Journal of Geotechnical and Geoenvironmental Engineering

A probabilistic method is presented to estimate the differential settlements of footings on cohesionless soils, considering the uncertainties in both the load and capacity sides of the design equation. A random field approach is employed to characterize the inherent soil variability. This method is first compared to typical limit values from the literature to denote critical combinations of design parameters that can lead to exceedance of tolerable differential settlements. Then, reliability-based design equations are developed for the serviceability limit state (SLS) design of footings on cohesionless soils. The key parameters controlling the SLS are the allowable angular distortion, site variability, and footing spacing. The results are given in a straightforward design format and indicate that currently suggested deformation factors (resistance factors for SLS) equal to 1.0 are likely to be unconservative for most design situations.

Most of the current design methods for driven piles were developed for closed-ended pipe piles driven in either pure clay or clean sand. These methods are sometimes used for H piles as well, even though the axial load response of H piles is different from that of pipe piles. Furthermore, in reality, soil profiles often consist of multiple layers of soils that may contain sand, clay, silt or a mixture of these three particle sizes. Therefore, accurate prediction of the ultimate bearing capacity of H piles driven in a mixed soil is very challenging. In addition, although results of well documented load tests on pipe piles are available, the literature contains limited information on the design of H piles. Most of the current design methods for driven piles do not provide specific recommendations for H piles. In order to evaluate the static load response of an H pile, fully instrumented axial load tests were performed on an H pile $\left(\text{HP}310\times 110\right)$ driven into a multilayered soil profile consisting of soils composed of various amounts of clay, silt and sand. The base of the H pile was embedded in a very dense nonplastic silt layer overlying a clay layer. This paper presents the results of the laboratory tests performed to characterize the soil profile and of the pile load tests. It also compares the measured pile resistances with those predicted with soil property- and in situ test-based methods.

Torpedo anchors are used as foundations for mooring deep-water offshore facilities, including risers and floating structures. They are cone-tipped cylindrical steel pipes ballasted with concrete and scrap metal and penetrate the seabed by the kinetic energy they acquire during free fall through the water. A mooring line is usually connected at the top of the anchor. The design of such anchors involves estimation of the embedment depth as well as short-term and long-term pullout capacities. This paper describes the development of a computational procedure that leads to prediction of torpedo-anchor embedment depth. The procedure relies on a computational fluid dynamics (CFD) model for evaluation of the resisting forces on the anchor. In the model, the soil is represented as a viscous fluid and the procedure is applied to axially symmetric penetration of the seabed. The CFD approach provides estimates of not only the embedment depth but the pressure and shear distributions on the soil-anchor interface and in the soil.

This paper presents the results of a series of plain-strain model tests carried out on both clean sand and oil-contaminated sand loaded with a rigid strip footing. The objectives of this study are to determine the influence of oil-contaminated sand on the bearing capacity characteristics and the settlement of the footing. Contaminated sand layers were prepared by mixing the sand with an oil content of 0–5% with respect to dry soil to match the field conditions. The investigations are carried out by varying the depth and the length of the contaminated sand layer and the type of oil contamination. A plain-strain elastoplastic theoretical model with an interface gap element between footing and the soil is carried out to verify the test results of the model. It is shown that the load-settlement behavior and ultimate bearing capacity of the footing can be drastically reduced by oil contamination. The bearing capacity is decreased and the settlement of the footing is increased with increasing the depth and the length of the contaminated sand layer. The agreement between observed and computed results is found to be reasonably good in terms of load-settlement behavior and effect of oil contamination on the bearing capacity ratio. A comparison between the model results and the prototype scale $\left(B=1.0\text{m}\right)$ results are also studied.

This paper presents a simplified model for evaluating damage potential of a building adjacent to a braced excavation. New developments include several component models for estimating the lateral ground movement profile, angular distortion and lateral strain in a building, and damage potential of a building adjacent to an excavation. These new developments along with an existing model for the vertical ground movement profile are integrated into the proposed simplified model. Case histories are used to verify the developed simplified model. Furthermore, the uncertainty in the developed model is quantified, which enables a probabilistic assessment of the excavation-induced building damage potential. An example probabilistic assessment of building damage potential is presented.

