PSI - Issue 5

H. Lopes et al. / Procedia Structural Integrity 5 (2017) 1205–1212 H. Lopes et al. / Structural Integrity Procedia 00 (2017) 000 – 000

1209

5

transitions from black to white and from white to black. However, this cannot always be clearly seen due to the loss of contrast of fringes, which is due to the presence of high levels of noise. Therefore, by comparing the contrast of fringes between phase maps it is possible to identify the phase map with higher levels of noise. This comparison, however, does not allow the quantification of these levels. For instance, from the comparative visual analysis of the phase maps presented in Fig. 3, it may be possible to conclude that the first and second images present higher levels of noise than the other two images. Therefore, one needs to find better ways to characterize the levels of noise. In the present work, the levels of noise are evaluated by computing its RMS and the SNR. The former is a measure of absolute accuracy of a signal, whereas the latter is usually a quantity relating variables with different magnitudes, thus measuring the relative accuracy.

Fig. 3. Phase maps corresponding to the first four modal rotation fields.

4.1. Shearing amount

A first set of measurements was carried out by fixing the vibration amplitude and setting the shearing amount to 2 mm, 5 mm, and 10 mm. Since the determination of each one of the four modal rotation fields was performed twice, we have a total of 24 measurements. In order to simplify and improve the consistency of the analysis, the RMS of noise and the SNR of each pair of modal rotation fields were averaged. Figure 4 shows the RMS of noise in the first four modal rotation fields as a function of the shearing amounts. It is clearly observed that there is a non-linear decrease of the RMS with the shearing amount. Although only three different shearing amounts have been considered, one can see that the rate of the decreasing of the RMS diminishes with the