GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 2.1 GNGTS 2023 The two-dimensional analysis is performed using the commercial code LSR2D from STACEC, which is based on an equivalent linear finite element method. For the 3D analysis, we used the open-source Cartesian code SPECFEM3D instead (Peter et al., 2011). For both the 3D and 2D analyses, we assumed linear viscoelastic and nonlinear behaviour, respectively, with seismic waves propagating at frequencies of engineering interest (i.e., 1-10 Hz). To achieve computational cost balance, the 3D model is discretized into 25 m hexahedral elements. Since the area to be analysed in 2D is only 2 km 2 (the 3D model is 4 km 3 ), a discretization of 10 m and a higher level of detail can be achieved. The results of the numerical simulations were compared with available experimental data (Laurenzano et al., 2019). The 2D numerical analysis was performed on an about N-S oriented section passing through the Old Town of Arquata del Tronto located on the hill and the neighbouring hamlet of Borgo, located on a sediment-filled valley at the foot of the hill (Fig. 1). Results The results were analysed in terms of the amplification functions (the ratio between the amplitude spectra of the ground motion and the seismic input) and the Housner amplification factor (the ratio of the Housner spectral intensity of the simulated ground motion and the corresponding seismic input). The comparison between the amplification functions, obtained by 2D and 3D simulations along the section, is shown in Figs. 2a and 2b. The 2D simulation predicts the largest amplification in the sedimentary valley, with an amplification of more than 2 between 4 and 6 Hz, but in the old hamlet, over the hill, the expected amplification is lower than in Borgo, between 2.5 and 3.5 Hz. On the other hand, the 3D simulation predicts lower amplification in the Borgo valley and strong amplification values over the whole frequency range on the hill, which correlates well with the topographic relief. We can conclude that the 2D simulations underestimate the effects of topography and overestimate the effects of valleys compared to the 3D simulations. These results are also evident from the analysis of Housner amplification factors obtained from 2D and 3D simulations along the section (Fig. 2c) in two period ranges: FA1 = [0.1-0.5] s (solid line), FA2 = [0.5-1] s (dashed line). The amplification values obtained from the 3D simulation are in better agreement with the observed damage distribution in the study area than the values obtained from the 2D simulation.