Gage R&R

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Gage R&R

    Does anyone have any info on the 10-1-1 method for CMM's? 10 runs 1 part 1 operator. I am trying to determine which method is best.

  • #2
    With a load and unload in between each test?

    Useful for proving the part Program but not the machine.
    Results are a "GR" without the Reproducibility.
    It will give you the data for a P/T ratio (does the gage fit in the 10% of the tolerance ratio, NOT to be confuesed with 10% GR&R)
    Links to my utilities for PCDMIS

    Comment


    • #3
      GRE = 5.15*STDDEV
      % of GRE = GRE/total tolerance

      Generally, the % of GRE needs to be less than 10% for the fixture to pass the Gage R study. Once it passes this, you can then do the full Gage R&R study (5 parts, 3 trials, 2 operators minimum).
      sigpic
      Originally posted by AndersI
      I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

      Comment


      • #4
        I've got an XL spreadsheet for the 10-1-1.

        PM with your email and I'll send it to you.
        Xcel & MicroVal Pfx & Global 37mr4 thru 2012mr1sp3
        Contura Calypso 5.4

        Lord, keep Your arm around my shoulder and Your hand over my mouth. Amen.

        Comment


        • #5
          What you are referring to is called a Gage R (meaning just repeatability).

          This is the best way to start out, its quick, guess what ? If it doesn't pass, then no sense carrying on to full blown GRR.

          I have a 10-1-1 spreadsheet if Dwade renegs....LOL
          Jim Jewell

          Comment


          • #6
            I understand that it is only repeatability. I have always thought that R&R was pointless on a CMM. Operator variation should be close to nothing therefore all you would need to verify is repeatability. I am in that very argument with one of our engineers. I did the 10-1-1 and that wasn't good enough so I did the full blown R&R. The results were that operator variation was basically none and the percent of tolerance was almost identical between the two methods. So I wasted about 6 hrs running 3 trials 15 parts 3 operators. When I told him it was good to begin with based on the 10-1-1.

            Comment


            • #7
              Originally posted by Ceeph View Post
              I understand that it is only repeatability. I have always thought that R&R was pointless on a CMM. Operator variation should be close to nothing therefore all you would need to verify is repeatability. I am in that very argument with one of our engineers. I did the 10-1-1 and that wasn't good enough so I did the full blown R&R. The results were that operator variation was basically none and the percent of tolerance was almost identical between the two methods. So I wasted about 6 hrs running 3 trials 15 parts 3 operators. When I told him it was good to begin with based on the 10-1-1.
              Are you aligning every part you check? If so, you are wasting your time. The Gage R and Gage &R studies are for fixturing a part. If you are aligning the part everytime, you are not doing either type of study, except on the CMM itself and you are wasting your time for the part you are working on. If you ahve a holding fixture or a checking fixture, you align the fixture ONCE and once only, then place parts, check, repeat. You WILL see gage and operator error when done this way, which is the proper way. The GAGE is the thing holding the part, NOT the CMM aligning the part.
              sigpic
              Originally posted by AndersI
              I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

              Comment


              • #8
                No I am not manually aligning each part and it is a machined housing going into a fixture. There is a dcc alignment for every run in the program. That is why operator variation should be almost none as long as the part does not move in the fixture after the alignment. I am pretty good a designing fixtures so that does not happen.

                Comment


                • #9
                  Originally posted by Matthew D. Hoedeman View Post
                  Are you aligning every part you check? If so, you are wasting your time. The Gage R and Gage &R studies are for fixturing a part. If you are aligning the part everytime, you are not doing either type of study, except on the CMM itself and you are wasting your time for the part you are working on. If you ahve a holding fixture or a checking fixture, you align the fixture ONCE and once only, then place parts, check, repeat. You WILL see gage and operator error when done this way, which is the proper way. The GAGE is the thing holding the part, NOT the CMM aligning the part.
                  Matt,
                  Excellent explanation!!! I have printed that so that I can explain the differences to people. I have a hard time giving examplesquite often. Again thanks for your insight.
                  sigpic

