Calibration Determination decision

[Ron Rompen]

"The OEM doesn’t have anymore idea on how frequently a device is used - they often recommend 12 months!"

Yesterday I had an external 9001:2015 auditor inform me that we should be calibrating our tools (case in point a digital caliper) in accordance with manufacturers recommendations.

Serveral years ago while I was with this organization as a consultant for a brief time, we made the decision to look at our list of several thousand tools and decide whether calibrating the majority of them every year was necessary. We reviewed years of data regarding the frequency of out-of-calibration events (extremely rare) and the potential impact of using an out-of-calibration tool to inspect product that would go to a customer (surprisingly low risk due to multiple levels of verification along the way) and made the decision to take a lot of our tools from one year to two year intervals resulting in a savings of approximately 10K USD yearly.

I recall finding somebody here on the Cove who had done some math regarding this issue (way over my head, I can cipher and count on my fingers) and I contacted them and shared them with my boss.

Anyway, I don't expect the auditor to make a big deal of it, but now that I'm an RFT employee of said organization I may as well begin stockpiling evidence that would support this policy, since it's now my task to defend it.

Norm: I have done much the same in my current position. When I took over here, everything (and I do mean EVERYTHING) was on a 12 month calibration cycle. I have changed almost everything based on 30 years of history, risk assessment, and common sense. As an example, digital calipers are done on an 18 month cycle - in our work conditions, they either work or they fail to function at all. NO history of calibration faults ever found. Analogue calipers, however, are done on a 12 month cycle, as they work strictly by mechanical transfer, and it is much easier for components to wear/gum up. Similar changes were made to all the other TYPES of gauges in the plant. Yes, there are a few unique exceptions, but there is a reason for those.
I have shared this plan (documented as a part of my lab scope) with customers and with external auditors (both ISO and IATF) and have never had an issue with it.

 

[Mike S.]

Yesterday I had an external 9001:2015 auditor inform me that we should be calibrating our tools (case in point a digital caliper) in accordance with manufacturers recommendations.

Maybe if you show him a response mathematically:

should ≠ shall

 

[QAhib63]

If the auditor has any issues with your approach after you have had a chance to show the data and explain the rationale behind the decision and insist on following “OEM recommendation” ask him/her/it where is the shall. Finito.

Next question

Correct - absolutely. ISO 9001 and AS9100 provide for flexibility based on what the Organization feels is suitable to ensure conformance. Just back up what you do with the appropriate data (Statistically quantifiable), and be prepared for thorough and conclusive product impact investigations should an incident ever occur.

 

[QAhib63]

I should add a footnote. In everyday language, the word Calibrate is used by different people in various contexts, but doesn't always mean the same thing. In an industrial quality setting, my definition of Calibrate versus Verify is that Calibrate includes either or both, comparison to a recognized reference standard and the ability to adjust (as necessary) to improve measurement accuracy. If you cannot adjust or it is more economical to replace, I call that Verify. I often use the example of a bathroom scale as illustration to explain. If you turn the little knob (on a mechanical scale) or step on/off and let the digital scale zero itself, taring is a form of calibration because it actually changes the measurement which results, before and after. The reference standard is nothing on the scale besides the parts of the scale.

Verification of accuracy and the ability to calibrate (adjust to spec.) can only be done with certified weight standards. An adjustable scale in your bathroom is just reference only, unless you can define the required accuracy and verify using the physical standard. If you throw away a gauge due to inaccuracy, the standard practice (For proper verification or calibration programs) is to verify and document how far out of tolerance the gauge is. External auditors have a field day in poorly managed calibration departments because that is the most likely place in an organization where violations of the standards in place will occur. It gets worse if there is no follow through for lack of determining how that issue may have affected your products, and for how long. I've had auditors try to pull me through a knothole when I was serving as a calibration tech, thankfully they never found much other than a single minor or OFI. I was very thorough because I had great Quality Directors that insisted on a well structured calibration program. It pays to know what you're doing - don't give any one an opportunity to take you down the rabbit hole. It's a bad place to be.

 

[mattador78]

We do a lot of thickness testing and the devices and probes remain in tolerance a hell of a lot longer than the foils they are calibrated against. We have moved all our machine calibration further out and brought all our foil changes closer together. Some of the foils are so thin they can be damaged when calibrating the meter after one use, we know this as it involved a written warning for an operator who somehow managed damage a set on the day it arrived through carelessness. We have had multiple customer and AS audits where this has been perfectly acceptable and never even been questioned

 

[QAhib63]

We do a lot of thickness testing and the devices and probes remain in tolerance a hell of a lot longer than the foils they are calibrated against. We have moved all our machine calibration further out and brought all our foil changes closer together. Some of the foils are so thin they can be damaged when calibrating the meter after one use, we know this as it involved a written warning for an operator who somehow managed damage a set on the day it arrived through carelessness. We have had multiple customer and AS audits where this has been perfectly acceptable and never even been questioned

fOILS ARE VERY FRAGILE

 

[KDenton]

I have a slightly different question to ask. My company does automotive parts manufacturing. My question is: If a gauge is calibrated on 1/1/2023 (as an example) and sits in the contolled lab until 8/1/2023, does it really need to be calibrated again on 1/1/2024? Or can it wait until 8/1/2024? Assuming the gauge was never used between 1/1 and 8/1. Customer requires it t be calibrated annually.

 

[Jen Kirley]

If the customer says annually, I would assume that is January because the gage is intended to be ready for use.

 

[John Predmore]

I acknowledge the argument that gauges unused in (environmentally controlled!) storage do not degrade. I suppose the best answer depends on failure modes and criticality. I am aware that some devices might degrade, even unused. That is, for example, why torque wrench instructions recommend spring tension preload be backed off in storage. I suppose the lowest-risk solution to your question would be to re-calibrate after 8 months, compare the recent result to what it was 8 months ago, repeat that action 10-20 times for a statistical justification, before concluding storage doesn't affect calibration.