Title Study of the sensitivity of existing urban buildings to settlements produced by tunnelling
Acronym SUBTIS
Reference BIA2009-13233
Contractor Spanish Ministry of Science and Innovation
Initial date January 2010
Final date April 2013
Fundings 204.000,74€
Professor in charge Climent Molins Borrell

The project contributed to the knowledge in the field of prediction of the damage to buildings produced by tunneling induced settlements. State of the art techniques (based on the equivalent beam method) have been appraised by applying them to a real building of masonry affected by the construction of the tunnel of the L9 Barcelona metro. This method has been extended to the case of buildings which are not aligned to the tunnel axis. In the field of numerical modeling, 2D models of façades of buildings affected by induced tunneling works were developed producing successful prediction of crack patterns caused by subsidence. In addition, complex 3D numerical models of buildings including soil and tunnel were used to analyze the effect of TBM advance and the subsequent evolution of damage. Results were also compared with simplified models.

At a most basic level, substantial developments in the characterization of mechanical properties of existing masonry fabrics and the determination of its strength from numerical micromodels were also performed. These micromodels allowed to analyze the behavior of brick masonry infill RC frames subjected to differential settlements. The numerical results were compared with the experimental results of other authors, resulting in a correct prediction of the crack initiation and the location of the failure mechanisms of the structure. Finally, a statistical tool for predicting damage to buildings based on the equivalent beam method was also produced. This tool allows to consider the uncertainty in the governing parameters (terrain, soil loss volume or the uncertainty of the models used) describing the damage of a wall of a building subjected to differential settlement produced by tunneling.