Logo of the Physikalisch-Technische Bundesanstalt

Metrological assessment of the 1 N·m torque standard machine completed


For very small static torques of up to 1 N·m, Working Group "Torque" has developed a standard facility which works according to the established principle – with a lever supported by an air bearing and load masses. The metrological investigations of this facility have recently been completed. These consisted in determining the quantities having an influence on the measurement result and in demonstrating the measurement uncertainties the facility is able to attain.

For the best possible realization of the unit of torque, the torque standard machines of the Working Group "Realization of Torque" are based on lever/mass systems supported by bearings. Hereby, air bearings are used, by means of which the forces required to generate torque ("force at the lever arm") are compensated for, without having a significant influence on the torque generated. This procedure allows a range from 1 mN·m to 20 kN·m to be realized. The smallest measuring facility has a measuring range of up to 1 N·m, and the investigation consisted in finding out to what extent different influence quantities may limit the measurement uncertainty which can be attained with this facility. From other machines we know that the lever arm length measured with coordinate measuring machines determines the best achievable measurement uncertainty to be 0.002 %. For the 1 N·m facility, however, 0.01 % were aimed at. Here, however, the influence of the residual friction in the air bearing had to be determined to be able to make a realistic statement on the measurement uncertainty in the range of 1 mN·m.

Figure 1 shows a result of these investigations which clearly demonstrates that a change in the effective torque by 1 mN·m can be detected with the aid of an adapted torque transducer. One can, however, also see the fluctuation of the amplifier's indication as well as a slight drift of the signal. The fluctuation of the indication is due to the fact that it is extremely difficult to manufacture torque transducers with this small measuring range which supply a high output signal but are, at the same time, not so sensitive that they are overstrained when they are normally handled during transport and installation. The drift of the signal shows that the influences of the ambient conditions (temperature and relative humidity) on such small transducers are clearly stronger than in the case of larger transducers.

Figure 1: Output signal of a torque transducer with a nominal torque of 0.1 N·m after repeated changes of the acting torque by ± 1 mN·m.

All known influence quantities were compiled and assessed in a publication. The measurement uncertainties were calculated using the known procedures described in the "Guide to the Expression of uncertainty in measurement" (GUM). It turned out that the smallest achievable expanded (k = 2) measurement uncertainty in the lower range (from 1 mN·m to 10 mN·m) amounts to 0.1 %, while otherwise 0.02 % or even 0.01 % (from 0.1 N·m to 1 N·m) can be attained. As expected, the dominant influence in the lower range was the friction of the air bearing. In the upper range, the uncertainty of the determination of the lever arm length represented the main contribution to the uncertainty. This article was published in the scientific journal "Metrologia" in January 2014.


[1] D Röske 2014 Metrologia 51 (2014) 87-96, Opens external link in new windowhttp://iopscience.iop.org/0026-1394/51/1/87/

Contact person:

Dirk Röske, Dept 1.2, WG 1.22, e-mail: dirk.roeske@ptb.de