New limit states for the seismic fragility assessment of circular tunnels: Application in case of tunnels in clayey soil deposits.

• A novel framework to define limit states for the seismic fragility assessment of circular tunnels is introduced. • Nonlinear static pushover analyses are employed to examine the seismic capacity of tunnels. • Nonlinear dynamic analyses are employed to examine seismic demand of tunnels. • The proposed framework is applied to 27 ground-tunnel configurations. • PGA -based fragility functions are derived, which are compared with the existing fragility curves. This paper introduces a novel framework to define limit states for the seismic fragility assessment of circular tunnels in soil. A numerical framework is developed for this purpose, focusing on the response of tunnels subjected to ground seismic shaking in the transverse direction. New limit states are defined based on the ovaling deformation of the tunnel, corresponding to different levels of liner stiffness degradation caused by seismic shaking. The latter is evaluated via nonlinear static pushover analyses of the examined ground-tunnel configurations. Nonlinear dynamic analyses are performed to evaluate the demand of examined tunnels and develop Probabilistic Seismic Demand Models (PSDMs). The uncertainties related with the definitions of capacity and demand are thoroughly evaluated based on the results of the nonlinear static pushover and dynamic analyses, respectively. The proposed framework is applied to a 6 m diameter circular tunnel embedded in uniform clayey soil deposit at a burial depth of 15 m. Various assumptions are made regarding the thickness and mechanical properties of the liner and the soil, leading to the investigation of 27 ground-tunnel configurations. A suite of ground motions is selected to perform dynamic analyses of each examined configuration. Based on the results of the analyses new PSDMs and PGA -based fragility functions are derived. Comparisons of the proposed fragility curves with existing, empirical, and analytical fragility curves for tunnels, reveal differences, which in some cases are significant and are mainly attributed to the different definitions of Engineering Demand Parameters (EDPs) and limit states between the compared curves, as well as to different assumptions in the analytical frameworks proposed by various studies. The proposed framework may be applied to other ground-tunnel configurations to develop fragility functions for a more rigorous risk and resilience assessment of these types of systems. [ABSTRACT FROM AUTHOR]