International Journal of Geomechanics

A new three-dimensional cone-cap limit-state surface (LSS) model is proposed for cross-anisotropic, saturated, cohesionless, or cohesive remolded soils in consideration of anisotropic yielding and kinematic hardening. Matsuoka-Nakai failure criterion equations are adopted as the cone yield functions, and a new ellipse cap associated with Matsuoka-Nakai cone is developed. The anisotropic vertical and horizontal yield stresses (${\sigma }_{\text{aL}}^{\prime }$ and ${\sigma }_{\text{rL}}^{\prime }$) are used as hardening variables to describe evolution of the fabric anisotropy. A smart kinematic hardening law is suggested without any addition of material parameters. With a nonassociated flow rule and because of the cone-cap connection on a constant $p\prime$ critical state plane, a smooth transition of a plastic strain increment vector at the cone-cap intersection points is ensured for the convenience of numerical calculation. There are very few parameters involved in the proposed model, and they are the same as those in the Cam-clay model, except one parameter for cross-anisotropy [slope of the anisotropic line (AL) $K_{\text{AL}}$ or horizontal yield stress ${\sigma }_{\text{rL}}^{\prime }$]. They can be simply determined from conventional laboratory odometer and undrained triaxial compression experiments. The soil sample is simplified as a stress element, and the element numerical analysis results validate the test data very well. The model-predicted results well illustrate the size and shape modification of the cone-cap LSS of isotropic and cross-anisotropic soils as well as the kinematic hardening effect on the stress-strain behavior. Based on this model, other important features of natural soils, such as viscosity, microstructure, and partial saturation, can be further incorporated by extending the vertical and horizontal yield stresses (${\sigma }_{\text{aL}}^{\prime }$ and ${\sigma }_{\text{rL}}^{\prime }$) in consideration of the effect of viscosity, microstructure, and partial saturation, respectively.

Unbound aggregate base and subbase in layered pavement structures provide an essential function of wheel load distribution through proper angular or crushed particle interlocking of coarse aggregate particles in the granular matrix. This paper used a validated model utilizing a particle image-aided discrete element method (DEM) to evaluate aggregate gradation and shape/morphological properties on pavement granular layer packing characteristics and load-carrying performance. The DEM packing simulations for different granular material gradations and particle shape categories were first performed to investigate the influence of the No. 4 (4.75-mm) sieve, separator of sand-sized and gravel-sized particles according to the Unified Soil Classification System (USCS), for typical dense-graded aggregates specified by the Minnesota Department of Transportation (MnDOT). Aggregate particles were modeled in the DEM as three-dimensional (3D) polyhedral discrete elements with both low and high angularity categories quantified by image analysis. According to the DEM simulation results, the No. 4 sieve (4.75-mm opening) size could be reasonably regarded as the breaking-sieve size that separated the load-carrying coarse fraction and void-filling fine fraction for the studied gradation band. The concept of gravel-to-sand (G/S) ratio, introduced in previous research studies by the authors, appeared to be working well for optimizing aggregate gradations for packing. A G/S ratio of around 1.5 was obtained from the DEM simulations using the 3D polyhedral particles according to the quantified imaging-based aggregate morphological indices to give the highest coordination number due to the achieved densest packing (indicated by the lowest porosity). The previously reported acceptable ranges of G/S ratio for the densest packing hence appeared to be sufficient criteria to optimize unbound aggregate performance. Next, DEM simulations were performed on different gradations with varying G/S ratio values to investigate an optimal gradation proposed for unbound permeable aggregate base (UPAB) materials. The verified G/S ratio–based gradation framework could be potentially applied to engineer aggregate gradations for improved particle contact and packing characteristics—especially when the overall rutting potential influenced by shear strength characteristics needs to be minimized—yet still without significantly compromising desired drainage requirements.

