GRR for Coordinate Measuring Machine
Six Sigma – iSixSigma › Forums › Old Forums › General › GRR for Coordinate Measuring Machine
- This topic has 5 replies, 4 voices, and was last updated 19 years, 1 month ago by
David Oakley.
-
AuthorPosts
-
July 18, 2003 at 7:47 pm #32828
Just looking for any suggestions on conducting a GR&R study on a coordinate measuring machine (CMM). Should we include several different operators measuring the parts, or does a CMM pretty much eliminate operator variation?
Any recommendations would be greatly appreciated. Thanks.0July 18, 2003 at 8:49 pm #88127
Louie WrightParticipant@Louie-WrightInclude @Louie-Wright in your post and this person will
be notified via email.Chris,
If you program your cmm for automatic programs (parts that are aligned using machine coordinates) or you auto align after manually touching off, then the operators shouldn’t make a difference. If it is a manual cmm then it is definitely recommended to use all three operators. All of our cmm’s are DCC and we auto align after touching off manually or program from machine coordinates because of a study that was done with only manual alignments. It showed cosine error does have a factor in your measurements. The tolerance of the products that you are measuring makes a difference in the decision of whether or not the cosine error is large enough to effect the outcome of your gage R&R. It is also recommended to place your parts in different area’s of the surface plate to factor in volumetric accuracy of your machine. (unless you use a safe plate and fixturing that always locates the part in the same area of your machine.) If you have any more questions we have a great staff of cmm programmers that would be more than willing to share their thoughts or idea’s with you. Please feel free to e-mail me at [email protected]0July 18, 2003 at 9:42 pm #88130
Mike CarnellParticipant@Mike-CarnellInclude @Mike-Carnell in your post and this person will
be notified via email.Louie,
I agree with everything you said – obviously you know CMM’s. There are a couple things I would add. Rather than placing parts in various parts of the plate I would either specify by quadrant or run a separate test (hypothesis test by quadrant) before I ran the MSA.
The other thing is that if you have a machine that is so large you actually stand on the surface plate you need to have different people stand in different locations. I have run across machines where it can make a difference. We ended up marking positions on the surface plate where you had to stand when you made a particular measurement.
Good luck.0July 20, 2003 at 1:51 am #88144
David OakleyParticipant@David-OakleyInclude @David-Oakley in your post and this person will
be notified via email.Chris
I would like to make two comments. The first being that one time I gave the same data to two different quality groups and one said that the gage was capible and the other said that it wasn’t. Different people have used different assumptions when developing the GR&R system and this can cause the results to be different. This could be a problem if you share the data with others.
My second comment would be that I have seen cases where the thing that “everyone” is sure couldn’t be the problem was. With my last company, in ten years no one had realized that there was an annual cycle to the reject rate. Once we discovered that, we tried temperature, humidity, and barametric pressure and didn’t get a correlation. We finally decided that it was the annual cycle that the managers went through.
David0July 20, 2003 at 2:46 am #88147
Mike CarnellParticipant@Mike-CarnellInclude @Mike-Carnell in your post and this person will
be notified via email.David,
We had a similar situation. The first MSA was done on the ball plate and went really well. We realized we had an issue when the measurements the machinists took by hand were not matching the CMM. Of course everyone knew that the calipers the machinists were using couldn’t possibly be correct if the CMM was as expensive as it was. Turns out that the CMM would lose track of its location when it had to travel long distances (it was broken).
Thank you.
Good luck.0July 20, 2003 at 3:41 pm #88149
David OakleyParticipant@David-OakleyInclude @David-Oakley in your post and this person will
be notified via email.Mike
Thanks for your post. Yours is a very good example. I expect that everyone that has been doing this for very long has an exmple but I thought it was worth mentioning to some one that seemed to be just getting started.
David0 -
AuthorPosts
The forum ‘General’ is closed to new topics and replies.