Failure Mode And Stability Analysis Of Large Karst Cave In Jointed Rock Mass

In recent years,with the construction of high-speed railroads,highways,urban subways and other basic transportation projects,it is inevitable that large caverns will be encountered in the construction of railroad tunnels and other underground engineering,and how to quickly judge the stability of large caverns is very important for the smooth implementation of the relevant underground projects.Usually the mechanism of large cavern formation is very complex and is the result of many complex geological factors.The analysis of the stability of the surrounding rock of large caverns is the basis for the design of cavern disposal plans,and the determination of the stability of large caverns depends on the determination of the failure mode of the surrounding rock of caverns.Based on the project of "Research on the genesis mechanism and engineering risk evaluation method of oversized karst caves in southwest China",this paper mainly analyzes the stability and failure mode of large caves under different combinations of penetrating joints,and qualitatively summarizes the failure mode of surrounding rocks of large karst caves and the characteristics of corresponding influencing factors,and based on this,BP neural network is used to make a rapid judgement of the stability of large caves.The following main research results were achieved in the paper.(1)Through single-factor test,it is concluded that the joint spacing,cavern crosssectional form,cavern burial depth and occurrence of single set of penetrating joints have certain influence on the joint opening zone,joint shear zone,maximum vertical displacement and plastic zone of surrounding rock in large caves.(2)The influence of the combination of joint occurrence on the stability and failure mode of the surrounding rock of large caverns under two sets of penetrating joints is studied,and the influence of joint occurrence on the failure degree of large caverns is quantified by counting the volume of blocks in different vertical displacement ranges.The increase or decrease of only the joint dip angle does not necessarily cause the corresponding increase or decrease of the stability of large caverns,and with the different joint dd angle,the change of the joint dip angle will show different changes on the unstable block volume and fallen block volume of the surrounding rock of large caverns.(3)Through orthogonal test design,several failure modes of large caves are qualitatively summarized,and the sensitivity magnitude of the stability of large caves to each influencing factor is derived by ANOVA,which provides a new idea for the factor sensitivity analysis of large caves under the joint action of complex factors.(4)BP neural network is used to train the computational samples of large caves,and the trained network is used to quickly judge the stability of large caves,and good prediction results are obtained,which verifies the feasibility of artificial neural network to judge the stability of large caves and provides new ideas for the judgement method of stability of large caves in practical engineering.