Soil Conditioning for EPB Shield Tunnelling in Sand-Cobble Strata: An Integrated Laboratory-DEM Approach

In large-particle sandy cobble strata, laboratory slump and impermeability tests can evaluate the workability of conditioned soil but do not adequately resolve its mechanical behaviour or the surface-settlement response during shield tunnelling. Thus, this study proposes a method for determining a soil-conditioning scheme by combining laboratory tests with full-scale discrete element simulations. For a Beijing metro shield project, the injection parameters were determined through conditioning-agent performance tests and slump tests. A full-scale discrete element model was established and calibrated using triaxial compression and slump tests. Cutter forces, cutterhead loads and surface settlement were used as indices to determine the preferred conditioning scheme. The results show that a 5% foaming-agent solution and an 8% bentonite slurry exhibited favourable preparation performance. When the bentonite injection ratio (BIR) was 12% and the foam injection ratio (FIR) was 30% or 40%, the conditioned soil exhibited favourable flowability and workability. At a BIR of 12%, increasing the FIR from 30% to 40% reduced the mean cutting force of the ripper cutters, mean total thrust and mean cutterhead torque by 19.10%, 22.29% and 8.09%, while the maximum settlement decreased by 25.3%. Field and simulated settlement trends agreed, with a final relative difference of 15%.