In the new year, I’ll begin with a new icon, which is :
That’s another tool, which the PEO license offers us. To describe it very generally, it inspects mutual interaction of input parameters, and compares their iterations with output parameters. But to make it a bit more clear, I’ll start with a simple example, i.e., a concept of a method for an unconventional design of semi-trailers. At the beginning carried out was the Target Value optimization, its purpose was to determine an optimum height of a semi-trailer, in order to carry the right weight. In other words, we define the weight, which is supposed to be conveyed, while the optimization calculates optimal trailer dimensions (here’s the height) :
So we got an optimal value – 50 T can be transported. OK, but when you evaluate new licenses, you need to pick holes and do some experiments 🙂 Therefore, how can we quickly get a notion of how will the weight of the whole trailer change depending on the density of goods? That’s where we can use Design Of Experiment :
We received in Excel format a comparison of mass-to-density correlations (with the same volumes) – a ready output, in order to prepare a diagram. This happens to be a trivial case, but imagine one, where you have to experiment with many input parameters and higher amount of output (not necessarily linear correlations, as in this case)… isn’t that an experimental approach to a case???
What I’ve described is only a scenario of having to use the PEO again, it should be dwelt on a little and tested, so as to justify that approach. Right, but figuring out the non-standard ways is a human thing 🙂
And what else can we use DOE for? I always look for inspiration on various forums and websites, that way I came upon a very cool description of the use of PEO combined with an analysis of FEA, in which DOE was used in a way opposite to mine, i.e., the “experiment” was conducted first, in order to locate the right area for optimization… A very nice introduction to PEO by the way :