Journal of Geotechnical and Geoenvironmental Engineering

Pressuremeter testing was conducted for the State Route (SR) 99 Bored Tunnel project in Seattle, Washington, to estimate in situ soil stress-deformation parameters along the tunnel alignment. Many of the tests were conducted in a very stiff to hard glaciolacustrine clay known as Seattle clay. This unit is historically known for deep-seated slope failures and many of these failures have been attributed to the release of high, locked-in lateral stresses. Estimation in situ lateral stresses along the tunnel alignment was a primary focus of the exploration program. Due to the hard consistency of this unit and the potential for cobbles, neither self-boring pressuremeter nor dilatometer testing was feasible; therefore, prebored pressuremeter testing was used. Using several lateral stress estimation techniques, including a novel in situ creep testing approach, the in situ lateral stresses in the Seattle clay were estimated to be significantly higher than what would be expected by assuming a simple, laterally constrained, vertical loading and unloading stress path due to glaciation. Deformational features commonly encountered in Seattle clay indicate its stress history also has included significant lateral shearing. The memory of this shearing within the fabric of the clay may influence the in situ stress state and response to lateral unloading.

The potential for incompatibility between soil-bentonite (SB) backfills of vertical cutoff walls composed of conventional sodium bentonite (CSB) and contaminated groundwater has led to the evaluation of chemically modified bentonites for improved chemical resistance. In addition, the stress conditions and consolidation behavior represent important considerations with respect to the performance of SB vertical cutoff walls. Accordingly, the consolidation behavior of SB backfills amended with a polymerized bentonite known as bentonite polymer nanocomposite (BPN) were evaluated and compared with those for a backfill composed of CSB. The three backfills that were evaluated included clean silica sand amended with either 2 or 5% dry BPN and mixed with 2% BPN slurry (i.e., 2BPN2 and 5BPN2, respectively), and the same sand amended with 5% dry CSB and mixed with 5% CSB slurry (5CSB5). The 5BPN2 backfill displayed bisemilog linear stress-strain behavior and was more sensitive to stress conditions than either the 5CSB5 backfill or the 2BPN2 backfill, both of which displayed semilog linear stress-strain behavior. The relatively low $C_c$ of 0.015 for the 2BPN2 backfill reflected a relatively incompressible backfill, probably the result of the low bentonite content for this particular backfill and the dominance of the sand matrix in terms of compressibility, although the hydraulic conductivity, $k$, of this backfill generally was less than ${10}^{-9}\hspace{0.17em}\text{m/s}$. Despite a lower bentonite content for the 5BPN2 backfill relative to the 5CSB5 backfill, the 5BPN2 backfill exhibited a lower $k$ as a result of the superior hydraulic properties of the BPN. The results illustrate that the consolidation behavior of chemically modified bentonites such as the BPN that are used in SB backfills to overcome incompatibility concerns may be somewhat different than that typically associated with more conventional sodium bentonite-based backfills.

A new approach is proposed for modeling thermal conductivity dryout curves (TCDCs) using measurements of soil-water characteristic curves (SWCCs) for coarse-grained unsaturated soils. The approach is based on observation of similitude between a bimodal formulation of the SWCC in the form $S\left(\psi \right)$ and the TCDC in the form $\lambda \left(\psi \right)$. The model may be used in either an empirical curve-fitting capacity to best-fit measured thermal conductivity data or in a predictive capacity to estimate a TCDC from a measured SWCC. Applying the model in a curve-fitting capacity facilitates numerical modeling of heat transfer problems by providing a continuous function for the TCDC. Applying the model in a predictive capacity potentially has broad applications because the SWCC is readily measured or estimated in practice, but the TCDC is not. Performance and characteristics of the model are evaluated by comparison with SWCCs and TCDCs measured for a suite of sands. Predicted thermal conductivity is within 10% of measured values over the complete range of saturation.

