Study On Rock Mass Failure Mechanism And Reliability Analysis Of Tunnel Engineering Based On Random Field Theory

For the fine construction and high-quality development of tunnel engineering construction,reliability methods should be adopted in the design and construction of tunnel engineering that can consider the uncertainty of key design factors as much as possible.At present,the natural spatial variability of rock mass materials is not fully considered,so that the failure risk of tunnel engineering cannot be truly and accurately evaluated.Most of the spatial variability studies are limited to the influence of the spatial randomness of rock mass properties on the engineering failure probability,and there is a lack of systematic research on the mechanical properties of spatially variable rock mass,and the failure mechanism of spatially variable rock mass is not clear.Therefore,this thesis aims at the abovementioned two key scientific issues,the failure mechanism of spatial variability rock mass and tunnel engineering reliability are investigated based on theoretical analysis,numerical simulations and engineering cases.The main research contents and results of this thesis are as follows:(1)An efficient stochastic finite difference calculation program is developed through random field theory and finite difference software FLAC 3D built-in FISH language,which avoids the generation of huge calculation code in the traditional stochastic finite difference calculation process and saves the storage space effectively.The scales of fluctuation of rock mass parameters were summarized from three aspects of rock physical and mechanical parameters,rock mass discontinuity parameters and rock mass evaluation indicators through literature survey and rock mass structure classification.The scales of fluctuation statistics of rock mass provided reasonable values for the uncertainty characterization of rock mass parameters.(2)The 3D random field theory was used to describe the horizontal and vertical spatial variability of rock masses based on uniaxial compression test results of small-scale rock in laboratory and Hoek-Brown theory,and the mechanical properties and failure mechanism of the large-scale spatially variable rock masses under uniaxial compressive stress was simulated by the stochastic finite difference method.The expansion of the plastic zone of the spatially variable rock mass has obvious directionality,and the plastic zone expands towards the direction with strong spatial correlation of rock mass parameters.As the scale of fluctuation increases,the mean values and standard deviations of the uniaxial compressive strength and elastic modulus of rock pillar gradually increase.(3)A random phase field method that can simulate the expansion process of hydraulically driven fractures in an anisotropic spatially variable rock mass was established combined with the random field theory and the phase field method.The random phase field method established in this paper was adopted to systematically study the irregular expansion of hydraulic fractures in spatially variable rock masses and the influence of the spatial variability of rock mass parameters on the fracture morphology,fracture center liquid pressure,fracture length and damage area.The mechanism of irregular expansion of cracks in spatially variable rock masses was discussed.Fracture propagation in spatially variable rock mass mainly extends along the weak region of rock mass.The influence of horizontal scale of fluctuation on the propagation of single horizontal fracture is greater than that of vertical scale of fluctuation.(4)A stochastic limit equilibrium method for failure probability analysis of tunnel roof wedge considering the spatial variability of rock mass discontinuities was established.The effect of the spatial randomness of the rock mass structural plane properties on the tunnel roof wedge failure probability was studied through the established stochastic limit equilibrium method,and a systematic parameter analysis was carried out to study the influence of the uncertainty of the tunnel geometry and geological parameters on the failure probability of the tunnel roof wedge.The failure probability of tunnel roof wedge considering the spatial variability of rock mass discontinuities is less than that of rock mass treated as homogeneous material in traditional reliability analysis.(5)A conditional random field generation method for large-scale engineering sites was established through Kriging theory and covariance matrix decomposition method(step-wise decomposition method).This method was successfully applied to the description of the 3D large-scale spatially variable rock mass parameters of the Lijiaping subway station tunnel in Chongqing.The failure probability of the Lijiaping subway tunnel during the excavation process was calculated combined with numerical results and reliability theory to investigate the effect of the random fields of rock mass properties on the probability of the tunnel failure under the condition of known geological survey information.The tunnel failure probability calculated by conditional random field based on site survey information is less than that calculated by traditional unconditional random field without considering survey information.