Zhen-Dong Cui

Urban metro tunnels and surrounding soft clay deposits may remain exposed to train-induced cyclic loading after being affected by tunnel fires or other high-temperature events. However, the coupled effects of thermal history, bidirectional cyclic stress path and microstructural evolution on saturated soft clay remain insufficiently clarified. In this study, remolded Shanghai soft clay was heated to 25 °C, 200 °C and 500 °C, and then tested using consolidated-undrained bidirectional cyclic triaxial loading to represent metro-type vibration. The axial dynamic stress amplitude was fixed at 20 kPa, while the radial dynamic stress amplitude was varied to obtain radial-to-axial cyclic stress ratios of 0, 0.25 and 0.5. Loading frequencies of 0.5, 1.0 and 2.5 Hz were applied for 5000 cycles. Macroscopic responses, including cumulative plastic strain, excess pore water pressure and dynamic elastic modulus, were linked to SEM-based pore morphology, equivalent pore diameter, pore shape factor, fractal dimension and directional frequency. The results show that increasing the radial-to-axial cyclic stress ratio and loading frequency generally reduced cumulative plastic strain and excess pore pressure. The 200 °C treatment produced a more loading-sensitive response, whereas the 500 °C treatment reduced strain and pore pressure accumulation and helped maintain stiffness under some bidirectional loading conditions. SEM observations indicate that the 200 °C treatment produced a transitional fabric sensitive to cyclic disturbance, while the 500 °C treatment formed a denser and less loading-sensitive pore structure. These findings provide experimental evidence for evaluating post-fire metro operation and settlement risk in heat-affected soft clay.

The interaction between tunnelling and retaining walls represents a complex soil–structure interaction problem that can significantly influence wall deformation and stability. This study investigates and compares the behaviour of cantilever and soil-nailed retaining walls subjected to tunnel excavation beneath the wall in dry Toyoura sand. A series of three-dimensional finite element analyses was performed using an advanced hypoplastic constitutive model to simulate nonlinear sand behaviour and stress-path dependency. Tunnel depth was varied using cover-to-diameter (C/D) ratios of 1.83, 3.33, 4.83, and 6.33. The computed results show that both retaining systems exhibit similar settlement patterns due to tunnelling, with maximum settlement reaching approximately 22.8 mm in the case of C/D = 4.83. However, soil nailing has limited influence on reducing tunnelling-induced settlement. In contrast, the soil-nailed wall develops larger rotation and overturning response due to tensile forces mobilized in the soil nails by tunnelling-induced ground movement. For the shallowest tunnel case (C/D = 1.83), maximum tensile forces reach approximately 72 kN and 60 kN in the top and middle nails, respectively. Tunnelling also causes significant redistribution of contact pressure and shear stress beneath the wall base, including partial loss of contact for shallow tunnels. In addition, lateral earth pressure increases substantially, resulting in total lateral forces up to 3.3 times the pre-tunnelling values. The results demonstrate that tunnel depth governs the wall response, while soil nailing primarily affects rotational behaviour and internal force development rather than settlement mitigation.

Land subsidence is the geological hazard caused by natural or human factors, resulting in a regional decrease in ground elevation due to the compression of the surface soil of the earth’s crust, which has brought huge losses to the national economy. The physical model tests were conducted to study the land subsidence induced by triple pumping including Pumping I, Pumping II and Pumping III. A total of 41 LVDTs were installed to monitor the settlement of the ground, and pictures of the front of the model were taken to obtain the section settlement via the particle image velocimetry (PIV) software. On Path 1, the subsidence is −1.40 mm, −1.50 mm, −1.86 mm, and −2.36 mm after Pumping I; it is −3.15 mm, −3.56 mm, −3.45 mm, and −4.57 mm after Pumping II; and it is −1.29 mm, −0.68 mm, −0.86 mm and −1.65 mm after Pumping III. The closer the soil is to the pumping well, the more severe the settlement after pumping. In the confined aquifer, the pore pressure in the soil experiences a process of initial decrease followed by an increase, which is the manifestation of the drawdown cone on pore pressure. In the layered settlement of the soil, compression in the upper and lower clay layers is significant, accounting for the majority of surface subsidence, while the confined aquifer experiences almost no compression. The results can offer a reference for the prevention and control of land subsidence in soft soil areas.