I have a procedure for calibrating dial calipers using traceable gauge blocks. Pretty straightforward yet assessor is requesting documents that indicate the procedure is validated. Traceability isn't sufficient? I have deviations & uncertainty. What am I not understanding about validation?

I'm not the ISO expert (more broad technical issues). So ISO folks, please correct me if needed.

My understanding that a procedure used in an ISO17025 context must be adequate. That is, there must be some basis that it correctly checks an instrument's specifications. If you use one developed by an OEM, military (GIDEP/MIDAS), ANSI, ASTM or other nationally or internationally recognized organization with their approval process, it is assumed that it was validated to assure it correctly and completely checked the instrument.

When you develop one yourself, you must somehow verify that it fully and correctly checks specifications.

For example, if verifying a high accuracy digital indicator with a 12 inch gage block, temperature is a significant source of error. If you wrote a procedure that did not include a proper method to deal with temperature implications, although the tolerance limits and other instructions may be correct, that missing detail (temperature correction) could invalidate the procedure. I don't imply that is a problem with yours, or that your procedure has any deficiencies that would invalidate it. Your procedure may be fully correct and complete. What I believe the assessor wants to know is what method you used to assure that it was complete and correct.

What constitutes adequate validation? I have numerous thoughts, but I'll leave that level of detail to more qualified folks.

One alternative may be to locate a published procedure for your mfr/model or it's equivalent. There are definitely GIDEP procedures to cover it (if you are eligible to participate). There may also be other acceptable published procedures.

:thanks: for your help and insight.

True enough. We cite the standard we are following in our written procedure. For calipers- I believe we cite a Navair standard, but I will have to double check that when I get to work.

Edited to add: Are you sure you can get away with only using gage blocks? How are you checking the ID jaws and the depth of stem etc?

I have a procedure for calibrating dial calipers using traceable gauge blocks. Pretty straightforward yet assessor is requesting documents that indicate the procedure is validated. Traceability isn't sufficient? I have deviations & uncertainty. What am I not understanding about validation?

First question is, did your procedure come from an accepted procedure (e.g., NA 17-20 series)? If so, reference that in your procedure, and depending on your AB you should be OK.

If this was developed in-house without referencing an accepted procedure then you must validate it under ANS/ISO/IEC 17025. The validation includes a plan to describe what the procedure is supposed to deliver, some number of actual calibration to obtain the information, and a report at the end matching results to expectation, and a statement that the procedure is fit for intended use.....now.....you must also calculate your measurement uncertainty as a part of the validation, and you should include basic checks such as the jaw and dirt checks (Light between closed jaws; run the calipers all the way out/in slowly and a sticky point indicate dirt.

Traceable gage blocks are a necessity, but remember the so-called NIST numbers by themselves are not valid for traceability.

Hope this helps.

I think you have received some good responses thus far.

I recently had an L-A-B auditor tell me that NAVAIR and T.O. procedures still had to be validated as non-standard methods for ISO 17025. He was OK with ASME procedures being taken as is without the validation. To be honest, I did not really agree, but the solution was more simple than the arguement.

Since we had been using these procedures for several years the validation was pretty simple. I documented 10 examples of calibrations we had performed using the procedure and made a statement that it was suitable for use. Certainly the fact theat it was a procedure acceptable to the military also gives you some confidence in its suitability. I calucalted an uncertainty budget. I stapled it all together, put a nice "verfied" stamp on the procedure and put it in a file.

One more thing about calipers. There are actually 4 measurement for a caliper: OD, ID, Depth and the less know Step. Most procedures (including the one I use) ignore this function on the caliper.

Thanks SO much, Hershal and jfgunn for your helpful posts! What a treasure to have found such a great website! :D