Journal of Geotechnical and Geoenvironmental Engineering

The morphology of two types of complex calcareous sand was investigated in this study. The materials were selected owing to their different geologic and biologic origins. Ledge Point is a bioclastic coastal sand, while Browse #1 is a hemipelagic sand. These two sands fall outside the range of common data sets used to correlate mechanical properties to particle shape parameters. Morphologic analysis of these calcareous sediments can aid with understanding the engineering behavior of calcareous soils. Moreover, sediments source tracing information could also be inferred from particle shape analysis. Two-dimensional Dynamic Image Analysis (DIA) was employed to capture five million and eight million particle images of each sand, respectively. A number of size parameters including diameter of equivalent projected circle (EQPC), Feret minimum, and Feret maximum diameter were efficiently obtained for each captured image using DIA, and used to investigate particle size distribution of these sediments. In addition, samples of over 800,000 particles were used to assess statistical distributions of various particle shape parameters including Aspect Ratio, Convexity, Sphericity, and Roundness-DIA (by volume). A Johnson family of distributions was found to provide a better fit to particle shape parameter distributions than the normal distribution for both sands. It is also shown that the Sphericity and Aspect Ratio are size independent, while Convexity and Roundness-DIA are correlated with particle size. Convexity is likely correlated with Sphericity for both calcareous sediments owing to their biogenic origin. Correlations of Roundness-DIA, Sphericity, Convexity, and Aspect Ratio and particle size are also observed but need more analysis.

The main objective of this study was to investigate the primary consolidation and creep characteristics of two unsaturated sand–bentonite mixtures subjected to drying–wetting cycles. An extensive array of laboratory testing was performed to examine index physical, mineralogical properties, soil-water retention characteristics, and compressibility of the tested materials. A modified osmotic oedometer was used to measure compressibility, creep, and swelling indexes of the sand–bentonite mixtures under three sets of drying–wetting at matric suction values of 200, 400, and 800 kPa. The results showed that the elastic compressibility indexes of the tested sand–bentonite mixtures were not affected by hydraulic loading. The results demonstrated kinematic hardening under hydraulic and mechanical loadings. Hydraulic loading stimulated hardening behavior and a decrease in creep index at all suctions. Results from the two test sets performed in the current study were used along with three other sets of laboratory measured data from the literature to calibrate and validate a model to determine the creep index of unsaturated soils by incorporating the role of stress state and matric suction.

In this study, the coefficient of lateral earth pressure at rest ($K_0$) for sand subject to freezing and thawing was investigated, focusing on the effect of pore water volume. Unfrozen (UF), frozen (FR), and thawed (TH) conditions were all addressed and considered in the investigation. Experimental testing programs were established and conducted to characterize the values of $K_0$ for different degrees of saturation ($S_r$) and relative densities. The effects of freezing and thawing on $K_0$ were significant for the fully saturated condition of $S_r=100\%$, whereas they were negligible for partially saturated or unsaturated conditions. For FR condition, the values of $K_0$ were low during the early loading stage and increased gradually as ${\sigma }_v^{\prime }$ increased due to the breakage of pore ice. The lower $K_0$ values for FR condition were more significant for higher $S_r$. After thawing, a net volume increase was observed for $S_r=100\%$, thereby an increase in $K_0$ took place. This phenomenon was suggested as an important aspect for the stability of retaining structures during thawing periods. The computerized tomography images and the shear wave velocities for UF and TH conditions confirmed the effect of $S_r$ on $K_0$. A $K_0$ estimation method considering the effect of freezing and thawing was proposed, showing an improved prediction of $K_0$.