                  James Mannes

                  Comment


                  • #10
                    Originally posted by Ceeph View Post
                    No I am not manually aligning each part and it is a machined housing going into a fixture. There is a dcc alignment for every run in the program. That is why operator variation should be almost none as long as the part does not move in the fixture after the alignment. I am pretty good a designing fixtures so that does not happen.
                    If that is what you are doing, DCC alignment on each part, then you are doing the CMM, not the fixture. One alignment only for the entire run, on the fixture, or you are 'studing' the CMM.
                    sigpic
                    Originally posted by AndersI
                    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                    Comment


                    • #11
                      OK, I have a question. I have been programming CMM's for twelve years and I have never ever done an alignment on a fixture. My alignments are always on the part datums and report features relative to the datums. So it should not matter how you load a part you are always aligning the part to itself. So why would you ever align to a fixture that you know will introduce variation into your measurements?

                      Comment


                      • #12
                        Originally posted by Ceeph View Post
                        OK, I have a question. I have been programming CMM's for twelve years and I have never ever done an alignment on a fixture. My alignments are always on the part datums and report features relative to the datums. So it should not matter how you load a part you are always aligning the part to itself. So why would you ever align to a fixture that you know will introduce variation into your measurements?
                        Do the 'guys' on the line, in production, on the floor, have a CMM to align and check the part every time or do they put a part on the fixture, throw the clamps, and do the feeler/stab/flush checks of the part on the fixture, that is being held only by the fixture and not aligned by a CMM?

                        Checking fixtures are, by nature, designed and built for an operator to do MANUAL checks of the part during produciton runs. If that fixture can NOT hold the part in the same way, every time and give the same results from one operator to the next, then it is a useless fixture, that is what the Gage R and Gage R&R studies are for. They prove that a fixture can hold the part the same way, everytime for every operator.
                        sigpic
                        Originally posted by AndersI
                        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                        Comment


                        • #13
                          Originally posted by Matthew D. Hoedeman View Post
                          Do the 'guys' on the line, in production, on the floor, have a CMM to align and check the part every time or do they put a part on the fixture, throw the clamps, and do the feeler/stab/flush checks of the part on the fixture, that is being held only by the fixture and not aligned by a CMM?

                          Checking fixtures are, by nature, designed and built for an operator to do MANUAL checks of the part during produciton runs. If that fixture can NOT hold the part in the same way, every time and give the same results from one operator to the next, then it is a useless fixture, that is what the Gage R and Gage R&R studies are for. They prove that a fixture can hold the part the same way, everytime for every operator.
                          Very well stated Matt...
                          sigpic
                          Xcel 15-20-10 - PFXcel 7-6-5 - Merlin 11-11-7 - Romer Absolute 7525SI
                          PCDMIS 2012
                          Windows Office XP

                          Comment


                          • #14
                            The only fixtures we use are the ones in the cnc holding the parts while they are machined. We do not do hard gaging. Inspection on the floor is with gage pins, bore mics, etc.. nothing requiring a fixture. My fixtures are only for holding the part on the cmm. The only time I would align to a fixture is if I was actually measuring one of our cnc fixtures. I have found that with our tight tolerances that fixtures to check position never pass R&R unless they are jig ground to extremely tight tolerances and therefore are not cost effective. I am simply trying to prove to this engineer that the CMM is repeatable which I know it is as long as my part does not move during inspection.
                            Last edited by Ceeph; 02-14-2007, 11:00 AM.

                            Comment


                            • #15
                              Ceeph ,

                              Clearly you are aligning each part each time, of course the F' ing thing repeats !

                              What you did as a 10-1-1 was MORE than enough to give your Engineer that warm fuzzy feeling about the measurements coming off the CMM.

                              Ya know, with all this 6 sigma, Kizan and all this other garbage Engineers are supposed to know about, why can't some of them use their minds for logic ???

                              What a TOTAL WASTE OF TIME MR. ENGINEER !! A total GRR, jeesh, get your head outta your **** !

                              All these companies are screaming about saving money all the time and this guy had you WASTE all that time...oh my...
                              Jim Jewell

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X