Pull-force estimation is important for the design of static pipe-bursting projects. Among the three components of the pull force, soil expansion force has a significant impact. In this study, a numerical model was employed to determine the soil expansion pressures acting on the expander during pipe bursting. The expansion pressures were used to calculate soil expansion forces, which were then compared to those calculated on the basis of a large-strain analytical cavity expansion solution and those measured from laboratory experiments. Additionally, a parametric study, using numerical modeling and the analytical solution, was conducted to examine the influence of depth of cover on the calculated expansion forces. Typical pipe geometries and soil properties in Edmonton, Alberta, Canada, were used for the parametric study. The results reveal that although the soil expansion force obtained from the analytical solution was higher than that from numerical modeling, the difference decreased when the depth of cover increased. The analytical solution can provide a conservative estimation with a maximum discrepancy of 30%.

This paper describes kinematic seismic interaction analysis of square pile groups in homogeneous soil deposits, focusing on bending moments induced by the transient motion. Analyses were performed by means of a three-dimensional (3D) numerical procedure able to account for both pile–soil–pile interaction and radiation damping. The seismic motion was defined by an artificial accelerogram at the outcropping bedrock, and one-dimensional (1D) propagation analyses were performed to define the free-field motion within the deposits. An extensive parametric study was conducted to determine the effects of different variables, such as the soil properties, the bedrock location, the number of piles, and the pile spacing, on the dynamic response of pile-group foundations. Bending moments obtained from the analyses of the pile group, both at the pile head and at the interface separating soil layers, were normalized with respect to the single-pile bending moments, allowing for the proposal of a new design formula for the estimation of the kinematic bending moments in the most stressed pile of the group, starting from the knowledge of the single-pile response. The proposed formula was used, in conjunction with some simplified approaches that allow estimation of the single-pile response, to evaluate bending moments in the analyzed pile groups. The adequacy of the formula for design purposes is demonstrated.

In this study, laboratory tests were conducted to obtain the geomechanical properties for the coal matrix and coal discontinuities. Computed tomography (CT) scanning technology was then used to detect the pre-existing fracture networks of the cubic coal blocks. Fracture tensor–based methodology was used to quantify the fracture geometry network that exists inside the cubic coal blocks. The same cubic coal blocks were subjected to the true triaxial tests to obtain the jointed coal mass strength (JCMS) values under different confining stresses. The fracture network constructed from the CT scanning was incorporated into the numerical model of jointed coal block to simulate the laboratory true triaxial tests, and to first calibrate the parameter values of the numerical model and then validate them. More numerical true triaxial tests were performed on some jointed coal blocks with selected fracture networks and five additional artificial fracture networks under different confining stress combinations to consummate the JCMS data bank. The obtained data bank was finally used to develop a new three dimensional (3D) coal mass strength criterion that can capture the scale effect and anisotropic strength behaviors.

Despite a body of research carried out on thermally coupled processes in soils, understanding of thermo-osmosis phenomenon in clays and its effects on hydromechanical behavior is incomplete. This paper presents an investigation on the effects of thermo-osmosis on hydraulic behavior of saturated clays. A theoretical formulation for hydraulic behavior was developed, incorporating an explicit description of thermo-osmosis effects on coupled hydromechanical behavior. The extended formulation was implemented within a coupled numerical model for thermal, hydraulic, chemical, and mechanical behavior of soils. The model was tested and applied to simulate a soil heating experiment. It is shown that the inclusion of thermo-osmosis in the coupled thermohydraulic simulation of the case study provides a better agreement with the experimental data compared with the case in which only thermal expansion of the soil constituents was considered. A series of numerical simulations are also presented, studying the pore-water pressure development in saturated clay induced by a heating source. It is shown that pore-water pressure evolution can be considerably affected by thermo-osmosis. Under the conditions of the problem considered, it was found that thermo-osmosis changed the pore-water pressure regime in the vicinity of the heater when the value of thermo-osmotic conductivity was larger than 10⁻¹² m²·K⁻¹·s⁻¹. New insights into the hydraulic response of the ground and the pore-pressure evolution due to thermo-osmosis are provided in this paper.

