The nonlinear behaviors of rock–lining and the lining–lining interfaces significantly influence the seismic response of tunnels with composite linings. Therefore, a closed-form analytical solution was developed for the dynamic response of tunnels with composite linings subjected to SH (horizontally polarized shear)-waves, and the two interfaces were modeled by the Coulomb-slip model. A case study was then conducted to investigate the effects of the two interfaces on the seismic response of tunnels. The results show that the imperfections of the two interfaces lead to decreases in the stresses of the composite linings and an increase in the deformation of the final lining. However, the two interfaces play opposite roles in influencing the deformation of the initial lining. As the shear stiffness of the shotcrete–cast concrete interface increases, the peak stresses in the composite linings increase rapidly at first and then tend to be stable. The effect of the shear strength of the interface, however, depends on the stress distribution in the interface. The mechanical parameters of the interfaces should be carefully determined for the seismic design of tunnels with composite linings.

Real‐time monitoring of three‐dimensional stress in the field is an effective approach to detect evolving stress in roof rock and to evaluate rock bursts risk. However, the sensors or data transmission cables may be damaged due to the volatile environment found in coal mines, which can lead to the loss of relevant monitoring data, and some critical information for rock burst prediction may be missed entirely. A number of methods that use historical data to predict missing data or future structural states have been proposed. However, the performance of these methods is poor when the training data are insufficient owing to lack of data. To address this issue, a methodology framework is proposed to predict the mining‐induced stress when some monitoring data are missing. The framework uses a long short‐term memory neural network integrated with the transfer learning method. The proposed method can transfer the knowledge learned from complete monitored data of adjacent sensor to target sensor to boost forecasting. A case study has been conducted to evaluate the method. The results show that the developed model can significantly improve the prediction performance for the target domain, which can be improved further by increasing the size of the target domain training data available.

Fiber-reinforced concrete (FRC) has a great advantage in earthquake-resistant structures, as compared with regular concrete. However, there are many difficulties in the construction and maintenance of concrete structures due to the high density and easy corrosion of the steel fiber in commonly used steel FRC. With the development of polymer material science, polyvinyl alcohol (PVA) fiber has been rapidly promoted for use in FRC because of its low density, high strength, and large elongation at break value. Dynamic uniaxial compression and splitting tensile experiments of FRC with PVA fiber were carried out with two matrix strengths (i.e., C30 and C40), which were blended with PVA fibers with a length of 12 mm in different volume contents (0, 0.2, 0.4, and 0.6%), at the age of 28 days, under different strain rates (i.e., 10⁻⁵, 10⁻⁴, 10⁻³, and 10⁻² s⁻¹). The results show that PVA has an obvious enhancing and toughening effect on concrete, which can improve its brittle properties and residual strength. With increasing strain rate, the compressive strength, split tensile strength, and elastic modulus increase to a certain extent, while the toughness index and the peak strain decrease to a certain degree. The post-peak deformation characteristic changes from a brittle failure of sudden caving to a ductile failure with dense cracking. The effect of PVA is different when enhancing the concrete with two different matrix strengths. The lower the matrix strength, the more obvious the enhancement effect of the fiber, showing characteristics of a higher compressive strength and low split tensile strength in FRC with low strength and a smoother post-peak stress–strain curve.