Reading a Torque-Wrench Calibration Certificate: As Found, As Left and Scope

A calibration certificate may finish with a passing result even though the wrench arrived at the laboratory outside its stated requirement. There is no contradiction. “As found” describes performance before adjustment; “as left” describes performance after the laboratory’s intervention. The sequence matters when a maintenance lead is deciding whether earlier work deserves review.

The certificate does not certify the joints previously tightened with the wrench. It reports measurements of an identified tool, over a stated scope, at a stated time. Reading it well means connecting those limits to the actual tool-use history.

First prove which wrench the certificate covers

Match the certificate’s manufacturer, model, serial or asset number to the physical wrench and the internal tool record. Similar wrenches can share a storage location or nominal range, so a handwritten case label is not enough when the serial number differs.

Check the calibration date, certificate number, laboratory identity and any page count or attachment reference. Confirm that the document is complete and that later amendments have not replaced a result. The next-due date, if stated, is an administrative interval; it does not tell you how the wrench behaved on every day inside that interval.

Identity is the first gate because a detailed result from the wrong wrench has no relevance to the work under review.

Read the tested range, directions and functions

A certificate rarely proves every possible use of a torque wrench. Review the units, indicated settings or test points, direction of operation, loading method and reported quantity. A wrench tested clockwise over part of its scale has not thereby been demonstrated counter-clockwise, below that range or in an untested operating mode.

This matters when tools are used on heavy bolt and nut connections. Compare the settings in the job records with the calibrated test points and scope. Do not use a familiar accuracy percentage as a universal substitute for the requirement stated by the applicable procedure, standard or certificate decision rule.

Also distinguish calibration from adjustment. Calibration compares indications with traceable reference values and reports results. Adjustment changes the instrument to improve its performance. A laboratory may perform both during one service visit, but they are different events in the chronology.

As-found data preserve the condition before intervention

The as-found section is the closest available evidence of how the wrench performed when it entered the laboratory. It may list reference torque, wrench indication, error and an assessment at several points. Read the sign convention and units given on that certificate rather than assuming whether a positive value means over- or under-delivery.

An out-of-tolerance as-found point raises a look-back question. It does not prove that every previous fastener received the same error. Tool behaviour can vary across the range, direction and operating conditions; job technique and the fastened joint add their own variables. The result should trigger a bounded review, not an automatic declaration that every joint failed.

If the certificate provides no as-found data, an as-left pass cannot recreate them. Ask the service provider what was recorded and whether adjustment occurred. Identical as-found and as-left numbers may mean no adjustment was needed, but that conclusion should come from the record rather than the coincidence alone.

As-left data describe the released tool

After adjustment or service, the laboratory may repeat the calibration and report as-left results. These values support the tool’s condition when returned, within the tested range and stated uncertainty. They do not erase the earlier findings.

Keep both datasets with the service record. If the as-found results were acceptable and the as-left results remain acceptable, the history is straightforward. If as-found was unacceptable and as-left is acceptable, the document records a correction and creates a defined period to examine. If the tool could not be brought into the required condition, the certificate or service report should describe that outcome rather than being read as a routine release.

Error, uncertainty and the decision rule work together

The reported error is not the whole result. Measurement uncertainty describes the quantified doubt associated with the calibration result. The certificate may also state how uncertainty was considered when making a conformity statement.

For that reason, avoid reading a test point by comparing only its central number with a borrowed tolerance. Use the requirement that actually governs the wrench and read the laboratory’s stated decision rule. If the certificate is unclear, the laboratory or the organization responsible for the tool program should interpret the result before it is used to release work.

Scope remains decisive. A satisfactory point at one setting does not fill an untested gap elsewhere on the scale, and a result in one direction does not automatically transfer to the other direction.

Bound the look-back review with real use records

When an as-found result is unacceptable, identify the last known acceptable calibration or functional check and retrieve the wrench’s use log since then. Map actual jobs to the affected direction, settings and functions. Give priority according to the consequence of the joint and any independent evidence, such as installation records or observed movement.

This process may reveal that some jobs used a different range and are outside the affected scope. It may identify a smaller group that requires engineering or manufacturer-directed assessment. The appropriate response for those joints depends on their design and service procedure; the certificate alone does not prescribe loosening, retightening or replacement.

A well-read certificate therefore produces a defensible question set: Was this the wrench? Was this use within the calibrated scope? What was its condition before service? What changed? Which recorded jobs could reasonably be affected? Those answers turn a page of calibration data into a controlled maintenance decision without claiming more than the laboratory measured.

References

• TorqueLab: As found and as left torque-wrench calibration
• Norbar: UKAS-accredited laboratory

This article is part of our excavator undercarriage parts guide, which covers the five wear components, the order they fail in and how to confirm fitment.