Research On Uncertainty Of The Regional Stress Field Inversion In Lhasa-Nyingchi Section Of Sichuan-Tibet Railway And Tunnel Rockburst Prediction

The in-situ stress measurement data are the basic data for tunnel and underground engineering construction.However,due to the influence of external factors such as geological tectonic movement and fault zone,as well as the technical and economic constraints such as measurement methods and the number of measured samples,the local in-situ stress measurement data have great uncertainty.Rockburst is an unavoidable risk disaster in underground engineering of hard rock with high geostress.The occurrence mechanism and prediction of rock burst have always been the focus of research on rock burst disaster.High ground stress is the driving factor of rockburst,and the uncertainty of high ground stress measurement data will lead to the uncertainty of rockburst prediction.Starting from the inversion and uncertainty analysis of the in-situ stress measurement data,this paper analyzes the distribution characteristics of in-situ stress in southwestern China based on the in-situ stress data along the Lhasa-Nyingchi section of Sichuan-Tibet Railway.The linear fitting inversion model of in-situ stress and the regional inversion model based on in-situ stress tensor are established,and the influence of different measurement depths,different tectonic plates and different measurement samples on the uncertainty of the regional inversion model based on in-situ stress tensor is studied.Based on the in-situ stress inversion method and its uncertainty research results,the rockburst criterion based on the uncertainty of in-situ stress field is proposed.Through the relationship between rock mass quality index and Hoek-Brown yield criterion empirical parameters,the rock burst criterion of underground rock engineering based on rock mass quality index is established.The main results of this paper are as follows :(1)Based on the measured ground stress data along the Lhasa-Nyingchi section of Sichuan-Tibet Railway,a linear fitting inversion model of regional ground stress based on hydraulic fracturing method is established from two aspects of administrative region and geological structure division.According to the uncertainty and Bayesian theory,the linear fitting model uncertainty evaluation and rationality test.(2)The multivariate non-independent normal distribution inversion model based on the measured geostress tensor is established.From the perspective of gravity field and tectonic stress field,the geostress measurement data are divided according to the measured depth level and tectonic plate.The inversion analysis of geostress under the multivariate non-independent normal distribution model is carried out by using the Monte Carlo simulation method.The number of measurement samples required for in-situ stress measurement data at a certain error level under different measurement depths and different tectonic plates is analyzed,and the uncertainty of local in-situ stress field inversion is quantitatively analyzed.(3)Based on the research results of the uncertainty of local in-situ stress field inversion model,the uncertainty analysis method of rockburst prediction is proposed from the three aspects of initial in-situ stress field inversion,elastic tangential stress solution of tunnel wall and in-situ stress inversion results of far-field tunnel.Through engineering cases,the prediction results of several rock burst criteria based on stress intensity ratio are compared,and a new combined rock burst criterion suitable for the uncertainty analysis of ground stress field is proposed.(4)Based on the elastic-plastic stress solution of tunnel wall under Hoek-Brown yield criterion,the rock mass quality index is introduced,and a new rockburst criterion considering the rock mass quality index is established.The correlation and sensitivity analysis of the parameters such as the critical stress value of original rock,the saturated uniaxial compressive strength of rock and the integrity coefficient of rock mass are carried out,and the influence of parameter uncertainty on the uncertainty of rockburst criterion is analyzed.The rationality of rockburst criterion is verified by engineering examples.The results show that the prediction results of the two newly established criteria are ideal.