A prediction framework for tunnel seismic instability threshold based on loading/unloading response ratio theory

The prediction of a tunnel seismic instability threshold is a key challenge in seismic design and disaster reduction. Current methods have shortcomings in effectively integrating the synergistic effects of surrounding rock and tunnel linings. To this end, a new prediction framework based on the loading/unloading response ratio (LURR) theory is proposed. First, three core components of the LURR model are introduced to control the dynamic response of tunnels. Subsequently, the evolution of LURR under different seismic intensities through elastic-plastic finite element time-range analysis is quantified. This approach enables us to elucidate the correlation between LURR characteristics and structural stability and validate it by analyzing the distribution of plastic zones. The results show that it is feasible to define the loading/unloading state using the absolute change of seismic acceleration; the difference in extreme values of LURRs at critical points surrounding the tunnel is negatively correlated with structural stability; the extreme values of LURR follow a unimodal distribution relative to the peak ground acceleration, where the peak represents the ultimate bearing capacity threshold, thus establishing a clear failure threshold.