This paper presents a critical evaluation of the model factor M = q_u,m/q_u,c for Eurocode 7 calculating the bearing capacity of circular footings on dense sand, where q_u,m = measured capacity and q_u,c = Eurocode 7 calculated capacity. Regression analysis is required to remove the dependency of M on the input parameters. Because the input parameters cannot be varied systematically in load tests, previous studies showed that finite-element limit analysis (FELA) can be used as an alternative to load tests for regression. This is further verified from the model factor M_FELA = q_u,m/q_u,FELA with a mean of 1 and a coefficient of variation (cov) of 0.1, where q_u,FELA = FELA predicted capacity. A correction factor (M_s = q_u,FELA/q_u,c) is next defined, which can be decomposed as a product of a systematic part f and a residual part η (i.e., M_s = fη), which is modeled as a lognormal random variable with mean = 1 and cov = 0.11. Finally, a new model factor (M′ = q_u,m/q′_u,c = q_u,m/fq_u,c) is defined. The model statistics of M′ = ηM_FELA can be obtained from those of η and M_FELA, where mean = 1 and cov = 0.11. This is consistent with the results (mean = 1.02 and cov = 0.15) characterized using the load-test database directly, because M′ no longer depends on the input parameters.

The present paper provides the results of a study of the propagation of torsional surface waves in an initially stressed self-reinforced layer lying over a dry sandy Gibson half-space in the presence of initial stress and gravity. The closed-form dispersion relation was obtained analytically. The influences of initial stress, reinforcement, inhomogeneity, gravity, and sandy parameter on the torsional wave were investigated by plotting the dispersion curves. It was observed that the rigidities, initial stress, and sandy parameter of the half-space have a favoring effect on the phase velocity, whereas gravity has the reverse effect on the velocity of the torsional surface wave. It also was noted that as reinforcement increased in the medium, the phase velocity also increased significantly.

This paper discusses how small-strain stiffness and some aspects of the yield surface affect the finite-element predictions of excavation responses, including wall and ground deformations. By using the in-house-developed constitutive model SC1SS, it is possible to isolate and quantify impacts brought about by selected aspects of the soil behavior. This soil model has been validated to reasonably model Taipei silty clay and deep excavations in such soil. This study shows that ignoring small-strain stiffness can overestimate deformations by as much as 80%, leading to conservative and costly design. Inclined yield surface and Lode-angle dependency (on deviatoric planes) have lesser effects on the prediction results. In addition, the mechanisms behind these impacts are investigated through numerical triaxial tests.

A variety of structures similar to marine dolphins, dock-fendering systems, tower foundations, inundated platforms, and abutments are installed on deep foundations, which are subjected to uplift loads. This paper introduces a modern pile-modification strategy known as finned pile that increases pile-tension capacity by providing an affirmative anchorage close to the pile tip. The alteration consists of four fins welded at the bottom of a pile around the circumferential area. Comparative small-scale model uplift tests were performed on normal piles without fins and on piles with fins. Investigations were done by changing the fin-width ratio (b/D), fin inclination angle (β), pile length-to-diameter ratio (L/D), and soil density. Results indicate that there is a considerable increase in uplift capacity of the piles when using fins at the end of the piles. When fins were installed with effective width equal to the pile diameter and at an inclination angle (β) of 90° for sand, with relative density (Dr) of 50%, the improvement in uplift capacity reached 1.82, 3, and 6 times that of the normal pile with L/D stiffnesses of 15, 20, and 30, respectively. It was also found that fins should be installed with an optimum fin inclination angle (β) equal to or greater than 45° to achieve the beneficial effect. The existence of such fins at the lower part of the tested piles provided an ideal anchorage system because of the significant locking-up effect of the soils within the fins, resulting in increased uplift capacity.

