To fill the gap that the impact of segmental joints and water-saturated soil is commonly considered separately in the existing research, this paper proposes a fluid-solid coupling FEM to examine the seismic behaviour of an underwater segmented tunnel, considering the weakening effect of the joints and the dynamic characteristics of the two-phase media simultaneously, with different input ground motions, buried depth, and soil stiffness. The presence of segmental joints leads to an increase of pore water pressure in the surrounding saturated soil of the tunnel, implying a greater probability of liquefaction. It also leads to an increase in the tunnel deformation due to a reduction of stiffness. Furthermore, although it beneficially decreases the overall lining forces, it is found to significantly increase the compressive stress at the lining edges near the segmental joints, inducing stress concentration accordingly. Finally, the compressive stress of the tunnel lining and pore water pressure of saturated soil have the most and the least sensitivity to the weakening effect of joints, respectively.

To investigate the effect of fluid -solid coupling on the seismic performance of underground structures in watersaturated soil, a comparison study is conducted in this paper on three-dimensional (3D) nonlinear seismic behavior of a 3 -story 3 -bay subway station obtained using two different finite element methods (FEM), i.e., the generally used simplified method with equivalent single-phase soil model and a newly developed 3D numerical approach capable of considering the dynamic behavior of saturated two-phase media. A 3D user -defined element embedded in ABAQUS is first introduced to simulate saturated soil's dynamic fluid -solid coupling effect. Then, more essential demonstrations are presented for establishing and validating the two FEM. Based on the two methods with and without incorporating fluid -solid interaction, 3D nonlinear seismic response analysis is performed on the subway station considering three different input seismic waves. Discussions are conducted in terms of accelerations, lateral displacements, inter -story drift ratios, rotation of columns, damage characteristics, and internal forces, based on which the limitations of the simplified method are quantitatively interpreted. The results show that neglecting the fluid -solid coupling effect can bring about conservative evaluations of the seismic behavior of underground structures in saturated soil. The effect of fluid -solid coupling on the seismic performance of underground structures is quite sensitive to the peak ground acceleration. It is significant to consider the fluid -solid coupling effect during the performance -based seismic design of underground structures enclosed in saturated soil to gain realistic seismic responses, especially for those subjected to major earthquakes.