We in the middle of getting our CMM calibrated, and per our quality manual we are supposed to get before and after readings for all calibrations. The calibrator was telling me this is a complete waste of money since whatever standards he is using may be off and he says it will always show out of calibration. He said it is probably something in our quality manual and not the actual standard (which it is) and that we should get rid of it. I did not write nor am I in charge of any of the content of our quality manual, but I am pretty sure the reason it was put in there if you calibrate a gage and get before readings that are out, you may have measured parts and got incorrect readings and sent them to the customer. Does anyone have any comments on the need for before and after readings? Is there something in the MSA manual that says it is necessary?
Send to friendSign In to get started!
Comments
firebrew 9/26/2001
Operating Chracteristics Curves use to be included in the full text of the old Mil 105 standard. These should help you answer your question.
Be very careful however, that you don't use too sharp a pencil when dealing with these issues.
The Dodge Romig sampling schemes (from which the standard was derived) are very dependent on several statistical assumptions. Some of these are:
1)Purely random sample selection.
2)Statistical stability of the stream of product being presented for sampling.
3)A continuous, homogeneous, stream of product from one source.
etc., etc. etc.
If any of these conditions cannot be verified, you need to investigate a different approach.
Shewhart told us that only in cases of statistical stability, could the results of sample nspections be extrapolated to those products that weren't inspected or tested. Sampling plans are no different. If Harold Dodge were around, I'm sure he'd agree.
How about it Don, did I mis-state something?
JB
gandydancer 10/2/2001
I recognize the inherent need for statistical stability when making "inferences" about the population based on a sample of that population. Given that, I'm not sure that I can get my answer from OC curves - there doesn't seem to be any treatment of this subject anywhere but I am sure I have missed something. Appreciate your response, however.