This paper proposes a performance-based damage assessment procedure for reinforced concrete (RC) box tunnels subjected to earthquakes, employing a pseudostatic approach and a ductility-based damage index that incorporates the relative stiffness between the structure and surround soil, widely denoted as flexibility ratio (F). Distributed plasticity frame elements and discretized spring elements were used to model tunnel structures (slabs, walls, and columns) and the reactions of surrounding soil, respectively. Two damage-state descriptors were investigated: one based on the number of yielding in the tunnel members and another on the material state. Results show that the number-of-yielding based descriptor captures global structural capacity only for specific F ranges, while drift ratio lacks consistency as a damage index across all F ranges. In contrast, the material-state descriptor and damage indexes based on curvature ductility provide effective capacity estimation and are independent of F. Therefore, combining both descriptors is recommended for seismic performance evaluation of RC box tunnels. Additionally, higher F leads to brittle failure due to better load distribution and increased yielding before the strength degradation, while lower F results in concentrated damage with less yielding. These findings highlight the necessity of seismic design considering flexibility ratio for earthquake-resistant tunnels.

This study investigates the seismic response of a reinforced concrete (RC) tunnel using two-dimensional plane strain finite element models calibrated and validated against experimental results. A comprehensive parametric study is then conducted to explore the influence of tunnel-soil flexibility ratio, soil relative density, Arias intensity of the input motion, and ground motion components on the seismic soil-structure interaction (SSI). The results demonstrated that the flexibility ratio and racking coefficient increase with overburden height, while soil deformations decrease. Acceleration amplification factors rise from the bottom soil to the ground surface, with dense soil showing higher amplification especially in the regions at and near the tunnel field. The horizontal amplification factor exhibits greater variability with increasing seismic energy intensity, and the effect of the vertical motion becomes more pronounced near the structure. The vertical amplification factor is lowest for the horizontal component, while the vertical and combined components exhibit higher values influenced by the presence of the tunnel with lower earthquake intensity. Soil relative density significantly influences the vertical and lateral pressures on the tunnel, with dense sand causing maximum vertical pressures on the top slab and walls. The vertical earthquake component has a greater impact on the tunnel's top slab pressure distribution than the horizontal component. Seismic bending moments are influenced by earthquake components, with the vertical component leading to the greatest positive bending moment values in the middle of the roof slab. Vertical soil deformation is significantly affected by the horizontal input motion component, whereas the vertical component minimally affects lateral soil deformation. These findings underscore the importance of capturing stress-strain response under cyclic loading, particularly near the tunnel crown, where complex stress interactions lead to increased variability in behavior.