The deformation of a deep foundation pit in Ulanqab city was analyzed by in situ monitoring and numerical simulation data. First, a limit equilibrium method was used to evaluate the stability of foundation pit under non-supporting and ribbed plate anchored retaining wall support. The results show that the ribbed plate anchored retaining wall support could control the soil deformation effectively in deep foundation pit. Additionally, a three-dimensional finite difference model of excavation zone and reinforcement structure was established using FLAC3D software to simulate the excavation of deep foundation pit. The simulation results were verified by monitoring field data. The simulation results show that all the bolts are in the tension state, and the maximum axial force is located at the end of the bolt, with a value of 46.4 kN. Foundation pit excavation is a continuous stress release process, making the rock and soil of side wall of foundation pit slide towards the free surface. The largest deformation is 4.0 mm, near the side wall of foundation pit, consistent with the deformation monitoring results.

Water level is decreased during foundation pit excavation to avoid water inrush under confined water pressure. Cut-off wall is often used as waterproof curtain to partially cut off the dewatered aquifer. When a foundation pit is located in a built-up area and the underlying confined aquifer is not cut off, the drawdown must be minimized outside the pit to avoid land subsidence in buildings and pipelines. The coupling effect of the cut-off wall and pumping well is used to control the drawdown outside the foundation pit. However, the coupling mechanism is not intuitively well understood because of the limitations of existing experimental methods. In this study, transparent soil was introduced to model the coupling mechanism in the physical model test. High-purity fused silica and mixed paraffin oil were used as skeleton and fluid to simulate the confined aquifer and groundwater. Industrial solid dye and paraffin oil were used as tracers. A camera was used to collect flow information. Tests were performed for the combinations of cut-off wall and partially penetrating pumping wells. The insertion depth ratio of the cut-off wall most effectively influenced the drawdown. The layout of the pumping wells in horizontal direction influenced water level distribution and flow rate. The optimal depth of the pumping wells was 1–5 m above the bottom of the cut-off wall, and the optimal horizontal distance between the cut-off wall and the pumping wells was 25% of the pit width. Non-Darcy flow was observed within the range of 0–10 m around the bottom of the cut-off wall. These results were significant in understanding the cut-off wall and pumping well coupling effect on foundation pit dewatering.