Errors in dynamic force measurement

Abstract

Loadcells for materials testing machines are calibrated statically by comparison with a transfer standard, traceable to the National Physical Laboratory. There is evidence to suggest that, under certain conditions, this static calibration may be insufficient for dynamic testing. The possibility of an error being generated by the inertia of the mass between the loadcell and the specimen is discussed and the results of experiments to measure this error are presented. For the case chosen it is shown that the inertia error may be predicted by calculation. Errors may also arise from the method of dynamic force measurement. The conventional DVM and the analogue peak hold voltmeter are widely used but are subject to errors in a practical machine situation due to their method of operation. Bandwidth limitations are also illustrated for a number of instruments with different sampling rates. A better, although more expensive and complicated solution is to digitise the signal and to use some form of spectral analysis such as the Fast Fourier Transform or the Cross Correlation Integral. The theory for the latter is presented and its merits discussed.