The work presents an improved approach to the determination of magnetic properties of weights, which have a shape according to OIML R111. The main stress is laid on a measurement of the volume magnetic susceptibility and an evaluation of its uncertainty. The method is presented, how the volume magnetic susceptibility of the OIML weight can be determined taking into account an actual shape of the weight instead of using only an approximation of its shape with "outer" and "inner" cylinders. Monte Carlo simulation was used to evaluate the measurement uncertainty. The susceptibility of 1 kg weight was calculated by the method of "outer" and "inner" cylinders and the method presented in the article. Results of both methods are compared and discussed.

The paper presents a practical example for the evaluation of the uncertainty in calibration using a comparison force standard machine with a maximum load of 1000 kN, tension and compression mode. The uncertainty of measurement associated with the input estimates is evaluated. It is also presented a way to minimize the uncertainty of the applied force with the view to attaining the necessary value for the calibration of class 1 force transducers, in accordance with ISO 376:1999. For this purpose an evaluation of the loading regime of the force machine is presented. There was identified the main input components influencing the uncertainty in calibrations, such as the force stability during measurements and the metrological characteristics of the reference transducers. Further are presented the practical results of the investigation of the force comparison machine after the optimization, including the technical and economical benefits.

The highest level of torque calibration is achieved by direct loading and, due to the use of mass stacks, is associated with the step-by-step method. In order to fit the demand of applications especially in industry, calibration devices with the ability of performing continuous loading were developed which are working as reference machines. A benefit of continuous methods is that the time for a loading-unloading cycle can be ten times or more shorter than using the step method. This fast change in torque leads to conditions of use that cannot be represented in the national torque standards. Therefore the appointed reference- and transfer-transducers should have negligible sensitivity to fast loading effects. With such transducers the results of slow absolute calibrations in the national torque standards are valid also in the fast continuous-working machines. In Germany a guideline treating this problems of continuous calibration is in preparation at a working group of the German Calibration Service (DKD). In this guideline special qualification-tests are proposed which can be performed in existing direct loading machines and which can deliver information about the suitability of transducers as reference or transfer for continuous working machines. In this paper we describe these qualification-tests, discuss the results of the tests in comparison to the behaviour of the tested transducers in continuous measurements and recommendations for the selection of transducers for continuous use are given.

Measuring of large forces up to 10 MN with high accuracy is a complex subject, blending theory, practice and applications, with a special accent in the field of Metrology. It is simply to generate complex shapes of mono-block elastic structures starting from simple shapes of strain gage measuring sections, the axisymmetrical ones being the best suited in this respect. The body of the force transducer is easy to design by 2D axisymmetric FEM and easy to manufacture. The paper have in view a unified approach of square (type A) and rectangle (type B). Square section is among the few with an analytical but complicated formula, having two strain gauges tangentially located on the outside of the ring torsion and the other two strain gauges, which complete the Wheatstone bridge, being diametrically opposed. The rectangular section (with different ratios between the two sides but without analytical formula), is obtained by a minimum modification of the previous one (making two slots), the strain gauges being circumferentially located on the upper and the lower faces. We formulate a standard FEA procedure for axisymmetrical elastic elements of strain gauged force transducers using ANSYS Mechanical program. The starting model is very “flexible”, so that, changing one by one different parameters (e.g. modifying the keypoints coordinates), a lot of variants could be studied. A special attention is necessary to obtain proper strain diagrams on the superior, lateral and inferior sides of the elastic element measuring section, because it is essential to compare these diagrams in order to establish the best strain gauges positioning. In this respect, appropriate paths were conceived and plotted each time on graph, more precise and suggestive than plotting on geometry. An original grouping of all kinds of strain diagrams on the same plot of the deformed and undeformed bodies is presented. A lot of interpretations are possible based on the multitude of data and having in view: the influence of the axisymmetrical elastic element shape (varying different parameters) on the strain gauge sensitivity; the best combination of conflicting design criteria: strain, stress (determining the overload) and displacement (determining the stiffness); the possible correlation between the nominal load and the dimensions of a particular variant, e.g. with square measuring section. In the next stages will be very useful to unite FEM and CAD, following the elastic elements parametric modeling for their best constructive optimization.

A revised draft of OIML R111 has been approved recently. Among the many notable changes to the previous edition are specific requirements for both the magnetic susceptibility and residual magnetization of weights used in legal metrology. The highest class of such weights is E1, for which a 1 kg weight has a manufacturing tolerance of ± 0.5 mg. It therefore follows that better magnetic properties might be needed for weights having mass uncertainties substantially better than E1 tolerances. The work presented below shows how this problem was approached at the BIPM and supports our conclusion that alloys meeting the new Class E1 specifications for 1 kg weights have magnetic properties sufficient for the needs of the BIPM.