15 Designing for quality

The improvement of the quality of a design is seen as the primary need of industry, but to facilitate this we need appropriate methods for predicting quality and evaluating the long-term quality of an engineer's design (Morup, 1993; Russell and Taylor, 1995; Shah, 1998; Taguchi et al., 1989). However, there is relatively little work published in the field of Design for Quality (DFQ) compared to Design for Assembly (DFA), for example, and little methodology exists as yet (Bralla, 1996). One possible reason for this is that DFQ methods should have the objective of selecting the 'technically perfect' from a number of alternative solutions (Braunsperger, 1996). The word quality, therefore, implies a relative rather than a precise standard from which the designer has to work (Nixon, 1958). This requires a cultural shift of thought in design activities.

Techniques such as FMEA, DFA and Quality Function Deployment (QFD) can enhance the success of a product, but alone they will not solve all product development issues (Andersson, 1994; Jenkins et al., 1997a; Klit et al., 1993). They provide useful aids in the process of quality improvement, but they do not ensure product quality (Andersson, 1994). There exists an important need for DFQ techniques to aid design and support the product development process (Andreasen and Olesen, 1990). In addition, it has been cited that in order to make further reductions in product development time requires new progress in these techniques (Dertouzos et al., 1989).

A substantial review of DFQ and the framework for its application has been proposed by Morup (1993). This identifies eight key elements in DFQ which are placed under the headings of preconditions, structured product development and supporting methods/tools and techniques as illustrated in Figure 1.20. In thinking about DFQ, Morup states that it is convenient to divide product quality into two main categories:

• 'Big Q' which is the customer/user perceived quality

• 'Little q' which relates to our efforts in creating big Q.

Product quality is a vector with several types of quality elements. The Q vector relates to issues including reputation, technology, use, distribution and replacement. The term q can also be considered as a vector with elements related to variability in component manufacture, assembly, testing, storage, product transport and installation. The notion of little q can also be expressed as an efficiency, related to efforts in meeting Q. The issues of q are met when the product meets those systems that are used to realize quality Q. The maintenance of Q relies upon the ability of a business to understand and control the variability which might be associated with the process of product realization. The quality in a product is not directly connected to cost. Every single Q element Qj has corresponding q elements that contribute to cost. Q is fundamentally connected to selling price.

As can be seen from the above, the DFQ and the Q/q concepts are extremely broad in perspective. The general model may be used to drive the considerations of the important issues throughout the stages of production development and in the design of individual components and assemblies. The q element of quality described by Morup is adopted in the CA methodology presented in Chapter 3 of this book.

The link between customer wants/perceived quality, Q, and quality of conformance, q, is a vague area. A large number of problems created at the design/manufacture interface are also caused by technical quality problems, for example wrong material specification, wrong dimensions, etc. These are essentially design communication deficiencies and so are amenable to an appraisal by a methodology of sorts. The flow of information through these quality disciplines is shown in Figure 1.21 where an analogy is made to the design of a simple hole in a plate. Further investigation of technical

DFQ preconditions


-v Strategy i \ deployment

-v Strategy i \ deployment


Measuring system

Measuring system

Structured product development

Figure 1.20 Preconditions for and main elements of DFQ (adapted from M0rup, 1993)

quality, Qq, should be also performed if we are to gain a further understanding of DFQ. For example, the measuring and monitoring of design drawing errors using SPC attribute techniques has been a major step forward in reducing the design changes for an aerospace company.

It is also possible to categorize the different types of DFQ techniques that are required to analyse the several types of quality highlighted above. These are (Morup, 1993):

• Specification techniques which aid product developers in formulating quality objectives and specifications ^ Q

• Synthesis techniques which aid the designer in generating ideas and in detailing solutions ^ Qq

Figure 1.21 Hole in plate analogy to quality and the Q/q concepts

• Verification techniques which verify and evaluate the quality of solutions in relation to the specification ^ q.

There is a need for verification techniques in DFQ that can be used in the early and critical product development phases, where the quality is determined, i.e. can be applied on abstract and incomplete product models (Morup, 1993). The CA methodology is largely a verification technique that aims to achieve this.

The DFQ approach at the verification level has many elements in common with Design for Manufacture (DFM) techniques. DFM helps create a product design that eases the task of manufacturing and lowers manufacturing cost. This is achieved by invoking a series of guidelines, principles and recommendations at the design stage and providing an understanding of the characteristics, capabilities and limitations of the manufacturing processes employed (Bralla, 1998; Kalpakjian, 1995). These design rules, or 'producibility' guidelines, are more generally applied at the component level than the assembly level, although DFA is sometimes, rather confusingly, associated with DFM (Leaney, 1996b; Russell and Taylor, 1995). Producibility guidelines are commonly developed by companies for designing products that are similar in nature to the ones for which the guidelines were written. They are therefore limited in their application because they may not apply to innovative design or where process capabilities are taken to the extreme limits by customer requirements.

Designing for robustness has also been associated with the DFM guidelines (Russell and Taylor, 1995). Robust design has different meanings to different engineering communities. For example, the three descriptions below focus on three different but connected aspects of product design:

• Robust design creates performance characteristics that are very insensitive to variations in the manufacturing process, and other variations related to the environment and time (Lewis, 1996; Sanchez, 1993).

• Robust design is the design of a product or process that results in functionally acceptable products within economic tolerances (Taguchi et al., 1989).

• Robust design improves product quality by reducing the effects of variability (Phadke, 1989).

The first definition focuses on a process orientated design, the second the economic aspects of the design, and the third the impact of variability on the product in use. Although robust design is mentioned as a DFM guideline, no guidance for how to achieve 'robustness' is given. The definition of robust design must be made clear, and more importantly detailed guidance must be given to the designer on what to do to achieve robustness in a practical way. DFM techniques do not specifically answer this question. However, a robust design can be defined as a capable design in the context of the work presented later.

As stated earlier, awareness is growing that cost and quality are essentially designed into products (or not!) in the early stages of product engineering. The designer needs to know, or else needs to be able to, predict the capability of the process used to produce the design and to ensure the necessary tolerance limits are sufficiently wide to avoid manufacturing defects. Furthermore, the designer must consider the severity of potential failures and to make sure the design is sufficiently robust effectively to eliminate or accommodate defects. The major benefit of doing this is to reduce the potential for failure costs. Alternatively, in seeking to control or reduce cost, the safety of the product can be jeopardized, for example allowing production cost to dominate a design decision to the extent that the product does not meet customer expectation. The design phase of a product is crucial because it is here that the product's configuration and, therefore, much of its potential for harm are determined. Those engaged in product design must give a high priority to the elimination or control of hazards associated with the product. This may have to be to the detriment of ease of manufacture, styling, user convenience, price and other marketing factors (Wright, 1989).

The experiences of industrial collaborators and surveys of UK businesses suggest that failure costs are the main obstacle to reducing the costs of quality. There is, therefore, a need for design methods and guidelines to give businesses the foresight to identify product characteristics that depict potential costly failures, particularly in the case of bought out components and assemblies. Designers need models to predict costs at the various design stages. The provision of these techniques is essential for future business competitiveness. Through DFQ we can change over from merely preventing and eliminating quality problems to actively incorporating the level of quality expected by the consumer into the product. At the same time we can see that the target of high product quality is to a great extent compatible with the target of low costs, and thus with the creation of good business (Morup, 1993). We need a culture, professionalism and techniques for Design for Quality.

0 0

Post a comment