Journal of Geotechnical and Geoenvironmental Engineering

Forum papers are thought-provoking opinion pieces or essays founded in fact, sometimes containing speculation, on a civil engineering topic of general interest and relevance to the readership of the journal. The views expressed in this Forum article do not necessarily reflect the views of ASCE or the Editorial Board of the journal.

Plate anchors have become an attractive technology for anchoring offshore floating facilities such as floating renewable energy devices because they provide high holding capacity relative to their dry weight. This allows for the use of smaller anchors (relative to a driven or suction-installed pile), which provide cost savings on production, transport, and installation. Loads delivered to the anchor via mooring lines may increase pore water pressure in fine-grained soils. This excess pore pressure will dissipate with time, resulting in a local increase in the undrained shear strength of the soil surrounding the anchor, increasing the capacity. There may be opportunities to consider these capacity increases if the consolidation process occurs over time periods that are short relative to the lifetime of the facility. This paper considers the use of drainage channels in a plate to make the anchor permeable and quicken consolidation times. Experimental data generated from model-scale experiments conducted in a geotechnical centrifuge show (for the anchor design tested) that excess pore pressure just above the anchor dissipated almost an order of magnitude faster for a permeable anchor, and that after full consolidation, the permeable anchor capacity was higher. The latter finding was not anticipated and is believed to be due to changes in load distribution resulting from the rapid reduction in negative excess pore pressure underneath the permeable anchor.

Axial forces can be artificially adjusted in deep excavations braced with servo struts to reduce excavation-induced deformations. The interactions among the servo strut, retaining wall, and retained soil must be considered to predict the excavation-induced deformations and axial force adjustment. According to the beam–spring model, this study advances a simplified method to calculate the wall deflections generated by excavation with adjustable axial forces in struts. The proposed method incorporates the coupling relationship between the alterations in the earth pressure and wall displacement to consider soil–wall interactions. A mixed boundary, including force boundary and spring boundary, is introduced in the equilibrium equation to simulate the adjustable forces in the servo struts. The recommended method accounts for the soil–wall–strut interactions. Two real excavation cases supported by traditional and servo steel struts, respectively, validate this study’s findings. Comparisons between the predictions and measurements demonstrate the effectiveness of the proposed method. Projections and measurements indicate that the servo struts reduce the maximum wall deflection, which decreases significantly with the increased axial force. Importantly, the retaining wall moves outside the excavation if excessive axial forces are applied in the servo struts.

An instrumented test section was constructed and monitored for 14 years to characterize the hydrology of rock-armored earthen final covers used for uranium mill tailings disposal facilities in semiarid and arid environments. The test section included two different test areas constructed with the same earthen materials and methods employed at an adjacent disposal facility. Both test areas contained a large ($10\times 20\mathrm{m}$) instrumented drainage lysimeter to monitor the water balance at near-field scale. Herbicide was applied to the surface of both test areas to preclude vegetation for the first 7 years, simulating conventional practices used by the DOE. After 7 years, herbicide application ceased on one test area (“naturalized” test area), and vegetation was allowed to establish. Treatment of the other (“conventional”) test area continued as before. Percolation rates increased in both test areas during the initial 7 years while herbicide was being applied. Percolation rates ranging from 10 to $47\mathrm{mm}/\mathrm{year}$ were measured in wetter years, which is considerably higher than anticipated ($3\mathrm{mm}/\mathrm{year}$). Increases in saturated hydraulic conductivity of the protection layer and the radon (Rn) barrier contributed to the higher percolation rates. Ceasing herbicide application in the naturalized test area had a profound effect on the hydrology. As vegetation established, evapotranspiration increased, volumetric water contents in the protection layer and Rn barrier decreased, and the percolation rate decreased significantly. In contrast, the conventional test area continued to perform as before, with elevated volumetric water contents and higher percolation rates in wet years.

