International Journal of Geomechanics

This paper reports numerical analyses of shallowly embedded hemispherical and toroidal penetrometers under torsional and vertical load. These novel penetrometers are a new design suited to the assessment of near-surface soil strength and are aimed at the analysis of pipeline embedment and axial pipe-soil interaction. The geometry of the penetrometers avoids the complication of end effects that arise if a short pipe segment is used, as is the current practice. The operation of these devices involves vertical penetration typically of up to half a diameter, followed by rotation about the vertical axis, whereas the corresponding loads are recorded. The FE analyses explore the undrained bearing capacity and stiffness factors required for the measured loads to be converted to soil strength and stiffness for application in design. On the basis of these analyses, the geometry of the toroidal penetrometer has been optimized to minimize the size of the instrument and limiting interference across the toroid, which would hamper comparisons between the penetrometer response and a pipeline. It is shown that a relatively compact toroid can be used.

The effects of particle-size distribution on a granular assemblage’s mechanical response were studied through a series of numerical triaxial tests using the three-dimensional (3D) discrete-element method. An assemblage was formed by spherical particles of various sizes. A simple linear contact model was adopted with the crucial consideration of varying contact stiffness with particle diameter. Numerical triaxial tests were mimicked by imposing axial compression under constant lateral pressure and constant volume condition, respectively. It was found that an assemblage with a wider particle grading gives more contractive response and behaves toward strain hardening upon shearing. Its critical state locates at a lower position in a void ratio versus mean normal stress plot. Nevertheless, no obvious difference in the critical stress ratio was shown. Model constants in a simple but efficient phenomenologically based granular material model within the framework of critical-state soil mechanics were calibrated from the numerical test results. Results show that some model constants exhibit linear variation with the coefficient of uniformity whereas others are almost independent of particle grading. This investigation provides an opportunity to better understand the implications and meanings of model constants in a phenomenologically based model from the microscale perspective.

In this paper the writers examine a borehole failure model that is based on fracture mechanics and layer buckling theories and compare its predictions with experimental data. The model assumes that the main failure mechanism of borehole collapse takes place in the form of (preexisting or formed) layers buckling. The model introduces a combination of fracture mechanics parameters with length dimension that scales the size of the holes, thus allowing for size effect predictions. The writers review the model and compare its predictions with available experimental data of hollow cylinder tests on weak rocks. The writers found for the model predictions and for the experimental data a strong correlation between hollow cylinder strength normalized by the rock strength and hole size normalized by the square of the ratio of fracture toughness over tensile strength. These findings can be used for constructing efficient mechanical models for predicting or interpreting failure of boreholes and perforations in mining and petroleum engineering.

The Industrial Revolution has led to a substantial increase in the emission of greenhouse gas (GHG) such as carbon dioxide (${\mathrm{CO}}_2$) into the atmosphere, exacerbating the global-warming phenomenon. One option to mitigate GHG emission is by capturing and safely storing the ${\mathrm{CO}}_2$ in suitable deep underground geological formations. This paper provides a literature review on the potential changes in permeability and geomechanical properties of the caprock and sedimentary reservoir rock caused by the injected ${\mathrm{CO}}_2$ under in situ conditions, highlighting the factors that can potentially hamper the caprock’s integrity. A two-dimensional (2D) axisymmetric numerical model of an ideal ${\mathrm{CO}}_2$ reservoir with 30 years of injection and 70 years of monitoring phase is also presented. The model results suggest that, for the parametric values and reservoir conditions used in the model, the injection pressure leads to an increase in pore pressure, thereby reducing the effective stress in the formation. An induced vertical displacement of less than 3 mm is observed at the caprock-reservoir interface at the end of the 100-year period. The observed relationships between injection pressure, effective stress, and total displacement are also discussed.

The characterization of the excavation-damaged zone (EDZ) around an underground excavation is a major research topic for deep geological disposal of medium- to high-level radioactive waste. Rock fracturing because of excavation and thermal loading and its resultant rock mass permeability change in the EDZ are important for construction projects and long-term safety. A new function to predict rock mass permeability change in fractured rocks has been developed and added to the existing fracture mechanics code FRACOD. The new functions in FRACOD have been applied to predict the extent of EDZ and permeability change in the vicinity of the tunnel sealing experiment (TSX) tunnel of the Underground Research Laboratory (Canada), the zone of excavation disturbance experiments (ZEDEX) tunnel of the Äspö Hard Rock Laboratory (Sweden), and the deposition tunnels in crystalline and sedimentary rocks (Japan). The predicted EDZ and its permeability are consistent with the measurement data of the TSX tunnel and ZEDEX tunnel. The results from both tests indicate that FRACOD with the new function is capable of realistically predicting the EDZ and permeability change.

