DOE Resources

[ubrben]

Could anyone recommend a good reference on Design of Experiments? I'm particularly interested in fractional factorial screening experiments.

Regards,

Ben

Senior Design Analyst
Dunlop Tyres
Motorsport Division

 

[plasgears]

Design of Experiments is a buzz phrase that came out of QS9000 programs. The experiments chosen should have basis in sound engineering. If failure is predicted after a long series of load cycles, then the testing should be performed in the same manner. Loads should not exceed the actual load in service. Test temperature should be realistic. Rest time should be realistic unless a cooling fan is introduced to moderate temps during rapid cycling.

I have toyed around with concepts of degradation factors. How do you demonstrate required life if the lifetime is 20 years or more? The answer can be arrived at in a reasonable time if you: increase the oper temp, reduce the lube, increase the load, etc. It is pure conjecture at this point how to arrive at the degradation factors. This is good material for advanced study.

 

[Hydroformer]

Sounds like you're in search of Taguchi methodology ??

http://www.itl.nist.gov/div898/handbook/pri/section5/pri56.htm

*Without data, you're just another person with an opinion.*

Hydroformer

 

[GregLocock]

"Design of Experiments is a buzz phrase that came out of QS9000 programs."

Gosh QS9000 must have been around for an awful long time, say 1950.

http://en.wikipedia.org/wiki/Factorial_experiment

for an overview of factorial experiments, Hydroformer's reference is good if you follow all the links, but does not seem to include much information about exactly which interactions and confoundings occur with a given experimental design.

I use proprietary stuff from the Taguchi institute

In particular I do NOT recommend this book (nothing wrong with it, but it is not what you need)

http://www.amsup.com/publications_software/description/08.htm

This one is useful, but not very explanatory

http://www.amsup.com/publications_software/description/05.htm

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Underdamped]

You may also want to check out Statistics for Experimenters by Box, Hunter, Hunter. The stat guide in Minitab is also pretty good if you have access to it.

 

[IRstuff]

Also:

http://www.statsoft.com/textbook/stexdes.html

from:

http://www.statsoft.com/textbook/stathome.html

TTFN