Time averaging interferometry is one of the techniques that allows to investigate a dynamic behaviour of MEMS elements. The information of the maximum amplitude of vibration is encoded in the argument of Bessel function. Many different approaches enable evaluation of this value. Due to the fact that accuracy of the result depends on the quality of the input data, Bessel function of a good quality must be calculated first. Wide range of modulation distribution calculation methods enable to obtain absolute or square value of J0. These results are more difficult in processing than pure Bessel function because of gradient function discontinuity or poor SNR. For ensuring smaller errors one may normalize absolute values of J0 by the sign of the function. In this paper, several different algorithms for determining the sign of Bessel function were investigated and compared. For each approach accuracy of the method was calculated. In the end, the best solution was found.
Time averaging technique applied to different interferometric methods is one of the most commonly used measurement technique in vibration testing. The information on amplitude of vibration is encoded in so called fringe envelope function described by the J0 Bessel function in case of harmonic motion. In the presented paper the author proposes novel solution for amplitude distribution evaluation. It is based on the analysis of the Bessel pattern modulation distribution. For modulation evaluation and Directional Spatial Carrier Phase Shifting (DSCPS) method are used. Conducted error analysis prove usefulness of the proposed approach.
Classical time-averaging is widely used for MEMS/MOEMS dynamic behavior investigations. In order to evaluate the information on maximum amplitude at a given load of vibrating object one needs to evaluate the argument of Bessel function that encodes the useful information. For this purpose many different approaches were presented. Among them Temporal Phase Shifting applied to Bessel fringes is of special interest since it provides most accurate results. It, however, requires additional cumbersome pixelwise correction routine via specially designed look-up-table. In this paper we investigate, through extensive numerical simulations, the possibility of reduction of phase evaluation error (without correction routine) by different strategies of phase shifting. Two different 5 step algorithms are investigated for that purpose. Additionally, quick and robust correction procedure based on evaluated phase distribution is presented.