"Subset" DOE

Hello,
I have a question regarding a screening design (resolution IV to get information on interactions). We consider a process which is run in two steps (factors A, B, C, D for step 1 and E, F, G, H for step 2; all factors are quantitative). I have the possibility at the output of the first step to split the product from one experiment in 10 samples which can then be submitted independently to the second step. At the end of the second step, I can do my analysis on final samples to have my responses.


What would be your recommendation? To run a 24-1 fractional factorial design (8 corner + 2 central points) for the first step and for each experiment to run a 24-1 fractional factorial design (8 corner + 2 central points) for the second step. It will represent 10*10 experiments, is it worth to do it? Samples issued from central points of the first DOE should be submitted to only central point of the second DOE (no second DOE for these two points)? It would lead to 82 experiments.

Thanks in advance for your advice.

 

Thanks a lot for your answer, I didn't know this kind of design. I'm going to do my homework and try to create my worksheet. I work with MODDE 13 but there is no split plot design wizard available.