Groundwater quality can be deteriorated by leachate that seeps through the landfills. Placing a low permeable barrier beneath the landfill is a cost-effective approach to protect the subsurface environment. Recently local soil and/or waste materials (fly ash and flushed silt) amended with clay have emerged as an effective barrier to restrict and delay the leachate migration to the subsurface environment. One of the most critical selection criteria for landfill liners is hydraulic conductivity (${\mathrm{K}}_{\mathrm{s}}\le {10}^{-7}\mathrm{cm}/\mathrm{s}$), which often varies during the field construction of the liners owing to nonuniform compaction efforts. Failure to achieve the desired compaction leads to the random distribution of hydraulic conductivity within the liner. The movement of contaminants through a landfill liner becomes more complex due to the random variations in the hydraulic conductivity within the liner. To address this, a probabilistic analysis has been conducted to investigate the movement of contaminants in the landfill liner, considering the stochastic hydraulic conductivity scenario. Stochastic hydraulic conductivity realizations of four different composites (bentonite amended sand, flushed silt, fly ash, and locally available natural soil) were generated by varying the standard deviation ($\sigma$) of lognormal distribution in hydraulic conductivity and horizontal (X) and vertical (Z) correlation lengths. The performance of the landfill liner is evaluated by conducting a series of numerical experiments in HYDRUS 2D/3D in terms of breakthrough time, peak concentration, and solute mass distribution in the liner. $\sigma$ is identified in an inverse correlation with the breakthrough time, and a value between 0.25 and 0.75 should be maintained to achieve a breakthrough $\mathrm{time}>30\mathrm{years}$. The ratio of $\mathrm{X}/\mathrm{Z}$ majorly governed the performance of the liner, and it is concluded that a higher ratio leads to better service life. At least a value of $\mathrm{X}/\mathrm{Z}=5$ must be maintained while ensuring that X should not be less than 2. A smaller value of Z would be favorable for both peak concentrations and the breakthrough time. Amongst all composites and all stochastic combinations, natural soil amended with 30% bentonite outperforms with a delayed breakthrough time and lower peak concentrations.

The design of rock-socketed pile foundations requires careful consideration of critical construction aspects such as the socket roughness and smear for accurately estimating the shaft load capacity. Despite the effect of smear being investigated extensively over the years, previous studies are limited in accurately characterizing the effect of smear on the shaft response. This paper presents a comprehensive investigation on the load-bearing mechanisms of model smeared rock-socketed piles with different smear fabrics (configurations) at the pile-rock interface. These model piles were loaded into synthetic soft rocks while being intermittently subjected to three-dimensional (3D) X-ray computed tomography (CT) imaging. Interpretations from the X-ray CT images were correlated with the pile-head load-displacement behavior to gain insights into the load transfer attributes between the pile, smear, and rock. The various interactions between these components were deduced by monitoring the micromechanics at the pile-rock interface through shaft energy utilization, interface void volume, smear volume, and sand density mapping. Based on the observations, it was inferred that the load transfer between the smear and the rock is primarily dependent on the percentage and position of smear occupancy at the interface. Results indicated that for the partially smeared interfaces, the interface mechanics are a combination of smeared and clean shafts with the residual resistances dependent upon the combined compression of smear and rock debris at the interface. Moreover, it was observed that the continuous presence of smear for more than half area of the leading faces of the pile asperities results in the smeared regions of the pile taking precedence over the unsmeared regions in governing the load-carrying capacity of the shaft. The discussions presented in this study can provide a strong base to further study the smear effect, subsequently aiding in enhancing the existing design guidelines to economically construct piles in soft rock.

This paper presents a comparative study on shear responses at the phase transformation, peak, and critical states together with the evolution of particle breakage of uniform-, gap- and fractal-graded carbonate soils. A total of 90 sets of drained and undrained triaxial tests were conducted at relative densities of 30%, 60%, and 90%, under a wide range of confining pressures. The test results show that the phase transformation friction angles ${\phi }_{pt}^{\prime }$ during drained and undrained shearing are approximately the same, and share the same trend with increasing relative density and confining pressure. The stability of the sample for the undrained test (quantified by the ratio of mean effective stress at phase transformation state $p_{pt}^{\prime }$ and initial confining pressure ${\sigma }_c$), peak friction angle ${\phi }_p^{\prime }$, and peak dilatancy angle ${\psi }_p$ for the drained test decrease with increasing confining pressure with the reduction rate decreasing sequentially for the uniform-, gap- and fractal-graded samples. The intercept and slope of the critical state lines (CSLs) of the carbonate soils vary with the initial particle size distribution (PSD) in the $e\text{-}\log p^{\prime }$ space, while the CSLs in the $p^{\prime }\text{-}q$ space appear to be unique regardless of the initial PSD. The crushing capacity, characterizing the crushing potential of a soil sample, was proposed and quantified by a newly proposed parameter $a$. It is also found that the value of parameter $a$ is well correlated with the parameters corresponding to the stability, strength, and deformation responses of carbonate samples.

Abstract

Low-pressure injection of nanosilica aqueous suspensions is often adopted to either waterproof or increase the liquefaction resistance of granular soils. The basic principle behind this ground improvement technique consists in filling the soil pores with a low-viscosity fluid that changes its consistency with time, first into a gel, then into a solid. From an application point of view, the simulation of the time-dependent permeation process is crucial to relate the in situ distance covered by the grout to the operational parameters. A comprehensive investigation was performed, combining laboratory experiments with theoretical approaches, to characterize the phenomenon and then derive predictive relations useful for designing treatment executions. The time-dependent rheological properties of different nanosilica aqueous suspensions were first quantified by means of rheometric tests, then described with Bingham’s law. Grout permeation in granular media was then simulated by suitably modifying Darcy’s law to incorporate the temporal evolution of Binghamian grout rheology. After validating the modified Darcy’s law employment for nanosilica grout flows with respect to laboratory experimental data, simplified analytical equations, capable of predicting the temporal evolution of the distance covered in situ by the grout and the flow rate–injection pressure relation, are provided.

