Aviation Tool Calibration: Records, Intervals and Traceability

Calibration is the part of tool control that is entirely about records. A wrench is either accurate or it is not, and nobody in a hangar can tell by looking. What an auditor can check is whether you can prove it, and that is a paperwork question with a technical core.

This guide covers the standards involved, what a certificate needs to say, how intervals are set and defended, and what to do when something is found out of tolerance. As with our Part-145 guide, work from the current official text for anything that matters, and treat this as orientation.

Three standards, and what each one is for

These get used interchangeably in conversation and they do quite different jobs.

ISO 10012 is about your measurement management system. It is the reference for how an organisation manages measurement processes and measuring equipment, and it is what applies when a manufacturer has published no calibration recommendation for a tool. It describes how you decide and control, not how a laboratory measures.

ISO/IEC 17025 is about the laboratory. It is the competence standard for testing and calibration laboratories, and a laboratory accredited to it has been assessed as technically competent to perform the specific calibrations listed in its scope. Note that word: a laboratory is accredited for a scope, not in general, so the scope is worth reading rather than assuming.

The ILAC mutual recognition arrangement is about whether anyone else has to accept it. Accreditation bodies that are signatories recognise each other's accreditations, which is what makes a certificate issued in one country meaningful in another. In a Part-145 context the expectation is calibration by a laboratory accredited to ISO/IEC 17025 by a body that is an ILAC MRA signatory.

Put together: ISO 10012 governs how you manage it, ISO/IEC 17025 governs whether the laboratory is competent, and the ILAC MRA is why the certificate travels.

Traceability, which is the thing being checked

The concept underneath all of it is metrological traceability: an unbroken, documented chain of comparisons from the reading on your instrument back to a national or international measurement standard, with a stated uncertainty at each step.

That is why a certificate saying only "calibrated, passed" is not enough. The chain has to be reconstructible from the paperwork. A calibration performed with an uncalibrated reference breaks the chain regardless of how carefully it was done.

What a certificate has to tell you

When a certificate arrives, check it rather than filing it. Auditors read them, and a certificate that is missing something is found at the worst moment.

• What was calibrated, identified unambiguously: make, model and serial number. The serial has to match the tool in your hand and the row in your register.
• Who did it, and their accreditation, so the ISO/IEC 17025 and ILAC chain can be checked.
• When. The date of calibration, not the date of printing.
• Against what. The reference standards used, and their own traceability.
• The results, as measured values with the measurement uncertainty, not just a pass. A reading near the tolerance limit is information you want.
• The condition it was received in, and whether any adjustment was made. An as-found reading out of tolerance is the trigger for the impact review below, and if the certificate does not mention as-found condition you cannot do that review.
• Environmental conditions where they affect the result.

Note what is not on that list: a due date. Many certificates carry one, and it is the laboratory's suggestion rather than your interval. The interval is yours.

Setting an interval you can defend

There is no universal number, and anyone who gives you one without asking about your usage is guessing. Part-145 requires a frequency that ensures serviceability and accuracy, which puts the decision and its justification on the organisation.

The hierarchy that is easiest to defend:

1. The manufacturer's recommendation, where one exists. Simplest to justify, because the responsibility for the number sits with the people who built the tool.
2. An interval derived under ISO 10012, where no recommendation exists, based on how the equipment is actually used and what the consequence of an error would be.
3. An interval adjusted from calibration history, which is the mature version. A tool that comes back in tolerance for five cycles running is telling you something, and so is one that drifts every time.

What genuinely moves an interval, in rough order:

• Consequence. A wrench used on a flight control attachment is not the same risk as one used on an access panel, even if it is the same wrench.
• Usage rate. Torque wrench recommendations are often expressed in cycles as well as months, and a wrench on a production line reaches a cycle count far faster than one in a crib.
• Handling and environment. A tool that gets dropped, or lives in a vehicle, drifts sooner than one on a bench.
• History. Documented stability over several cycles justifies extension. Documented drift shortens it, and doing that before an auditor points it out is worth a great deal of credibility.

Whatever you choose, write down why. An interval with a rationale is a defensible position; the same interval with no rationale is a finding waiting for somebody to ask.

Storage and handling, which affects intervals

Two habits do more for measuring equipment than any interval policy.

Store a click type torque wrench wound back to the bottom of its range, never at setting, because a spring held under load takes a set. This is the single most common reason a wrench that has never been abused comes back out of tolerance.

Quarantine anything that has been dropped or overloaded, immediately, and have it checked rather than assuming. A wrench that has been used as a breaker bar is out of calibration whatever the sticker says, and the person who did it is not always the person who reports it, which is an argument for a culture rather than a procedure.

When a tool is found out of tolerance

This is the procedure most organisations have written and never exercised, and it is where an audit gets uncomfortable.

The question that follows an out of tolerance finding is not about the tool. It is: what was this used on since it was last known to be good, and does any of that work need review?

Which means you need three things before it happens:

1. As-found readings on your certificates, so you know the magnitude and direction of the error rather than only that it failed.
2. A usage record good enough to bound the exposure. If a wrench is issued to a bay rather than a task, the honest answer to what it was used on is everything that bay did, which is a large and expensive answer.
3. A decision route, with somebody named who evaluates the impact and decides whether any maintenance action needs revisiting.

The connection back to tool control: the tighter the link between a tool and the work it did, the narrower the review. Serialised tools and issue records turn a hangar-wide question into a task-level one. That is not the reason to serialise tools, and it is the reason you are glad you did.

What we do and do not do

Plainly, because this is a page about evidence.

Measuring equipment we supply comes with a calibration certificate stating the laboratory, the standard and the date, so it can enter your controlled equipment register without waiting on paperwork. Torque wrenches in a kit are supplied calibrated.

We are not a calibration laboratory and we hold no ISO/IEC 17025 accreditation. Where calibration is performed it is performed by a laboratory, and the certificate names them rather than us. We do not set your intervals, we do not own your recall dates, and etching a serial on a tool proves identity and nothing about accuracy.

What we can help with is the state of the crib: whether every measuring item has a current certificate, whether the register matches reality, and whether the infrequently used, leased and loaned tooling has quietly fallen outside the system. That is a tool crib audit.

Questions