Study on the Seismic Response of the Tunnel in Silty Clay Based on Shaking Table Test

ABSTRACT The seismic response of shield tunnels in Yellow River silty clay was examined through 1:20 shaking table tests and corroborating three-dimensional finite element (FE) simulations. Remolded site soil was compacted in a 3-m-diameter flexible container, and two representative motions—a low-frequency, long-duration Wenchuan record and a high-frequency Jinan synthetic wave—were imposed at peak ground accelerations (PGAs) of 0.143–0.918 g. As PGA increased, the soil exhibited pronounced nonlinearity: the acceleration-amplification factor and natural frequency declined, whereas the damping ratio rose. A frequency-dependent tunnel-soil interaction emerged: below a PGA-controlled critical frequency, the tunnel and ground moved coherently, but above it, they decoupled; crown decoupling initiated when PGA exceeded 0.3 g. The greatest circumferential tension consistently developed at 45° to the tunnel axis. Under the Wenchuan wave at 0.918 g, the diameter-change rate reached 11.93 ‰, nearly twice the Chinese code limit, whereas the Jinan wave at 0.6 g induced marked torsion with lesser ovalization. Lateral displacement amplified nonlinearly with depth, and the sand-clay interface (∼90 cm) behaved as a slip plane when PGA surpassed 0.4 g. FE results reproduced the measured responses at low PGA but overpredicted stiffness at high PGA because interface slip was not modeled. The findings indicate that seismic design in such deposits should reinforce the 45° sectors, employ ≥C60 concrete plus flexural stiffeners in low-frequency regions, strengthen antitorsion joints in high-frequency regions, install ≥10-cm rubber-particle isolation layers at sand-clay boundaries, and pressure-grout the upper 1-m cover to mitigate shallow deformation.