Journal of Geotechnical and Geoenvironmental Engineering

A probabilistic cone penetration test (CPT) based liquefaction triggering procedure for cohesionless soils is derived using a maximum likelihood method with an updated case history database. The liquefaction analysis framework includes revised relationships for the magnitude scaling factor (MSF) and for estimating fines contents from CPT data when laboratory test data are not available. The updated case history database and methodology for developing the liquefaction correlation are described. Measurement and estimation uncertainties, the potential effects of false positives and false negatives in the case history database, and the effects of the choice-based sampling bias in the case history database are accounted for. Sensitivity analyses showed that the position of the most likely triggering curve and the magnitude of the total error term are reasonably well constrained by the data. The sensitivity study provides reasonable bounds on the effects of different interpretations, from which probabilistic and deterministic relationships for practice are recommended.

The mechanical behavior of sands treated using microbially induced carbonate precipitation (MICP) has been investigated at the macroscale and the microscale. Triaxial and confined compression tests with embedded shear and compression wave (S-wave and P-wave) sensors were conducted on two MICP-treated sands, Ottawa $50/70$ and $20/30$ silica sands. Triaxial compression tests were conducted at three different confining pressures (25, 50, and 100 kPa). Tests were also performed at calcium chloride (${\mathrm{CaCl}}_2$) concentrations of 0.1 and 0.3 M, resulting in specimens with average calcium carbonate (${\mathrm{CaCO}}_3$) content ranging from 1.5 to 2.5% for the $50/70$ sand and from 1 to 1.6% for the $20/30$ sand. In contrast to previous research, the results of triaxial tests presented in this paper show an increase of the soil strength even at 1% calcium carbonate content. After the tests, samples taken from the specimens were utilized to measure the ${\mathrm{CaCO}}_3$ content and to perform analysis using scanning electron microscopy (SEM) and energy dispersive X-ray spectroscopy (EDS). The SEM and EDS images were used to assess the morphology and spatial distributions of ${\mathrm{CaCO}}_3$ at the microscale.

Installing stone columns is a convenient method of soft ground improvement. Although several analytical and numerical solutions are available to predict the performance of stone column–improved soft ground, these models are incapable of capturing the influence of cyclic loading on transport corridors, such as highways and railways. The authors developed a novel finite-difference model adopting modified Cam clay theory to analyze the response of stone column–reinforced soft soil under static and cyclic loadings. Apart from predicting excess pore water pressure dissipation and the resulting settlement, the model also captures the effect of cyclic loading by considering yield surface contraction. The solutions developed were validated using available field observations and laboratory test results. The model was successfully applied to a selected case study at the Australian National Field Testing Facility at Ballina, New South Wales, Australia. The limitations of the proposed model are highlighted as well.

Seismic design of geotechnical structures often requires estimates of how shearing resistance of liquefiable soil is reduced to its minimum value (residual strength, $s_r$) as pore pressures build up and is subsequently regained as pore pressures dissipate. It was envisioned that the shear strength of liquefying sands could be measured in-flight in a seismic geotechnical centrifuge model by pulling thin coupons (plates) horizontally through the sand models before, during, and after shaking to simulate the large strains and strain rates associated with liquefaction flow failures. This paper presents the results of seismic centrifuge tests that were used to make such measurements. Companion ring shear (RS) tests also are described. Although centrifuge and RS residual strengths were generally similar and increased with relative density, the centrifuge $s_r$ and $s_r/{\sigma }_{vo}^{\prime }$ ($s_r$ divided by preshaking effective vertical stress) increased significantly with small changes in relative density, while the RS test $s_r$ and $s_r/{\sigma }_{vo}^{\prime }$ increased only slightly with changes in relative density. Furthermore, many measured $s_r$ and $s_r/{\sigma }_{vo}^{\prime }$ values fell below penetration test-based design curves used in practice. Dilative tendency during shearing is believed to have resulted in a dramatic increase in $s_r$ and $s_r/{\sigma }_{vo}^{\prime }$ in the centrifuge tests that consisted of soil dense of the critical state. In contrast, RS tests likely resulted in shear band development, where intense shearing and particle damage occurs, leading to suppressed soil dilative tendency and divergence from the centrifuge test results at higher relative densities. The centrifuge tests provided evidence that, as expected, postliquefaction strength recovery is linearly proportional to effective stress as excess pore pressures dissipate.

