Integrated design is based on concurrent engineering principles.
All stakeholders have to work together throughout the whole design process. So they can exchange and take common decision.
The process emerges step by step by the introduction of constraints by stakeholders.
Emergence is the way complex systems and patterns arise out of a multiplicity of relatively simple interactions
In the design process, we asked the different actors to introduce their different constraints in order to obtain at the end a global product. It is for that we speak about emergence where the details of the final solution are revealed during the process.For such a design process, we consider that we need two different steps:
-The first one is the step of function-structure decomposition
-The second step is really the detailed design done with multiple stakeholdersIn his thesis, Philippe Belloy proved that the first step ends when we get the functional surfaces, these functional surfaces being the consequences of choices for technological elements, esthetic forms or aerodynamic ones. And it is when we get these functional surfaces that we have to ask all the people being concerned by the life cycle of the product to be present to introduce their constraints as soon as they get them.”Philippe Belloy, “Intégration de connaissances métier dans la conception : un modèle pour les pièces mécaniques», PhD thesis”
In order to understand the interest of the first step, let take this exercice:
To take a pen and a sheet of paper, and to draw in one minute what can be a heater for a baby-bottle.
95% of the answer are like this, as the normal thinking is to reproduce a well known object doing such functionality.
But in fact, what is the question ? To draw a heater for a baby-bottle !
We request here about a functionality for heating the baby bottle. And the first question we have normally to answer is: what are the physical principle in order to heat a container ?
What are the possible physical principles:
-By conduction: to heat water with gas (in a pan), or with electricity (as the previous heater)
-With the use of microwave oven using the molecular agitation
-With the use of solar energy
-With the use of chemical reactionOf course the choice of one or an other physical principle depends on the context for the use of this baby bottle !It is absolutely necessary to think about the functionalities which we wish to solve before thinking to a physical solution.Such a step can make it possible to find unusual solutions to solve the problem and thus to think innovation.
The first step of Function-Structure decomposition consists in:
-To focus on a functionality
-To choose the possible physical principles
-For each of them to evaluate the admissible solutions
-To chose one solution or some solutions if we want to prepare variants
-For a chosen solution, to determine if we have new induced functions (the choice of a cooling by oil for an engine requires to envisage a seal, which is not necessary for a cooling by air)
-For each of the solutions, to determinate the functional surfacesThis work is mainly done by technologist, or specialists who implement technology within a field of engineering.
The mechanical engineering field is not the same as the electronic field. Why? Because we have many hundred of experiment and the mechanical science has been optimized and so get a lot of complexities.In electronic technology, we create the science in the same time we create the tools to use it: the computer. And the integration of the components on a chip is such that we increase every year the capacity on a chip in the same time we reduce the size of the connection. In this field, we can think about functionalities, and we transform each of the functions in a specific package of transistors. A specific controller is specified as a tree of functions transformed in the same tree of solutions.In mechanical engineering, we can start with a first function, and with the choice of possible physical principles to get a list of possible solutions. For a function 1 we can choose a structure 1.Perhaps a function 2 will need 2 different structures 2a and 2b to be solved. But the choice of the structure 2b can induced the need to add a new function.If you want to refresh the crankcase of a motor of a bike, you can do it using the air flowing around the motor, or you can do it using oil inside the crankcase. But in this case, you need to assume a new function that is the sealing of the case.Sometimes, the same structure can be used to solve two functions: the design of the crankcase of a gearbox is use to fix the relative position of the shafts but also contain the oil and to permit the transfer of heat between the oil and the external air.But what happens if we want to suppress one of the two function?We saw that in our case the function –structure representation is not a tree but a graph. And it is more complicated to work with a graph than with a tree !
source : ECQA
EAGLE – Standard 2 year: PCB design made easy, Autodesk EAGLE.