A modified method is proposed for predicting the lateral displacements of the ground caused by installation of soil-cement columns. The method is a combination of the original method derived on the basis of the theory of cylindrical cavity expansion in an infinite medium and a correction function introduced to consider the effect of the limited length of the columns. The correction function has been developed by comparing the solutions obtained using the spherical and the cylindrical cavity expansion theories for a single column installation. Both the original and the modified methods have been applied to a case history reported in the literature, which involves clay soils, and the predictions are compared with field measurements. The advantage of the modified method over the original method is demonstrated. Finally, the modified method has also been applied to a case history involving loose sandy ground, and the calculations show that the method can also be used for this type of soil provided that appropriate consideration is given to the volumetric strain occurring in the plastic zone of soil surrounding the soil-cement columns.

This paper addresses the interpretation of plate load tests bearing on double-layered systems formed by an artificially cemented compacted top soil layer (three different top layers have been studied) overlaying a compressible residual soil stratum. Applied pressure-settlement behavior is observed for tests carried out using circular steel plates ranging from 0.30 to 0.60 m diameter on top of 0.15 to 0.60-m-thick artificially cemented layers. The paper also stresses the need to express test results in terms of normalized pressure and settlement—i.e., as pressure normalized by pressure at 3% settlement $\left(p/p_{3\%}\right)$ versus settlement-to-diameter $\left(\delta /D\right)$ ratio. In the range of $H/D$ (where $H=$ thickness of the treated layer and $D=$ diameter of the foundation) studied, up to 2.0, the final failure modes observed in the field tests always involved punching through the top layer. In addition, the progressive failure processes in the compacted top layer always initiated by tensile fissures in the bottom of the layer. However, depending on the $H/D$ ratio, the tensile cracking started in different positions. The footing bearing capacity analytical solution for layered cohesive-frictional soils appears to be quite adequate up to a $H/D$ value of about 1.0. Finally, for a given project, combining Vésic’s solution with results from one plate-loading test, it is possible (knowing of the demonstrated normalization of $p/p_{3\%}\text{-}\delta /D$ , where the pressure-relative settlement curves for different $H/D$ ratios produce a single curve for all values of $H/D$ ) to estimate the pressure-settlement curves for footings of different sizes on different thicknesses of a cemented upper layer.

Soil reinforcement with discrete fibers is a viable technique to reduce desiccation cracking in compacted clay soils. The reduction in cracking is attributed to an increase in the tensile strength of the fiber-soil composite. A theoretical model is developed to describe the mechanism of the increased tensile strength due to fiber inclusion of soil undergoing desiccation. The model includes a distinctive effective stress combination acting on the fiber strings due to the generated matric suction by desiccation. Model formulation makes use of Mohr-Coulomb failure criterion at the interface area between fibers and the surrounding soil. The desiccation process of the soil generates matric suction within the soil mass, under given stress condition. The basic elements used in the model formulation include soil-water characteristic curve, Mohr-Coulomb parameters, and unsaturated soil parameters. Fiber inclusion increases significantly the tensile strength of the fiber-soil composite. This increase in tensile strength is expressed as a function of fiber content and soil-water content in this paper. Comparisons are made to published data regarding changes in tensile strength with variable water content.

The contribution of transverse ribs to the soil-geogrids interaction under pullout mode has been well documented. However, the contribution of transverse ribs to the soil-geogrid interaction under direct shear mode is, at best, unclear. Consequently, this paper presents the results of a comprehensive direct shear testing program aimed at evaluating the contribution of transverse ribs to the interface shear. The direct shear tests involved Ottawa sand and several polyester geogrids with a variety of material tensile strength, percent open area, and aperture pattern. The test results show that the shear strength of sand-geogrid interfaces under direct shear mode is significantly higher than that of sand-geotextile interfaces. Analysis of shear displacement-strength response of the interfaces indicates that, in addition to interface shear components due to sand-rib friction and sand-sand shear at the location of the openings, the transverse ribs provide additional contribution to the overall sand-geogrid interface resistance. Specifically, analysis of the results reveals that the transverse ribs of the geogrid used in this study provide approximately 10% of interface shear resistance. This contribution is positively correlated with the tensile strength and the stiffness of geogrid ribs, but is negatively correlated with the percent open area of the geogrid. A simple model is proposed to quantify the contribution of transverse ribs to the interface shear strength under direct shear mode.

