by David A. Vossbrinck
Traceability of accuracy for measuring or calibration equipment forms the
very foundation of any quality program. Without traceability, all other
aspects of quality control, from simple calibration procedures through statistical
process control, can be worse than meaningless-they can be misleading.
Tracing the accuracy of any and all measuring equipment to a central control
authority is essential for accurate communication within and between industries.
One one-thousandths of an inch must mean the same in New York City as in
San Diego. Without traceability to an absolute standard, there would be
What is traceability?
Traceability is the comparison of any measuring device or system to a standard
of greater accuracy. Typically, a minimum of four-to-one greater accuracy
is required, although 10-to-one is not uncommon. This ultimately has a known
comparison to a national standard such as those kept by the National Institute
of Standards and Technology. They, in turn, periodically send their standards
to Sevres, France, to the Bureau International des Poids et Measures in
order to achieve international agreement.
Although a standard should be at least four times more accurate than the
device it is testing, it need not have the same principle of operation.
For instance, a spring scale with an accuracy of ±1 percent could be
certified by a load cell with an accuracy of ±0.25 percent, which in
turn could be certified by dead weights with an accuracy of ±0.0625
The load cell is a secondary standard, while the calibrated weights are
considered a primary standard because of the direct link between the scale
(measuring in pounds) and the known weights. Secondary standards are used
for ease and convenience but must ultimately be traceable to primary standards.
Any valid certification of accuracy has all the information necessary to
trace the standards used to NIST.
Sometimes it is necessary to provide traceability of two standards to qualify
the calibration of one device; a torque device for example. Torque is defined
as a force multiplied by a known distance from a center point. In the case
of torque, both weights and calibration arms must have independent certifications
traceable to NIST, and the final accuracy is a combination of each component's
uncertainty. Hanging weights and arms on every torque tool would be an unmanageable
procedure, thus the domination of strain gauged torque analyzers used to
calibrate torque tools.
Periodically, calibration equipment needs to be certified. In some cases,
the frequency of recertification is determined by the manufacturer's recommendation,
but ultimately it is determined by each company's quality assurance program.
Frequency of usage plays a large part in frequency of recertification, as
does the difference in accuracy between the calibration equipment and the
Unfortunately, recertification may be useless if the agencies performing
the recertification don't follow equipment procedures. Proper procedures,
while essential to any meaningful recertification, are often omitted in
quality audits. For example, auditors pay attention to the traceable certification
of an arm used as a primary standard for measuring torque, but often overlook
errors introduced during its use. It is unrealistic for quality auditors
to be experts in procedures for every type of calibration equipment used,
so they rely on the agencies doing recertification work to generate and
properly use procedures.
Periodic recertification, traceable to NIST, is the core of any quality
assurance program when measuring devices are an essential part of a company's
manufacturing or inspection process. Without it, liability increases while
product quality suffers, with a resultant increase in the overall cost of
About the author . . .
David A. Vossbrinck is CEO of Torque & Tension Equipment Inc.
in Santa Clara, California. For more than 25 years, Torque & Tension
Equipment has supplied both primary standards and calibration equipment
to all industries for the measurement of static and dynamic torque.