Deformation and failure of hard rock under laboratory and field conditions

Sammanfattning: The understanding of the fracture mechanisms and failure processes of the rock is an important requirement for the design of mining excavations and civil engineering constructions. The fracture process is necessary for the excavation and fragmentation of rock, but fracture of rock must be avoided and controlled to preserve the integrity of the construction. This licentiate thesis work is focused on hard rock masses and conditions typical on Fennoscandia. This work was initiated with a review of literature. Following the literature review, information about real underground excavations with deformation monitoring was collected. Laboratory test data was also collected after the survey of cases. The last task for the licentiate work was to perform numerical analysis simulation using the program Phase2 and evaluate the strains due to possible failure. The literature review showed that fracture of brittle rock is the process by which new surfaces in the form of cracks are formed in rock-like material, or existing crack surfaces are extended. Five stages of deformation are distinguished in the fracture process of brittle rock: crack closure, linear elastic deformation, fracture initiation, fracture propagation and post-peak behaviour. The most common reason for stability problems in underground excavation is structurally controlled failure and stress-induced failure. The ground response curve is a technique for describing the response of rock under parameters such as deformation and stress. Thus, the response of the rock mass response can be evaluated and related to the distance to the face of the excavation. The failure criteria reported in the literature are formulated in terms of stresses and include one or several parameters that describe the rock mass properties. Only a few failure criteria were formulated in terms of strains. Since macroscopic failure surfaces are characterized by strain concentrations, fallout criteria should be expressed in terms of strain quantities. Further studies have to be done in orderto be able to formulate strain-based fallout criteria. The four underground cases with hard rock mass and conditions typical of Fennoscandia are: Mine-by Experiment, Instrumented drift at the Kiirunavaara mine, Arlandabanan tunnel and Äspö Pillar Stability Experiment. These cases contain very good information regarding rock properties, geology and stress state. These cases are a good example of in situ deformation measurement. For some cases, the failure occurred and the measured deformation is related to the failure. Laboratory tests of hard rock specimens were performed at Luleå University of Technology and by Posiva Oy. The tested rocks are Fennoscandian types such as limestone, quartzite, diorite, norite, gabbro, diabase, syenite porphyry, mica gneiss, tonalite gneiss and a variety of granites. In these tests, the rock properties and stages of deformation (crack closure, crack initiation and crack damage) were measured and determined. The evaluation of the laboratory tests showed that the stages of deformation vary between rock types and depend on factors such as grain size and mineral composition. Therefore, it may be better if each rock type is treated individually. Failure (i.e., intersection of shear bands forming a v-notch) of a real case and fictitious case was simulated using Phase2. The evaluation of predicted quantities such as maximum and minimum principal, volumetric and maximum shear strains along the depth of the v-notch showed good agreement with the point where the v-notch ended. Furthermore, the comparison between the predicted lateral strains related to failure and the critical lateral strains of rock tested in laboratory showed good agreement as well.