ANAB ISO 17025 Accreditation - Uncertainty and systematic error?

[greif]

We are working toward ANAB ISO 17025 accreditation and are wondering about the scope of measurement we need to send to ANAB eventually.

Does the scope contain only uncertainty, or should it also contain systematic error? On our current calibration reports we usually list uncertainty at k=2 plus potential x.xxx systematic error.

 

[Sidney Vianna]

Have you read ANAB's AR 2251?

 

[greif]

Thanks for the pointer to ANAB AR2251. I did not see the answer to my question in there.
If I calculate uncertainty by including repeatability, user variation, temperature, and resolution (thus covering the major influences), then use the instrument to measure a traceable artifact, coming up with an error curve.

My question ; is this error supposed to go back into the uncertainty (maybe using the U-shaped distribution like the example in lines 4 and 5 of the Canadian NRC uncertainty spreadsheet)?
or
Do I report the uncertainty and also report potential systematic error for various ranges.?
Thanks!

 

[Sidney Vianna]

Thanks for the pointer to ANAB AR2251. I did not see the answer to my question in there.

My point was:

Section 4 of the document lists the minimum requirements for MU reporting. Whatever you want to do above and beyond the minimum requirements is up to you. But, I am sure you can approach ANAB with the question before the audit and they would answer it for you.

 

[brucefon]

A good source for references to understand MU is found at isobudgets com /documents

 

[itsbiodiversity]

Systematic error would be included in the uncertainty budget with the appropriate distribution set. Rectangular is always safe. It would not be an "add on" to the uncertainty, rather it would be a factor in the uncertainty. From reading you may be making this more difficult that it truly is. Let me know if I can clarify any further.

 

[dwperron]

When you are generating the CMC (best measurement uncertainty)for ANAB you are stating the uncertainty of the measurement for the particular class of measurement in your scope.
It will have all of the standard components of the uncertainty calculations, including environmental contributors, repeatability, reproducibility, resolution, etc. Since we don't know what is included in your "x.xxx systematic error" we can't tell what you have accounted for.
It is also good for you to check with ANAB to see how they expect to see the uncertainty reported on your scope. Different AB's have different rules.