Gage R&R lite? This form only uses two operators measuring five parts one time

[Qaware]

I recently came across a form for Gage R&R which I feel quite skeptic of. This form only uses two operators measuring five parts one time per operator. R-mean is then calculated, by using differences between operators, and multiplied with a constant=4.33, to get R&R-value. I think this looks strange. Can anyone help me understand???

 

[Atul Khandekar]

I think this is the 'Short Method' or "Range Method' given in the MSA manual. The Short method provides a quick approximation or the overall picture of measurement system variability. The measurement system variation is not decomposed into repeatability and reproducibility.

 

[Qaware]

constant?

Thank you for your answer, I still think that the amount of data is a little thin. Where does the constant come from, and how is it derived? How will it change if I would use the same method but added one operator?

 

[Atul Khandekar]

Well, this is just a quick approximation. In the words of the MSA manual:

"This approach has the potential to detect an unacceptable measurement system (i.e. GRR > 30%) 80% of the time with a sample size of 5 and 90% of the time with a sample size of 10."

If you have the MSA manual, you'll find the factors in the d2* table depending upon the number of operators and number of parts used in the study.

 

[Qaware]

Thank you

Thank you very much for that answer, helps me alot. I don't have the MSA manual you are referring to. Who has published it?
_______________________________________________________________

I think I found it myself. MSA 3rd edition- published by AIAG. Correct?

Appreciate your help.

 

[Atul Khandekar]

Correct!. Here's the link:
(broken link removed)

 

[Casana]

Atul,

May I ask where in the MSA does it define the 'Range Method"?
Thanks!
-Ana

 

[Atul Khandekar]

Atul,

May I ask where in the MSA does it define the 'Range Method"?

Ana,
This method is described on Page 97,98 of the AIAG MSA Manual- 3rd edition.
AIAG recommends GRR% of Process StDev.

Also see the section 8.3 of this Daimlerchrysler document:
https://gsp.extra.daimlerchrysler.com/mfg/amedd/gages/index.htm
This document recommends GRR% of Tolerance. I think this is the reference Qaware is talking about in the first post.

Hope this helps.

 

[Jim Wynne]

Ana,
This method is described on Page 97,98 of the AIAG MSA Manual- 3rd edition.
AIAG recommends GRR% of Process StDev.

Also see the section 8.3 of this Daimlerchrysler document:
https://gsp.extra.daimlerchrysler.com/mfg/amedd/gages/index.htm
This document recommends GRR% of Tolerance. I think this is the reference Qaware is talking about in the first post.

Hope this helps.

I think it's important to stress the difference between process variation and within-study variation; imo the former is almost always preferable, if process variation is reliably known when the study is done.
Whether % of tolerance or % of variation (or both) is used is a matter of choice (that might, as Atul suggests, be dictated by the customer) but if you choose one or the other, you should be able to tell why.

 

[Atul Khandekar]

I think it's important to stress the difference between process variation and within-study variation; imo the former is almost always preferable, if process variation is reliably known when the study is done.

Whether % of tolerance or % of variation (or both) is used is a matter of choice (that might, as Atul suggests, be dictated by the customer) but if you choose one or the other, you should be able to tell why.

Typically based on what the measurement system is being used for: Process control/improvement or inspection.