AbstractsBusiness Management & Administration

Tunneling construction represents an alteration to the distribution of soil pressures that will almost inevitably generate ground subsidence, which can endanger the adjacent buildings in urban areas. The risk of building damage can be substantially reduced with a comprehensive preliminary damage assessment during tunnel design phases, in combination with excavation techniques that minimize ground subsidence. The present research aims to increase the knowledge concerning the analytical and numerical techniques for building damage prediction related to tunneling. The thesis starts with the study of a real case of masonry building affected by the construction of the L9 metro line tunnel in Barcelona. Data available made possible to develop 2D and 3D numerical models of the building. The latter includes also the soil, the lining and interface models to simulate the contact between the building and the ground. The predicted crack patterns and opening widths in walls were verified by comparison to real damage reports. The case study also allowed a back analysis of the classical analytical prediction techniques based in the equivalent beam concept from Burland and Wroth. Analytical predictions of building damage are typically done for building walls aligned transversally or longitudinally with respect to the tunnel axis. These buildings are statistically representative, since many urban tunnels follow the tracks of avenues or streets. However, there is a significant number of buildings randomly aligned with respect to tunnel axes, in particular when using a Tunnel Boring Machine. For these buildings, the application of the classical analytical methodology can be done only with approximations, which can lead to unrealistic damage assessments. For this reason, a novel equation for the determination of ground strain has been developed. This equation allows the application of the classical settlement Gaussian profiles and the equivalent beam method in 3D, i.e. for buildings located in whichever position with respect to the tunnel axis. In addition, the model allows considering the position of the tunnel heading, which increases the realism of the settlement trough generated by tunnel construction. Another detected issue during the present research was the high sensitivity of both analytical and numerical damage predictions to certain parameters related to the characterization of ground. In the case of analytical predictions, the modeling of settlement troughs by Gaussian curves offers numerous mathematical advantages. However, the simplicity of this approximation leads to substantially different estimations of damage for small variations of the governing parameters. For this reason, the use of reliability-based methods can be useful for the assessment of building damage. In this way, the present thesis shows the development of a probabilistic model for the prediction of tunneling-induced damage. A procedure to determine the maximum allowable settlements that are used as monitoring threshold values of the construction process is…