In this research, a novel meshless numerical method, termed general particle dynamics (GPD), to simulate crack initiation, propagation, and coalescence in rock-like materials subjected to impact loads was developed. In the GPD code, interaction among discrete particles was formulated with use of the virtual-bond method. Fractures of virtual bonds among particles were determined through a damage evolution law of rock-like materials. An elastobrittle damage model was applied to reflect the initiation and growth of cracks and the macrofailure of the rock-like materials. A damage coefficient, determined by the number of fractures of virtual bonds around one particle, was introduced into the constitutive equations to modify the damaged particles. The damaged virtual bonds were considered the initiation of cracks. The growth path of cracks was captured through the sequence of the damaged virtual bonds. Three numerical cases are presented here. The numerical results were in good agreement with the experimental results.

Characterization of geomechanical parameters in naturally fractured reservoirs remains one of the most challenging tasks in civil, mining, and petroleum engineering. Extended leak-off tests (XLOTs) are generally carried out in new wells to obtain in situ stresses for hydraulic fracture-treatment design and well-trajectory optimization in petroleum engineering. The largest and smallest principal in situ stresses can be calculated by shut-in/closure pressure and breakdown/reopening pressure of XLOTs. However, in situ stresses obtained from XLOTs in the traditional theoretical framework are not completely correct because XLOTs still keep the same test collocations as leak-off tests. In addition, the traditional method cannot be used to simultaneously calculate other parameters beyond in situ stresses. Given these challenges, a hybrid artificial neural network (ANN)–genetic algorithm (GA) method was tested for identification of the principal in situ stresses and joint parameters. First, XLOTs were performed to generate samples for an ANN. The ANN model was then applied to map the nonlinear correlation between geomechanical properties and pressures. Finally, a GA was used to identify geomechanical properties on the basis of the fitness function established using pressures of XLOTs. The results indicate that the inverse-analysis model of pressure established by the ANN–GA provides a powerful and effective tool for multiparameter identification, and it is also a cost-saving and time-saving method.

This paper reports the results of a numerical investigation of the behavior of a natural soft clay deposit under the installation of a case study pile. The case study problem included installation of an instrumented close-ended displacement pile in a soft marine clay deposit, known as Bothkennar clay, in Scotland. The site has been used for a number of years as a geotechnical test bed site, and the clay has been comprehensively characterized with both in situ tests and laboratory experiments. The soft soil behavior, both after pile installation and subsequent consolidation, was reproduced by using an advanced critical-state-based constitutive model that accounts for the anisotropy of soil fabric and destructuration effects during plastic straining. Furthermore, a time-dependent extension of the model was used to study soil creep and the significance of its consideration in the overall pile-installation effects. The simulation results were compared against field measurements; furthermore, for comparison, the pile installation was also analyzed using the well-known isotropic modified Cam-clay model to highlight the importance of considering inherent features of natural soil, such as anisotropy and structure, in the simulations. A series of sensitivity analyses was also performed to evaluate the influence of initial anisotropy and bonding values on simulation results and to check the reliability of the numerical analyses.

In this study, the ultimate bearing capacity and settlement of square and strip foundations in the presence of upward seepage were assessed using physical modeling. The experiments were conducted in two series with different loading values of excess pore-water-pressure paths. The experimental results demonstrate significant degradation of bearing capacity resulting from excess pore-pressure development; however, a remarkable amount of soil strength was observed when the excess pore-pressure ratio was unity in the free-field ground. A correction coefficient (f_γ) for the bearing-capacity factor N_γ was estimated based on the results. This factor is a function of upward-seepage magnitude. Based on the experimental results, dimensionless charts are presented to predict the settlement of square and strip footings rested on saturated sand with varied amounts of excess pore-water pressure. Results show strong dependency of settlement on the bearing-capacity safety factor presumed for the foundation design.

In underground mining practices, cemented paste backfill (CPB), a mixture of tailings, cement, and water, is widely adopted to fill extracted stopes. During and after the filling of the underground mine excavations or stopes with CPB, complex multiphysics processes (including thermal, hydraulic, mechanical, and chemical) take place in the CPB mass. These multiphysics processes and their coupling govern the consolidation characteristics and behavior of CPB. As a result, conventional soil mechanics consolidation theories and models are not appropriate for the evaluation and prediction of the consolidation behavior of CPB. Therefore, based on the principles of the continuity of pore space and conservation of mass, energy, and momentum, a three-dimensional coupled multiphysics consolidation model for CPB was developed in this study. The prediction capability of the proposed model was verified by comparing the predicted results with experimental data. Good agreement between the predicted values and experimental data was obtained. Furthermore, some of the important features of the developed model are highlighted by a comparison between the simulated results and Gibson’s solution.

