hexagon logo
Register Sign in
  • Replies 3 replies
  • Subscribers 17 subscribers
  • Views 59 views
  • Users 0 members are here
Options
Related

CMM Capability

kmcmm
We all know that every CMM has a manufacturer stated accuracy and repeatability. We also all know that the procedures used to establish these numbers are performed using very specific hardware and programs. I want to establish a capability based on what we are using in the real world.

My idea is to use an artifact with known values and using the hardware that is currently utilized on the CMM, perform a repeatability study and calculate the standard deviation of each combination of hardware (probe). Calculate the 3 sigma limits of each combination and use the highest value to compare to my part tolerances to establish capability.

Does anyone have an opinion or method they currently use to evaluate capability? I understand that a gage R&R will tell me if the CMM can repeat and reproduce the measurement as compared to part tolerance however this doesn't take into account the accuracy of the measurement.

Or have I just been looking at numbers too long...
Parents
  • Your approach seems to be interesting, but I don't understand what you call "an artifact with known values".
    If it's a classical part, how can you give "known values",and if it's a calibrated standard, you will ignore the part piece of uncertainty.
    The good point of your method is taking into account probes (different tips and different angles).

    Good luck !
Reply
  • Your approach seems to be interesting, but I don't understand what you call "an artifact with known values".
    If it's a classical part, how can you give "known values",and if it's a calibrated standard, you will ignore the part piece of uncertainty.
    The good point of your method is taking into account probes (different tips and different angles).

    Good luck !
Children
  • in reply to JEFMAN
    The artifact that I'm referring to is a piece that I had designed and manufactured by a local gage and fixture shop. I have their uncertainties and have them calculated for each dimension on the artifact.

    I was planning on doing a proficiency test using the reference values from the artifact and the manufacturer stated accuracy in place of the uncertainties to avoid calculating the measurement uncertainty of all of the CMM's at each of our facilities but I may end up doing that in the end anyway. I haven't decided yet, I was hoping someone already had a method for a estimating accuracy for reviewing tolerance compared to accuracy.
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement