Minitab - Design combinations (DOE)

[dykkms]

Hello,
I am planning experiment in the field od welding.
I am going to weld two different materials together. Factorial design is almost good for me, but it produces some unnecessary fields.
Example:
Factor: Material 1: levels: A, B, C
Factor: Material 2: levels: A, B, C

I need to make experiments for all combinations, but
AB = BA
AC = CA
BC = CB
(because it does not matter, which material is first and which is second).

How can I create this kind of DOE?

Thank you,
Michal

 

[Miner]

Re: minitab - design combinations

Are the materials the only factors in your experiment?

If so, try a 1-way ANOVA of the material combinations in which you are interested (i.e., AA, BB, CC, AB, AC, BC).

 

[dykkms]

Re: minitab - design combinations

Hi, thank you for answer!

No, there are another 3 factors.

I was also thinking to create only one factor for welding joints (2 materials), but then it would be very hard to analyze results, e.g. which material is best for welding.

Thank you!

 

[Miner]

Try using a General Full Factorial. One factor would have 6 levels for the material combinations, while three additional factors would have 2 levels (presumably). The best material would still be discernible by the best level of that factor as well as in any interactions that might exist.

 

[dykkms]

Thank you, that sounds as a very good plan to me. I hope it will show the results nicely afterwards.