International Journal of Geomechanics

Developing reliable and versatile constitutive models for geomaterials has been a challenging task for nonlinear finite-element analysis of geotechnical engineering systems. This paper is mainly concerned with the elastoplastic constitutive relationship of unbound granular materials under various stress paths. The fuzzy set plasticity theory has been developed to account for the nonlinear soil stress-strain behavior. The theory in many ways resembles the bounding surface plasticity, where the plastic modulus is a function of the distance between the current and image stress states in the stress space. Based on the concept of fuzzy set theory in mathematics, a membership function whose value ranges from 1 and 0 is introduced to specify the plastic modulus at a given stress state. Hollow cylinder and cubical test results on Hostun sand from the CASE database were selected to assess the fuzzy set plasticity model. The numerical results showed that the fuzzy set model is capable of mimicking and simulating the nonlinear behavior of loose sands under various loading paths.

An analytical model dealing with the problem of an expanded cylindrical cavity in an elastoplastic medium is presented. The constitutive material is assumed to be weightless, homogeneous, and isotropic; its resistance is governed by the Mohr-Coulomb’s strength criterion. Elastoplastic behavior, characterized by a variable plastic potential of flow, is then considered to determine the limit pressure of the expanded cavity. Such a constitutive model takes into account the volume variation, which depends, in particular, on the horizontal distance between the applied load and each point of the expanded medium. Potential applications in soil mechanics are suggested to perform the proposed model.

In the design practice for sand compaction piles (SCPs), it is usually assumed that the state of in situ soil does not change during installation. In fact, there is evidence to suggest that extensive remolding of the preferred soil fabric can occur within the zone immediately surrounding the SCP during its installation. Properties of the soil within this heavily disturbed zone can effect pore pressure dissipation and hence the time after which the ultimate strength of the composite ground is available. In this study, the effect of the smear zone around SCPs is modeled using triaxial tests. SCPs of 25–80 mm diameter were installed in 100-mm-diameter cylindrical samples. The sand columns were installed by predrilling a hole and then backfilling it with well-compacted sand. The smear zone was created by using a rough casing to drill the hole. The effect of the smear zone on SCP was investigated by observing the change in pore pressure during consolidation and undrained shear strength of the composite ground. Although both the shear-induced pore pressures and the undrained shear strength of the composite samples could reasonable be predicted, wide scatter still existed in the relationship between the stress concentration factor and the area replacement ratio.

This article uses the relevance vector machine (RVM) for the prediction of the overconsolidation ratio (OCR) of fine-grained soils based on piezocone penetration test data. RVM provides an empirical Bayes method of function approximation by kernel basis expansion. It uses the corrected cone resistance ($q_t$), vertical total stress (${\mathrm{\sigma }}_{\mathrm{v}}$), hydrostatic pore pressure ($u_0$), pore pressure at the cone tip ($u_1$), and the pore pressure just above the cone base ($u_2$) as input parameters. An equation has also been developed for the determination of OCR. The developed RVM model gives the variance of the predicted data. Sensitivity analysis has been conducted for determining the influence of each input parameter. The results are also compared with some of the existing interpretation methods. Comparisons indicate that the developed RVM model performs better than the existing interpretation methods for predicting OCR.

On the basis of the axisymmetric consolidation model, the governing equation and the corresponding solution were developed for the consolidation of a composite foundation with an impervious column by incorporating an arbitrary stress increment. The consolidation behavior was then investigated as part of the parameter analysis for a composite foundation with an impervious column. The results show that the consolidation rate for a composite foundation with an impervious column was slower than that for a composite foundation with a granular column but was more rapid than that for a natural soil foundation. The consolidation rate accelerated with increasing values of the column-soil constrained modulus ratio or the top-to-bottom stress increment ratio and with decreasing values of the loading period or the radius ratio of the influence zone to the column. The column-soil total stress ratio increased with consolidation and approached the value of the column-soil constrained modulus ratio.

Soil nailing is an in situ soil reinforcement technique used for enhancing the stability of slopes, retaining walls, and excavations. Nails work predominantly in tension, but may also work in bending and/or shear under certain circumstances. To accurately simulate the interaction mechanism between soil mass and nails, it is necessary to study the behavior of their contact interface. The Goodman element is one of the interface elements used extensively in geotechnical contact problems. However, larger normal stiffness values are needed in the application of the Goodman element, which often leads to larger numerical error. In this paper, the Goodman element model was modified by adding two rotational degrees of freedom. Then, a finite-element program was written based on the modified Goodman element model and was applied in analyzing a foundation pit. Compared with the actual monitoring results, it shows that the modified Goodman element model developed in this paper is reasonable and feasible.

An analytical solution for the advection-dispersion equation, including sorption effects, was first proposed by Akio Ogata and Robert Banks in 1961. This solution allows the computation of the percentage of dissipation of the contamination plume at a given point in space and time. However, environmental engineers generally wish to determine the time period necessary for this dissipation to take place to reach a minimum acceptable level established in environmental laws. This paper presents a simplified solution that closely matches the exact solution but has the advantage of allowing a numerical-analytical computation of the flow time and/or thickness of the clay liner, using a scheme also developed in this paper. The proposed solution is compared with the simplified solution known as first-order approximation. The comparison shows that, contrary to the method proposed here, the first-order approximation does not satisfy the limit case of purely diffusive flow and gives unsafe estimates of the solute concentration, by as much as 50% lower than the exact values.