This paper presents an innovative experimental approach for simultaneous measurements of the suction head, volumetric water content, and the acoustic admittance of unsaturated sands. Samples of unsaturated sands are tested under controlled laboratory conditions. Several types of uniform sand with a wide range of particle sizes are investigated. The reported experiments are based on a standard Buchner funnel setup and a standard acoustic impedance tube. It is a novel, nondestructive, and noninvasive technique that relates the key geotechnical parameters of sands such as volumetric water content, density, and grain-size distribution to the acoustic admittance and attenuation. The results show a very sensitive dependence of the acoustic admittance on the volumetric water content controlled by the value of suction head applied. Analysis of the obtained data demonstrates that the relationship between the volumetric water content and the real part of the surface admittance in the frequency range of 400–1,200 Hz can be represented using a logarithmic equation. It is found that the coefficients in the proposed equation are directly related to the uniformity coefficient and the acoustic admittance of the dry sample, which can easily be measured or predicted for a broad range of sands. A validation exercise is conducted to examine the accuracy of the proposed equation using a sand sample with markedly different properties. The results of the validation exercise demonstrate that the proposed relations can be used to determine very accurately the volumetric water content within the porous specimen from the acoustical data. The error in the acoustically measured volumetric water content is found to be ±2.0% over the full range of volumetric water contents ( $0\le \theta \le n$ , where $n$ is the sample porosity).

In soils with high electrical conductivities, time domain reflectometry (TDR) principles fail to measure the apparent dielectric constant by travel time analysis. This limits the application of water content measurement of existing TDR technologies on those materials. This paper describes a new approach for determining the dielectric constants in highly conductive soils from surface reflections of TDR signals. The multiple reflections at the interfaces of impedance mismatches in the ASTM standard probe were studied. Extension rods were used to avoid the overlap of the reflections along the probe. A relationship between the reflection coefficient at the soil surface and the apparent dielectric constant of the soil was established theoretically and validated by laboratory experiments. Results indicate that the dielectric constant can be determined with reasonable accuracy from the surface reflection coefficient even for soils with high electrical conductivities, where the conventional travel time analysis fails. Compared with alternative approaches, such as based on inversion model, this new method offers high time efficiency in extending TDR applications in highly conductive soils.

It is well established that the main mechanism for the occurrence of liquefaction under seismic loading conditions is the generation of excess pore water pressure. Most previous research efforts have focused on clean sands, yet sand deposits with fines are more commonly found in nature. Previous laboratory liquefaction studies on the effect of fines on liquefaction susceptibility have not yet reached a consensus. This research presents an investigation on the effect of fines content on excess pore water pressure generation in sands and silty sands. Multiple series of strain-controlled cyclic direct simple shear tests were performed to directly measure the excess pore water pressure generation of sands and silty sands at different strain levels. The soil specimens were tested under three different categories: (1) at a constant relative density; (2) at a constant sand skeleton void ratio; and (3) at a constant overall void ratio. The findings from this study were used to develop insight into the behavior of silty sands under undrained cyclic loading conditions. In general, beneficial effects of the fines were observed in the form of a decrease in excess pore water pressure and an increase in the threshold strain. However, pore water pressure appears to increase when enough fines are present to create a sand skeleton void ratio greater than the maximum void ratio of the clean sand.