This paper presents a representative engineering case, the Shiziping dam, to illustrate deformation controls of highrise earth core rockfill dams in seismic areas. Techniques for controlling dam deformation, including structure design, foundation treatment, material properties, and construction are discussed in detail. Field monitoring data and deformation characteristics were reviewed to verify control effectiveness. The deformation during the construction and initial impoundment period accounted for 70–85% of the overall deformation; the deformation was primarily influenced by weight, wetting behavior, and creep. Creep stabilized 7 or 8 months after the completion of construction. The total deformation is caused by fill and foundation materials. Uneven deformation is mainly caused by foundation shape, foundation compression symmetry, and uniformity of fill materials. Uneven deformation may be improved by using reasonable dam zone and fill material designs and appropriately increasing core wall compaction.

A reliable estimate of secant undrained Young’s modulus ($E_u$) is necessary to quantify the undrained settlement of structures underlain by fine-grained soils. A large number of ${\mathrm{CK}}_0\mathrm{UC}$ triaxial tests have been carried out on eight resedimented clays to establish the effects of soil composition, applied shear stress, overconsolidation, and vertical consolidation stress on the magnitude of $E_u$. These tests were performed at effective stresses ranging from 0.02 to 100 MPa. Low plasticity soils tend to have a higher $E_u$ compared with high plasticity soils. The effect of composition diminishes as the applied shear stress increases. Overconsolidated (OC) soil has a higher undrained modulus than normally consolidated (NC) soil, though a change in the overconsolidation ratio of OC soil does not have a significant effect on $E_u$. Regression analyses performed on the experimental data define relationships for $E_u$ as a function of stress level for both NC and OC soil.

Granular column ground-improvement methods are widely used to improve bearing capacity and provide a drainage path. However, the behavior of granular columns depends on the confinement provided by the surrounding soil, which limits their use in poor soils. A new ground-improvement method is proposed using pervious concrete piles to provide high permeability while also providing higher stiffness and strength, which are independent of surrounding soil confinement. Building on prior research on the behavior of vertically loaded pervious concrete piles and granular columns, this paper investigates the behavior of laterally-loaded pervious concrete piles and the effects of installation on their response. Two fully-instrumented lateral load tests were conducted on a precast and cast-in-place pile using different installation methods. Advanced sensors measured the soil–structure interaction during installation and under lateral loading. Test results confirmed that laterally-loaded pervious concrete ground-improvement piles behave as flexible laterally-loaded piles, while still providing permeability similar to granular columns. Although the ultimate load capacity of the two piles was not affected by installation, the cast-in-place pile had 55% of the displacement of the precast pile at ultimate load, which indicates soil densification and lateral compression due to installation. The directly-measured $p\text{-}y$ curves for the cast-in-place pile also better matched the existing procedures developed for driven piles and showed a higher stiffness and ultimate soil reaction than the precast pile. Results also determined that the zone of soil affected by installation extended to 2.5 D from the pile surface.

