Proceedings of ICAD2009
The Fifth International Conference on Axiomatic Design
Campus de Caparica – March 25-27, 2009
ICAD-2009-XX
Copyright © 2009 by ICAD2009
ABSTRACT
Nowadays, due to its critical role regarding product cost and
performance, as well as, time to market, product design is
considered to be at the new frontiers for achieving
competitive advantage. Therefore, to face today’s rapidly
changing business environments, it is extremely important to
adopt a systematic approach to product design, in order to
avoid errors and consequently achieve shorter time-to market
performances. In this context, we will describe a new
approach to support product design, which links Axiomatic
Design (AD) and Multidisciplinary Design Optimization
(MDO), applied in an integrated way at the conceptual design
and the detailed design stages, respectively. Firstly, the
conceptual design stage is undertaken by AD, which is used to
map Functional Requirements (FRs) with the corresponding
Design Parameters (DPs). Even though we try to guarantee
the Independence of FRs, as established by Axiom 1 of AD,
if some remaining coupled relations subsist that is not
prohibitive. Afterwards, the detailed design is carried out by
MDO, considered to be an appropriate methodology to
design complex systems through an adequate exploitation of
interacting phenomena. The proposed approach is applied to
the design of metallic moulds for plastic parts injection, since
the mould makers sector involves constant design and
production of unrepeatable moulds, where uncoupled designs
solutions, mostly due to technological and time reasons, aren’t
common (this sector is strongly influenced by customers who
place enormous pressures on lead-time and cost reduction).
This application points out the high potential of improvement
that can be achieved through the simultaneous improvement
of mould quality, reliability and time to market.
Keywords: Axiomatic Design, Multidisciplinary Design
Optimization, moulds design, coupled designs.
1 INTRODUCTION
In general product development can be described, as an
iterative process where some recursive and repetitive tasks,
dominated by empirical knowledge, are performed until an
acceptable solution is achieved. Due to today’s market
pressure to reduce costs and time-to-market of products, as
well as, to increase its quality, new design approaches must be
adopted allowing faster and efficient product development.
One of these methodologies is Axiomatic Design (AD),
which establishes a systematic and scientific basis for product
design process with the final goal of determining the best
design solution [Suh, 1990]. The basic postulate of AD
approach is that there are two fundamental axioms that must
govern the design process: the Independence axiom and the
Information axiom. The first axiom, states that the
independence of Functional Requirements (FRs), which are
the specific requirements translated from customer’s needs,
must always be maintained. The second axiom establishes that
the best design, amongst designed solutions that satisfy the
independence axiom, is the one that has the smallest
information content.
In addition, AD establish that the design process must
progress by mapping the FRs into Design Parameters (DPs),
which characterized each design solution, in a top-down
hierarchical manner. For each level of decomposition, the
relationship between FRs and DPs can be described
mathematically as a design matrix [A]. According to the
structure of this matrix, there are three types of design:
Uncoupled, Decoupled and Coupled. The Uncoupled design
(most preferred) is characterized by a diagonal matrix, which
indicates the independence of all FR-DP pairs [Jang, Yang et
al., 2002]. The Decoupled design (second choice) is
characterized by a triangular design matrix. Therefore, the FRs
can be answered systematically from FR1 to FRn by
considering only the first n DPs. Finally, the Coupled design
(undesirable) is characterized by a design matrix with no
specific structure. Therefore, a change in any DP may
influence all FRs, simultaneously, meaning that the
independence axiom is not accomplished. Although this type
of design is not promoted by AD, because it does not
guarantee the first axiom , some authors (e.g. [Crawley, Weck
et al., 2004]) believe that there are some cases where it should
be applied, especially when performance, efficiency and
packaging constraints dominate, where uncoupled/decoupled
solutions might not be feasible. This is the case of metallic
moulds for plastic parts injection [Ferreira, Cabral et al., 2006],
where, by technological and time reasons, mould designs are
generally a coupled solution, or has at least some coupled
areas. In this sense, Multidisciplinary Design Optimization
(MDO), which is considered appropriate to design complex
systems trough exploitation of interacting phenomena, can be
A NEW AD/MDO APPROACH TO SUPPORT PRODUCT DESIGN
Irene Ferreira.
iferreira@estg.ipleiria.pt
Centre for Rapid and Sustainable Product
Development,
INDEA, Campus 5
Rua das Olhalvas,
2414-016 Leiria, Portugal
José A. Cabral
jacabral@fe.up.pt
Department of Industrial Engineering
and Management,
Faculty of Engineering,
University of Porto
Rua Dr. Roberto Frias,
4200-465 Porto, Portugal
Pedro M. Saraiva
pas@eq.uc.pt
Department of Chemical
Engineering,
University of Coimbra,
Pólo II - Pinhal de Marrocos,
3030-290 Coimbra, Portugal