According to the ISO/IEC Guide 99—“International vocabulary of metrology—Basic and general concepts and associated terms (VIM),” the traceability of a measurement result is demonstrated through a documented unbroken chain of calibrations, each contributing to the measurement uncertainty. This means that not only is it necessary to use a reference standard of length that has an unbroken chain of calibrations, but the uncertainty of the resulting measurements must also include the uncertainty of all the links in the calibration chain.
ADVERTISEMENT |
While the VIM definition is clear about what is needed for the traceability of a measurement result, it does not specify what the end of the calibration chain should be. Some may see this as a national metrological institute, such as the National Institute of Standards and Technology (NIST). Others may assert that a single calibration certificate is sufficient if, for example, it’s from a lab accredited by an organization like the International Laboratory Accreditation Cooperative (ILAC). Depending on the needs of an organization, either of these traceability interpretations might be sufficient to the users of their measurement results.
There are various interpretations of the VIM definition of traceability. In this article I’ll briefly review a few of them. We first look at traceability from the perspective of a national metrological institute like NIST that does not rely on external calibrations. This is followed by an interpretation used by accreditation bodies for testing and calibration labs, such as ILAC. Finally we highlight some key points of a technical report on traceability from ASME B89, which is focused on describing traceability for metrology practitioners.
NIST’s website directly addresses their policy toward traceability. It states: “NIST does not define, specify, assure, or certify metrological traceability of the results of measurements except those that NIST itself provides.” Which basically means that NIST makes no claims on the traceability of measurement results unless they were directly measured by NIST. The policy further states that supporting traceability claims is the responsibility of the measurement provider. Taken together, this implies that no organization may claim that NIST will confirm a measurement result to be traceable to NIST. Rather, it is up to the organization that performed the measurement to prove through calibration documents that traceability is linked to NIST. For example, at the end of the calibration chain, a length standard may have been used that is calibrated by NIST and has a document of calibration.
Another interpretation of traceability is to rely on accreditation. International accreditation bodies for testing and calibration labs, such as ILAC, have a policy to say that traceability is assured by the use of calibration services from accredited labs. Accreditation is typically achieved by complying to ISO 17025—“General requirements for the competence of testing and calibration laboratories” and passing an external audit of that compliance. When measurements are performed by an ILAC-accredited lab, they may claim that their results are traceable. To support the claim, they need only show a calibration certificate for their length standard because the accreditation implies that the chain of calibrations is maintained upstream. While accreditation provides a sense of reliability and robustness, it can be an expensive process that consumes a lot of time and resources. It may not be appropriate for organizations that are the primary users of their own measurement results. For example, a manufacturer that would simply like to know whether the measurements made on the shop floor are traceable.
The technical report, ASME B89.7.5—“Metrological traceability of dimensional measurements to the SI unit of length,” provides practical steps for an organization to self-identify their measurement results as traceable. The report describes traceability in the same spirit as the VIM definition, but it also clarifies several key aspects relating to the calibration chain, uncertainty, and quality assurance. The report states that the calibrated length standard used by an organization should have documentation traceability back to the meter. For example, there should be a chain of calibration certificates for the reference object used that culminates in a calibration from a national metrological institute like NIST. A second requirement of the B89 traceability is that every measurement result must have an uncertainty that is consistent with the ISO/IEC Guide 98—“Uncertainty of measurement—Part 3: Guide to the expression of uncertainty in measurement (GUM:1995).” Finally, a quality assurance program is needed to monitor the status of the measurement system and calibrated reference object. For example, an organization should implement a process control system that periodically recalibrates their length standard and re-verifies their measurement instrument. These are just a few of the key requirements of B89 traceability, refer to the technical report for more information and details.
Depending on the needs of an organization, each of these interpretations of VIM traceability might be appropriate. Whether the measurements are based on self-calibration and for internal use only, or if they are produced by an accredited testing and calibration lab, understanding the chain of calibrations and their uncertainty is key to knowing the traceability of a measurement result.
Add new comment