iSixSigma

Gauge RR but with only one R…….

Six Sigma – iSixSigma Forums Old Forums General Gauge RR but with only one R…….

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #50492

    boggled BB
    Participant

    Evening all,
    Looking for some advice; If i want to analyse a measuremnt system for a cmm machine am i right in presuming the model still fits (Mintab gauge study crossed) with out the differing operators. The product is placed on a measuring fixture then the cmm does its thang!
    Im also finding that in some cases im getting Distinct catagories <5 however i think this is related to pretty low part to part variation. Ifi remeber my trainng for a good study we should be aiming to have part to part as high as possible.
    Could my low distinct catagories be due to no operator analysis?
    Interested to hear from anyone else how they would handle this.
    Thanks!
     
     
     

    0
    #173597

    Taylor
    Participant

    Boggled BB
    Your parts should range across the specification from low to high in order to assure distinct categories are present, or at least be a representitive sample of the parts being measured that you would expect to see on any given day. Placing parts in a fixture sometimes is not as easy of a task as it should be, Therefore you should also have Different operators. My past experience has shown me that the CMM is not really that reliable of a measuring instrument. Also don’t forget to perform a calibration prior to doing the study.
    I have seen some really frustrated QA guys after they see the results from an R&R, so good luck

    0
    #173598

    boggled BB
    Participant

    Chad,
    Thanks for the feedback. I have been playing with data since the post and im getting some good results with DC’s >5. My only worry now is the impact on validity by only showing repeatability. However it is the reality of how we would operate this set up.
    The new measuremnt system is going to cost some hard earned $$$ to create and needs customer buy in so i just want to be sure my test still makes sense.
    Thanks again

    0
    #173599

    Mikel
    Member

    Bad studies for CMM’s are usually set up and procedure issues. To get repeatability, make sure the entire set up is broken down and redone
    each time.I disagree with Chad’s advice a little – the parts should be
    representative of the process capability, not the spec.I have also seen lots of bad results like Chad, and all have nothing to
    do with the CMM and everything to do with fixturing and procedures.

    0
Viewing 4 posts - 1 through 4 (of 4 total)

The forum ‘General’ is closed to new topics and replies.