This study presents a comprehensive investigation into the monotonic, cyclic, and postcyclic response of a medium stiff, plastic, lightly-overconsolidated, alluvial clayey silt deposit. The laboratory investigation is performed on natural, intact, and reconstituted specimens derived from high-quality samples to establish the role of the stress history and soil fabric on cyclic resistance, excess pore pressure generation, and postcyclic behavior. The results of cone penetration tests (CPTs), vane shear tests (VSTs), and downhole and crosshole geophysical tests are used to compare the monotonic, stress-controlled direct simple shear (DSS) test data and interpret differences in the cyclic response of natural and reconstituted specimens. Constant-volume, stress-controlled cyclic tests indicate that the naturally-overconsolidated (OC), artificially normally-consolidated (NC), and reconstituted OC silt exhibit the cyclic mobility-type failure mechanism. The cyclic resistance of the intact, natural silt was determined to be sensitive to overconsolidation, with greater pore pressure generation potential observed in the NC silt for the same loading conditions. Despite significantly denser conditions, the reconstituted OC silt specimens exhibited a cyclic resistance approximately 44% lower than the natural, intact OC specimens at the cyclic resistance ratio corresponding to 15 cycles of loading. The magnitudes of postcyclic volumetric strain in the natural, intact specimens were independent of the overconsolidation ratio (OCR) and depended on the maximum excess pore pressure generated during cyclic shear. Strain-controlled cyclic DSS tests were used to identify the cyclic threshold shear strain, ${\gamma }_{tp}$, to generate residual excess pore pressure, which is approximately 0.012% in the naturally OC silt, and the cyclic degradation indices for various shear strain amplitudes. The various stress- and strain-controlled cyclic response of this natural silt deposit are compared to similar plastic soils reported in the literature to provide a well-characterized soil deposit for reference by geotechnical practitioners.

Engineers have long integrated predicted volumetric strains in liquefied soils to predict free-field, postliquefaction settlements. These free-field settlements have commonly been used to define a lower bound estimate of settlements beneath structures or embankments. Semiempirical methods for predicting volumetric strains during liquefaction require characterization of the seismic loading impacting the soil. The way in which probabilistic estimates of seismic loading is characterized can significantly impact the predicted postliquefaction settlements. This study evaluates three potential approaches for characterizing probabilistic seismic loading and their impact on the computed free-field, postliquefaction volumetric settlement hazard. The pseudo-probabilistic approach, which is arguably the most common approach in engineering practice in the US, shows considerable bias in its predicted volumetric settlements, including an average underprediction of settlement equal to 38% when compared to a fully probabilistic approach at a return period of 2,475 years. A simple linear relationship to correct pseudo-probabilistic estimates of free-field, postliquefaction volumetric settlements at return periods of 475 and 2,475 years for two common semiempirical volumetric strain models is presented.

Pipe piles are increasingly used to support embankments of high-speed railways. This study presents a series of static and cyclic axial loading tests on two industrial prestressed-concrete pipe piles installed on a field site with soft clay in Huzhou, southeastern China. The 29.5-m-long preinstrumented piles were subjected to different loading combinations of static load ($5–400\mathrm{kN}$) and cyclic load ($50–400\mathrm{kN}$) to maximum numbers of 50,000 cycles. The geotechnical details of the testing site and the instrumentations were explored before characterizing the accumulated displacement at the pile head. Preliminary results reveal that pile compressive capacities increased significantly after cyclic loadings. The cyclic behaviors of the pile-head displacement are classified into three patterns: quickly stabilized, progressively developing, and dramatically failed. Correspondingly, cyclic stability criteria are identified as stable zone, metastable zone, and unstable zone. Stable utilization line, critical utilization line, and ultimate utilization line are proposed to divide these three zones. For some cases under the same total load, the accumulated settlements of the piles are, however, different, depending largely on the cyclic loading level.

An extensive experimental program of constant-volume (undrained) cyclic simple shear tests was undertaken on Ticino, Italy, sand with different contents of nonplastic fines, ranging from 0% to 40%. The samples were reconstituted by moist tamping and tested with different initial states, including void ratios and effective vertical stresses. Test results confirmed that the concept of equivalent granular void ratio $e^{*}$ is appropriate for the interpretation of the undrained cyclic behavior of sand with different amounts of fines up to the limiting fines content. Because a single trend for critical state (CS) data points was observed in the $e^{*}\text{-}\log (p^{\prime })$ plane (EG-CSL) for different amounts of fines, the cyclic simple shear test results were analyzed within a unified critical state soil mechanics (CSSM) framework in terms of an alternative state parameter, ${\mathrm{\Psi }}^{*}$. A unique correlation between undrained cyclic strength (CRR) and ${\mathrm{\Psi }}^{*}$ was found, irrespective of the fines content and initial state. Although a correlation between the cyclic resistance ratio and the conventional state parameter $\mathrm{\Psi }$ works as well, the procedure based on ${\mathrm{\Psi }}^{*}$ has the advantage that the cyclic behavior of a certain sand with different contents of non plastic fines is described by a single reference curve (EG-CSL). In contrast to previous investigations in the literature, which mainly used triaxial tests, the CRR-${\mathrm{\Psi }}^{*}$ correlation proposed in the present study is based on cyclic simple shear tests, which better represent the real ground conditions under seismic loading.

