| Surrounding rock is the main research object of tunnel engineering. Surrounding rock deformation and failure mode is the issue that tunnel engineering research and design personnel have been focus on, and it is of great significance for the load of surrounding rock of tunnel design, the selection of supporting parameters and the determination of construction scheme. Sandy tunnel surrounding rock is mainly composed of sandy soil, with loose, broken and low cohesion properties. Collapsing destruction of surrounding rock is sudden and hard to predict, and has uncontrollable spread rapidly when disasters cause great engineering accidents. The current widely used research method based on continuum theory is no longer well applied to large discreteness of sand surrounding rock, so it is necessary to adopt the discontinuous media theory to study.According to this problem, we can select the particle flow code in two dimension who based on discrete element theory and overcome the limit of small deformation and allow bedded slip and fall. Particle flow code in two dimension can simulate damage of particles bond caused by particle separation and the large deformation, to help us research material deformation and mechanical properties from the perspective of mesoscopic. Around the subject of sandy tunnel surrounding rock failure mode to research, main work and achievements are as follows:(1) Based on PFC2D biaxial test to trial sand mesoscopic parameters of PFC. Establishing database of trial samples to find the corresponding mesoscopic parameters to sandy soil macro physical and mechanical parameters. Conducting impact analysis of microscopic parameters on the macroscopic parameters, perceptively recognize the effects of various microscopic parameters on the macroscopic parameters.(2) Grey relational analysis was carried out on the trial sample database, to associate degree sequence between sand macro parameters and microscopic parameters. The sensitivity analysis of sand macro parameters and microscopic parameters is given from a quantitative point of view. For grading the mesoscopic parameters, find out the main mesoscopic, secondary mesoscopic parameters and not sensitive parameters of various macro parameters.(3) The BP artificial neural network is applied to the mesoscopic parameters calibration to build and test a neural network that can predict with mesoscopic parameters of PFC corresponding to the target macro parameters. Based on trial database design and train the BP neural network, the predicted results meet the range of allowable error upon examination. Illustrate that a well designed and trained BP neural network can be applied to the prediction of mesoscopic parameters.(4) Obtained using the calibration microscopic parameters of sand to simulate the excavation of sand surrounding rock in the tunnel to study it’s failure mode. First simulated, using the method of AC/DC set different initial in-situ stress and the lateral pressure coefficient and stress of surrounding rock after excavation monitoring. Then, we also simulate runnel excavation under the weight of strata stress state, compare the depth of buried tunnel collapsing damage process and describe cave arch and pressure arch dynamic development process.(5) Poor self-stability of sandy tunnel surrounding rock collapse immediately after excavation, its deformation and caving is a gradual process of development. It will be a greater range of deformation with the caving in the development of its own deformation and stress redistribution occurs within a short time after the initial stability of surrounding rock. This process will loop again after surrounding rock stability within a short time, and surrounding rock continued to develop a more stable state. In the whole process, the development trend of the pressure arch and cave arch development are coordinated. |
PDF Full Text Request