This paper analyzes the possibility of reducing the uncertainty of the cyclic resistance ratio (CRR) estimates by incorporating stress history into the liquefaction correlations. A way of obtaining this objective stems from the combination of two well-recognized notions: (1) sensitivity of the flat dilatometer test (DMT) parameter $K_D$ to stress history, and (2) necessity of stress history information to obtain better estimates of the liquefaction resistance. The main aim of this paper is to develop a framework providing CRR estimates based not on the one-to-one correlations $\mathrm{CRR}\text{-}{Q}_{cn}$ or $\mathrm{CRR}\text{-}{K}_D$, but on a correlation based at the same time on both $Q_{cn}$ and $K_D$. A $Q_{cn}\text{-}{K}_D\text{-}\mathrm{CRR}$ correlation has been constructed by combining the current $\mathrm{CRR}\text{-}{Q}_{cn}$ and $\mathrm{CRR}\text{-}{K}_D$ correlations. It is expectable that an estimate based at the same time on two measured parameters is more accurate than estimates based on just one parameter. 'A chart is presented providing estimates of CRR based at the same time on both $Q_{cn}$ and $K_D$.

Nearly all soils are comprised of mixtures of coarse and fine particles. Behavior under mechanical and thermal loading of soil is strongly influenced, and in some cases governed, by the ratio of coarse to fine particles. A better understanding of the fundamental behavior of soil mixtures will provide insight to design decisions for new and emerging geotechnologies. In this work, behavioral threshold fines fractions were identified by experimental methods, where the threshold was defined as the point where changes in the coarse/fine mixture ratio result in abrupt behavior changes. Binary mixtures of sand and kaolinite clay ranging from 0 to 100% fines content were subjected to consistency and undrained shear strength testing with the fall-cone apparatus, compressibility tests using an oedometric cell, thermal conductivity tests with a thermal needle probe, and stress-strain-strength testing in undrained triaxial shear. Results indicate that behavioral thresholds exist at a critical fines content where a minimum void ratio occurs and at a percolation threshold where continuous force chains are present. The behavior changes are explained using theories of effective properties and percolation. Interpretations of these results lead to a clearer understanding of soil behavior.

Tests were conducted to assess how permeation with actual recirculated leachate from municipal solid waste (MSW) landfills affects the hydraulic conductivity and exchange complex of geosynthetic clay liners (GCLs). Long-term hydraulic conductivity tests were conducted ($\sim 1\mathrm{t}\mathrm{o}5\mathrm{years}$) on GCL specimens with conventional sodium bentonite using seven recirculated MSW leachates and one conventional (nonrecirculated) MSW leachate as permeant liquids. Effects of stress and temperature were also assessed. Hydraulic termination criteria were achieved for all tests, and chemical termination criteria were achieved for nearly all. The long-term hydraulic conductivity of the GCL specimens to recirculated leachates ($1.0–2.0\times {10}^{11}\mathrm{m}/\mathrm{s}$) was slightly less than the long-term hydraulic conductivity to the conventional leachate ($2.0\times {10}^{-11}\mathrm{m}/\mathrm{s}$). Furthermore, the ratio of the hydraulic conductivity to recirculated MSW leachate to the hydraulic conductivity to DI water for GCL specimens (1.4–2.9) fell within the range reported in other studies where GCLs were permeated with conventional MSW leachate (0.05–2.6). The long-term hydraulic conductivities to recirculated and conventional MSW leachate at 70 kPa were low ($<2.0\times {10}^{-11}\mathrm{m}/\mathrm{s}$) because of the relatively low ionic strength of the leachates ($<249\mathrm{mM}$) and the preponderance of monovalent cations in the leachate. Geosynthetic clay liner specimens tested at higher stress (70 versus 20 kPa) had lower hydraulic conductivity (2.3–3.0 times lower), which slowed mass delivery of exchangeable cations and thereby delayed exchange processes and corresponding changes in hydraulic conductivity. Gas production from biological activity can be minimized during leachate compatibility testing by reducing the temperature to 4°C without compromising cation exchange processes and related effects on hydraulic conductivity. Leakage data from two landfills showed that leakage rates from composite liners with a GCL can be very low ($<0.34\mathrm{L}/{\mathrm{m}}^2\text{-}\mathrm{year}$) when the landfill is operated with leachate recirculation.