Mode I fracture is common in geomechanics in desiccation cracking, hydraulic fracture, and pressuremeter testing. The cohesive crack model has been used extensively and successfully in numerical modeling of such fracture in concrete and steel but has not been applied in modeling of soil fracture to the same extent. It is argued that the cohesive crack model may be more appropriate than linear elastic fracture mechanics (LEFM) for soils because it takes into account finite tensile strength and any likely plasticity during fracture. With special reference to the Universal Distinct Element Code (UDEC) computer program, a methodology of using interfaces in the distinct element method (DEM) of analysis to model fracture has been validated herein, and this approach is considered to be useful in geomechanical modeling applications. The methodology is based on the cohesive crack approach and shows how softening laws could be back-calculated from load-displacement curves of test specimens. It has been validated using three geometries: a tension test with a rectangular cross section, a notched three-point bend beam, and a compact tension test. Approximate softening laws for St. Albans clay from Canada are proposed.

A viscoplastic cap model was previously developed to address the high strain rate effect on soil behaviors. Although the model is an improvement over the inviscid cap model, it does not update soil density and bulk modulus as the shock wave propagates through the soil. Further, soil should be modeled as a three-phase porous media to accommodate various degrees of water saturation. This is especially true for the soil mass surrounding the source of energy release because each of the three phases responds differently to shock loading. A revised cap model comprising a Gruneisen equation of state for each of the three phases has been developed. These equations of state for solid, water, and air have been integrated with the viscoplastic cap model to simulate behaviors of soil with different degrees of water saturation. Numerical results from this revised soil cap model compared closely with experimental data from explosive tests in both dry and saturated soil.

Nonlinearity of the soil medium plays a very important role on the seismic behavior of soil-pile-structure interaction. The problem of soil-pile-structure interaction is further complicated when the piles are embedded in liquefiable soil medium. A finite-element code was developed in MATLAB to model three-dimensional soil-pile-structure systems. Frequency dependent Kelvin elements (spring and dashpots) were used to model the radiation boundary conditions. A work-hardening plastic cap model was used for constitutive modeling of the soil medium. The pore pressure generation for liquefaction was incorporated by a two-parameter volume change model reported in the literature. In this paper, a $2\times 2$ pile group in liquefiable soil is considered and a parametric study is conducted to investigate its seismic behavior. The effects of loading intensity and stiffness of the soil on the seismic behaviour of the soil-pile system are investigated, considering nonlinearity and liquefaction of the soil medium for a wide range of frequencies of harmonic excitations. The inertial interaction attributable to a structure is analyzed for a system consisting of a four-storied portal frame on the pile group-soil subsystem. The responses of the structure are investigated for harmonic excitation and transient excitations. The importance of consideration of nonlinearity and liquefaction of the soil medium for analysis and design of a pile-supported structure is highlighted. Results from an analysis considering a practical soil-pile problem are presented to demonstrate the applicability of the developed algorithm for a practical problem.

Two design charts for computing the safety factor of soil slopes are presented here. The first one is for an undrained (${\varphi }_u=0$) soil slope, similar to the one proposed by Taylor, but with significant differences. Taylor’s work is based on three types of failure circles: toe circle, slope circle, and midpoint circle. It appears that there can also be compound circles that are made of two circular arcs separated by a straight line at the interface with the stiff stratum. These are incorporated in the proposed design chart. The second chart is for drained ($c^{\prime }-{\varphi }^{\prime }$) soil slope that enables the users to compute the safety factor of the slope without any iterative procedures that are required with the Taylor’s chart. In $c^{\prime }-{\varphi }^{\prime }$ soils, Taylor assumed that the failure occurs along toe circles. The analysis presented herein shows that when the slope is very shallow, it is possible to have midpoint circles. Both charts are quite simple and straightforward to use in engineering analysis of homogeneous slopes. Numerical examples are presented to illustrate the use of the two design charts.