How do you check your CMM ?

  • Filter
  • Time
  • Show
Clear All
new posts

  • How do you check your CMM ?


    how do you check your CMM ?

    There is a Norm (VDI/VDE 2617 Blatt 5) which is talking about a Ball Plate (Kugelplatte) and a Probe Testing Unit (Tasterpr├╝feinheit).

    Somebody use these for checking the CMM ? (for example :

    Can somebody send me a copy of the Norm VDI/VDE 2617 ?

  • #2
    I have a question? Are you doing something like GR&R on your cmm?


    • #3
      Do you mean Gage R&R? Or just a general accuracy check if you think something looks weird? Or do you mean probe qualification? Or yearly calibrations?
      SF7107(PCD), SF454(PCD), 152614(Quindos), 9159(Quindos), 7107(Quindos), B&S Manual, M&M Gear Checker


      • cvelisetty
        cvelisetty commented
        Editing a comment
        You covered most of the scenarios.

    • #4
      Looks like this is something like doing a tuning or calibration to the CMM with similar equipment like slip gauges.

      Krishna Chaitanya Velisetty
      Metrology Software Applications Specialist,
      Hexagon Manufacturing Intelligence.


      • #5
        Just for checking everything is fine.


        • #6
          I know that for NADCAP you have to have an artifact for checking your CMM between calibrations. We bought an artifact from Hexagon that came with a program and results from their measuring the artifact that has met our NADCAP requirements for the last two years. The cost was only between $500-$600 for it and well worth to us.
          DeWain Hodge


          • B. Jacobs
            B. Jacobs commented
            Editing a comment
            Do you know what the artifact was and how it was purchased? I am looking at the Hexagon Store and don't see anything other that the test blocks.

        • #7
          duschi Get some calibrated ring gages and some calibrated gage blocks. Measure the diameters and/or heights (whatever you decide to do) to see if the CMM is reading accurately.
          SF7107(PCD), SF454(PCD), 152614(Quindos), 9159(Quindos), 7107(Quindos), B&S Manual, M&M Gear Checker


          • #8
            RIght or wrong... We made a Cube with bored holes on the top and on all 4 sides, created a program to check the size and location of each of the 5 holes with all of the probes on the cmm. Then we ran some studies on a freshly calibrated machine, calculated the out the noms and then set them in the program as such. We run that at the begining of each shift , after calibration, or if we have a hard probe hit. Probably depends on what type of product you make as to how far you want to go with this.


            • #9
              CMM accuracy depends on linear readings (scales) and global (without wordplay )geometry (straighnesses, rotations - roll, pitch, yaw on each axis).
              Just measuring some blocks or rings can be enough, but can be insufficient.
              For example, measuring a gauge block alongan axs gives a idea of the linear defect of the scale, measuring the saame gauge blocks along the same axis (x or y) at two different heights gives a idea of the pitch defect.
              Along Y on the right and on the left gives an idea of the jaw.
              If you have time enough, you can check the cmm with a Koba gauge (for example) on a mini of 19 locations (1 by axis- 3 is better ! ;2* 6 "flat" diagonals, and 4 3D diagonals).
              You will obtain a good idea of dimensional defects, and by calculating the standard deviation of all the deviations, you will obtain a credible term for the uncertainty of the cmm (without taking into account the method or the part - cmm only !).
              You can also search about different tools (ball bar, hole plates or standards - I like this one : )


              • #10
                Our shop makes one type of automotive component, for multiple customers. Our customers are happy with us doing a ten piece R&R across our 3 CMMs. Making basically one "type" of product makes examining repeatability very easy. We recently had a problem that was traced back to the controller map files being corrupted. I noticed the problem in our reports almost immediately. True position of a bolt pattern is measured on every part. One of our CMMs was consistently reporting the Z position of one bolt hole almost 100 microns different than the other 2 CMMs. It really matters what kind of product you're inspecting. For automotive, it seems R&R is sufficient. For medical or aerospace, I imagine the requirements are different


                • #11
                  This is what we use. It's a great diagnostic tool as well to see the error in each axis.
                  GSG Manufactures Thread Gages, ID & OD Cylindrical Gages in Carbide, Steel & Chrome along with Design & Build Services. Helping Companies Measure Success!
                  Negative, Ghostrider...the pattern is full...


                  • #12
                    Ball-bar, 29 positions
                    Step-bar, 4 orientations (sq to X, sq to Y, 45 degrees to XY, both ways)
                    Table "flat" (every 2" in X & Y, makes for over 1800 points on the big Global)
                    Head Check (9 angles, 10 times, to verify reputability)

                    Run the checks right after calibration of the machine for base-line values. These values SHOULDN'T ever change, and on my machines, they haven't.

                    Ball-bar & Step bar are used to verify that the machine is still tracking square over the course of the year and the step-bar checks the linearity over the course of the year.

                    Head check, pretty obvious, ensures that the head still seats correctly

                    Table "flat". Can't really check the flat of the table with the machine, BUT, nothing in this world is "rock solid", not even the rock of the table. This lets me know that the table IS supported correctly and has not drooped due to an improper support base and that vibrations (can't eliminate them all!) are not causing settling in some strange way.

                    Combine them all, if all are still within 'spec', then the machine is just fine & dandy for continued use.

                    Still working on a way to stand the step-bar on end to do the Z axis....
                    Originally posted by AndersI
                    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.


                    • louisd
                      louisd commented
                      Editing a comment
                      Typically I would get a big 90┬░ Knee-block and use a C-clamp to restrain the scale onto the knee. Use surface plate + other parallels or blocks to make sure the clamping is square and at zeroed-height when clamped and setup vertical.

                  • #13
                    The same
                    with the "user's manual"....
                    Attached Files


                    • KIRBSTER269
                      KIRBSTER269 commented
                      Editing a comment
                      Must be Friday, Mr. helpful, at his best.

                  Related Topics