The one-dimensional (1D) Richards equation is used to model infiltration flow in an unsaturated two-layer soil. Imposing continuities of both flux and pressure head at the interface yields a nonlinear equation for determining interface conductivities. The authors show that multiple solutions of this nonlinear interface equation may exist if the spatial discretization is not fine enough around the interface, in particular as sharp wetting fronts pass through the interface, or for flow across highly dissimilar materials. Three hydraulic models, the Gardner model (G), the Mualem-van Genuchten model (MvG), and the Fredlund-Xing-Leong-Rahardjo model (FXLR), are investigated to demonstrate the nonuniqueness of solutions to the interface problem. For the simplest G model, a full mathematical analysis of the interface problem in terms of two parameters depending on local hydraulic conditions and mesh size is possible. For more advanced models, the interface equation can only be analyzed numerically. In all cases, the interface equation exhibits multiple solutions under suitable conditions, and the impact on the overall numerical solutions is studied. In particular, both a fixed-point iteration and a Newton iteration are used to solve the interface equation numerically.

Cemented paste backfill (CPB) is widely applied in underground mines around the world. Prior to stope backfilling, barricades need to be built in drifts near drawpoints to retain the flowable fill. A number of reported barricade failures have shown that barricade stability is critical for ensuring successful and safe application of backfill. The barricades are usually made of high-strength materials such as bricks, concrete blocks, or reinforced shotcrete. Alternatively, barricades made of waste rocks are becoming popular because of their simple and low-cost construction. A simple solution was proposed for sizing waste rock barricades (WRBs) by considering the limit equilibrium of a three-dimensional (3D)-rectangular block. More recently, the authors modified this solution by considering the global stability of trapezoidal barricades, but the local stability was not taken into account. Consequently, the size of the crest can be underestimated, leading to a nonconservative design. In this paper, a more complete solution is proposed, considering both the global and local stabilities of trapezoidal WRBs. The analytical solution was calibrated and validated using numerical modeling. The flexibility and validity of the proposed solution were further tested with complementary simulations. Sample calculations are also performed here to show the application of this solution and to illustrate the effect of key influencing factors on barricade design.

In this article, the influence of gluing-solution layout on the gluing-solution distribution at the final stage of solidification process in a realistic railway ballast is investigated. A unified Cauchy stress model, with an asymptotic time-dependent viscosity, is proposed to account for the coupled antithixotropic and shear-thickening features of gluing solutions. Realistic railway ballast is accomplished by using the vertex-identification technique, in which regulations for gravel composition in ballast practice are followed. A network and matrix layouts of gluing solution are applied. Flow characteristics and time-dependent solidification of the gluing solution in the multiconnected conducts spanned by the gravels are simulated by using a finite-element code. Although more concrete lateral and vertical gluing adhesions among the gravels can be accomplished by spreading gluing solution in the network arrangement, the gluing-solution distribution obtained by pouring gluing solution in the matrix layout is relatively localized and limited. For a better pointwise gravel–gravel adhesion, gluing solution in a network arrangement is pronounced to be an a priori gluing practice.

This paper proposes three-dimensional (3D) simplified nonlinear methods for examining the soil–building interaction for nonlinear behavior of soils based on an input seismic wave field. A seismic wave field is defined as seismic waves propagating in a 3D medium. The proposed 3D methods were developed on the basis of the 3D linear method, which was recently proposed to adequately treat seismic surface waves trapped by a several-kilometer-deep underground structure. To demonstrate the feasibility of the proposed methods, interaction analyses of a midrise RC building and a wood building were performed in the reclaimed zone of Tokyo Bay in the cases of soils with linear, nonlinear, and liquefaction behavior for the 1923 Kanto earthquake. These interaction analyses provide a reasonable evaluation of building performance. In particular, building responses became excessively large, following extremely large increases in the amplitudes of surface waves in liquefied soils. The building responses provide significant clues for interpreting a typical damage pattern in which Japanese RC building damage is concentrated on the first story.