Utilizing a pair of building–foundation models placed in strategic configurations, two centrifuge tests were designed to evaluate nonlinear soil–foundation–structure interaction (SFSI) and structure–soil–structure interaction (SSSI) effects. The models were designed to represent a low-rise inelastic frame building founded on individual spread footings and a midrise elastic shear-wall building founded on a mat foundation designed to rock during seismic excitation. Four experimental cases are considered: (1) the baseline response of the low-rise frame structure; (2) the response of two structures placed adjacent to each other along the direction of earthquake shaking; (3) the response of two structures placed adjacent to each other perpendicular to the direction of earthquake shaking; and (4) the response of the frame structure when two wall structures are placed adjacent to it, such that both effects are promoted. Using spectral-analysis methods, it is demonstrated that the baseline frame structure responds independently from the influence of other structures being tested simultaneously and that wave-based SSSI effects have little practical influence on the SFSI response of an inelastic frame structure with highly nonlinear foundation responses. Measurements indicate that footing-force coupling contributes significantly to the seismic response of a frame structure with individual footings, as the nature of the oscillating vertical footing loads from frame action results in dynamic footing overturning moment and base shear capacities that vary significantly from their static values. The proximity of the dynamic footing-force combinations to the theoretical footing-force bounding surface is observed to significantly influence hysteretic footing response.

Drilled shafts have been widely used as an effective means for slope stabilization. In this paper, a design procedure for stabilizing an unstable slope with a row of equally spaced drilled shafts is presented. The limiting equilibrium–based slope-stability analysis method, modified to incorporate the drilled-shaft-induced arching effects through a semiempirical load-transfer factor, is formulated and coded into a general slope-stability computer program, UASLOPE, for complex slope geometry, soil profile, and groundwater table conditions. The UASLOPE program is validated with three-dimensional (3D) finite-element simulation results concerning factor of safety of the drilled-shaft/slope system and net force imparted on drilled shafts. Regression analysis performed on more than 40 cases of 3D finite-element simulation results has led to a semiempirical predictive equation for the drilled-shaft-induced load-transfer factor, from which the phenomenon of soil arching in a drilled-shaft/slope system is elucidated by identifying the influencing factors, such as soil-strength parameters, shaft dimension and spacing, slope geometry, and shaft location on slope. With an objective to achieve the required FS and minimized drilled-shaft quantity and construction cost, two design examples are given to illustrate an iterative process seeking optimization of design outcome in terms of drilled-shaft location on slope, dimension of drilled shafts (diameter and length), spacing between adjacent drilled shafts, and reinforcement requirements.

The soil-structure interaction of piles used to stabilize failing slopes (i.e., subjected to lateral soil movement known as passive piles) was experimentally investigated using a state-of-the-art soil-structure interaction facility. A 102-mm diameter, 1.58-m-long precast concrete pile was installed in well-graded sand. The pile was instrumented with displacement and tilt gauges at the pile head and strain gauges, a flexible shape acceleration array, and thin tactile pressure sheets along the pile length. Furthermore, the three-dimensional (3D) movements of the soil surface and the top of the pile were monitored using two stereo digital image correlation (DIC) systems. By monitoring the relative movement of the soil surface and pile top, the DIC systems track the progression of the soil-pile interaction that occurs as the lateral displacement of the soil increases. The use of advanced sensors allows simultaneous measurement of the pile lateral movement and the soil-pile interaction pressures along the pile length. The measured pressures indicate that the moving soil applies a linearly increasing pressure along the pile length above the sliding surface. Most importantly, the pile displacements and soil-pile interaction pressures along the pile length, which were measured directly, were used to develop the soil-pile interaction force versus displacement relationships (i.e., $p\text{-}y$ curves) for the passive loading condition.

Desaturation is a method for the mitigation of liquefaction of sand. This method has gained increasing interest in recent years as it may become a more cost-effective solution than many conventional methods. In this paper, a microbial denitrification process was adopted to generate nitrogen gas for desaturation of sand. This approach offers several advantages. First, nitrogen gas does not dissolve in water easily. Second, nitrogen gas is chemically inert. Third, as the gas is generated in situ, its distribution is also more uniform. Fourth, the gas bubbles generated by the microbial process are tiny, and thus the potential of escaping from the ground is less. In this paper, undrained triaxial compression and extension tests were adopted to evaluate the liquefaction behavior of sand desaturated using nitrogen under static monotonic loading conditions. Test results show that, under an undrained axisymmetric condition, when the degree of saturation of loose sand is reduced from 100% to a range of 95–88%, the undrained shear strength can increase by more than two times. The reduction in the degree of saturation also leads to a transition from strain softening to strain hardening in the stress-strain behavior in compression tests on sand with relative density $D_r\approx 10\%$ and extension tests on sand with $D_r\approx 30\%$. Moreover, the slope of the instability line that specifies the condition for instability in sand also increases as the degree of saturation reduces. This study proves that the use of biogenic nitrogen gas is effective in enhancing the liquefaction resistance of sand under static, monotonic loading conditions.

