Two- and three-dimensional bonded-particle discrete element modeling of mechanical behavior of transversely isotropic rock

다국어 초록 (Multilingual Abstract)

This paper presents the two- and three-dimensional bonded-particle discrete element modeling of mechanical behavior of transversely isotropic. Systematic verifications of the elastic and strength anisotropy were performed by comparing the numerical results with the analytical solutions, which indicated that there was good agreement with appropriate consideration of overlapping ratio of smooth-joint contacts.
Two validation cases are presented in this study. In the first case, the bonded-particle DEM model successfully captured the monotonic and concave variations of elastic and strength anisotropy and Brazilian tensile strength anisotropy observed in the laboratory tests of Asan gneiss, Boryeong shale, and Yeoncheon schist. Improved match in Brazilian tensile strength was attributed to higher average coordination number in three-dimensional numerical model than that in two-dimensional numerical model whereby leading to higher interlocking.
The second validation case was conducted against CIU (isotropically consolidated undrained triaxial) and Brazilian tensile strength tests of a Tertiary shale from the Norwegian Continental Shelf (NCS). The bonded-particle DEM modeling of the triaxial tests with different initial effective confining pressures showed that the peak strength at failure provided by the numerical model reasonably simulated the strength variation as well as the various elastic moduli with respect to the inclination angles. Brazilian tensile strength of the numerical model was in range of the laboratory data. Furthermore, the numerical model produced similar post failure patterns as exhibited on the disk-shaped NCS shale, e.g., axial splitting and crushing failure.
Three application cases are considered. Two-dimensional bonded-particle DEM model for TI rock was applied to both foundation and borehole stability analyses. As to the elastic response, stress redistribution occurred by the line load exerted on the top surface or the borehole excavation observed from the DEM model was compared with that calculated from analytical solution so that it confirms that the DEM model can effectively present such cases. In addition, continuum based finite element method (FEM) modeling for foundation and borehole problems was performed. Consequently, the overall values obtained from the FEM models were practically the same as those of analytical solution. The large-scale borehole model successfully captured borehole breakout patterns typically observed in both isotropic and TI rock.
A series of hollow-cylinder tests, a representative experimental approach for investigating borehole stability problems, were performed on these transversely isotropic models that have five different inclined bedding planes, which were then compared to the laboratory observations. The result obtained from three-dimensional bonded-particle DEM model was able to capture the instability of the inner hole, i.e., a considerable reduction of effective confining pressure when the hole axis is sub-parallel to bedding planes. Furthermore, the DEM model manifested the overall failure patterns in the vicinity of the hole such as shear failure or spalling, which highly depend on the inclination angle of bedding planes.
The results demonstrated that the bonded-particle DEM model with embedded smooth-joint contacts is a viable model for emulating the mechanical behavior of transversely isotropic rock with the potential of enhanced predictive capability of anisotropic numerical model.

목차 (Table of Contents)

• 1. Introduction 1
• 1.1. Anisotropy of rock 1
• 1.2. Numerical approach 3
• 1.3. Objectives and outline of the dissertation 7
• 2. Methodology 11
• 2.1. Bonded-particle DEM model 11
• 2.2. Generation of intact rock 13
• 2.3. Representation of weak cohesive planes 19
• 3. Verification of bonded-particle DEM model for transversely isotropic rock 24
• 3.1. Verification of elastic anisotropy 24
• 3.1.1. Analytical solution of elastic modulus of TI rock 24
• 3.1.2. Variation of the elastic modulus in the TI model 30
• 3.2. Verification of strength anisotropy 38
• 3.2.1. Analytical solution of strength of layered rock 38
• 3.2.2. Variation of the strength in the TI model 39
• 4. Validation of bonded-particle DEM model for transversely isotropic rock 42
• 4.1. Asan gneiss, Boryeong shale, Yeoncheon schist 42
• 4.1.1. Laboratory data 42
• 4.1.2. Determination of micro-parameters 42
• 4.1.3. Elastic and strength anisotropy 47
• 4.2. Tertiary Norwegian Continental Shelf (NCS) shale 54
• 4.2.1. Laboratory data 54
• 4.2.2. Determination of micro-parameters 59
• 4.2.3. Elastic and strength anisotropy 61
• 5. Application of bonded-particle DEM model for transversely isotropic rock 70
• 5.1. Large-scale foundation modeling in transversely isotropic rock 71
• 5.1.1. Analytical solution of stress distribution under line load in TI rock 71
• 5.1.2. Verification of large-scale bonded-particle DEM model against analytical solution 74
• 5.2. Large-scale borehole modeling in transversely isotropic rock 83
• 5.2.1. Analytical solution of stress redistribution induced by borehole excavation 83
• 5.2.2. Verification of large-scale borehole model against analytical solution 84
• 6. Conclusions 104
• References 108
• Abstract (Korean) 114