In order to meet the requirements for a state-of-the-art metrology institute and in an endeavour to build a fully automatic mass laboratory, the Austrian Federal Office of Metrology and Surveying (BEV) has developed and realized in cooperation with Sartorius AG in Goettingen and the Vienna University of Technology three different handling systems for automatic testing weights on high-precision mass comparators.
Fully automatic System for alternating weights on mass comparators with different weighing capacity have been realized; one from 1 mg up to 10 g, one from 10 g up to 1 kg an the third one from 1 kg up to 20 (50) kg using two comparators. The resulting load alternator "robot" holds between ten and eighty test weights and a corresponding number of reference Standard weights. All weights are transported without pallets. The Systems are based on Computer control and linear drive trains, and are used at the BEV to disseminate mass and for the calibration and verification of weights.

In the research a possibility was studied how to reduce a number of repetitions of measurements when performing dissemination of 1 kg mass standard and at the same time maintain the level of confidence of the measurement results.
The method considered is designed for weighing schemes where a pair of support plates for weights is required to carry out comparisons of combinations of weights on comparators with automatic weight handler and the same pair of support plates is used for many of the compared weight combinations.
The results of analysis indicate to minor relation between the changes in mass differences of the compared weights and the changes in mass differences of the support plates. It can be seen that there is no influence on the mass of compared weights due to improper handling of the weights. The main reason for the deviations of the measurement results when using the support plates might be the performance of the comparator.

The presented study suggests a method that allows estimating the reproducibility error in torque calibrations without the need of performing several measurement series with the torque transducer being mounted in different mounting positions. The aim is to enable uncertainty estimations for simplified calibration procedures, too. The approach is to use statistical data gained from previous calibrations of transducers of the same type. The study shows how to select and classify past calibrations in order to allow prognoses for future calibrations. Furthermore it introduces the method of statistic evaluation and gives special considerations about the way in which the reproducibility error influences the overall measurement uncertainty in the case of a simplified calibration.

Force transfer standards allow high-accuracy measurements of forces to be carried out – but they measure only one component of the force vector. The spatial direction of this component is given by the orientation of the transducer. A new force transducer is now designed to measure additionally the direction of the force vector, i.e. all its three components, as well as additional bending moment components and the torque, with high accuracy. For use as a transfer standard, the force vector sensor has to be traced back to standard machines in national metrological institutes. For calibration in deadweight machines, new methods have to be developed.

This paper presents first results of experimental investigations performed with mechanical springs made of single-crystalline silicon (S-Spring) for load cells (LCs) with sputtered-on thin film strain gauges. The investigations are carried out on a specially manufactured S-Spring load cell (S-Spring-LC) with a nominal load of 6 kg and geometry parameters which are optimized by numerical simulations.
This paper deals with load depending deformation measurements of the S-Spring to verify the expected low mechanical after effects. For the first time the investigations are carried out with a Fizeau interferometer, which delivers 3-D topology data of surfaces. This interferometer actually used at PTB features the grand adventure to determine the deformation of a complete surface of a load cell under load within a short time interval and one single measurement.
The measurements reveal information about the fraction of deflection generated by tilting effects of the force introduction and which fraction is generated by deformations of the S-Spring. The investigations point out that tilting of the force introduction is not negligible but can be compensated analytically. Only with analytical compensation of tilting effects mechanical after effects of the S-Spring can be determined by recording a multiplicity of 3-D topology data within 30 minutes. As a result the S-Spring investigated shows mechanical after effects which value is about a factor of 100 lower in comparison to conventional metallic material LCs and therewith confirm values given in literature for mechanical after effects of silicon.