Practical Applications

Nanosilica aqueous suspensions are environmentally nontoxic materials with the consistency of a low-viscosity fluid, suitable for injections into fine-graded soils, but, when mixed with a sodium-chloride solution, they transform with time into a gel of solid consistency. Thanks to these properties, they are frequently adopted to provide a fast remedial against piping induced by excavation, seal contaminants or reduce the liquefaction potential of sands. Nanosilica grout is commonly injected at low pressure, leading to filling soil pores during seepage. The previously mentioned transient evolution of the suspension rheological properties, controlled by nanosilica and sodium-chloride proportions, starts during the injection-seepage phase, playing a paramount role in affecting the geometry of the treated soil domain. The present work provides an accurate description of the time-dependent grout rheology and a predictive seepage analytical tool to simulate the diffusion of nanosilica grouts, characterized by variable compositions and injected from sources of different geometrical layouts into homogeneous soils with different grain size distributions. This tool allows tailoring of the injection parameters (pump pressure, nozzle spacing, injection time) on in situ soil hydraulic properties and rheology of the selected nanosilica suspension.

Dynamically loaded soils can exhibit large-deformation flow liquefaction or limited-deformation cyclic mobility mechanisms, depending on the initial state of the soil. Undrained cyclic triaxial tests were performed on saturated calcareous and silica sand specimens prepared with different relative densities and subjected to various effective confining pressures and cyclic stress ratios to study the flowability of viscous liquefied sand. The cyclic shear stress–strain rate relationship for calcareous and silica sands transitioned from an elliptical shape to an asymmetric Lamé curve shape as excess pore pressures accumulated under cyclic loading. The asymmetric Lamé curve–shaped relationship demonstrates that the saturated sand exhibited low shearing resistance and high fluidity under elevated excess pore pressures for the conditions evaluated. The average flow coefficient, $\overline{\kappa }$, defined as the maximum shear strain rate triggered by the unit average cyclic shear stress, and the flow curve defining the variation in $\overline{\kappa }$ with the number of loading cycles, describes the flowability of the saturated sand and is used to quantify the cyclic failure potential of the saturated sand under a proposed viscous fluid flow failure criterion. The effect of relative density, effective confining pressure, and cyclic stress ratio on the flow curves and the number of cycles to failure under the proposed viscous fluid flow failure criterion is discussed and compared with the cyclic resistance determined from widely used excess pore pressure– and strain-based cyclic failure criteria. The viscous fluid flow cyclic failure criterion is more stringent than these alternative criteria, and the corresponding axial strains are consistent with those associated with liquefaction triggering under cyclic strain approach.

This study investigates the geotechnical characteristics of a soft tidal mudflat in the Great Bay Estuary, New Hampshire. Laboratory testing of surficial sediment samples of the upper 10 cm and field observations from a portable free-fall penetrometer (PFFP) were used to characterize soil strength properties (coefficient of consolidation and undrained shear strength). Pore pressure measurements from the PFFP were analyzed using a square-root of time method to estimate $t_{50}$ and empirical correlations from the literature to estimate the coefficient of consolidation. Undrained shear strengths were estimated from the PFFP deceleration using cone factors of 10–12 and a strain rate correction factor of 0, resulting in undrained shear strengths of 0.5–1.1 kPa. Laboratory testing using miniature vane shear testing found the undrained shear strength to be 0.6–1.8 kPa. Both techniques indicate soft soils. The PFFP deceleration profiles also suggest the presence of layering within the seabed, a situation at the field site that could arise from environmental factors, such as winter ice formation and seasonal erosion. The coefficients of horizontal consolidation estimated with the PFFP ($5.4\times {10}^{-6}–1.7\times {10}^{-5}{\mathrm{m}}^2/\mathrm{s}$) were 1–2 orders of magnitude greater than the coefficients of vertical consolidation measured during oedometer testing ($9.51\times {10}^{-8}-1.78\times {10}^{-6}{\mathrm{m}}^2/\mathrm{s}$ across all samples and load steps). Effects impacting the coefficient of consolidation such as to mismatches between sample depths and the resting depth of the penetrometer, sample disturbance, anisotropy, differences in loading conditions, and limitations associated with deriving values from literature charts are discussed. The results suggest that PFFPs are able to well-characterize the geotechnical properties of tidal mudflat soils and contribute to improving the understanding of the geotechnical parameters of undrained shear strength and coefficients of consolidation.

