Supporting design knowledge exchange by converting models of functional decomposition

Abstract

The concept of technical function is used with different meanings in the engineering domain. This ambiguity hampers the exchange of design knowledge across functional modelling approaches. This paper investigates three issues relevant to handling this communication problem. First, it provides a clarification of engineering meanings of function and functional decomposition (fd) models. Secondly, it presents an assessment of the benefits of co-existing functional conceptualisations. Thirdly, following the positive outcome of this assessment, it presents a conceptual strategy to exchange design knowledge across modelling approaches. This strategy does so by converting models of fd across approaches. It is shown that the proposed conversion strategy is able to retain design knowledge that is lost when using another methodology for the conversion of fds. This paper ends by identifying a future research agenda for model conversions.

Keywords:

• design knowledge exchange
• technical function
• functional decomposition
• model conversion

Acknowledgements

I thank the guest editors, four anonymous reviewers and Pieter Vermaas for their valuable comments. This research was supported by the Netherlands Organization of Scientific Research (NWO).

Notes

I do not consider user functions – intentional behaviours carried out by an agent using a device – in this paper. For an in-depth analysis of such functions, see van Eck (2010a).

In these approaches, the concept of behaviour is introduced alongside the concept of function. The behaviour concept takes care of physical laws in these approaches.

Of course, the constraint that all design solutions are compatible with one another (and, hence, all functions are realised) is a crucial constraint in every functional modelling framework. Modelling frameworks differ, however, in which design phase this constraint must be satisfied. In some approaches, fds should satisfy this constraint (Chakrabarti and Bligh 2001), whereas in others, it should be satisfied in later design phases (Deng 2002). Thus, only in some approaches is it a constraint that applies to fds.

Figure 1 gives the same adaptation of the FB fd as Ookubo et al. (2007) present. This adaptation consists in excluding several operations on flows which are described in the original FB fd. The vertical lines intersecting the ‘human force’ flow and the ‘staples’ flow represent this exclusion.

FCO fds do not take physical laws into account. This makes perfect sense, since in the FCO approach the concept of behaviour takes care of physical laws. An example given by Sasajima et al. (1996) illustrates the point clearly. They describe the behaviour of a particular device as dividing an input saline solution into pure salt and a saline solution. The function ascribed to this behaviour is specified as ‘producing salt’.