iSixSigma

Gage RR

This topic contains 4 replies, has 3 voices, and was last updated by  Jim Shelor 12 years, 4 months ago.

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #47226

    pedro
    Participant

    Hi ! I know the methodology of the Gage R&R is to take 2 to 3 readings(random), to several pieces showing variation (usually 10)by 2 to 3 different operators, then the measuring system is evaluated.. But ,, how could this be applied in cases where several readings may not be taken,, like pull test in a cable, when by just performing the test is a desctructive test.. no way to take a second reading.
    Or in the case of mesuring the distance in a catapult process,, where a shot is unique. I am attending a six sigma class and this issue was arised.. So far no advice or explanation. I would appreciate yor opinion. thanks
    pedro

    0
    #157266

    CT
    Participant

    Pedro
    If all appraisers are using the same process to measure samples from a well-known and established process, then we’d expect the averages to be the same and the variation represented by the UCL and LCL to be equal.
    If the averages or standard deviations are different from appraiser to appraiser (i.e., tech to tech), then the measurement system needs improvement.
    Another way is to test against a known “calibrated” standard. Although expensive sometimes, is very effective.
    CT

    0
    #157276

    pedro
    Participant

    Let’s say that the process is not pretty well stable, and we would like only and for the moment just to concentrate in the measuring system. But to run a study the appraissers would need to measure each piece at least twice,, and imagine that is not possible, pull test case. thanks.

    0
    #157278

    CT
    Participant

    Pedro
    In order to concentrate only on the measuring system, you need to be able to compare the process to a known Standard. Suggestion-Fishing line from a single spool has a very consistant break point. You should be able to get in a range very close to your pull specifications. Another is locate Calibrated Pull Blanks that can give very specific break points. Run the Gage R&R for known standards technique, this will put the focus on the gage and not the product. Some engineering may have to be done to fit the blanks to the gage, but is really the only way to take the process variation out.
    CT

    0
    #157286

    Jim Shelor
    Participant

    Pedro,
    For a gage R&R for a destructive test, use the Gage R&R (nested) method.
    If your process is not stable, any gage R&R could be a problem (inaccurate) because the Part-to-Part variation can swamp your study.
    Jim Shelor

    0
Viewing 5 posts - 1 through 5 (of 5 total)

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