The postcyclic undrained monotonic shear behavior of Mississippi River Valley (MRV) silt is investigated by carrying out triaxial tests under various initial consolidation conditions. Results show that the variation of cyclic shear strength of MRV silt against initial overconsolidation ratio (OCR) and mean effective consolidation pressure (${\sigma }_c^{\prime }$) was consistent with previous studies, i.e., it increased with both increase in OCR and decrease in ${\sigma }_c^{\prime }$. The cyclic stress ratio (CSR) had no apparent effect on postcyclic shear behavior for MRV silt. At higher OCRs, the postcyclic reconsolidation behavior was more similar to recompression. Unlike the specimens not previously subjected to cyclic loading, the initial OCR did not influence effective stress path of the postcyclic reconsolidated specimens at the early stages of shearing. The normalized undrained shear strength ($S_{u,OC}/S_{u,NC}$) of MRV silt reconsolidated after cyclic loading increased with OCR. Stress-strain curves of MRV silt reconsolidated after cyclic loading can be normalized by ${\sigma }_c^{\prime }$ as those of the specimens not previously subjected to cyclic loading. Without reconsolidation, the liquefied specimens tend to dilate at the early stage of postcyclic undrained monotonic shearing phase. The undrained shear strength of the specimens not reconsolidated after cyclic loading did not approach to zero, and it increased with an increase in OCR. The undrained shear strength of postcyclic MRV silt can be predicted based on the equation $S_u=M/2\times p_0^{\prime }\times {10}^{-\psi /\lambda }$. However, cyclic loading changes the critical state line (CSL) of MRV silt, and the prediction will be better if the state parameter is computed based on the postcyclic CSL.

An alternative approach to replacing corrugated steel culverts is to insert a new pipe inside the existing culvert and grout the space between them, a process known as sliplining. Though sliplining is a preferred rehabilitation approach among departments of transportation, very little research has been done to investigate the capacity enhancement provided by sliplining and how the load is shared between components in a sliplined culvert (i.e., the existing pipe, the liner, the grout, and the surrounding soil). A series of experiments was conducted on a deteriorated corrugated steel culvert at two different burial depths (600 and 900 mm) under surface loading before and after the culvert was rehabilitated with a grouted high density polyethylene (HDPE) slipliner. The rehabilitated culvert was found to be considerably stiffer and the diameter changes under surface load were reduced by more than 90% compared with the unrehabilitated culvert suggesting that the negative arching between the pipe and the soil has increased. The strains in the existing pipe were also reduced by more than 70% with negligible strain being measured in the invert of the culvert where the corrosion was concentrated. The results also indicated that when a neat grout with a compressive strength of approximately 30 MPa is used (what could be considered to be high strength in comparison with low density grouts featuring entrained air), the existing pipe and the grout carry most of the load, and the liner’s main role is to improve the hydraulic conductivity. Additionally, because of the increased stiffness of the rehabilitated pipe, the surrounding soil carries less of the surface load applied above the pipe.

Understanding and assessing the spatial extent of liquefaction requires that the spatial dependence of soil properties to be taken into account. In this work, a cone penetration test (CPT)-based approach for the evaluation of liquefaction potential is presented where the soil spatial variability is explicitly considered through internally-consistent probabilistic models developed at multiple scales. The novelty of the proposed work comes from the integration of the classical empirically-developed liquefaction criteria with tools in geostatistics and novel multiscale random-field models. A unique feature of the proposed work is its ability to refine and obtain higher resolution random fields for soil properties in critical areas, such as those adjacent to important infrastructure or in areas with detailed small-scale field data. An illustrative example assessing the liquefaction potential at various shaking levels in the Marina District of San Francisco is used to demonstrate the capability of the proposed method.

