To study the dynamic response law of large-section cracked lining structures under seismic waves, comparative tests of large-scale shaker tunnel models of non-destructive lining structure (model 1), a crack in the vault of the lining structure (model 2), and two parallel cracks in the vault of the lining structure (model 3) were carried out by applying 0.1-1.0 g progressively increasing the peak acceleration of the input waves. This paper visually showed the distribution of cracks in three groups of the lining structures. In addition, the acceleration response of the lining and surrounding rock, dynamic soil pressure, the dynamic strain on the inner and outer surfaces of the lining, and dynamic internal force variation were obtained, and the seismic performance of three groups of lining structures was discussed. The results showed that the seismic weak positions of model 1 were the arch shoulder and the arch foot, the seismic weak positions of model 2 were the arch shoulder, the arch foot, the initial damage area, and the inverted arch, and the seismic weak positions of model 3 were the positions of the arch foot, the cracks of the vault, the inverted arch, and the arch wall. The soil pressure values at the vault of three groups of models were model 2 > model 1 > model 3 in turn. The surrounding rock amplified the input seismic waves. With the gradual increase of the peak acceleration, the seismic energy was gradually consumed due to plastic damage to the lining structure or the loosening and destruction of the overlying soil, resulting in the acceleration amplification coefficient value of the surrounding rock in the upper part of the lining structure showing a changing trend of first increasing and then decreasing. When the peak acceleration was 0.2 g, the crack propagation phenomenon occurs in the initial crack position of model 2 and model 3. When the peak acceleration was 0.4 g, the cracking phenomenon occurs at the right arch foot of model 1. The above phenomenon confirmed the conclusion that cracks can weaken the seismic performance of the structure. When the peak acceleration was 0.8 g, the peak values of the amplification coefficient of the lining at the inverted arch and near the filled soil surface were about 1.2 and 1.6 respectively. The research results can provide a reference for the seismic performance evaluation of cracked tunnels.

To explore the mechanical properties of cracked lining subjected to seismic loads based on wave theory and the extended finite element method, the dynamic viscoelastic boundary of a seismic wave and equivalent nodal force load were generated by MATLAB programming software to establish a simulation model of cracked lining structure. The internal force state change law of the lining structure was studied by varying the crack depth, crack length, secondary lining thickness, and other parameter values (including the layout between multiple cracks). Also, the safety factor of the lining crack section was obtained, and the functional relationship between the safety factor and parameter variables was established. Results show that the crack depth and secondary lining thickness were the main factors affecting the internal force of the crack section. Based on the least square method, the calculation formula and 95 % confidence interval between the minimum safety factor (Kmin) and each parameter of the crack section were obtained. Meanwhile, the Kmin prediction model was obtained via multiple nonlinear regression. When the crack depth value was 30 % of the lining thickness value, Kmin reduced 2.5 times. At 57 % crack depth of the lining thickness, the Kmin was less than the specification value, indicating that the lining structure’s safety reserve was low. Compared with the arrangement of the vertical distribution of the two cracks, the stress concentration generated when the two cracks were arranged in parallel would more likely affect the structure adversely. The findings can provide a reference for the safety study of cracked tunnels.

Geological faults impair tunnel stability during earthquakes. This study establishes a tunnel dynamic stability evaluation index based on load/unload response ratio (LURR) theory. It considers a seismic wave as a load/unload parameter and tunnel structure strain response as a response parameter. The rationale behind this evaluation index and the factors affecting tunnel stability across fault zones under seismic conditions are investigated. Compared to the traditional dynamic instability criterion, the LURR accurately measures the degree of structural deviation from the steady state and better determines the potential destabilization region of the structure. As the peak value of the input seismic wave increases, the LURRs of the more unstable parts increase, while the LURRs of the stable parts remain unchanged. According to LURR theory, the size of the range affected by the fault on the tunnel during an earthquake depends mainly on inherent fault properties (i.e., the dip angle, strike, and thickness), independent of the earthquake intensity. Because the LURR can theoretically be infinite, its dynamic instability threshold cannot be determined accurately.

No abstract