Excavations near an existing tunnel are frequently encountered in underground construction. The excavation–soil–tunnel interaction mechanism is not yet fully understood, notably the long-term behavior. This study carried out three-dimensional centrifuge test of an existing tunnel subjected to a nearby excavation in saturated kaolin clay. Emphasis is put on the long-term behaviors of the ground and tunnel regarding the undrained shear strength, excess pore-water pressure, horizontal earth pressure, compression behavior, settlement, and bending moment. T-bar penetration tests showed that the retaining wall installation leads to a soil strength increase, whereas the excavation causes a reduction. Excess pore-water pressure below the excavation base gradually dissipates to zero. In comparison, the excess pore-water pressures above the tunnel crown and nearby the right springline (close to the excavation) experience continuous increases, although those close to the left springline (away from the excavation) exhibit a different pattern. Reduction amplitudes in horizontal earth pressures around the tunnel differ in both burial depth and horizontal relative position. The lateral earth pressure coefficient nearby the right springline decreases, and the stable value is higher than the Rankine’s active earth pressure coefficient. The compressibility of soils at the left side of the excavation (tunnel exists) was lower than at its symmetrical side, and the ground surface settlement is compared with empirical estimates concerning ground conditions and excavation geometries. Long-term responses show that the excavation base exposure leads to continuous and appreciable increases in ground settlement, tunnel settlement, and bending moment.

This paper describes the material properties along with the inverse limit-equilibrium and permanent deformation analyses used to investigate the 2015 reinforced slope failure at the Yeager Airport near Charleston, West Virginia. Inverse two-dimensional (2D) limit-equilibrium analyses were first performed to evaluate laboratory-derived strength parameters, slope geometry, and soil reinforcement configuration that would reproduce the observed critical failure surface. Because of the shape of the reinforced soil slope (RSS) (outside radius), the impact of the direction of the uniaxial geogrid reinforcement, varying from parallel to almost perpendicular to the direction of sliding, was analyzed using a three-dimensional (3D) limit-equilibrium analysis. Finite-difference permanent deformation analyses were also conducted to understand the internal stresses and deformations of the RSS prior to failure and kinematics of the slope failure. The results of these various analyses are consistent with postfailure field observations and demonstrate the value of performing multiple types of analyses, e.g., 2D and 3D limit-equilibrium and permanent deformation analyses, when analyzing a complex slope failure.

Because only limited research has been done on base geometry effects on pile base resistance, design methods used in practice for full-displacement, partial-displacement, and nondisplacement piles in sand do not consider these effects on base resistance calculations. In this paper, the effect of base geometry on pile resistance is studied by performing a series of tests in a half-cylindrical calibration chamber that allows observation of the sand domain through the chamber symmetry plane during installation and loading. Two model piles, one with a flat base and one with a conical base with an apex angle of 60°, were installed and load-tested in dense and medium-dense sand samples. Digital images taken during the loading of the model piles were processed using the digital image correlation (DIC) technique. Detailed displacement and strain fields obtained with the DIC technique and base resistance measurements demonstrate that the addition of a conical base to a pile changes the deformation pattern in the sand below the base during loading. Furthermore, the base resistances of both the jacked and preinstalled model piles with a conical base were less than those for the piles with a flat base for the same conditions, with the corresponding ratio ranging from 0.64 to 0.78 at relative displacement levels less than or equal to 10% of the diameter $B$ of the model pile and from 0.70 to 0.84 at greater relative displacements. The results have implications also for the use of cone resistance from the cone penetration test (CPT) in the pile base resistance estimation.

The objective of this study was to evaluate variability in reinforcement fibers, peel strength, and internal shear strength of needle-punched reinforced geosynthetic clay liners (GCLs). A methodology was proposed to characterize the reinforcement fibers via the number of reinforcement fiber bundles per specimen length, number of monofilament fibers per fiber bundle, and fiber diameter. Variability in reinforcement fibers was evaluated for six GCLs using the proposed methodology. Peel strength and internal direct shear tests were performed on a single GCL to assess the effect of reinforcement fiber variability on shear strength. No pattern was observed in the spatial variability of peel strength along the machine or cross-machine directions. However, peak internal shear strength correlated with manufacturer reported peel strength, and this correlation improved when peel strength was computed from specimens located adjacent and in machine direction relative to direct shear specimens.