In this paper, a mechanical model to idealize the load-settlement response of the load transfer platform (LTP) on column-improved soft soil is proposed. This model simultaneously considers the nonlinear and time-dependent stress-strain behavior of soft soil and the negligible tensile strength of the granular material in LTP. The reinforced Timoshenko beam is adopted to model LTP to consider the shear and flexural deformations. Soft soil is idealized by a spring-dashpot system that includes nonlinear and time-dependent behaviors. The columns and geosynthetics are modeled with linear Winkler springs in the applied range of stresses and rough elastic membrane, respectively. The response function of LTP has been derived for distributed pressure loading in the plane strain condition. The principle of superposition is used to solve the fourth-order differential equations. Parametric studies indicate that the spacing of columns, thickness of LTP, degree of consolidation of the soft soil, and tensile stiffness of the geosynthetics significantly affect the behavior of LTP. This study also evaluates the accuracy of using reinforced Timoshenko theory by comparing the results with Pasternak and Euler-Bernoulli theories.

A mathematical model of propagation of the Love wave in an intermediate isotropic vertical layer placed in between two prestressed half-spaces with different elastic properties is proposed. The upper half-space is orthotropic in nature. The density and rigidity of the lower half-space are assumed to be space dependent. The proposed model is solved to obtain the 10 different dispersion relations for the Love-type surface wave in the elastic medium of different properties. The effects of the initial stress parameter and heterogeneity parameter of half-spaces on the dispersion of Love waves are shown numerically by plotting various curves of nondimensional phase velocities against nondimensional wave number. Love wave velocities are derived analytically. In the absence of the considered technical constants, the obtained results validate the formulation. Some special cases of interest are also discussed. This analysis may be useful to interpret the nature of propagation of seismic waves in elastic media.

For the past several decades, researchers around the world have used nanoindentation techniques to characterize different materials for biological, medical, and polymer science applications. In recent years, nanoindentation techniques have been explored by some pavement professionals to characterize asphalt materials. This study used the atomic force microscope (AFM)–based PeakForce quantitative nanomechanical mapping (PFQNM) technique to evaluate nanoscopic properties of asphalt binders modified with four different types of industrial and agricultural waste materials, which included ground tire rubber (GTR), reclaimed asphalt pavement (RAP), reclaimed asphalt shingles (RAS), and rice husk ash (RHA). Three scan sizes (20 × 20, 10 × 10, and 5 × 5 µm) were used to map the properties of the aforementioned binder samples. Mechanistic properties such as adhesion, energy dissipation, deformation, and Derjaguin, Muller, and Toropov (DMT) moduli of modified binder samples were estimated at different morphological phases. The PFQNM analyses revealed distinct microstructures and grain distributions among tested additive-modified binder samples. The GTR-modified samples exhibited two distinct phases, namely, Catana and Peri phase, but the other additives provided another called the Perpetua phase. Mechanistic properties of the asphalt samples in the Catana and Peri phases were found to be similar but different for the Perpetua phase. The DMT moduli varied from 70 to 800 MPa, and the adhesion ranged from 8 to 12 nN. Among the four additives, RHA reduced the hardness of the asphalt binder unlike the others. The dissipation energy of the binder samples was found to be related to the adhesion, whereas the hardness depended on the depth of deformation. Demonstrations of quantitative mappings of the nanomechanistic properties of asphalt binders are expected to help build an understanding of the underlying sciences.

