Calibration Frequency of Measurement and Test Devices

[noboost4you]

We have numerous Test Gauges, Digital Calibrators, Multimeters, Thread Gages, Step Gages, etc. within our facility. Test Gauges are used on a daily basis and have a calibration date of every three months. Other gages, such as thread gages, are rarely used and have no calibration date. We only use the thread gages, step gages, and certain other measuring devices, when a pallet of parts come in to check a very small sample.

As far as I'm concerned, these particular gages have never been recalibrated/reverified since their acquisition when they were brand new years ago. They are not electrical devices and are handled and stored properly. The only label(s) on each gage range from the manufacturer, manufacturer number, size of thread, etc. There is no mention of calibration status, calibration intervals and date of next calibration.

Since these gages are rarely used, are non-electrical, and handled/stored properly, what could an external auditor do about them? Having them all recalibrated now will cost thousands of dollars. It's not cost effective just to have the auditor pleased.

Any suggestions?

[andygr]

As long as you have the history you can extend the intervals based on observed stability.
The other approach you might concider is a calibration cycle based on use for those tools only used once every now and then.
Your approach taken would have to be documented and the calibration actuals on file to suport your practice.
You will always deal with the auditors view that the risk is greater with the longer intervals but if you have the data to suport your practice you should be fine.

Search the terms calibration frequancy, calibration use, ect and you will find plenty of info to concider.

[bobdoering]

If you very rarely use them, you can pull them out of service, and calibrate them when you need them. You may opt to calibrate on a number of checks frequency rather than a time frequency (tracking possible wear) - but you need a valid system to track it. You can even have an extended period, such as 2 or 3 years, as long as you can sensible justify it. It all boils down to the audit question: How do you know the gage is still reading correctly? As long as you can answer that question in a solid, logical manner, you win.

[noboost4you]

Our building comprises of 2 companies, only one of which is getting certified. The other company uses the thread gages, step gages, etc. 90% of the time while the company getting certified may only use them 10% of the time.

Right now the way we know the gages are reading accurately is by a couple ways. Our suppliers use the same gages we use. They need to check their product before the send it to us. Once we receive it, we'll check a small sample. If any discrepancies are found, we'll ship the defective parts back and have them recheck them. If they come back and say everything is fine, we'll then look at the gages. Also, certain gages we use more than others, such as 1/4 inch NPT gages. We have approximately 4 of those gages. If all 4 read the same, then we know they are all good. If one reads differently than the other 3, and so on, we know something isn't right.

We have certifications for all the gages when we bought them. However, we have not had them recertified. If they are really out of tolerance, we would rather buy new gages than pay someone to tell us we need to buy new gages anyways.

[bobdoering]

Do you have thread gage masters?

[noboost4you]

No we don't.

[BradM]

Good posts on this.

Also, if you get a chance, you might want to surf the other threads on calibration frequency. There is a fountain of information. Some of the threads are at the bottom of your post (at the very bottom).

What calibration frequency you set (barring industry/regulatory/customer requirements) is yours to choose. You will need to assess risk. Basically how risky is it should this item be out of tolerance? The lower the risk, the longer you can go between intervals.

I would recommend having your identified instruments on a cycle of some sorts for calibration. I understand that it seems wasteful paying for a "failing" calibration. However, it is important to know the magnitude of the calibration error. It may not matter, but then again, it will. If you just toss them, you don't know how much error they were contributing to your measurement process.

If I understand your posts correctly, there are several areas for potential error. You have sampling error, measurement error on the part of the customer, and measurement error at your facility. If you do not assess the uncertainty of your measurement instruments, then you have just added another uncertainty. This could be increasing your probability of acceptance/rejection errors.

[bobdoering]

Originally Posted by noboost4you

If they are really out of tolerance, we would rather buy new gages than pay someone to tell us we need to buy new gages anyways.

I'd go for that too, if it wasn't a chicken and egg thing. For the facility that is being registered, comparing gages as you described will not qualify as calibration. Using your supplier to verify your gaging, won't pass muster, either. Calibration is a very specific process, with traceability to a national standard. How often you calibrate is your call - but how you calibrate is going to be easily determined to be correct or not.