It is generally accepted that the use of void ratio and bulk density as measures of soil state have limitations in the case of gap-graded soils because the finer grains may not transmit stress. However, no research has explored systematically whether this issue also emerges for soils with continuous gradings. Building on a number of experimental and discrete-element method (DEM) studies that have considered the idea of an effective void ratio for gap-graded or bimodal soils, this paper extended consideration of this concept to a broader range of particle-size distributions. By exploiting high-performance computers, this study considered a range of ideal isotropically compressed samples of spherical particles with linear, fractal, and gap-graded (bimodal and trimodal) particle-size distributions. The materials’ initial packing densities were controlled by varying the interparticle coefficient of friction. The results showed that even for soils with continuous particle-size distributions, a significant proportion of the finer particles may not transmit stress, i.e., they may be inactive. Drawing on ideas put forward in relation to gap-graded soils, both a mechanical void ratio and mechanical bulk density that consider the inactive grains as part of the void space were determined. Even for the linear and fractal gradings considered here, the difference between the conventional measures and the mechanical measures was finite and density dependent. The difference was measurably larger in the looser samples considered. These data highlight a conceptual/fundamental limitation of using the global void ratio as a measure of state in expressions to predict granular material behavior.

Deep mixing columns commonly are used to reduce settlement under road and railway embankments on soft soils. However, the structural behavior of the soil volume improved with columns is difficult to predict, due to the existence of considerable uncertainties in the mixing process and the structural interaction between the columns and the untreated soil. This paper probabilistically investigated two serviceability limit states of deep mixing columns from a system reliability perspective. A design framework employing the observational method is proposed that considers allowable residual settlements, excessive settlement from column yielding, and the curing time of the columns. The design framework facilitates an effective reduction of the geotechnical uncertainty during construction and promotes risk-aware decision-making during both design and construction of the embankment.

Induced partial saturation (IPS) is a novel method to suppress the generation of excess pore-water pressure and increase the liquefaction resistance of loose granular soils. Mechanical benefits associated with IPS are linked to the persistence of entrapped bubbles. Civil infrastructure operates for decades, often longer than a century, and thus the longevity of gas is a salient consideration for adoption of IPS in practice. Modeling the physical and chemical mechanisms that influence the persistence of entrapped bubbles is a practical avenue to address gas durability on these time scales, a limitation of physical experiments. The governing aqueous-phase advection-diffusion processes and interphase gas kinetics associated with bubble dissolution are simulated in a finite-difference numerical framework, validated with elemental and bench-scale experiments, and then extended to address soil resaturation rates under different subsurface conditions. The study demonstrates that emplaced gas is durable to the extent where diffusion-induced and groundwater seepage-induced dissolution should not discourage advancement of IPS, but will not remain indefinitely. Potential solutions to mitigate the decay of a gassy soil layer are discussed.

Microbially induced calcite precipitation (MICP) was used to treat several sandboxes filled with naturally occurring beach sand collected from Atlantic Beach, Florida. A surface-spray/percolation technique was used to treat these sandboxes where a relatively high-concentration bacteria solution and high-concentration calcium chloride/urea solutions were applied directly to the boxes’ surfaces. Several different treatment combinations were tested whereby bacterial optical density, bacteria/urea/calcium chloride volume relative to pore volume, and bacteria/urea/calcium chloride ratio were manipulated. Treated sandboxes were tested for erodibility using a pocket erodometer. In addition, sandboxes were dissected after erosion testing to examine crust depth. Results showed that higher optical densities, higher bacteria quantities relative to void volume, and higher bacteria quantities relative to urea led to lower erodibility and greater crust depth. When MICP constituent quantities were maximized to give the best erosion/crust-depth results, erodibility improvements began to approach what may be considered adequate erosive resistance. Further investigation is required to better classify this behavior more quantitatively.

