Vibrational interaction between urban agglomerates and geological system with heterogeneous composition

This PhD thesis evaluates Site-City Interaction (SCI - Guéguen et al. 2002, Bard et al. 2005, Guéguen and Bard 2005, Kham et al. 2006, Semblat et al. 2008), i.e., the influence of buildings on ground motion and local seismic responses. The free-field conditions of the ground surface, which consider the absence of structures that can generate vibrations, are widely used to analyse the local seismic response for both scientific and technical issues. Nevertheless, this assumption leads to a very strong approximation in urban areas, in which the dynamic interaction between the urban agglomerate and the soil cannot be neglected. Many bibliographic data showed the transmission of vibrations from buildings to the soil, i.e., the amplification of ground shaking during earthquakes and the generated wave field, which can propagate far from the city centre (Wirgin and Bard 1996, Guéguen et al. 2000, Kham et al. 2006, Semblat et al. 2008). Previously published studies mainly focused on engineering, while the geological component was strongly simplified. Thus, complex 1D amplification effects from the soil layering and 2D effects from the lateral heterogeneity, topography and shape of the seismic bedrock have been underestimated. The Fosso di Vallerano valley case study was chosen because this area is characterized by a highly heterogeneous geological setting and has recently experienced a massive expansion of urbanization during the last decade, which completely perturbed the free- field conditions of the original alluvial valley that was created by the tributaries of the Tiber River during the Holocene. In particular, the Fosso di Vallerano valley hosts the “Europarco Business Park”, i.e., the highest buildings (120 m) in Rome. A preliminary phase of this research was dedicated to the reconstruction of the engineering-geological model of the valley, the 1D numerical modelling of the seismo-stratigraphic setting of the alluvial body (Bozzano et al. 2015, 2016) and the calibration of an absorbing layer system to remove spurious wave reflections at the model boundaries (Varone et al. 2014). Several geophysical investigations have also been conducted by considering both seismic events and noise measurements. In recent decades, great effort has been dedicated to numerical approaches to evaluate the local seismic response. Numerical modelling actually represents the main tool to estimate local seismic responses, particularly in urban areas, where geophysical measurements are often not suitable. A proper 2D numerical modelling of the seismic response in free-field conditions that considers the city agglomerate according to an SCI approach is conducted through the CESAR-LCPC FEM code, which considers two geological sections across the Fosso di Vallerano valley. The structural and dynamic features of different building typologies in the selected study area are also considered. The urban agglomerate in this valley mainly consists of residential reinforced 11 concrete (RC) buildings, which are characterized by rectangular or square geometry and heights from 6 m to 25 m. The valley also hosts particular type of buildings that are part of the “Europarco Business Park”, including two skyscrapers (named “Europarco Tower” and “Eurosky Tower”) that are 120 m and 155 m high, respectively. These towers are characterized by a rectangular plan geometry and consist of steel that is coupled with a reinforced concrete structure. The eigenmodes are computed through the CESAR-LCPC code to evaluate the dynamic characteristics of the buildings. All the buildings are modelled by considering their super-structure, i.e., columns and beams, and by assuming the concentrated masses and stiffness values in 2D. The main periods of the urban expansion are defined to evaluate the variations in the local response because of the increasing urbanization that was observed over the last decade. Overall, 5 models that assume SCI conditions are simulated. Ricker wavelets (Ricker 1943, 1953) of order 0 with PGD = 1 m (synthetic wavelet) and three real weak motions are applied as seismic input within the seismic bedrock of all the models by assuming visco-elastic conditions. The results are analysed in terms of wave propagation along the models’ surfaces, the spatial variation in the amplification function and the distribution of the kinetic energy along the ground surface. The wave propagation maps show the effectiveness of the absorbing layer system on both sides and at the bottom of the models to dampen spurious waves. Indeed, no spurious contributions enter the valley through the boundaries or the bottom of the model. The presence of the buildings induces major changes in the propagated wave field, inducing a low ground motion at the building’s foundation level and increasing the ground shaking in the area surrounding the structures. The influence of these buildings is also very important in the amplification function: all the modes under free-field surface conditions are nullified in the portions that are occupied by the buildings, and larger amplifications are calculated laterally and near the buildings’ foundations. Additionally, a redistribution of the energy along the ground surface is shown by a strong reduction in the kinetic energy close to the building and an increase in the areas surrounding the buildings. These findings highlight that the presence of buildings significantly changes the seismic response of the alluvial valley at least at the local scale.