Ensayo

J. ENG. DESIGN,

2000, VOL . 11, NO. 1, 17–29

D esign for assembly: in uencing the design process
G. F. DA L G L E I S H † , G. E . M . JA R E D ‡ and K . G. S W I F T † *
There is extensive evidence to suggest that industrial products are being designed with an excessive number of parts and, invariably, many more costly and complex assembly operations than should be the case.A feedforward or proactive approach to the application of Design for Assembly, aimed at in uencing the design process to overcome these problems, is presented. This paper reports on progress towards the development of the Design for Assembly ‘Sandpit’, 1 where Design for Assembly procedures are integrated with advanced computer-aided design concepts to realize an assembly-oriented design environment. A casestudy is included to help illustrate the ways in which the approach can be used in industry.

1. Introduction Designers would rather have tools that can be used earlier within the product introduction process to quantitatively assess candidate design solutions and thus allow them to pursue those that are more likely to lead to competitive products. They do not always welcome tools that criticallyevaluate their completed designs after much development effort. In other words, designers would prefer to have such evaluations carried out at a point when they have the most opportunity to act on the results —the consequences in terms of reduced lead times and improved products are clear. There is extensive evidence to suggest that in industry, products are being designed with an excessivenumber of parts and, invariably, many more costly and complex assembly procedures than should be the case (Boothroyd 1994). The established Design for Assembly (DFA) methodologies (Miyakawa and Ohashi 1986, Boothroyd and Dewhurst 1989, Lucas Engineering Systems Ltd 1993) have had proven successes within manufacturing businesses, and are able to highlight these assembly issues. However, by relying onthe evaluation of completed designs, they are essentially reactive tools. While some design improvements can be implemented at this late stage in design, it is often too late to make any necessary radical changes, and many of the DFA bene ts are then lost. Therefore, the key issue is the integrated application of DFA during the design activity. This necessitates exploring the required DFA tools andthe ways in which they might be successfully applied. While the recognition of this need is not altogether new (Bäßler and Schmaus 1987), relatively little work has been carried out

Revision received November 1999. † School of Engineering, University of Hull, H ull, UK. ‡ School of Industrial and Manufacturing Science, Cran eld University, Cran eld, UK. * To whom correspondence should beaddressed. 1 Sandpit: literally, a shallow pit or container holding sand for children to play in; has additional connotations of an environment for ‘learning by doing’ in an inexpensive medium .
Journal of Engineering Design ISSN 0954-4828 print/ISSN 1466-1387 online © 2000 Taylor & Francis Ltd http://www.tandf.co.uk/journals /tf/09544828.htm l

18

G. F. Dalgleish et al.

to address theproblems associated with realizing the generation of assemblyoriented design. This paper reports on progress towards ‘preceptive’ or ‘proactive’ approaches to DFA. The paper introduced DFA procedures that can facilitate a feed forward approach, and describes an environment for their application —the so-called ‘DFA Sandpit’. A case study is included to help illustrate the ways in which the approach can beused in industry. The work presented is part of a collaborative research project involving RADAN Computational, Computer Sciences Corporation (CSC) and Rover Group.

2. DFA evaluation methods DFA techniques have been around for about 20 years (Redford et al. 1981, Boothroyd et al. 1982) and a number of the methods have been taken up widely in industry. Popular commercially available DFA...