A numerical study is presented for a rigid block supported through a frictional contact surface on a horizontal or an inclined plane, and subjected to horizontal or slope-parallel excitation. The latter is described with idealized pulses and near-fault seismic records strongly influenced by forward-directivity or fling-step effects (from Northridge, Kobe, Kocaeli, Chi-Chi, Aegion). In addition to the well known dependence of the resulting block slippage on variables such as the peak base velocity, the peak base acceleration, and the critical acceleration ratio, our study has consistently and repeatedly revealed a profound sensitivity of both maximum and residual slippage: (1) on the sequence and even the details of the pulses contained in the excitation and (2) on the direction ( $+$ or −) in which the shaking of the inclined plane is imposed. By contrast, the slippage is not affected to any measurable degree by even the strongest vertical components of the accelerograms. Moreover, the slippage from a specific record may often be poorly correlated with its Arias intensity. These findings may contradict some of the prevailing beliefs that emanate from statistical correlation studies. The upper-bound sliding displacements from near-fault excitations may substantially exceed the values obtained from some of the currently available design charts.

This paper presents an analytical solution for the consolidation of soft soil foundations reinforced by stone columns under time-dependent loadings. The differential equations of the foundations reinforced by stone columns are obtained including smear and well resistance under arbitrary applied loadings. The closed-form solutions of pore pressure and the overall average degree of consolidations are obtained for some common types of loadings, such as step loading, ramp loading, and cyclic trapezoidal loading. By solving the equations using a semianalytical method, the comparisons agree very well with the existing analytical solutions, which verify the correctness and accuracy of the proposed methods. Using the solutions obtained, some selected charts are presented and the relevant consolidation behavior is investigated and discussed.

The compression index is an important soil property that is essential to many geotechnical designs. Over the decades, a number of empirical correlations have been proposed to relate the compressibility to other soil index properties, such as the liquid limit, plasticity index, in situ water content, void ratio, specific gravity, etc. The reliability and thus predictability of these correlations are always being questioned. Moreover, selection between simple and complicated models is a difficult task and often depends on subjective judgments. A more complicated model obviously provides “better fit” to the data but not necessarily offers an acceptable degree of robustness to measurement noise and modeling error. In the present study, the Bayesian probabilistic approach for model class selection is used to revisit the empirical multivariate linear regression formula of the compression index. The criterion in the formula structure selection is based on the plausibility of a class of formulas conditional on the measurement, instead of considering the likelihood only. The plausibility balances between the data fitting capability and sensitivity to measurement and modeling error, which is quantified by the Ockham factor. The Bayesian method is applied to analyze a data set of 795 records, including the compression index and other well-known geotechnical index properties of marine clay samples collected from various sites in South Korea. It turns out that the correlation formula linking the compression index to the initial void ratio and liquid limit possesses the highest plausibility among a total of 18 candidate classes of formulas. The physical significance of this most plausible correlation is addressed. It turns out to be consistent with previous studies and the Bayesian method provides the confirmation from another angle.

The hydraulic conductivity $\left(k\right)$ of specimens from columns containing initially dry sands mixed with bentonite slurries was measured. The mixed specimens represented a range in void ratios $\left(0.672\le e\le 3.94\right)$ and bentonite contents ( $0.61\%\le \mathit{BC}\le 7.65\%$ , by dry weight). The measured $k$ values, which ranged from $2.4\times {10}^{-7}\text{cm}/\text{s}$ to $6.8\times {10}^{-4}\text{cm}/\text{s}$ , correlated poorly with the total void ratio $\left(e\right)$ of the specimens, due to the complicating effect of the bentonite in the sand-bentonite slurry mixtures. However, the measured $k$ values correlated better with the void ratio of the bentonite $\left(e_b\right)$ , which is consistent with the results of previous studies involving permeation of compacted bentonite and sand-bentonite specimens, even though the range in values of $e_b$ in this study $\left(42.5\le e_b\le 127\right)$ was much higher than that previously reported. The relatively large range in $e_b$ values for the sand-bentonite slurry mixtures was also consistent with the relatively large range in measured $k$ values, which are about one to seven orders of magnitude higher than values of $k$ commonly reported for compacted sand-bentonite mixtures, despite similar bentonite contents. In terms of bentonite content, addition of more than 3% bentonite via slurry injection and mixing with the sands was successful in reducing the $k$ of the unmixed sands $\left(9.4\times {10}^{-3}\text{cm}/\text{s}\le k\le 5.4\times {10}^{-2}\text{cm}/\text{s}\right)$ by as much as four orders of magnitude to values less than $1.0\times {10}^{-6}\text{cm}/\text{s}$ .

