Measuring customer parts on a CMM - How many decimals to report to the customer?

[greif]

We are getting ready to be a 17025 measuring lab. When we measure a part for a customer on our CMM, the CMM resolution is 0.002mm. If we measure the part 3 times and get; 50.000, 50.000, 50.002mm, is it ok to report 50.0007mm (along with the associated uncertainty)?

 

[Eredhel]

Are you taking multiple measurements of the same feature on the same part a lot or is this specific situation like first articles or something? A little bit off topic from your question, but the CMM might have a setting for how many decimals to show. I know in Zeiss Calypso you can choose how many decimals to display.

 

[greif]

I am looking to do this for routine measurements of customer parts (for which we then issue a 17025 Calibration Certificate), so yes it would be happening a lot. Yes, same feature multiple times, then report the average. We have an old digital display, no choices on resolution, just reads to 0.002mm.

 

[Jim Wynne]

We are getting ready to be a 17025 measuring lab. When we measure a part for a customer on our CMM, the CMM resolution is 0.002mm.

ISO 17025 describes requirements for calibration and testing labs. You seem to be referring to measuring parts rather than doing calibration. You also wrote:

I am looking to do this for routine measurements of customer parts (for which we then issue a 17025 Calibration Certificate)...

Please explain--how can you issue a calibration certificate for measuring parts?

 

[greif]

The customer parts in this case, are glass measuring scales, for which they would like calibration data of specific line to line distances.

 

[Jim Wynne]

The customer parts in this case, are glass measuring scales, for which they would like calibration data of specific line to line distances.

How can you do that with a CMM? Is it an optical device? What is the customer's stated tolerance for the graduations?

 

[greif]

A microscope hangs on the z axis of the cmm. Customer does not state tolerance, only wants measurement reported.

 

[RoxaneB]

We are getting ready to be a 17025 measuring lab. When we measure a part for a customer on our CMM, the CMM resolution is 0.002mm. If we measure the part 3 times and get; 50.000, 50.000, 50.002mm, is it ok to report 50.0007mm (along with the associated uncertainty)?

I do not have a 17025 background, so please take this for what it's worth...

During an ISO 9001 certification audit, the external auditor noticed that we were measuring (and recording our measurements) of a part to 1/8", however, the drawing measurement (for comparison) was to 1/16". We were issued a finding for not having the proper tools to ensure we were meeting the design requirements.

Common sense says to me that if you measure it to 3 decimal places, you cannot report to 4. Your device does not seem to have that level of accuracy.

 

[Jim Wynne]

A microscope hangs on the z axis of the cmm. Customer does not state tolerance, only wants measurement reported.

Curiouser and curiouser. Calibration is comparison of a device to a standard. You apparently have no standard for comparison, so you are indeed just measuring parts, not doing calibration. You shouldn't be issuing calibration reports for this activity. You can provide an inspection report, but that has nothing to do with ISO 17025 beyond the calibration status of your own device. Just report the information you've got to the customer.

 

[greif]

The standard in this case is my CMM (which is part of a calibration system, has properly documented uncertainty, and is checked with a standard scale calibrated by NIST.) So- when I sent my standard scale to NIST, they did what I am doing for my customer (just a whole lot more accurately!)
Calibration does not require adjustment, so calling it calibration seems proper. I just checked and my report from NIST is titled "Report of Calibration".