Anyone who wants to verify a tensile or compression test bench according to ISO 7500-1 needs a reference force transducer that has itself been calibrated according to ISO 376 and assigned to a class. That class, 00, 0.5, 1 or 2, indicates how well the instrument performs on a number of strict criteria. In this article you can read what the classes mean, how an ISO 376 calibration is carried out and which class you need in practice. How we perform such a calibration is described on the page ISO 376 calibration of reference force transducers.
What is ISO 376 about?
ISO 376 describes the calibration of force measuring instruments used for the static verification of uniaxial testing machines. Think of a reference force transducer with readout that a laboratory or machine manufacturer uses to check a tensile test bench. The standard lays down how the instrument must be loaded, in which positions, how many steps are required and which quantities are determined. On that basis the instrument is assigned to a class, for a given range and a given period of validity.
The standard is therefore not intended for every force gauge. A dynamometer in lifting applications or a load cell in a press is usually calibrated as a measuring chain for its application, without classification. For that, see what is force calibration?
The four classes and their criteria
ISO 376 assesses the instrument on a number of relative errors, expressed as percentages of the indicated force. The most important are:
- Reproducibility (b): the difference between series in different mounting positions of the instrument;
- Repeatability (b′): the difference between two consecutive series in the same position;
- Interpolation error (fc): the deviation between the measured values and the interpolation equation that describes the relationship between force and indication;
- Zero error (f0): the indication after unloading, as a fraction of the maximum force;
- Hysteresis or reversibility (v): the difference between increasing and decreasing loading at the same force;
- Creep: the change in indication over time at constant force, for instruments where this is relevant.
Maximum values apply per class. In broad terms:
| Class | Reproducibility b | Repeatability b′ | Interpolation error fc | Zero error f0 | Hysteresis v |
|---|---|---|---|---|---|
| 00 | 0.05 % | 0.025 % | ±0.025 % | ±0.012 % | 0.07 % |
| 0.5 | 0.10 % | 0.05 % | ±0.05 % | ±0.025 % | 0.15 % |
| 1 | 0.20 % | 0.10 % | ±0.10 % | ±0.05 % | 0.30 % |
| 2 | 0.40 % | 0.20 % | ±0.20 % | ±0.10 % | 0.50 % |
The exact requirements, including those for the resolution of the readout and for creep, are given in the current edition of the standard. Important to remember: the instrument is given the class of its worst criterion. A transducer with excellent repeatability but too much hysteresis therefore ends up in a lower class.
How is an ISO 376 calibration carried out?
The procedure is more demanding than an ordinary force calibration:
- Preloading. The instrument is preloaded three times to the maximum force, so that the measuring body settles.
- Several positions. The instrument is measured in at least three mounting positions, each time rotated about the force axis. In the first position two series are measured for repeatability; the other positions provide the reproducibility.
- At least eight steps. In each series the force is increased in at least eight load steps, distributed as evenly as possible over the range, and where necessary decreased again for the hysteresis.
- Calculation. From all the series the relative errors are calculated, the interpolation equation (usually a polynomial) is established and the measurement uncertainty is determined.
- Classification. The instrument is assigned to the class it meets over the entire calibrated range. The class only applies within that range; below the lowest calibrated force the class is not valid.
An ISO 376 calibration is generally carried out for the complete measuring chain: transducer and readout. If you change the amplifier or display, the classification lapses.
Which class do you need?
That follows from ISO 7500-1. It states that the reference instrument used to verify a testing machine must be of a class equal to or better than the class in which the machine is to be classified. If you want to verify a tensile test bench as class 1, you therefore need at least a class 1 reference; class 0.5 gives extra margin. For classes 00 and 0.5, a very stable readout with high resolution is also required.
If you do not use the transducer to verify machines but as a reference for your own force measurements in production or quality control, a classification is often not necessary. A regular calibration with clear deviations and measurement uncertainty is then sufficient.
Class and measurement uncertainty
A class is not a measurement uncertainty. The class says that all criteria remain within a limit; the measurement uncertainty on the certificate combines the actual, measured contributions of reproducibility, repeatability, interpolation, hysteresis and the uncertainty of the force calibration bench itself. A class 1 transducer can therefore have a measurement uncertainty well within the class limit. If you use the transducer as a reference, you work with the measurement uncertainty from the certificate, not with the class percentage. How that uncertainty is calculated is laid down in EURAMET cg-4.
Validity and interval
ISO 376 sets the maximum validity of a calibration at 26 months. Many laboratories choose a shorter interval, for example 12 or 24 months, matched to the use and history of the instrument. After an overload, damage or repair, immediate recalibration is required. How to justify an interval is explained in how often should a force gauge be calibrated?
In summary
ISO 376 is the standard for reference force transducers used to verify tensile and compression test benches. Classes 00, 0.5, 1 and 2 indicate how well the instrument scores on reproducibility, repeatability, interpolation error, zero error and hysteresis. Which class you need is determined by the class of the machine you want to verify. Do you have a reference transducer that is due for calibration? See ISO 376 calibration at MASSLAB.