Dam risk analysis is at the heart of dam failure prevention and mitigation. In order to assess dam risk, it is essential to conduct a quantitative analysis of the process of a dam breach, which can be described by such parameters as breach geometry, breaching duration, and peak outflow rate. The main objective of this paper is to develop robust empirical formulas with physical meaning for predicting dam breaching parameters based on past dam failure data. A database of 182 earth and rockfill dam failure cases has been compiled; among these cases nearly one-half are for large dams higher than 15 m. A multiparameter nonlinear regression model is recommended to develop empirical relationships between five breaching parameters (breach depth, breach top width, average breach width, peak outflow rate, and failure time) and five selected dam and reservoir control variables (dam height, reservoir shape coefficient, dam type, failure mode, and dam erodibility). The relative importance of each control variable is evaluated. The dam erodibility is found to be the most important factor, influencing all five breaching parameters. The reservoir shape coefficient and the failure mode also play an important role in the prediction models. Two case studies are presented to show the application of the empirical models developed in this paper.

A matrix formulation to determine the coupled response of rigid foundations where soil properties are modeled by springs and dashpots is presented. Location and orientation of springs and dashpots can be arbitrary, thus a general solution is determined. The response is given in terms of translational and rotational displacements considering coupled effects. Physics of the problem presented here has been extensively studied and a broad range of useful formulae to determine springs and dashpots properties for soil-structure interaction is available. These published formulae for springs and dashpots properties are an input to the approach presented. Thus, the novelty of this approach is a matrix manipulation that leads to a simple expression allowing coupling all degrees of freedom even when springs and dashpots are not orthogonally oriented. For validation purposes, finite element solutions are compared with the approach presented.

The hydraulic conductivity, the coefficient of consolidation, and the coefficient of volume compressibility play major roles on the pore pressure generation during undrained and partially drained loading of granular soils with fines. This paper aims to determine how much these soil parameters are affected by the percentage of fines and void ratio of the soil. The results of a large number of flexible wall permeameter tests performed on 60 specimens of two poorly graded sands with 0, 5, 10, 15, 20, and 25% nonplastic silt are presented and discussed. Hydraulic conductivity measurements were done at effective confining stresses of 50–300 kPa. The evaluation of the data shows that the hydraulic conductivity and the coefficient of consolidation of sands with 25% silt content are approximately two orders of magnitude smaller than those of clean sands. The coefficient of volume compressibility of the sand-silt mixtures is affected in a lesser degree by void ratio, silt content, and confining stress. The influence of the degree of saturation on the laboratory-measured $k$ values is also discussed.

Smectite is usually used as buffer clay for contaminant of high level nuclear waste canisters. Corrosion of the canisters will expose the buffer clay to ${\text{Cu}}^{2+}$ ions as one of the corrosion products. This study investigates the effect of ${\text{Cu}}^{2+}$ cations in a smectite, with and without addition of sodium carbonate in the clay, on its consolidation behavior and transmission characteristics. Characterization experiments conducted in the study included determination of changes in the buffering capacity and permeability properties of the clay in relation to the presence of varying concentrations of ${\text{Cu}}^{2+}$ , and consolidation experiments. Double-layer swelling forces and buffering capacity are seen to be dominant factors in the performance of the clays—apparently through their influence on formation and restructuring of microstructures in the overall smectite clay mixtures.

The liquefaction susceptibility of various graded fine to medium saturated sands are evaluated by stress controlled cyclic triaxial laboratory tests. Cyclic triaxial tests are performed on reconstituted specimens having global relative density of 60%. In all cyclic triaxial tests, loading pattern is selected as a sinusoidal wave form with 1.0 Hz frequency and effective consolidation pressure is chosen as 100 kPa. Liquefaction resistance is defined as the required cyclic stress ratio causing initial liquefaction in 10 cycles during the cyclic triaxial test. The results are used to draw conclusions on the effect of the extreme void ratios and void ratio range on the liquefaction resistance of various graded sands.