| Under the combined action of layered load and geological movement,there are many discontinuities such as faults,joints and fissures in the deep strata of coal mines.Under the action of complex environmental stress,the expansion and penetration of joints are easy to lead to the deterioration of the mechanical properties of the surrounding rock of the roadway,instability and damage,resulting in serious coal mine safety accidents such as rib spalling and roof falling.Therefore,the mechanical behavior and failure mechanism of deep jointed rock mass in coal mines have become one of the key scientific issues in the construction of deep strata roadways and the control of surrounding rock stability.In this paper,triaxial compression laboratory tests and numerical tests of standard cylindrical sandstone specimens with non through single joint under different confining pressures and joint dip angles are carried out by combining laboratory tests and numerical simulation.The research reveals the mechanical response law and deformation failure mechanism of jointed rock mass under triaxial compression.Based on BP neural network algorithm,a prediction model of triaxial compression strength of jointed rock masses is established,and the influence of joint parameters,stress paths and other factors on mechanical properties of jointed rock masses is further explored.The main research contents and conclusions are as follows:(1)Using sandstone to prepare complete cylindrical standard specimens and specimens with non through single joints,complete sandstone mechanical property tests were carried out,and the basic physical and mechanical parameters of sandstone were obtained;Triaxial compression tests of jointed rock masses with different dip angles were carried out.The effects of joint dip angles on deformation characteristics and critical stress deterioration of jointed rock masses were analyzed,and the crack failure characteristics of jointed rock masses with different dip angles were discussed.(2)The numerical model of discrete element method for jointed rock mass under triaxial compression is established,the meso parameters of rock are calibrated based on the basic physical and mechanical parameters of sandstone,the triaxial compression numerical tests are carried out,the mechanical characteristics,the crack fracture evolution process and the failure mode of jointed rock mass under different confining pressures are analyzed and studied,and the crack propagation evolution law of jointed rock mass and the failure mechanism of jointed rock mass under the condition of high geostress are revealed.It is found that the existence of joints degrades the peak stress and crack initiation stress of rock mass in varying degrees,and the confining pressure and joint inclination affect the final failure mode of rock mass.The applicability of numerical simulation method is verified by comparing the failure characteristics of jointed rock mass under the same conditions of laboratory tests and numerical simulation.(3)Based on the numerical test results of triaxial compression of jointed rock mass and BP neural network algorithm,randomly select training samples and test sample data,and construct a prediction model of mechanical properties of jointed rock mass after iterative training.According to the evaluation of test sample data,the error of predicted compressive strength is less than 5%,and the multiple judgment coefficient of all data is 0.98735.The model has strong generalization ability and can be used to predict the triaxial compressive strength of jointed rock mass under different conditions.Through the above research work,the numerical model of discrete element method for jointed rock mass and the BP neural network prediction model for mechanical properties of jointed rock mass are established,and the influence of confining pressure and joint inclination on crack fracture evolution and mechanical properties of jointed rock mass is revealed.It can provide theoretical basis and technical support for the stability control of surrounding rock in deep jointed strata of coal mines in the future.Figure 36 Table 14 Reference 80... |