Microbially induced calcium carbonate precipitation (MICP) is a promising biomediated soil improvement technique. Compared with traditional ground improvement methods, MICP has proven extensive applicability and ecological benefits. However, heterogeneity during MICP treatment is a major obstacle in engineering applications, causing a strong need for monitoring methods to investigate the process of treatment and optimize the subsequent design scheme. Electrical resistivity tomography (ERT) is a noninvasive spatially resolved monitoring technique for visualizing the distribution of both treatment solution and precipitated calcium carbonate in MICP treatment. In this study, three-dimensional electrical resistivity tomography was employed to monitor the biocementation process of a sand-packed column. The electrical response over different treatment stages and the final ${\mathrm{CaCO}}_3$ content were collected. Three-dimensional resistivity distribution intuitively characterizes the distribution of treatment solution in the sand column. Changes in the resistivity indicate the spatial variability of the urea hydrolysis and ${\mathrm{CaCO}}_3$ precipitation during the MICP treatment. The final resistivity distribution corresponds to the distribution of precipitated ${\mathrm{CaCO}}_3$, and the relationship between electrical resistivity and ${\mathrm{CaCO}}_3$ content is established. The study serves to explore the potential of ERT as a monitoring method for the MICP treatment process. Of particular interest are its possibilities to contribute to the monitoring of spatial heterogeneity of cementation because of the limitations of traditional techniques in this respect.

Interactions between adjacent piled-raft foundations may induce additional settlement and cause the tilt of the structure. This case study investigates the long-term performance of a large piled-raft foundation system supporting high and heavy alumina solution storage tanks based on field measurements and three-dimensional (3D) numerical simulations. To properly consider the interactions among rafts, piles, and soils; the elastoplastic behavior of clayey and sandy soils; and the consolidation behavior of soil after construction, a 3D finite-difference code was utilized for the numerical modeling. Two undrained mechanical analyses were conducted to simulate the installation of foundations/tanks and the filling of alumina solution into the tanks, while two coupled hydromechanical analyses were performed to model the period between the two undrained stages and the stage after tank filling, respectively. The predictive capability of the numerical model was checked by comparing the simulation results first with single pile load test data and then with in situ settlement measurements at different locations of the piled-raft foundations. The long-term behavior of the piled-raft foundations was further investigated by analyzing various aspects including the axial force variation along the pile at different locations, the time-settlement relation at monitoring locations, the vertical effective stress increment and displacement of soil, and the differential settlement between monitoring locations. Based on the simulations, the tilt of the platform was mainly due to the superposition of additional stresses near the centerline of the two rows of piled-raft foundations. Finally, parametric studies were conducted to investigate the interactions among the piled-raft foundations at different pile length configurations. The results show that by properly increasing the length of the piles near the centerline of the two rows of piled-raft foundations, the differential settlement of the platforms can be mitigated.

A thermal conductivity function (TCF) is proposed for unsaturated fine-grained soils describing the evolution of thermal conductivity with respect to degree of saturation at room temperature and having parameters associated with the different mechanisms of water retention. Calibration with data from different fine-grained soils reveals that the proposed TCF captures the sigmoidal evolution in thermal conductivity with degree of saturation with a better fit to data in the low and high saturation regimes compared to other TCFs. Correlations between the parameters of the proposed TCF with those of a soil–water retention curve (SWRC) that considers both capillarity and adsorption water retention mechanisms confirm the coupling between these thermo-hydraulic relationships. Thermal conductivity values at degrees of saturation of 1 and 0 can be obtained from experiments on saturated and dry specimens, and the parameters of the new thermal conductivity function correlate linearly with the degree of saturation at maximum adsorption and the SWRC pore size distribution parameter. A strong correlation was also observed between maximum suction and thermal conductivity in dry conditions, possibly due to effects of mineralogy and dry density on these parameters. A successful validation example for compacted bentonite indicates that consideration of the mechanisms of water retention permits deeper insight into linkages between the TCF and SWRC for fine-grained soils.

Using previously proposed values of standard deviation ($\sigma$) for the drained fully softened and residual strength correlations, the probability of failure ($P_f$) can be calculated to supplement the drained factor of safety (FoS). Using the standard deviation ($\sigma$) of soil strength, e.g., fully softened (FS) and residual, the standard deviation of FoS (${\sigma }_{\mathrm{FoS}}$) and coefficient of variation of FoS ($V_{\mathrm{FoS}}$) can be calculated readily using slope stability software to estimate $P_f$. Being able to calculate $P_f$ allows owners, regulators, and/or engineers to select a frequency of failure for a particular slope given the consequences of failure instead of a nebulous value of FoS. For example, it is easier for owners, regulators, and/or legal representatives to understand a frequency of failure, e.g., 1 failure in 1,000 years, than a drained or effective stress FoS of greater than 1.5 with some level of uncertainty. Even more confusing to owners, regulators, and/or legal representatives is the difference between various values of drained and undrained FoS, which is simplified via the selection of a suitable value of $P_f$.