Ground-improvement techniques, such as jet-grout columns, may have positive effects on structures founded on soft soil when subjected to earthquake loading. The main objectives of this paper are first to evaluate the effects of soil-improvement width and depth on the ground response of a layered soil and, second, to define critical zones in which cemented columns might crack. A series of numerical calculations are performed by means of the finite-element method, considering a linear viscoelastic model for soil layers and rigid bedrock condition, and also by applying an advanced constitutive model for simulating the behavior of jet-grout columns. The soil profile consists of a 10-m loose sand layer overlain and underlain by a 5-m dense sand layer. The acceleration time recorded during the Loma Prieta earthquake in 1989 is used as the input motion. The results show that the seismic load is decreased or may increase as a result of ground improvement. In addition, it is shown that by applying an advanced model, possible crack zones in cemented columns can be identified.

A novel bond-based peridynamic method is proposed to investigate the initiation, propagation, and coalescence of cracks in brittle rock materials subjected to compressive loads. To simulate the failure of brittle rock materials under compressive loads, tangent bonds are introduced into the bond-based peridynamic theory. By comparing the strain energy with peridynamic energy, the micro-peridynamic parameters can be expressed by the macromechanical parameters of rocks. The novel peridynamic model can solve the problem that the normal bond-based peridynamic theory is limited to constitutive models with a Poisson’s ratio of 1/4 for three-dimensional problems. A program code was compiled to demonstrate the validity of the novel bond-based peridynamic method. It was found that the numerical results of the initiation, propagation, and coalescence of cracks in brittle rock materials subjected to compressive loads are in good agreement with the experimental data.

Alkaline activation of fly ash creates a geopolymeric cement that can replace ordinary portland cement in several applications such as soil improvement, with the advantage of much lower carbon dioxide emissions and reusing an industrial by-product otherwise landfilled, which averts several environmental problems. In this paper, the behavior of a silty sand improved by the alkaline activation of fly ash is analyzed from small to large strains by presenting uniaxial and drained triaxial compression test results and seismic wave velocities measured throughout the curing period. The dynamic, cyclic, and static tests showed a significant increase in stiffness with curing time, even beyond the 28-day curing period. On the basis of the nondestructive wave-propagation technique, the increase of the shear and compression wave velocities with time were drawn, giving the evolution of the elastic shear modulus and the Poisson ratio values. The dynamic Young modulus was compared to the correspondent secant Young modulus obtained from the mechanical tests. In addition, the evolution of the properties of this stabilized soil with curing time was compared and confronted to that of soil cement on the basis of the elastic stiffness of both materials, which showed that the most significant difference lies on the curing rate.

The mechanisms that result in the settlement of structures constructed on peat foundations have been the subject of numerous investigations. A recent study of the behavior of the peat subgrade showed that gas bubbles within the peat strongly influence the pore pressure. The thermal expansion of these gas bubbles results in an increase in pore pressure during the warmer months. The gas bubbles remain trapped until the pore pressure is sufficient to push them through the pore constrictions and toward the drainage boundary. The expulsion of the gas bubbles is followed by a sharp drop in pore pressure and a rapid change in volume. This paper presents the analysis of field data conducted to quantify the impact of this mechanism on the settlement of an embankment constructed on a peat subgrade. Measured pore pressures and deformations in the field are analyzed to show the acceleration of vertical deformation during the pore pressure drops. An approach developed from laboratory isotropic consolidation testing results is presented to correlate the magnitude of the drop in pore pressure to the corresponding volumetric strain. This correlation is then extrapolated to estimate the settlement of peat subgrade corresponding to the expulsion of gases. The results suggest that approximately 15% of the annual vertical settlement of the embankment occurs due to the cyclic dissipation of gas bubbles from the peat subgrade.

