If you’re not maintaining good practice in calibration, you’re storing up problems for the future. Carl Hitchens, the Nuclear AMRC’s head of machining, offers his advice for effective calibration.
At a time when you might be operating with a fraction of your usual staff, or preparing to re-start production within safe-working guidelines, the calibration of your measuring instruments might be the last thing on your mind.
Something happened a few years ago which really made me think about the importance of calibration. I owned a car which had reached its tenth birthday, and I received a recall request – the issue was, apparently, due to the potential for the steering wheel to become detached and affect directional control.
I thought this sounded important, and tried to find out more. The dealer told me: “It appears that a torque wrench used to lock down the nut on the steering wheel was not applying the correct force.”
“Was it calibrated?” was my question.
Silence, then came the response: “What’s calibration?”
That recall affected more than 840,000 vehicles worldwide. The cost of the recall was significant, but the reputational risk would have been much higher.
As we think and plan about returning our manufacturing facilities to full operations, I implore you not to forget the calibration process – it was once part of normal good practice, but it might not now be at the forefront of your mind.
Make sure that your engineers and staff using measurement instruments are still familiar with their operation – operator variation is one of the largest sources of measurement error.
Remember that the purpose of calibration is to detect, correct and document the instrument performance and uncertainty. Instruments require calibration because their accuracy may drift over time, generating a systematic error and a bias in the recorded measurement, which will require correction.
There is no standard that dictates how frequently an instrument should be calibrated. To determine the time interval between calibrations, you first need to monitor the instrument’s stability.
I’d suggest that a micrometer which is infrequently used in a clean inspection department, for example, should be calibrated every three months. The result should be recorded and reviewed against previous calibration results to check for instrument stability. If, over time, no change is observed, then it could be acceptable to extend the interval between calibrations. If the instrument is in constant use in an area where cleanliness is an issue, on the other hand, a shorter interval would be more appropriate.
However, in my opinion, the most important consideration – as shown by the example of my car’s steering wheel – is the value, quantity and safety implications of the components which the instrument will have validated during the calibration interval.
If an instrument is found to be out of calibration, you will need to know the period when its performance might have been affected. To be sure of product quality, all products measured since the last calibration will need to be checked and validated again. This can be both expensive and damaging to your business reputation.
Here’s a checklist of questions to consider when defining the calibration period:
- How stable is the instrument over time?
- How often is the instrument used?
- What is the quantity and value of the components validated by the instrument?
- What kind of environment is it stored in?
- What is the manufacturer’s specification for the instrument?