Assigning a calibration tolerance - An x-y coordinate machine - Uncertainty as my verification tolerance

[greif]

We have a x-y coordinate machine that has had an uncertainty assigned to it, based on length of measurement k=2 0.0042mm + (0.0000135L) where L is in mm.
example; at 0 mm the uncertainty is 4.2um, at 400 mm the uncertainty is 9.6um.

We periodically calibrate/verify this machine by reading a master calibrated scale.

Question: is it proper to use the full above mentioned uncertainty as my verification tolerance? Or do I need to include some other factor?

Thanks!!

 

[dwperron]

You need to check with the manufacturer for the correct interpretation of the specification.
These days too many companies use accuracy, tolerance, and uncertainty interchangeably.

If they really mean uncertainty they will be able to state what uncertainty contributors are included, such as resolution, repeatability, hysteresis, environmental, etc. along with the basic accuracy specification.
If it is basic instrument accuracy, then you need to include the other contributors to determine the uncertainty of the machine.

For the purpose of calibrate/verify the machine, you would use the basic accuracy specification.

 

[Ninja]

Additionally, consider the accuracy/reliability/uncertainty (not interchangeable, but not additive either)...that you NEED.

One question is "Is the gage good as new?"...a whole different question is "Is this good enough for my needs?"
Answer both questions...

 

[greif]

My main focus is to accurately report uncertainty in reports that go to customers (getting set up to be 17025 accredited). The machine is quite old and does not have a useful tolerance from the former manufacturer.
So my original question; can I use the uncertainty that we calculated in -house as my machine periodic verification tolerance, or is that too circular?

 

[dwperron]

For the full uncertainty calculation that you report you will need include your "calibrate/verify" results (the accuracy that you determined the machine is capable of), the uncertainty of your calibration standard(s), any environmental factors, the resolution of your results, the repeatability / reproducibility of your results (based on your studies), and any other significant contributors in your process.

As long as your standard(s) all have accredited calibrations you are not too circular.