GNGTS 2014 - Atti del 33° Convegno Nazionale

316 GNGTS 2014 S essione 2.3 Fig. 2 compares the SSRs obtained for the two horizontal components separately (the SSR relative to the vertical component is also shown) with H/V spectral ratios calculated from the recordings at the third floor. The SSR based on the recordings of the Piacenza earthquake is also shown for comparison. The structure shows a fundamental frequency at around 3 Hz. Note that this peak appears shifted towards 4 Hz in the H/V curve determined from the radial component (X) of noise recordings. This is indicative of differential noise polarization. Besides the 3 Hz peak, two further peaks can be identified at frequencies of about 7.5 Hz and 11 Hz (this latter peak is more evident from ambient noise measurements). SSRs allow us to exclude that these peaks are generated by the resonance of the ground. As known, SSRs have the advantage of deleting the influence of the soil cover, which could mask some eigenfrequencies of vibration of the building. Although the cause of these peaks remains uncertain, the former could be related to the second mode of vibration of the structure while the latter could be indicative of the response of the building to traffic noise (Beijing Jiaotong University, 2007; Luo et al. , 2008). Analysis of filtering effects for the seismic assessment. Because of the seismic vulnerability of masonry structures, the seismic assessment and the protection of historical centers is a very important and complex topic. The seismic behavior of this kind of structures is influenced by numerous factors deriving from their articulated configurations, which are often the result of different spontaneous transformations stratified during their life time. Hence, they are often prone to the activation of out-of-plane mechanisms, which may cause structural failures for the loss of equilibrium of local portions. Since local mechanisms generally involve masonry portions placed in the upper part of a building, an important issue is that related to the adoption of proper spectra taking into account the filtering effect provided by the structure (namely floor spectra ). Starting from the analytical floor spectra originally proposed by Suarez and Singh (1987), some approximate formulations have been derived and proposed by Curti (2007) and Lagomarsino (2014). Also in MIT (2009) some analytical expressions are proposed. In such context, the measurements and experimental data discussed in the previous sections are particularly useful to validate such analytical formulations and to calibrate the filtering effect provided by the examined structures. To this end, by way of example, the recordings of the October 3, 2012 Piacenza earthquake are processed in order to compute the corresponding acceleration response spectra at the base and third level of building B3 (Fig. 3). It is possible to observe two main amplifications in the floor spectrum in correspondence of the fundamental period ( T 1 ≈ 0.33 s) and the second mode ( T 2 ≈ 0.13 s) of the structure, being these values in agreement also with those discussed in the previous section. In order to verify the reliability of the aforementioned expressions proposed in the literature, the floor spectrum obtained from the experimental results has been compared with the analytical Fig. 3 – Comparison between the analytical floor spectra and the acceleration response spectra relative to the October 3, 2012 Piacenza earthquake.