Prediction of tunnelling-induced effects on a historic building in Rome

Tunnelling in urban environments often poses serious concerns regarding potential detrimental effects induced on the surrounding structures and infrastructures, as in the case of the soon-to-be-built new Flaminio railway station. The station's project involves the conventional excavation of three adjacent shallow tunnels, whose roofs are located very close to the foundation of a historic building. To minimize the risk of damaging the building, the excavation is preceded by ground improvement works and the installation of a very stiff pre-support system made of concrete-filled steel pipes just above the tunnels’ roofs. This paper presents a Class A prediction of tunnelling-induced effects, obtained developing a three-dimensional numerical model that accounts for the complexities of the problem's geometry and construction phases. The soil mechanical behaviour is described by an advanced constitutive model (the Hardening Soil model with small strain stiffness). The results provide useful insights about the soil-tunnel-building system's behaviour. The expected settlements are very little and the potential damages to the building practically negligible. In addition to drastically reducing the risk of local instabilities, the heavy ground improvement works and the installation of the pre-support – whose effects are quantified through comparative analysis – further reduce settlements and distortions induced on the building. Finally, the results are compared with the predictions obtained adopting a simple linear elastic perfectly plastic constitutive model (Mohr-Coulomb model) both using the parameters employed at the design stage, calibrated assuming precautionary medium–high strain levels, and attempting to capture a more realistic level of induced strains. A comprehensive monitoring system will record a significant amount of data concerning the soil-structure interaction, enabling to better understand the behaviour of the system and to test the accuracy of the numerical predictions.