The appropriate estimation of seismic earth pressures acting on retaining walls supporting submerged backfill is an important area of research in earthquake geotechnical engineering. There is a scarcity of research work on retaining walls with submerged soil under seismic conditions in the literature. In this paper, closed-form generalized solutions for computing seismic active earth thrust and its distribution on nonvertical rigid retaining walls with fully submerged backfill are proposed using the modified pseudodynamic approach. The seismic analysis of retaining walls was carried out with a planar rupture surface in viscoelastic submerged backfill. The proposed modified pseudodynamic analysis satisfied the zero-stress-boundary condition on a free surface and considered the standing shear and primary waves, soil damping, amplification in the backfill, and an excess pore-pressure ratio in the backfill. A detailed parametric study was performed to gain an understanding of the effects of different parameters, including the horizontal seismic acceleration coefficient, excess pore-water-pressure ratio, period of lateral shaking, damping ratio, wall inclination, and soil-friction and wall-friction angles. A comparison of the present study results and the existing analytical and experimental studies resulted in good agreement.

A numerical protocol was developed for the evaluation of the passive earth-pressure coefficients (K_pq) for the case of an inclined rigid wall retaining an inclined cohesionless backfill loaded with a uniform surcharge. The shape of the proposed failure surface is assumed to be a complete log-spiral. The methodology developed herein uses Kötter’s equation, which enables the computation of soil reactive pressure distribution and the resultant soil reaction on the failure surface. Furthermore, all three limit-equilibrium conditions are effectively utilized to identify the unique failure surface for the given parameters, viz, soil frictional angle (ϕ), wall frictional angle (δ), wall inclination angle (λ), and sloping backfill angle (i). Design tables are also presented for several combinations of ϕ, δ, λ, and i. For validation, the K_pq values obtained are compared with those reported by several other researchers and are found to be in good agreement with most of the available theoretical solutions.

In chemically active rocks, pore pressure and stress state are not the only parameters that are governed by hydromechanical processes. Chemical osmosis is also responsible for changes in the effective stress caused by the presence of chemical potential between the reservoir and the borehole filled with drilling fluid. Subsurface rocks such as shales are highly anisotropic, and assuming an isotropic model will fail to provide an accurate depiction of the in situ stress state. In this work, governing equations for a fully coupled anisotropic nonlinear porochemoelasticity are developed. A numerical model using the FEM is constructed to simulate the drilling of an inclined borehole problem in an anisotropic and chemically active medium. The model accounts for solute transport contributed by diffusion and advection. Several sensitivity analyses are conducted to highlight different features of the chemohydromechanical model. Solute concentration, membrane efficiency, solute-diffusion coefficient, and advection are investigated. Results of this study show that the rock elastic anisotropy significantly enhances the stresses induced by the borehole excavation. Moreover, the presence of chemical potential has been shown to have a large bearing on the near wellbore stress state and pore pressure. It was demonstrated that this effect cannot be generalized and should be studied carefully on a case-by-case basis. Last, the model is used to analyze borehole stability for inclined boreholes under chemical, hydraulic, and mechanical loading.

In this paper, a synthetic rock mass (SRM) method is used to numerically characterize the effect of joint sets on rock pillars. The SRM model is constructed by explicitly inserting discrete fracture network into a particle assembly. Conceptual SRM models show that pillar-loading capacity is weakened by inserted joints. Pillar peak strength is lower when inserted joints favor shear sliding of rock blocks, and strength becomes higher when pillar failure is controlled by fragmentation of intact rocks. Meanwhile, loading capacity is weakened when longer joints are simulated. The effect of joint sets on pillar modulus is similar to the observed effect on peak strength. Pillar failure behaves as a continuous shear failure when the inserted joints are inclined and changes into intact rock splitting when the joints become vertical. The SRM method is then used to characterize the joint set effect on real pillars in the Doe Run mine. A pillar model with a width/height ratio of 0.8 is initially constructed on the basis of derived joint characteristics from photogrammetric mapping. Numerical results show that the pillar peak strength and deformation modulus were reduced by 68.1 and 44.8%, respectively, in comparison with the corresponding properties of the joint-free model. A series of additional pillar SRM models is also studied, and SRM pillar strengths agree with the empirical formulas in general. Finally, this paper presents a comparison study between continuum and SRM models. The coincidence in findings between the two numerical methods further validates the robustness of the SRM method for characterizing joint set effect on rock pillars.