Monopile foundations supporting offshore wind turbines are exposed to cyclic lateral loading, which can cause accumulated pile displacement or rotation and evolution of the dynamic response. To inform the development of improved design methods, the monopile’s response to cyclic lateral loading has been explored through small-scale physical modeling at $1g$ and in the centrifuge, as well as at large-scale in the field. There are advantages and disadvantages to each physical modeling technique, and the response may be most efficiently explored through a combination of modeling techniques. However, stress levels vary significantly between these techniques, and only centrifuge testing can simulate full-scale stress levels. This paper explores the effect of stress level on the response of a monopile foundation in dry sand to monotonic, unidirectional cyclic and multidirectional cyclic lateral loading with small-scale tests at $1g$ and in the centrifuge at $9g$ and $80g$. With an identical setup at each $g\text{-}\mathrm{level}$, stress-level effects were isolated. Qualitatively, the responses are similar across the stress levels, but some important quantitative differences are revealed. In particular, the rate of accumulation of pile displacement and the rate of change of secant stiffness under cyclic loading are found to reduce with increasing stress level. The results highlight the need to simulate full-scale stress levels to thoroughly understand foundation behavior, but also demonstrate the qualitative insight that can be gained through $1g$ physical modeling. The data and trends presented in this paper provide input for the modeling of monopile responses at different stress levels.

In recent years, the random field finite-element method (FEM) has been used increasingly in geotechnical engineering to carry out analyses that account for the inherent spatial variability in the physical and mechanical properties of both natural and treated soils. However, this method, which usually is performed in tandem with Monte Carlo simulation (MCS), requires significantly greater computational resources than deterministic finite-element analysis. Metamodeling is one of the techniques commonly adopted to alleviate the computational burden. This paper proposes a novel and computationally efficient metamodeling technique that involves the use of convolutional neural networks (CNNs) to perform random field finite-element analyses. CNNs, which treat random fields as images, are capable of outputting FEM predicted quantities with learned high-level features that contain information about the random variabilities in both spatial distribution and intensity. CNNs, after being trained with sufficient random field samples, could be used as a metamodel to replace the expensive random field finite-element simulations for all subsequent calculations. The validity of the proposed approach was illustrated using a synthetic excavation problem and a synthetic surface footing problem. The good agreement between the CNN outputs and the FEM predictions demonstrated the promising potential of using CNNs as a metamodel for reliability analysis in spatially variable soils.

Boulders entrained in debris flow induce high impact forces on a rigid barrier. In current design practice, the concentrated load from boulders is estimated using the Hertz equation with a load reduction factor ($K_c$). Separately, the distributed load from the debris is estimated using the hydrodynamic equation. The existing design practice is simply adding the estimated loads using the two equations. The interaction between debris flow and boulders during the impact process is neglected. In this study, physical tests were conducted using a newly developed $28\text{-}\mathrm{m}\text{-}\mathrm{long}$ flume to shed light on the impact dynamics of debris flows with and without boulders on an instrumented rigid barrier. Contrary to existing design practice in which the boulder and debris impact loads are added together, the debris provided a cushioning effect to attenuate the impact force of the boulders. This cushioning effect was governed by a reflection wave with a length scale $L_R/d$ (where $d$ = boulder diameter), which serves to cushion thickness on impact. $L_R/d$ from 0.4 to 2.0 can reduce the impact load by up to 80% compared with existing design practice ($K_c=0.1$).

This research investigates the influence of pile spacing on the dynamic behavior of pile groups by performing a series of specifically designed dynamic centrifuge experiments on pile foundations embedded in a two-layered soil profile. A single pile and two $3\times 1$ row pile groups with different pile spacing were used as model pile foundations, and the soil models consisted of a soft clay underlain by dense sand. The influence of earthquake frequency on the dynamic behavior of two-layered soils is discussed using the centrifuge data and 1D site response analysis from DEEPSOIL. Further, the results of these centrifuge tests agreed with the conviction that the group effects will be diminished with the increase in pile-to-pile spacing in a pile group due to reduced pile-soil-pile interaction. However, these reduced pile group effects can lead to larger kinematic pile bending moments in the widely spaced pile group compared with a closely spaced pile group. Moreover, the single pile always has larger bending moments than both the tested pile groups—an exception to this is when there is a significant phase difference between the kinematic and inertial loads for a single pile but not for the widely spaced pile group. The influence of pile spacing on the shadowing effects and location of peak bending moments in the piles of a group are also discussed in this paper. Lastly, an attempt is made to evaluate the individual contribution of inertial and kinematic loads for the seismic design of pile foundations considering soil-pile-structure interaction effects.

Dr. Rudolph Bonaparte delivered the 54th Terzaghi Lecture on March 8th, 2018, at IFCEE 2018 held in Orlando, Florida. This introduction briefly reviews Dr. Bonaparte’s contributions to geotechnical engineering and discusses some of the highlights of the corresponding paper.