This paper describes model tests used to investigate how ballast shoulder width and height contribute to a railway sleeper’s resistance to lateral movement for a range of shoulder widths and heights. Deflection and resistance were measured and photographs taken during the tests. The photographs were analyzed using a digital image correlation technique to identify the zones of ballast surface disturbance, which demonstrated that a bulbed failure volume was mobilized at the ultimate limit state. An idealized three-dimensional failure mechanism is proposed, and resistances are calculated using the limit equilibrium approach. The calculation provides a reliable estimate of the measured resistance. The work identifies the optimum shoulder width and height. The calculations are extended to demonstrate that when a number of sleepers are moved simultaneously, the sleeper end resistance may be one-third less per sleeper than that indicated in tests on an isolated sleeper. Image analysis and limit equilibrium calculations show that this is caused by overlapping of mobilized failure volumes from adjacent sleepers.

The effect of geomembrane thickness (1.5, 2.0, and 2.5 mm) on aging when immersed in a synthetic leachate is investigated over a period of approximately 7 years. Based on data at five different temperatures (55, 65, 70, 75, and 85°C), the predicted time required for a reduction in stress crack resistance to 150 h (half the typically specified value) at 35°C is 62% longer for the 2.5 mm than for the 1.5-mm geomembrane tested and 12% longer for the 2.0-mm than for the 1.5-mm geomembrane. Thus, other things being equal, the results suggest a longer time to nominal failure with increasing geomembrane thickness. It is also shown that the data from a proposed stage-parallel testing procedure collected over 2.5 years fit well with data from traditional incubation of virgin samples over almost 7 years and hence provides a viable means of obtaining good data in a reasonable period of time.

Studies of soil improvement by microbial-induced calcite precipitation (MICP) have focused primarily on fine sand. This paper explores the viability of the MICP technique for improving the engineering properties of a typical tropical residual soil. A species of Bacillus, B. megaterium, was used to trigger calcite precipitation. Four variables were considered in this study: the concentration of B. megaterium, the concentration of the cementation reagent, the treatment duration, and the flow pressure of the cementation reagent. The results show that the improvement in the engineering properties of the MICP-treated residual soils is comparable to those of treated fine sands. The preferable treatment conditions for the soil studied are B. megaterium concentration of $1\times {10}^8\hspace{0.17em}\text{cfu/mL}$, cementation reagent concentration of 0.5 M, flow pressure of 1.1 bar of the cementation reagent, and treatment duration of 48 h. Using this combination of parameters, the obtained shear strength increase and hydraulic conductivity reduction are 69 and 90%, respectively. A minimum calcite content of 1.0% ($15\hspace{0.17em}{\text{kg/m}}^3$) is required to provide measurable improvement in the engineering properties of the soil.

The three fundamental constitutive relations that describe fluid flow, strength, and deformation behavior of variably saturated soils are the soil-water retention curve (SWRC), hydraulic conductivity function (HCF), and suction-stress characteristic curve (SSCC). Until recently, the interrelations among the SWRC, HCF, and SSCC have not been well established. This work sought experimental confirmation of interrelations among these three constitutive functions. Results taken from the literature for six soils and those obtained for 11 different soils were used. Using newly established analytical relations among the SWRC, HCF, and SSCC and these test results, the authors showed that these three constitutive relations can be defined by a common set of hydromechanical parameters. The coefficient of determination for air-entry pressures determined independently using hydraulic and mechanical methods is $\text{>}0.99$, $\text{>}0.98$ for the pore size parameter, and 0.94 for the residual degree of saturation. One practical implication is that one of any of the four experiments (axis-translation, hydraulic, shear-strength, or deformation) is sufficient to quantify all three constitutive relations.

