Precursor Evolution Of Blocky Jointed Rock Mass Failure And Monitoring And Early Warning Method Of Block Collapse Disasters In Tunnel Engineering

With the continuous rapid development of my country's economy and society and the continuous improvement of people's requirements for quality of life,the overall scale and quantity of transportation projects have shown an increasing trend.Since the beginning of the 21st century,the construction of railways and highways has reached a climax,and the scale of tunnel construction has further increased.my country has become the country with the largest number of tunnels constructed and operating mileage in the world.The collapse of the surrounding rock structure of the tunnel is one of the most common geological disasters in jointed hard rock tunnels.It has the characteristics of strong concealment,strong suddenness,strong destructiveness,and strong disaster-causing characteristics.The catastrophe process involves the progressive destruction of jointed rock masses and dangerous blocks collapse.The large-scale collapse of the body group is extremely difficult to prevent and control,causing serious economic and property losses and casualties every year.Aiming at the monitoring and early warning of tunnel joint hard rock fracture and derived block collapse disaster,this paper focuses on the two important disaster stages of rock bridge fracture and rock block instability.By means of case analysis,laboratory test,precursor monitoring,machine learning,physical simulation and numerical simulation,the failure behavior and associated multiple precursor evolution of jointed rock mass under different stress states are systematically studied.Based on this,the cusp catastrophe model of rock block instability process is further explored,and the early warning criterion of rock block catastrophic instability based on the evolution of natural vibration frequency of rock block under static and dynamic load is proposed,which provides an important theoretical support for the prevention and control of block collapse disaster.The main research results are as follows:(1)This paper analyzes the controlling factors of the failure of tunnel jointed rock mass,and systematically reveals the typical failure mode and internal mechanical mechanism of tunnel jointed rock mass.Aiming at the three typical stress states of tension,tension shear and compression shear,a new multi-functional rock mechanics testing machine of "tension-compression-shear" is developed around the requirements of different failure process simulation and multiple precursors monitoring test of unified scale rock samples.The key is to solve the different types of test open loading aids design,small force stable loading in the tensile test,and other test technical problems,to provide equipment support for the follow-up experimental research.(2)This paper summarizes the common fracture and block falling phenomena of tunnel jointed hard surrounding rock under high,medium and low ground stress conditions,analyzes the internal driving factors and mechanical mechanism of fracture,and generalizes the block collapse disaster into two main stages:rock bridge fracture and rock block instability.Aiming at the research on rock bridge fracture behavior under three typical stress states of tension,tension-shear and compression-shear,a new "tension-compression shear" test system is developed.The multi-functional rock mechanics test system can meet the requirements of direct failure process simulation and associated multi parameter information monitoring of unified scale cubic rock samples under different stress states,and provide equipment support for intuitive and accurate comparison of different failure behaviors and precursory differences.It focuses on solving the experimental technical problems such as eccentric restraint of direct tensile test,weakening of end stress concentration effect,weakening of shear bending moment effect,stable loading of tensile stress with small force value,and design of new loading aids.(3)Aiming at the acoustic emission monitoring with more sensitive response,the acoustic emission signals of rock bridge breaking process under different stress states are analyzed from two aspects of characteristic parameters and waveform parameters.Based on the acoustic emission monitoring indexes of count,energy,amplitude,b value,dominant frequency and entropy,the rock bridge breaking process under different stress states is analyzed from the aspects of fracture number,fracture scale and fracture order.On this basis,the tensile and shear crack classification method using RA-AF value is further used to explore the evolution law of fracture types in the failure process of rock bridge.Different failure tests show that tensile cracks are dominant in the early stage,and the crack evolution mechanism of shear cracks near the failure time.Three typical stress states are established by integrating characteristic parameters,waveform parameters and crack identification types Three color identification method for safety state of jointed rock mass is proposed.(4)Aiming at the failure of jointed rock mass under different stress states,this paper explores a universal early warning criterion based on RA-AF value of tensile and shear crack classification method,introduces machine learning algorithm,proposes an automatic clustering analysis method of acoustic emission RA-AF value based on gaussian mixture model(GMM),establishes the optimal segmentation method of tensile crack and shear crack cluster combined with support vector machine(SVM)model,and analyzes the reliability of the model in automatic identification of crack type,which solves the problem of artificial experience and uncertainty existing in the diagonal segmentation method in JCMS-IIIB5706 specification.According to the engineering practice,a dynamic identification method of crack type based on acoustic emission data points and time periods is established,and the shear crack number in unit period is more than 20%and the shear crack data points cluster is close to the RA axis as a universal warning criteria.Furthermore,an automatic identification method based on likelihood ratio estimation for tension,shear and composite cracks is established.(5)In view of the possible secondary rock block instability and collapse after the rock bridge is broken,a simplified spring proton vibration model of the block is established to reveal the influence mechanism of the shear stiffness of the sliding surface on the natural vibration frequency of the block,and a large-scale physical simulation test of the rock block instability process is carried out to reveal the influence of stress and contact area on the natural vibration frequency of the block and the response characteristics of acoustic emission parameters in the process of sliding instability.Combined with the 3DEC numerical analysis,the effectiveness of the simplified vibration model considering the shear stiffness of the sliding surface was verified.This paper also establish the cusp catastrophe analysis model of the block instability,put forward the early warning criterion of the rock block catastrophe instability under static and dynamic load conditions.It is the first time to build a bridge between the natural vibration frequency and the stability analysis of the block through the shear stiffness of the sliding surface.