Reliability Analysis Of Jointed Rock Slopes Based On Bayesian Updating

With the expansion of Chinese water Conservancy and hydropower,open-pit mines,high-speed railway and highways,more and more slope projects need to be constructed,which leads to the problem of slope stability.However,the natural slope,especially the rock slope,will produce many joints and cracks in the process of rock formation and under the geological action,the existence of these joints and cracks has a great influence on the stability of rock slope.For this reason,the calculation accuracy and efficiency of the existing slope stability analysis methods are low,a systematic and in-depth study has been carried out on the reliability analysis problem of jointed rock slope under the condition of limited measured data,which includes the simulation of random fracture network and the consideration of the uncertainty of structural plane geometric and mechanical parameters,the major research work is as follows:(1)The basic theory and method of Bayesian updating method are introduced,the detailed procedures for determining the probability distribution types of structural plane parameters by chi-square test and K-S test are given,three discrete fracture network models are summarized,and the advantages and disadvantages of these three models are discussed and compared systematically,it can provide an important theoretical basis for stability analysis and reliability updating of rock slope.(2)Non-intrusive stochastic finite element method is used to analyze the reliability of jointed rock slope,and the interface program between Phase~2 software and MATLAB software are compiled,the automation of the parameter assignment of the fractured material and the calculation of the slope stability are realized,and the problems of the large amount of calculation and low efficiency in the reliability analysis of jointed rock slope calculated by MCS method at present are solved,taking an actual reservoir bank rock slope as an example,the failure probability of the rock slope under different working conditions(natural working condition,rainfall working condition and earthquake working condition)are calculated.(3)A reliability analysis method of complex jointed rock slope is developed,which can consider the variation of joint geometric and mechanical parameters simultaneously.The method combines half-factor experimental design and central composite design,through variance analysis,the structural plane parameters that have a significant influence on slope stability are identified,and the response surface functions of the factor of safety and the structural plane parameters are constructed,it provides an efficient way for reliability analysis and calculation of complex jointed rock slope.(4)The Bayesian updating method is proposed to optimize the probability distribution of structural plane parameters,and the validity of the Bayesian updating method is verified by comparing the a posterior probability distribution obtained by the analytical solution and the BUS method,in the case of a Xiaowan Dam slope,the effects of uncertainties of structural geometry and shear strength parameters(cohesion,internal friction angle)are considered,it provides a reliable way to get the network of rock mass fracture which is more realistical to the engineering practice.