PSI - Issue 64

Luigi Petti et al. / Procedia Structural Integrity 64 (2024) 637–644 Petti L., Lupo C., D’Angelo T., Dallocchio P. & Guizzetti D. / Structural Integrity Procedia 00 (2019) 000 – 000

642

6

The data analysis considers different time windows in order to analyse the response by taking into account daily, weekly and monthly evolutions. Longer time windows allow for the assessment of slow effects in order to evaluate the static behaviour of the structure. Shorter time windows allow for the better identification of rapid changes, such as those related to temperature and rain. The preliminary analysis of the raw data involved the correlation of the cinematic response of the structure with temperature. This was conducted by considering moving time windows of 6, 12 and 24 hours, with a step of one hour at a time. Figure 7 shows the trend of correlation coefficients obtained for a data sample of 120 days, considering a 24-hour moving time window. The graphs (b) and (d) respectively show the displacement and tilt response considering a delay between 0 to 8 hours with regard to the temperature. Graphs (a) and (c) describe the maximum correlation coefficients (blue points) respectively for displacement and tilt response, taking into account the rainy days (red rectangle). As described, the considered delays would to investigate the thermal inertia impact of the weather evolution on the structural response.

(a)

(b)

(d)

(c)

Fig. 7. Example of a correlation analysis on a sample of 120 day of raw data: maximum correlation coefficients (blue points) between displacement (a)/tilt (c) and temperature, taking into account the delay in the response and rainy days (red rectangle); correlation coefficients trends between displacement (b)/tilt (d) and temperature, considering delays in the response to the thermal variation from 0 to 8 hours. The correlation analysis allows to firstly assess the structural time response to the temperature variation. Figure 7 (b) and (d) show that the maximum correlation coefficients are reached with a delay of 1 hour (orange line). Additionally, the results highlight the effects of rain on the correlation analysis. It is evident that the rain affects the structural response by suddenly changing the stone surfaces as well as the device temperature. Moreover, while the correlation between tilt and temperature (fig. 7(c)) is relatively stable and close to one, a scattered pattern is evident for displacement (fig. 7(a)). To improve the assessment of the structural response evolution, the raw signals were processed through a Fourier based filtering methodology. As known, a function ( ) defined in the time domain, which is periodic and piecewise regular, can be expressed as an infinite sum of sines and cosines with increasing frequency (equation 1), see Brunton S. L. and Kutz N. J. (2019).