The parameters of underground structure have the characteristics of large spatial heterogeneity and strong uncertainty. Recently, people gradually begin to use the idea of probability to analyze the safety of underground structures. However, few articles analyze the dynamic response of structural reliability. In this paper, the reliability of tunnel roof under blasting disturbance is studied, and the influence of the number of disturbances and the distribution of random variables on the structural reliability is analyzed. The prediction model of structural reliability is proposed. The sensitivity of different parameters to the structural reliability is discussed. The results show that the traditional method of evaluating the structural safety status is more dependent on the design value of each parameter. However, the design value does not represent the true state of the structure in a completely accurate way, which will result in a certain deviation between the assessment of the structural safety state and the actual state. Single disturbance has a limited impact on the structural reliability, while repeated disturbances can still have a noticeable effect on the structural reliability. When the disturbance reaches a certain number of times, the reliability of the structure can drop sharply. This paper proposes the prediction function of structure reliability according to the decay of structure reliability after blasting disturbance. Only a small number of parameters are needed to predict the structure reliability at different times. The error between the forecast results and the actual results is about 1%. With the increase of the number of blasting disturbances, the sensitivity coefficient of the structural resistance factor gradually decreases, and the sensitivity coefficient of the loaded effect gradually increases. The control of the blasting strength should be gradually increased to ensure the structural safety. Even if the mean and mean square error of random variables are the same, different distribution forms will have a great impact on the calculation results.

Conventional seepage–heat transfer models in simulating the heat transfer between fluid and rock in fractures mainly involve one-way coupling and do not consider the influence of temperature on the seepage. Moreover, it is an enormous challenge to define parameters of the explicit heat transfer between the rock and fluid. In order to resolve these shortcomings, a two-way fully coupled model of the seepage–heat transfer in the fractured rock was established in the present study. Based on the original geometric structure of the experimental device, combined with the actual engineering scale, the local dynamic heat transfer coefficient of the fractured rock was established, which is related to the fracture aperture, flow velocity and thermal parameters. Then, the proposed model was verified through the experiment and excellent agreement was achieved in this regard. It was found that the dynamic heat transfer coefficient changes the temperature distribution of fluid and rock in the original static heat transfer coefficient fracture system. The proposed model simplifies the parameters required for the calculation of the heat transfer coefficient. These parameters are related to characteristic variables, such as velocity and rock temperature, and can be simply obtained from standard laboratory tests.

The hydraulic and mechanical coupling characteristics of fracture networks under normal stress and shear stress were studied in this paper. The hydraulic and mechanical coupling model of the fracture network comprehensively considers the normal stress, shear stress, seepage pressure and roughness characteristics. Based on the boundary conditions and reasonable assumptions, COMSOL Multiphysics software was used to develop the hydraulic and mechanical coupling finite element model of the fracture network with different intersection points under normal stress and shear stress, focusing on the study of the effect of normal stress and shear stress on the fracture permeability. The degree of permeability change caused by the normal stress and shear stress is different. The shear stress has a significant influence on the fracture permeability, and when the normal stress is low, the relationship between the fracture permeability and shear stress can be described by a linear relationship. Then, the influence of the number of intersection points in the fracture network on the average fracture width, average water pressure, average seepage velocity and seepage passage of the fractured rock mass was analyzed. The number of intersections in the fracture network has little influence on the average fracture gap width and average water pressure but has a great influence on the flow velocity. The analysis in this paper is very helpful to understand the seepage characteristics in rough fractures under normal stress and shear stress.