GR&R (Gage R&R) on a CMM (Coordinate Measuring Machine)

[spacemanspiff]

One of our customers is requesting a GR&R on our CMM itself. I have done GR&R on fixtures in the past but never the CMM itself. In my opinion the CMM should be as repeatable as your standard of dev. on the Calibration cert. Correct?? if not could someone please give me some insight as to how to do this properly.

Thanks in advance
Ed

[Ninja]

Howdy,

No promises that I'm not way off base here...if so I'll learn along with you...

The point of GRR is to measure how repeatable your measuring PROCESS is should an operator run the same process multiple times, and to measure how reproducable your measuring PROCESS is should someone else use that same process. This data is then compared to a specification range. Measuring the capability of the machine used a particular way makes sense. The comparing of that usage to a spec range also makes sense. Rating a piece of equipment to a spec range is where your customer's request ceases to make sense.

You might want to ask your customer what it is that they really want...they are asking for one thing, but what are they actually trying to achieve? Are you the manufacturer of the CMM?

It is common to ask for what you want poorly enough that the words of the question make no sense. Ask for clarification of what the goal is...in my experience, people ask for logical things but word the requests poorly enough that the requests make no sense. Funny thing is that I find myself doing the same thing!

[JAltmann]

You and Ninja are correct. The person asking for this most likely does not truly understand what a GRR is and isnot. I have encountered this in the past, best case is you can educate the requestor of the error in the request.

Most likely they are trying to check a box on a "one size fits all" form. Typically for this vague request i just give them a copy of the calibration certificate and point out the CMM's measurement uncertianty.

Worst case i supply them with a GRR for one of their parts measured on the CMM.

[Rand T]

The Gage R & R on a CMM is not an unusual request and one I have run into before. I have tried to argue against doing these in the past, and decided to just do it since it really wasn't worth the time and effort to debate against it.

The machine is part of the entire measurement system which includes the operator and any fixturing if applicable. I assume the customer that is asking for the R & R has a part which you currently measure on the CMM. I would do the traditional 10 x 3 x3 R & R on their part like any other R & R. I have been in the situation where we only had one dedicated CMM Operator and in that case, I tried to negotiate just doing a repeatability test.

[bobdoering]

In Reply to Parent Post by JAltmann

Worst case i supply them with a GRR for one of their parts measured on the CMM.

This is what I also suspect they need - to ensure the CMM measuring process (loading, unloading, tip choice, program (if used-technique if not), etc.) is adequate.

[spacemanspiff]

Thanks for the input Ninja!
We currently produce aluminum castings for this customer and we use many different gages to verify the parts we produce. We have done GR&R's on all the other gages using pass fail criteria as there is no variable data collected. But we also CMM 1 part per shift on the cmm. So far we have done a CAP study on the parts but now they want a GR&R not on the fixture used but the actual CMM.???

[Sturmkind]

Hi, Spacemanspiff!

I am in agreement with the above posters and that it would be worthwhile having additional conversation with your customer; they may want what Bob D. suggests.

I have run into situations with some critical parts where the CMM probe POSITION relative to the part was found crucial: radius on a small 1.2mm thick stamping. After running 5 parts twice with a complete set-up/fixture load & unload the measurement PROCESS consumed 469% of tolerance.

Upon investigation and running the probe in very slowly on 1 part it was discovered that the break and not the shear was being measured. Once this was corrected and the probe was changed to a ruby cylinder, the GRR dropped to 7%.

Best of luck with this!

[Rand T]

In Reply to Parent Post by spacemanspiff

Thanks for the input Ninja!
We currently produce aluminum castings for this customer and we use many different gages to verify the parts we produce. We have done GR&R's on all the other gages using pass fail criteria as there is no variable data collected. But we also CMM 1 part per shift on the cmm. So far we have done a CAP study on the parts but now they want a GR&R not on the fixture used but the actual CMM.???

I think you are trying to make this too complicated. The CMM on the aluminum casting is collecting variable data, correct? Just think of it as another measuring tool like calipers, micrometer etc.

Also, you said you had previously done a capability study using the CMM on the parts. How do you know the CMM is giving you reliable data without knowing if it passes R & R?