I have 2 CMM, 1 is still within calibration, the other is about due for calibration. I want to extend the due date but want to prove.
If 1 operator inspects the same part, on these 2 CMM. And the results were lightly different (example: CMM 1: 0.0040, CMM 2: 0.0048, both are still in specs).
Can I tell that the CMM 2 is good for use? Or do I need to do GR&R?
thank you so much!
If you want to extend the calibration, then you need to prove that it is still reading accurately. This is done by gaging it against a NIST traceable artifact. But this is all dictated by your quality manual and documented processes as Kat stated above.
Measure a bunch of parts on CMM1.and CMM2. Calculate the range and the standard deviation for each feature on both CMM's. You'll need a fairly complicated 3d part for this. If the bias and standard deviation are reasonably close to each other you 'should' be good. Create a SPC chart with control limits on the "Good" CMM. If the points on the Unknown CMM fall within control limits then you can reasonably conclude that the CMM's are measuring the same.