With much of rock engineering involving large scale projects, the effects of scale - both spatial and temporal - on rock deformability and strength are central issues that are still not well understood. Joints and similar large-scale geological discontinuities are assumed to be the primary factor in reducing strength, but there is evidence to support the existence of a significant size effect even in "intact" rock. Recent developments in numerical modeling allow the interaction of fractures and the bridges of intact rock between them to be analyzed more realistically than in the past. The deep underground science and engineering laboratory (DUSEL) in Lead, South Dakota provides an excellent location to carry out fundamental research to evaluate and improve the numerical predictions.
Large-scale in situ compression tests, in which pillars will be loaded to failure, are proposed for DUSEL Complete load-deformation curves will first be obtained in laboratory tests to assess the potential for brittle failure. Numerical predictions will be made of the pillar response to loading - using the Synthetic Rock Mass model - for all stages of pillar excavation. It is anticipated that the test specimens will range in diameter (or side width, if square) from about 2 m to 5 m, all with a final height to width ratio of 2:1. Knowledge transfer will be accomplished through a Grand Challenge, whereby research and consulting teams will be invited to submit blind predictions of the outcome of key stages of the in situ studies. This is intended to establish the state-of-the-art of predictive modeling, and to stimulate development. The educational mission will be to create and develop mechanisms and technologies that demonstrate, investigate, and enable the control of subsurface excavations through machine stiffness effects.
Recent trends of deep underground projects, whether it is mining, petroleum, or infrastructure related, are expected to continue. Two challenges that arise at excavation depths of thousands of meters are measuring the in-situ stress state and ensuring that rock properties found in the laboratory reflect those of the rock. Improper estimates of either can lead to unexpected failures. To sample the rock, boreholes must be excavated, changing the distribution of stress up to five diameters away. Due to large principal stresses relative to the uniaxial compressive strength, irreversible damage can occur in the core and surrounding rock mass. In extreme circumstances, this damage can be viewed from drilling-induced fractures called core disking. Yet, it is well-known from laboratory tests that microcracks form before failure, implying that the stress redistribution from borehole excavation can damage the core and surrounding rock mass even when no visual damage is observed. Also, experiments have shown that drilling-induced damage may lower the observed strength and stiffness in laboratory tests, misrepresenting the intact rock values assigned to the rock. The research investigated two different aspects of drilling-induced damage. The first is core damage from the excavation process itself. Core samples were extracted from a block of rock in both stressed and unstressed conditions and the effect on Young’s modulus was recorded (Figure 1). The second aspect is core disking, which involves fracture perpendicular to the core axis during the excavation process in a somewhat regular interval, leaving thin "disks" of rock (Figure 2). This often occurs in regions of high in-situ stress where traditional methods of measuring these stresses encounter problems. This has led researchers to suggest core disking as one of many potential means of indirectly measuring one component of the in-situ stress state. Due to the complex three-dimensional nature of the extraction process and multitude factors that may affect disking (bit geometry, bit pressure, fluid pressure, rock characteristics, thermal and fluid effects, etc.), analytical and empirical relations will most likely not be able to give a generalized criterion that can describe the relation between disk thickness and a component of the in-situ stress state. Thus, numerical methods are useful. The excavation and disking processes were simulated and the numerical results were compared with laboratory and field data, giving insight into the disking process itself.
