Quantcast
Channel: PTC Community : Discussion List - PTC Mathcad
Viewing all articles
Browse latest Browse all 2936

Analyzing Results of Experiments with DOE

$
0
0

Dear all,

I am completely new to DOE in MathCAD, my only experiences with DOE are related to a DOE-Module contained in a simulation software (ADAMS)... and that was a lot of years ago.

 

I am aware that there is a complete set of DOE-tools in Mathcad, I also tried to have a look at the quick-sheets, but I am not understanding how can I use this tools. I would need some kind of E-Book or the like, a "getting-started"-sort of thing.

Anyway I attach an example (without units or real references, but slightly simplified from a real experiment whose results I am requested to evaluate).

What I would like to do is very simple. I do not need to "design" the experiment, that was already done "by hand", so the different combinations of parameters are already given.

There are a few "digital" parameters (P1, P2,... up to HV4) that can take the value 0 or 1.  In my experiment they are pumps, heat-exchangers that can be on or off.  Then there are a few real parameters (F1, F2, ... PT2).  They can be volumetric flows or pressures.  Then there is one result (actually I would be interested in more than one, but for the sake of simplicity I stay with one in this example). It is named "Result".  It can be thought of as a percent output, a temperature, etc.

 

Now given that I would like:

1) a quantitative estimation of the importance of any parameter to the final result

2) a "best fit" function (response surface), maybe with a given degree, that tries to compute the result with a new combination of parameters.  And a quality indicator for this result.

 

Am I on the right way?  Or is what I am trying to do not the purpose of the Mathcad DOE-Module?

 

Thanks a lot for any hint.  I just need a starting point, or some hint where to find help.

Best regards

Claudio


Viewing all articles
Browse latest Browse all 2936

Trending Articles