This paper performed centrifuge load tests of a tripod bucket foundation preliminarily designed as a supporting system of wind turbines and compared the results to those obtained from a test of a monopod bucket foundation. The tripod foundation prototype studied in this study has three bucket caissons, each of which is 6.5 m in diameter and 8.0 m in length. The center-to-center distance between the buckets was 26.9 m. The site is composed of an 11-m thick layer of dense silty sand overlying a thick sandy silt layer. The horizontal load was applied at a height of 33 m from the seabed floor according to the design load condition, and the vertical load was simulated by the self-weight of the model. The moment-rotation angle curves for the tripod foundations were compared with that of the monopod bucket foundation. The moment-rotation curve of the tripod was nearly bilinear, whereas that of the monopod showed a gradual decrease in slope. The yield moment for the tripod foundation was half that of the monopod, but the rotation angle for the yield moment was only 20% that of the monopod. The behavior of the tripod foundation under a cyclic load with respect to the accumulated plastic deformation has also been examined in this study. When the resultant moment of cyclic loading was smaller than the monotonic yield moment, negligible accumulated plastic deformation was observed for both one-way and two-way loading. However, when the resultant moment was higher than the monotonic yield moment, significant cumulated deformation resulted.

Suction embedded plate anchors have been increasingly applied in deepwater projects because of their efficiency and limited cost. Current industry practice equips the fluke of the anchor with an outwardly rotating keying flap aiming at a reduction of the loss of embedment during keying. However, recent experimental and numerical research has questioned the efficiency of the keying flap currently used in practice because it has demonstrated that the flap does not open during keying as was intended. This paper presents a two-dimensional plain strain large deformation finite-element analysis by assuming the soil as homogenous to investigate the performance of plate anchors equipped with various keying flap designs. The results of the paper agree with previous centrifuge and finite-element analysis that the outwardly keying flap has no effect on the loss of embedment and results in a reduction of the bearing capacity postkeying. The performance of an alternative keying flap design where the flap is rotating inward is also discussed. The new design effectively reduces the loss of embedment during keying and exhibits a bearing capacity higher than the current design.

Simplified empirical liquefaction triggering models are commonly used throughout the world to assess the liquefaction initiation hazard. Significant differences between two popular liquefaction triggering models have precipitated a substantial amount of confusion in modern engineering practice. Past research to study these differences has generally focused on deterministic or scenario-based applications of the liquefaction models without directly accounting for uncertainties from the models or seismic loading. This paper investigates how accounting for these uncertainties through a performance-based liquefaction triggering assessment affects the differences between these two probabilistic liquefaction models. A comparative parametric study is performed in which a wide range of soil parameters, seismicity levels, and return periods is considered in assessing the disparity between the two models. This disparity is shown to vary significantly with variations in the parameters that were evaluated. Additionally, the performance-based approach is shown to reduce the disparity between the two liquefaction models between 35 and 54% on average over the conventional pseudoprobabilistic approach. The results from this study may prove useful to engineers seeking to explain observed differences between these two liquefaction models when applied in a performance-based manner.

Local tensile strains in a 1.5-mm-thick high-density polyethylene (HDPE) geomembrane induced from coarse overliner soil materials intended to simulate the physical conditions at the base of a heap leach mineral extraction pad under very deep burial are reported. They were obtained from physical experiments conducted in a 590-mm-diameter pressure vessel at applied vertical pressures up to 3,000 kPa for 100 h and a temperature of 21°C. An applied pressure of 3,000 kPa corresponds to a heap leach depth of around 150 m. Three different coarse-grained overliner materials placed directly above the geomembrane and a silty sand underliner beneath the geomembrane were examined. Both the grain size and grain size distribution of the overliner affected the maximum tensile strain in the geomembrane. At an applied pressure of 3,000 kPa, the largest strain of 27% recorded for the coarsest overliner tested, which had a maximum particle size of 50 mm and 20% sand, well exceeded one proposed maximum allowable strain limit of 6%. Finer overliners with both a smaller maximum particle size (25 mm) and much more sand (35 and 55%) reduced the geomembrane strains, but the maximum values still exceeded 6% by a factor of 2, even for the finest overliner examined. At these high pressures, a 150-mm-thick silty sand protection layer between the geomembrane and even the coarsest overliner examined was found to be very effective at reducing local indentations in the geomembrane from overliner particles and was able to reduce the tensile strain in the geomembrane to 2%.

