GNGTS 2015 - Atti del 34° Convegno Nazionale

The energy E ( x ) reached at the surface along the section was calculated according to the Eq. 2 as proposed by Kham et al. (2006): (2) where T is the signal duration, u the horizontal displacement, ρ the density, and t the time. These parameters represent a cumulated kinetic energy, for unit of volume, normalized to the signal duration. The seismically induced strain, in terms of maximum shear strain (MSS) along the geological section was calculated starting from the displacement values: (3) where Δ U is the difference between the displacement values corresponding to two points located at two difference depths and at the same distance along the section, Δ H is the difference between the two depths at which the displacement values are related to. Since it is was already proved that seismically-induced strain effects in a heterogeneous alluvial fills are significantly conditioned by 2D effects (Martino et al. , 2015), a numerical modelling of the valley with a clayey homogeneous filling (lithological unit 4 in Tab. 1) was also carried out. The maximum shear strains (MSS) obtained by the 2D numerical modelling, considering heterogeneous and homogeneous fill respectively were compared with the MSS obtained by 1D modeling carried out through the EERA code (Bardet et al. , 2000) by discretizing the numerical domain in soil columns distant 10 m each other. Results. The distribution of the E ( x ) values along the geological cross-section of the Fosso di Vallerano valley (Fig. 1 top) shows that the thickness of resonant body, i.e. recent alluvial deposits and part of the Paleo-Tiber 4 deposits, plays an important role on the energy reached at the surface. In fact, the highest values (up to 8000 J/m 3 ) are reached in the portion of the alluvial body where the thickness is lower. The wave propagation maps shown in Fig. 1 (middle) point out the efficiency of the absorbing layer system, in particular is possible to notice that, at the lateral boundaries, the portion of the model that correspond to the inner side of the absorbing layer system is characterized by the absence of spurious waves due to lateral reflections. Moreover, it is worth noticing that the wave propagation along the geological section is strongly influenced by the shape of the valley. The analysis of the A( f ) x (Fig. 1 bottom ) highlight a non-homogenous distribution of the resonance peaks along the valley; more in particular, a wide part of the section is characterized by a first resonance peak around 1 Hz while the upper resonance modes are due to the peculiar heterogeneity in each portion of the valley. Indeed the central and the eastern portion of the valley show a first resonance peak at a higher frequency value (around 3 Hz) while upper modes due to the peculiar heterogeneity in each portion of the valley are also present. The distribution of the MSS values were obtained along the geological cross sections by interpolating through a Kriging regression the MSS values computed with the 1D and 2D numerical modelling of the heterogeneous and of the homogenous geological sections. For each assumed condition, the results are reported in contour maps (Fig. 2). The MSS distribution in the heterogeneous model (Figs. 2a-2c), both in 1D and 2D condition, shows that the highest shear strain values are concentrated in the recent alluvial body, more in particular in the lithological unit 3 (peaty clay). In the homogenous model (Figs. 2b-2d), the MSS values show highest values in the western part of the valley that is characterized by a wide shape of the valley respect to the eastern portion; anyway, these values are lower respect to the ones resulted in the same location by considering a heterogeneous filling. GNGTS 2015 S essione 2.2 181