Fil Salustri's Design Site

Site Tools


Table of Contents

Function Modelling

General Notes on Function Modelling.

The terms purpose, function, and behaviour are often used interchangeably - or at very least with extremely ambiguous definitions. One man's function is another man's behaviour. The lack of precision in understanding these terms can lead to confusion during design. This project seeks to clarify the three terms are develop a model that crisply distinguishes between them. The model is then used to guide the development of design methods that should allow simpler treatment of design problems.

Think of a pump.

What kind of pump did you think of? And, perhaps more importantly, did you associate some particular use with that pump? Did you think of a reciprocating pump, or a centrifugal one. Did you think of a blow-dryer for hair (yes, Virginia, that's a pump in there although we call it a “fan”). Or did you think of a pump for fluids (e.g. a gas pump)?

When we think of an engineered artifact as vaguely labelled as “a pump”, we will think of some specific pump from our experience. This is because we cannot have a memory of something that does not exist. And the general concept of being a pump is not something that exists, so we can't have a memory of it. That is just how the brain works.

While this is perfectly natural human behaviour, it does pose a problem for designers. Design engineers need to develop artifacts from general, abstract ideas of what those artifacts do. That is, design engineers need to separate form from function, but function, being abstract, is not something we can think about as easily as we can think about form.

In order to assist in doing this, it helps to have a structured set of concepts which can be learnt and them applied via methods to particular problems.

In the literature, terms like purpose, function, and behaviour are often used vaguely, without crisp definitions. And when definitions are given, they are usually contentious at least in certain areas of application (or, at least, I think they are).

So the goal of this project is to develop an understanding of these terms that is operational - that can guide the actions of a designer.

Think of the pump again. Say it's a centrifugal pump used in a pipeline of some sort. What does the pump do? It forces fluid through the pipe. But if we think in general terms, we can't really refer to the pipe, because the concept of pump is not specific to /pipes/ alone. What about pumps that are not connected to pipes at all? And what if it's not a centrifugal pump?

What is the universal function that all pumps provide? They impart kinetic energy. Consider a pump that raises a fluid from a lower reservoir to a higher one. One might think that this pump imparts potential energy (by raising the fluid), but that is not correct. It is the /system/ containing the pump and the two reservoirs that impart potential energy to the fluid. The pump adds kinetic energy to the fluid, which converts it to potential energy as it moves up the gravity well into the higher reservoir. Again, the pump by itself provides only kinetic energy.

Since we're talking about functions, we can use the notion of a transformation; a function transforms its input into its output. In the case of the pump, the transformation is an addition of kinetic energy. Indeed, we could say that the (general) term pump is just a label for this kind of transformation (as well as a label given to any object that provides this transformation).

But where does this leave the centrifugal pump in the pipeline? Can we connect the general function pump to this specific case? The specific function is the purpose of the general function in a specific setting. That is, the specific case answers the question Why is the general function needed? Why do we add kinetic energy to a fluid?

We might be tempted to say because we need to move it through the pipeline. But again, that's too specific. We do not want to refer to the pipeline because even the specific pump might be used elsewhere besides a pipeline. Here's a better, albeit obvious, answer: we add kinetic energy to the fluid, to move it. The pipeline acts to contain the fluid and direct its movement in desired directions, which has nothing to do, transformationally, with the pump.

Conversely, the general function specifies how we achieve the specific function (i.e. purpose). How do we move the fluid through the pipeline? By pumping it.

You may think I'm cheating here: the specific function has changed a bit. We started by saying “pump a fluid through a pipeline” and ended up with “moving a fluid through a pipeline.” “Pump” and “move” are not necessarily the same. I will get around this problem by recalling that design starts with function, not form. The difference between “pump” and “move” is that “pump” makes commitments to /form/ that (a) “move” does not make1) and (b) are not necessarily warranted at the beginning of a design process, unless there are constraints on the problem that allow us to “decide” that pumps must be used from the outset. That is, we can replace “pump fluid through a pipe” with “move fluid through a pump” without loss of intention - we're still saying what we mean.

We can summarise all this in Figure 1.

**Figure 1:** Relationship between purpose and function.

Now let's look at another part of the problem: why do we need to move fluid in a pipe?

ATTENTION TBD …To Be Continued.

See Also


<refnotes> notes-separator: none </refnotes>

I don't like this. Pump is a specific type of movement and not a commitment to form. Must rethink this.
research/function_modelling.txt · Last modified: 2020.03.12 13:30 (external edit)