Abstract
Estimations of calibration uncertainty in vibration acceleration standards are described. The sensitivity of charge amplifiers and parasitic motions of a vibration exciter are measured, and then a method of analysis of uncertainty is proposed. In the calibration of charge amplifiers, the calibration procedure is explained in detail, and several CAs (charge amplifiers) are calibrated in terms of the frequency sensitivity gain, the long-term stability, the linearity for input charge, and the linearity for gain setting. The maximum discrepancy in the sensitivity gain is 4% among the tested CAs. A sensitivity difference of 1.5% still remains when same type CAs are compared. It should be mentioned that a difference of 1.5% is not negligible. Such a large difference is not well known up to the present time. As for the evaluation of the vibration exciter, its parasitic motion is investigated using a dual laser doppler vibrometer (LDV). The parasitic motion mainly consists of transverse, rocking and bending motions. In order to evaluate the transverse motion quantitatively, the ratio of the vibration amplitude in the transverse direction to that in the longitudinal direction is measured as a function of the frequency. The ratio curve shows clear peaks at the resonance frequencies of the exciter. A novel method of estimating the uncertainty due to the transverse motion is proposed. Additionally, the effects of rocking and bending motions on acceleration measurement are discussed. To reduce the effects, a simple compensation method is adopted, and improved calibration results are obtained.
