
S : The solution entity of
ij
: group ‘j’ that physically
FR i S i1 S i2 S ij S im translate function i
(e.g., hardware,
: software, field effect)
FR n S n1 S n2 S nj S nm
Figure 5.10 The design synthesis matrix.
of energy, material, and information is properly identified. These are
the requirements to conceive sound design structure. A structure is a
description of the design in the concerned mapping (see Sec. 5.3.8 and
Chap. 8). The design is first identified in terms of its FRs and then
progressively detailed in terms of its lower-level functional require-
ments and design parameters in the form of design matrices
(mappings). This hierarchy is an output of the zigzagging method
employment in the design structure detailing task. Normally, each
functional requirement can be delivered by several possible DPs in a
given structure hierarchical level within the structure. Therefore, the
synthesis matrix exercise should be conducted at all levels of design
structure.
Assume that we have a design array of n FRs, and FR i is the ith
row in the array. In addition, assume that an arbitrary functional
requirement, say, FR i (where i 1,2,3,…,n) can be delivered physi-
cally by j 1,2,…,m i DPs. A synthesis matrix cell, say, S ij , in the
matrix is the design parameter indexed j, DP j , of functional require-
ment indexed i, FR i . The identification of all possible alternative
solutions (DPs) per a functional requirement may be facilitated by
the use of the morphological approaches of Zwicky (1984) and TRIZ
methodology (Chap. 9).
Several feasible high-level and undetailed concepts are usually gen-
erated using the synthesis matrix. This generation of multiple concepts
poses a selection problem, specifically, which concept to select for
further detailing in the DFSS algorithm. The DFSS team must select
the best concept using the Pugh concept selection method.*

*The concept selection problem was formulated by El-Haik and Yang (2000a, 2000b)
as an optimization problem. They provided an integer programming formulation, both
fuzzy and crisp, for the concept selection problem and employed design axioms as selec-
tion criteria.