This paper reports $K_0$ of dry sand in response to different loading conditions and under secondary compression. The experiments were carried out using an oedometer and tactile pressure sensors were used to measure the horizontal stress for $K_0$ calculations. During loading, $K_0$ remains a constant as the vertical stress ${\sigma }_v^{\prime }$ increases. The corresponding discrete element method (DEM) simulations reproduced this experimental finding and suggest that the force transmission pattern is similar and the contact forces among particles are proportionally increased in both the horizontal and vertical directions to give a constant $K_0$ as ${\sigma }_v^{\prime }$ increases. $K_0$ increases as the overconsolidation ratio (OCR) increases during unloading because of the locked-in horizontal stress. Reloading makes the sample less sensitive to OCR and thus the sample behaves as if it were in a normally consolidated state where $K_0$ is a constant. Experimental results and DEM simulations demonstrate that the horizontal stress ${\sigma }_h^{\prime }$ and $K_0$ continue to increase during secondary compression on the loading path. Secondary compression is initiated by creep at particle contacts and exhibits similar responses like primary creep. The subsequent loading after secondary compression can gradually bring $K_0$ back to its initial value. During secondary compression on the unloading path, however, $K_0$ continues to decrease according to the DEM simulation results. The creep process tends to relax the locked-in horizontal stress and the sample expands in the vertical direction. Therefore, ${\sigma }_h^{\prime }$ and $K_0$ gradually decreases.

The monocaisson foundation has become an important choice with the increasing demand of offshore wind turbines constructed in silt and sand seabeds in China. To guarantee the normal operation of the wind turbine, the angular rotation of the foundation should be strictly controlled. It is necessary to predict the bearing capacity of caisson foundations at an allowable angular rotation. Several series of model tests were carried out in sandy silt under 1g gravity on suction caissons with different sizes subjected to overturning loading. The test results revealed the effect of the diameter, aspect ratio, and loading eccentricity on normalized moment capacity of caissons. The instantaneous rotation centers of model caissons were simplified to be at the depth of about four-fifths of the skirt length, almost directly below the lid center. Apart from the available ultimate moment capacity prediction approaches, a new calculation approach of deflection-based bearing capacity was proposed based on the mechanisms of soil-structure interaction and deflection of the caisson. Furthermore, determination of the model parameters is discussed. The present analytical approach can predict the moment-rotation curves of the model tests in silt and of the field test in a sand well. This deflection-based approach makes it possible to calculate bearing capacity at any allowable angular rotation. Further studies show that soil pressures and friction on the outer skirt at the passive zone contribute most to the moment capacity of the caisson, specifically the aspect ratio and the horizontal subgrade reaction constant. Finally, design charts of monocaisson foundations in silt and sand for offshore wind turbines are given.

Flow in porous media, such as geosynthetic and granular drains, is often nonlaminar. Nonlaminar flow can vary in character from semiturbulent to turbulent, depending upon the flow velocity, which is related to the hydraulic gradient. An exponent on the hydraulic gradient is used in the relationship among the hydraulic gradient, the apparent flow velocity, and the saturated hydraulic conductivity or transmissivity to quantify the degree of turbulence of the flow in porous media. Numerical values of this exponent are established for both granular and geosynthetic drains based on an analysis of published and unpublished experimental data. It appears that for typical hydraulic gradients encountered in practice, the degree of turbulence for flow in granular drains depends essentially on the particle size, with flow through sand drainage layers being laminar and flow through gravel drainage layers being nonlaminar. In geosynthetic drains, the degree of turbulence depends on the applied compressive stress and the material in contact with the drain.

In practice, the lateral loads applied to piles are often multidirectional. To examine the effect of the loading path on the behavior of piles embedded in clay under lateral cyclic loading, a series of model tests under various displacement paths was conducted using a computer-controlled biaxial motion platform. The displacement paths at the pile head were the unidirectional regular path, the cross path, the eight-shape path, the unidirectional irregular path, and the multidirectional irregular path. The experimental results revealed that under the same displacement amplitude, degradation of the lateral resistance with the number of cycles was more significant under the eight-shape path than under the cross path, and the degradation under the cross path was more significant than under the unidirectional path. The degradation factor was as high as 38% under the eight-shape path with a relatively large displacement amplitude, which was approximately 65% larger than the corresponding value under the unidirectional path. The results for the irregular paths also indicated that the resistances under multidirectional loading were lower than those under unidirectional loading.

The long-term displacement of thermoactive piles subjected to static load and thermal cycles is analyzed using the numerical solution of the one-dimensional load-transfer method modified to account for thermoelastic effects. Numerical results show that thermal cycles cause changes in load transfer and may lead to gradual plastic displacement accumulation due to the mobilization of side shear resistance with the number of thermal cycles. Displacement accumulation depends on the static factor of safety, the amplitude of the thermal cycles, and the ratio between the shaft resistance and the ultimate pile capacity.