Measurement Uncertainty (MU) in Dimensional Inspection in a QC Laboratory

[qamer]

Hi I work in a quality control lab as manager QC. We perform dimensional inspection of automotive parts. i want to implement measurement uncertainty in dimensional inspection, so that i keep in mind the uncertainty of measurement while taking the decision whether part is accept or reject. While calculating type A part has to be measured at least three times but it is practically impossible for us to write down three values for single dimension of a part. can some one guide me how to implement measurement uncertainty in a QC lab.

[Coury Ferguson]

Originally Posted by qamer

Hi I work in a quality control lab as manager QC. We perform dimensional inspection of automotive parts. i want to implement measurement uncertainty in dimensional inspection, so that i keep in mind the uncertainty of measurement while taking the decision whether part is accept or reject. While calculating type A part has to be measured at least three times but it is practically impossible for us to write down three values for single dimension of a part. can some one guide me how to implement measurement uncertainty in a QC lab.

I have moved this post to this forum, since it involves measurement/measurement uncertainty and should get better responses.

[Hershal]

Originally Posted by qamer

Hi I work in a quality control lab as manager QC. We perform dimensional inspection of automotive parts. i want to implement measurement uncertainty in dimensional inspection, so that i keep in mind the uncertainty of measurement while taking the decision whether part is accept or reject. While calculating type A part has to be measured at least three times but it is practically impossible for us to write down three values for single dimension of a part. can some one guide me how to implement measurement uncertainty in a QC lab.

First thing to do is get training, for you and your inspectors. That will help you and your folks with development and implementation of uncertainty.

Next is to figure out what all affects the measurements. The uncertainty of the instrument(s) you use as reported by the accredited cal lab, temperature and/or other environmental factors, coefficient of expansion of the part, the DIFFERENCE in temperature between the temp that instruments were calibrated at and the temp they are used at, the coefficient of expansion of the instruments (where known), and many other factors may also be present.

Separate them into what you can quantify, what cannot be quantified, and what needs more work to quantify.

Remember if you get uncertainties that are as small as the accredited calibrations or smaller then there is a mistake somewhere.

Take a few tries at it to get used to working with it, without being concerned about the numbers.

When you have identified and quantified what you need, and taken a few tries, then work your uncertainties for real.

Remember, uncertainties are in constant flux, they are never completely stable.....it is just how much resolution is required to see the change that determines whether in fact you will see it.

But first and foremost, get training for you and your inspectors.

Hope this helps.

Hershal