I have designed the optimization experiment using Box-Behnken approach.
What should I do if any of the factors combination fails, for example because the aggregation occurs.
Should I review whole optimization or is there any method to skip the particular factors combination?
And if I need to review the whole experiment, what method should I use to evaluate boundary factors values? Screening methods I have seen require at least 6 factors to be screened.
Any help is appreciated.
Greetings.
Hello,
DOE is more of a trial and error approach as far as choosing the boundary values of the factor levels is concerned, this depends explicitly on the domain knowledge, that is, the theoretical information involved about the experimental system.
Removing one or more factor combinations from the design usually makes the design unbalanced, resulting in loss of orthogonality among contrasts, additional confounding etc.... It is important to choose the experimental space such that the experiments can be actually performed within that space, if experiment fails or goes wrong at one or more experimental factor combinations, it's best to repeat the complete experimental setup and collect data.
About the DOE issues you've mentioned, I believe you will get further insights into them, from the articles on DOE experimentation by stat-ease team of experts available at their website, from the book on DOE experimentation by Anderson and Whitcomb and the book on Response Surface Methodology by Myers, Montogomery and Cook.
If particular factor combinations are leading to "loss of data" because you can not evaluate it, then you can set so-called constraints in your future DOE. As a result, you will work with D- (in screening) or I-Optimal (in optimization) Designs. You need these computer generated designs in your case, because the design is not orthogonal so classical factorial designs are not appropriate.
https://www.statease.com/docs/v13/tutorials/optimal-numeric/