In this study, a simple idealization of heat transfer by conduction at the interparticle contacts is utilized for the evaluation of thermal conductivity of granular soils when subjected to external loading conditions. Discrete element method simulations employing this idealization were performed to examine the impact of loading the soil in a consolidated drained triaxial test environment on soil thermal conductivity. Results of conducted simulations show that shear-induced anisotropy results in an anisotropic thermal conductivity tensor. The results also indicate that contact density affects the average thermal conductivity of granular materials. The larger the average number of contacts per particle, or the coordination number, the larger the thermal conductivity. During shearing, the coordination number tends to decrease, resulting in a reduction in soil thermal conductivity with dense soils, showing a larger decrease in thermal conductivity upon shearing.

The overall development of pore-water pressure in a deep, homogenous, and isotropic porous formation caused by wave-induced cyclic loading has been analyzed. A simplified equation for the estimation of the maximum possible liquefied soil (sediment) depth has been proposed for residual liquefaction in deep, homogenous, and isotropic soils with known soil parameters, water depth, and design wind-induced wave parameters.

This paper presents a case study of an underground crude oil storage project. This storage site is researched with detailed field and laboratory investigations. The layout, cross section, and elevation of the caverns are analyzed and designed based on geological and geotechnical information obtained during these investigations. In this study, two-dimensional (2D) elastoplastic analysis is performed using the finite-element code Phase 2D to study the rock mass behavior caused by staged excavation of the cavern ($20 \text{m wide}\times 30 \text{m high}$). The excavation of the cavern is simulated with different stages, i.e., excavation of a pilot tunnel and subsequent side slashing followed by excavation for a pilot bench and subsequent side slashing with a total of three benches. Displacement values obtained analytically are compared with the observed displacement as the excavation progresses. Because the 2D FEM code is based on the plain strain concept, it could not take into account the effect of excavation in other directions that is very obvious at the intersection with tunnels and shafts. Therefore, these intersections were analyzed using a linear three-dimensional (3D) boundary element code, Examine3D. The stability analysis of rock wedges formed from various joint patterns is also studied using the UNWEDGE software based on block theory for each section of the cavern.

Conduits buried in narrow trenches and covered with granular soil backfills are commonly used for several services. Because of soil arching action, a significant fraction of self-weight of the backfill is transferred to the wall. The load on top of the conduit is expressed as $C_d\gamma B^2$ where$\mathit{\gamma }$ is the unit weight of granular backfill, $\mathit{B}$ is the trench width, and ${\mathit{C}}_{\mathit{d}}$ is the load coefficient, which increases with depth. Values of ${\mathit{C}}_{\mathit{d}}$ for ditch conduits covered with granular backfills have been presented in the literature. Recent research has shown that providing a layer of geosynthetic above the conduit, anchored at both the walls, can reduce the overburden load transferred to the conduit. This paper is aimed at formulating a theoretical framework for computing the load coefficient ${\mathit{C}}_{\mathit{dR}}$ for ditch conduits covered with geosynthetic-reinforced granular backfills. The new ${\mathit{C}}_{\mathit{dR}}$-value takes into account the load reduction attributable to soil arching as well as geosynthetic arching but ignores the soil-geosynthetic frictional/adhesion interaction in the analysis for simplicity’s sake. It is shown that the ${\mathit{C}}_{\mathit{dR}}$-values depend on the stiffness of the geosynthetic layer and the rut depth, among other factors that govern the unreinforced case. The developed expression should be useful for field applications of geosynthetics in buried structures and for developing standards for such applications.

The magnitude of consolidation settlement is often calculated using Terzaghi’s expression for average degree of consolidation ($U$) with respect to time. Developed during a time of limited computing capabilities, Terzaghi’s series solution to the one-dimensional consolidation equation was generalized using a dimensionless time factor ($T$), where a single $U\text{-}T$ curve is used to describe the consolidation behavior of both singly and doubly drained strata. As a result, any comparisons between one- and two-way drainage are indirect and confined to discrete values of time. By introducing a modified time factor $T^{*}$ in terms of layer thickness ($D$) instead of the maximum drainage path length ($H_{\text{dr}}$), it is now possible to observe the effect of drainage conditions over a continuous range of time for a variety of asymmetric initial excess pore pressure distributions. Although two separate $U\text{-}T$ plots are required (for singly and doubly drained cases), the time factor at specific times remains the same for both cases, enabling a direct visual comparison. The importance of a revised time factor is evident when observing the endpoint of consolidation, which occurs as $U$ approaches 100%. This occurs at $T^{*}\cong 0.5$ for two-way drainage and at $T^{*}\cong 2$ for one-way drainage, an observation not possible using the traditional expression for time factor.

The design of stone columns is currently based on theories developed for a single column, ignoring the group interaction and therefore the group efficiency. Whereas single stone columns mainly fail by bulging, a group of stone columns may fail by bulging or shear of the entire soil/columns mass. A numerical model was developed to simulate the case of a single stone column and a group of stone columns installed in soft clay. The model establishes the level of interaction between individual columns and therefore determines the mode of failure of a given geometry, soil, and loading condition. The model was validated with the available experimental data in the literature and used to generate data for a potential mode of failure. This parametric study is conducted to examine the effect of the parameters believed to govern the mode of failure and includes modulus of elasticity of the stone column material and the clay, column diameter, and the spacing and angle of shearing resistance of the column material. The design procedure is presented to assist the designer in establishing the mode failure of a group of stone columns (bulging, general, local, or punching shear).