Ricerc@Sapienza

The expansion of cities often involves tunnel construction, which may result in detrimental effects on existing structures. This paper considers the interaction between tunnelling-induced ground deformations and framed buildings on raft foundations. The aim of the paper is to study the differences between scenarios where an equivalent plate and a real frame are used in the tunnel-building interaction problem. The paper includes experimental data obtained from geotechnical centrifuge tests and predictions of building response based on numerical analyses. The centrifuge experiments involved framed building models with walls extending in the direction of the tunnel axis. Digital image analysis was used to measure soil and building displacements during simulated tunnel volume loss. Results are compared to existing data from centrifuge tests that used an equivalent plate approach. Numerical simulations carried out with simplified and advanced Finite Element methods, performed respectively before (i.e class A predictions) and after (i.e. class C predictions) the centrifuge tests, are presented. Both methods effectively predict the soil and building deformation profiles determined experimentally. Finally, the potential for errors resulting from the equivalent plate approach are also discussed.