Modelling Of Strength And Deformation Behaviors Of Brittle Rocks

With increasing demands of underground resource exploitations, many large geotechnical engineering structures are constructed in and on ground with very complex geological conditions, such as high in-situ stress, high temperature, and high underground water pressure, etc. These geotechnical engineering structures include deep tunnels, boreholes for oil or gas production, underground caverns for storage of radioactive waste, and wells for injection of carbon dioxides, etc. A deep rock mass in high in-situ stress grounds can fail in a brittle manner. Therefore, it is important to study the strength and deformation behaviors of brittle rocks under complex geological and loading conditions. It is found from laboratory compression tests of brittle rocks that the deformation process of a rock can be divided into several stages, which include closure, initiation, propagation, and coalescence of microcracks in the rock. The development of microcracks influences the strength and deformation behaviors of brittle rocks significantly.The strength and deformation behavior of brittle rocks is a fundamental topic in geotechnical engineering and rock mechanics. Estimation of rock strength and deformation parameters is important for stability evaluation of engineering structures constructed in and on rocks. A better understanding of the strength and deformation behaviors of rocks will facilitate proper engineering design of such structures and maintain the long-term stability of these structures. The strength and deformation behavior of brittle rocks is researched in this study by collecting a large number of triaxial compression test data; some triaxial compression tests are also conducte. The research outcome is useful for better understanding the mechanical behavior of rocks. The main findings of the research are:(1) After studying various methods which have been proposed for identifying the crack closure stress, a new method, which is called Axial Strain Response (ASR) method, is proposed. An attractive feature of the proposed method is that it does not require subjective interpretation from the user and it can be easily programmed. The method is verified using some test data from uniaxial and triaxial compression tests. The crack closure stress determined from the ASR method is comparable with that from other methods in unconfined and confined conditions, indicating that the proposed ASR method is valid. In addition, the crack closure stress is used to evaluate the stress-induced microcrack damage of three rock types at the Forsmark and Oskarshamn sites, Sweden. By correlating the crack closure stress to sampling depth, it is found that the crack closure stress is closely related to the microcrack amount (or density) in the rock, and the crack closure stress can be used to estimate the stress-induced microcrack damage. A linear relationship between the crack closure stress and the differential in-situ stress is also observed. It is indicated that the magnitude of the differential in-situ stress may be a key factor that dictates the amount of stress-induced microcracks in the rocks.(2) The development of the Hoek-Brown (H-B) failure criterion is briefly reviewed and its application condition is discussed using a set of triaxial compression test data. It is found that the brittle-to-ductile transition boundary of rocks, which is defined by the Mogi’s line, could be considered as an applicability condition for using the H-B failure criterion. However, if more meaningful uniaxial compressive strength values are to be determined from the H-B model, a more strict condition of O< σ3< 0.5σc should be reinforced. The H-B failure criterion can fit triaxial compression test data well when its application condition is honored. By establishing a relation between the parameter mi and confinement, a new failure criterion for intact rocks is obtained. The proposed model is not constrained to the application condition of the H-B model, and is able to fit laboratory triaxial test data well over a wide range of confinements. In addition, the new model shows its flexibility to fit the test data of damaged rocks well, indicating that it has potential for being used to model microcrack damaged rocks. The new model can also fit well test data that include tensile test data.(3) By analyzing variation of strength component in the deformation process, a cohesion loss model based on the H-B model is proposed for determining residual strength of rocks. Triaxial compression tests of a coarse marble sample were conducted to verify the proposed cohesion loss model. It is found that the cohesion loss model can fit the residual strength test data well, indicating the proposed model is valid. In order to study the model parameter A,16 sets of triaxial compression test data are analyzed. Parameter A can be related to the origin process and microscopic structure of the rock and it is quite different for different rock types. The ratios of the parameter λ to the uniaxial compressive strength are 5.0 to 15.0,6.0 to 7.5, and 4.4 to 4.6 for sedimentary rock, metamorphic rock, and igneous rock, respectively.(4) By carrying out triaxial compression tests of several thermal-damaged rocks, it is found that thermal damage influences the rock strength and deformation significantly. The peak strength and the elastic modulus decrease and the post-peak ductility is enhanced due to the thermal damage. When the confinement is low, the strength degradation will increase with the increase of the temperature because more microcracks will reside in the rock specimen. On the other hand, when the confinement is high, the peak strengths of the rocks under different temperatures are basically the same, indicating that the main factor that influences rock strength is confinement. A strength degradation index is then defined, and a negative exponent model is established to represent the relation between the index and confinement. By modeling the strength behavior of thermal-damaged rocks, it is found that the strength degradation method can be used to represent the variation of strength with confinement under different temperatures.(5) Based on the framework of the effective medium theory, a quantitative model for charactering the crack closure behavior of rocks is proposed. Several sets of uniaxial and triaxial compression test data are examined to verify the proposed model, and it is found that the model can be used to capture the crack closure and the elastic deformation stages of rocks in both uniaxial and triaxial compressions. The crack closure strain increases with the increase of the microcrack damage level but decreases with increasing confinement. The crack closure strain may be used to estimate the microcrack quantity (or microcrack density) in the rock.(6) In the framework of the continuum damage mechanics, a phenomenological damage model, which uses a logistic function to describe damage evolution, is adopted and combined with the crack closure model to model the pre-peak and post-peak deformation stages of rocks. This yields a new phenomenological constitutive model. A uniform continuity condition is used to ensure that the stress-strain curve is smooth and continuous at the junction point of the crack closure model and the damage model. The proposed model is used to simulate uniaxial compression tests of the Carrara marble under different heating cycles and the uniaxial compression tests of a coarse marble under different temperature heating. The simulated stress-strain curves agree well with the test data, from initial crack closure to near peak and post peak, suggesting that the model can be used for simulating the entire deformation stage of brittle rocks with different degrees of initial microcrack damage.