Settlements Due to Single and Twin Tube Urban EPB Shield Tunnelling

Shield type Tunnel Boring Machines have been used for several decades; however little information exists about the actual mechanism of shield-ground interaction. The ground response mechanism induced by shields is difficult to explain, since it engages both reliable ground deformation measurements in the field, along with relative operation parameters records of the mechanized shield. Existing empirical and analytical relations that are used for the calculation of ground settlement do not take into account the operation parameters of the shield performing the excavation. Even numerical analyses consider generally only ground characteristics and no more than simple geometrical data are required for the simulation of the shield. It is the objective here to include the operation parameters records in the interpretation of surface settlements due to Earth Pressure Balance Shield (EPBS) tunnelling. Use is made of data on ground movement deformation and EPBS operations that have been collected for Metro Projects of Athens and Thessaloniki. Such data are the ground settlements, ground qualities and EPBS operation parameters measured at the single and twin tunnels of these Metro projects. The transverse settlement troughs are drawn and their shape discussed. The ground conditions are similar over long tunnel stretches, which allows one to establish relationships between ground characteristics and EPBS operations to surface deformations. A comprehensive tunnelling data acquisition system was developed to collect information from in situ data on a ring-by-ring basis, such as geological conditions, shield operation parameters, tunnel geometry. Having utilized the aforementioned information and correlated the operation parameters to the maximum influence on ground movements, an artificial intelligence system, based on Artificial Neural Networks (ANN), is employed to develop predictive relations. In order to minimize the risk of ground movement, the tunnelling engineer and the EPBS pilot need to be able to make reliable prediction of settlements based on the operational data. Such an ANN algorithm has the ability to map input to output patterns i.e. all influencing parameters are related to surface settlements. It serves as a means to understand and predict ground movement behaviour induced by EPBS shield tunnelling, and deals as an indicator for construction control and for the implementation of remedial measures where needed.