PSI - Issue 32

I.O. Glot et al. / Procedia Structural Integrity 32 (2021) 216–223

221

6

Shestakov A.P./ Structural Integrity Procedia 00 (2021) 000 – 000

Fig. 8. (a) spectrogram of the signal from the HFS5 sensor during dropping of ore (window size 0.05 s); (b) the spectrum obtained by averaging the spectrogram over time. 5. Calculation of spectrograms The spectrogram for the HFS5 sensor is shown in Figure 8a. This spectrogram is plotted for the time interval corresponding to the drop of the ore. The parameters of the windowed Fourier transform have the following values: window size 0.05 s, offset 0.02 s. The spectrogram clearly shows the unevenness in time of harmonic signals. This confirms the conclusion obtained in the previous section that technological operations, including dropping of ore, consist of a multitude of shock impact. The horizontal lines in the spectrogram are formed by oscillations at natural frequencies, since they have a greater contribution to the vibrational response of the structure and, as a consequence, have a higher intensity in the spectrogram. It also follows from the spectrogram that during the dropping, not all natural frequencies are well represented at all time intervals. Therefore, to determine the natural frequencies, the spectrogram should be averaged over time. The result of this operation is shown in Figure 8b. The spectrum obtained in this way makes it possible to unambiguously determine the natural frequencies of the structure. 6. Electrical noise fix Electrical noise problem can be solved with spectrogram processing. Figure 9 shows the spectrograms for the HFS3 sensor, calculated for the time intervals corresponding to dropping of ore. Before processing, the spectrogram

Fig. 9. Spectrogram of the signal from the HFS3 sensor: (a) without filtering; (b) with filtration.