Gage R&R Study with CMM

[D-Train]

I've been asked to do a Gage R&R project using the CMM as the gage, and a step gage as the control subject. We are trying to determine the R&R of the operators and of the CMM, so we are using the step gage, because we know there will be no variation in that object. I'm going to be doing 3 people and 1 part, so my question is...has anyone ever done a R&R with just one part instead of 3? So if the recommended Gage R&R is 3by3by30, being 3 people, running 3 parts, 30 times...what is your experience and advice for doing this with 3 people and only one part? How many times would this part be run by each person? How would you gather the data in Data Page and convert it into a R&R report? I know MiniTab is usually used, and there are directions for how to use this type of R&R on MiniTab, but I'm being told to use DataPage, so I'm hoping someone has some input on how this would be done.

Thanks everyone.

[bfire85]

Sounds like you want to do a Type 1 Gage R&R. One operator and run the part 30 times refixturing it every time.

As for Datapage, I'm not well versed in the software. We have it, but we mainly use it for data collection and use Minitab for studies.

[VinniUSMC]

They want you to do a gage RR to the CMM? With a step gage?

What is going to change from run to run? How is the CMM being tested?

Are you "loading" the part somehow? Gage RR is of the setup, not the CMM.

Rewriting the program? Maybe you'll get info about programming ability.

If you don't change the part, or the program, or the setup, then you are only doing a repeatability.

I think it's been argued, many times, here that a GRR of the CMM is not possible.

[D-Train]

That's what I thought to, setting it up the fixture is the exact same place, the exact same way, will not give you any operator variation. I guess we are doing this way to see if the operator can take it off and set it up the same way repeatedly. Also it's to check and see if the CMM can repeat the same data.

This is just how it was explained to me. And since i've only been part of running parts for a GR&R once before, i'm not real familiar with what kind works best and for what reason, this is why I was reaching out to see if anyone could fill me in on a "possibly" better way. Or easier way for that matter.

[VinniUSMC]

I don't think there is /any/ way to GRR "the CMM". The software is the operator for purposes of measuring. You can find out that a program isn't made very well. You can find out that a person can't put a part in the same place each time. But you can't find out anything meaningful about the CMM itself (beyond repeatability, which you can do with 1 operator, 1 part, 10+ runs, never moving the part, skipping alignments after the first run).

You can perform a Gage RR of a /gage/ on a CMM. But you can't perform a gage RR of "a CMM".

There are roughly half a million (give or take, lol) threads on this forum, and over at the forum of the Guys of CMM that discuss this...

[D-Train]

Thanks Vinni I'll go check into those.

[markybarky10]

What are you actually measuring on the cmm? hole? wall thickness?

[D-Train]

A step gage is being measured, and the distances between the steps.