Research On The Experiment And Meso-simulation Of Tensile Characteristics And Its Fracture Mechanism Of Brittle Rock

Research on fracture mechanism of brittle rock is one of the basic scientific issues of deep rock engineering. The failure of brittle rock includes spalling, rockburst, etc. It would have an impact on the schedule of project construction, cause huge loss of equipment or staff, and even result in the damage of the whole project, which is closely associated with National security and economy. However, the traditional experiment can not probe into the internal failure process of brittle rock, the strength criterion can not effectively explain the brittle failure, and classical discrete element model has three intrinsic problems in analyzing the fracture mechanism of brittle rock.Lots of laboratory tests and field tests have shown that tensile cracks play a predominant role in the failure process of brittle rock. Thus, this paper, from the view of experiment and mesoscale, did research on the tensile characteristics of brittle rock. First of all, an appropriate meso-model needs to be selected. And then, based on the experimental result, three attributes, Brazilian tensile strength and model I fracture toughness of brittle rocks are respectively analyzed based on corresponding meso-models. The mechanism of meso-tension fracture of brittle rock is summarized. Main content and results of this thesis are as follows:（1） Normally, in compression tests, intact rock presents three attributes:a nonlinear strength envelope with a high UCS/TS ratio and a high frictional angle. However, when using classical bonded-particle model （standard BPM） to calibrate the macro-properties of brittle rock, these three intrinsic problems are raised. By analyzing the composition and constitutive relationship of the standard BPM, and the advantage and disadvantage of the improved standard BPM model, the reasons that causing these three intrinsic problems are summarized. A new bonded meso-model named 3D flat-joint model （FJM3D） was introduced to analyze the fracture mechanism of brittle rock.（2） Combined with the calibration of the experimental result of Jinping Mable and the results of parametic study, the meso-parameters that having a major effect on the three attributes were determined, and a new method for calibration procedure was proposed, which could comprehensively reproduce all the macro-properties of brittle rock.（3） According to the results of typical Brazilian experiment and Brazilian test of Brisbane tuff, the meso-model of Brazilian disc was generated directly by the use of FJM3D model and with the idea of measuring circumference through polygonal approximation. Meanwhile, the circumferential smoothness of the meso-model is controlled by circumferential resolution. This copes with the problems associated with the standard BPM and the traditional method for constructing the meso-model of Brazilian disc. The factors that having major effect on Brazilian tensile strength （BTS） were determined through parametric study.（4） Based on the test of cracked chevron notched Brazilian disc （CCNBD） for determining mode I fracture toughness (K_Ic), the effect of specimen size and loading rate on K_Ic of granite was obtained. The meso-model of CCNBD was constructed by the combination of FJM3D and smooth joint model （SJM）. With use of moment tensor theory, several conclusions were obtained:b value of the acoustic emission （AE） of the granite was about 1.7242, and the frequency of AE events and the number of micro-cracks contained satisfied a negative exponential function. Meanwhile, it was found that the fracture toughness had a positive linear relationship with the bond tensile strength, an exponential positive relationship with the average grain diameter, and an exponential negative relationship with the heterogeneity of the grain size.（5） By collecting lots of the accessible material about brittle rock, it did find that there are inherent linear relationships among tensile strength （TS）, crack initiation stress （CV） and mode I fracture toughness (K_Ic). The linear coefficient between TS and CI is about 0.075 and the linear coefficient between K_Ic and TS is in the range of 0.1 and 0.15. However, the correlation coefficients of these relationships are not high. It is suggested that a standard and unique experiment is used for obtaining each property, and more datum are needed for modifying the empirical relationships. These can be the foundation for the design and assessment of similar brittle rock engineering.