Suffusion occurs when soil particles are loosened, detached, and transported away by seepage flow through a series of pores and pore constrictions. While traditional suffusion assessment methods are often based on particle size distribution analysis only, modern assessment methods are focused on constriction size distributions, which are derived from particle size distributions based on certain assumptions. This paper provides a new assessment method, which employs the probability of loose particles being transported to the next pore. The new approach can introduce the influence of an overlapping zone between the two fractions of loose particles and soil primary fabric. This overlapping zone was often overlooked in prior studies. The constriction size distribution of the primary fabric can be calculated approximately by two-dimensional or three-dimensional methods. The three-dimensional method used in the presented study even offers the opportunity to study the influence of particle arrangements on suffusion. The results of the new assessment methods show satisfactory agreement with experimental results.

A significant difference between predicted and measured installation resistance of stiffened suction caissons was identified due to the existing uncertainty regarding the mobilized soil-flow mechanisms. This paper describes an extensive investigation of stiffened-caisson penetration in nonhomogeneous clays undertaken through large deformation finite-element (LDFE) analysis to provide insight into the soil behavior during installation of a caisson. The soil-flow mechanisms around and between stiffeners, and inside and outside of the caisson, and the corresponding penetration resistances were presented from a parametric study, exploring a range of dimensionless parameters related to stiffened-caisson geometry, caisson roughness, and soil strength nonhomogeneity. The LDFE results were compared with centrifuge test data in terms of the soil-flow mechanisms and penetration resistance profile, with good agreement obtained. Three interesting features in the mobilized soil-flow mechanisms inside the caisson were observed: (1) soil started to infill the gap between the bottom two stiffeners when the bottom stiffener penetration reached $H_{\mathrm{r}}$ ($H_{\mathrm{r}}$ is the critical depth of rotational soil flow around bottom stiffener); (2) soil started to infill the gap between the second and third bottom stiffeners when the second stiffener penetration reached $H_{\mathrm{c}}$ ($H_{\mathrm{c}}$ is the critical depth of soil backflow into gaps above the 2nd stiffener); (3) profound soil heave due to significant inward flow and the presence of stiffeners and gap between stiffeners. Both $H_r$ and $H_{\mathrm{c}}$ were shown to be a function of the stiffener-width to caisson-diameter ratio, normalized soil strength at the mudline, and soil strength nonhomogeneity. To predict the penetration resistance profile in the field, a rational analytical penetration model, based on the revealed soil-flow mechanisms, was proposed. The LDFE data were used to calibrate the model. Very good agreements were obtained when the proposed model was validated against measured data from field installation and centrifuge tests.

The interdependency of microsoil fabric-related properties and swell/shrink behavior of expansive soils needs to be identified to achieve a more thorough and accurate prediction of heave of expansive soils. In a recent research study, two microsize-related properties of a soil—pore size distribution and specific surface area—were identified and examined in an effort to address their synergistic effects on swelling behavior of soils. To study this aspect in detail, eight natural expansive clayey soils from known expansive soil regions were sampled and studied by using two test methods: conventional one-dimensional vertical swell tests and novel three-dimensional volumetric swell strain tests. Microinternal structural details, including pore void distribution, were obtained from mercury intrusion porosimetry (MIP) studies on compacted soil specimens. Specific surface area (SSA) details of the same soils were determined using the chemical ethylene glycol monoethyl ether (EGME) method. Attempts were made to predict macroswell properties, using both pore and surface area properties. Modeling analyses and results showed that the swelling behavior of the clays was dependent on a newly introduced parameter that accounted for both clay mineralogy constituents and internal pore distribution of the soils. The new swell property models showed accurate predictions of measured swell strains of the present soils, depicting the importance of including mineralogy and pore fabric details in a given soil.

A unique simulation was developed within the large-scale atomic/molecular massively parallel simulator (LAMMPS) open source molecular dynamics environment to model kaolinite clay during one-dimensional consolidation. The simulation was composed of a coarse-grained representation of the mineral with long-range and short-range interaction potentials. Additionally, the simulation allows for the edge-to-edge and edge-to-face attractions to develop through use of two subparticle types. Periodic simulation boundaries were created and thus allowed for consolidation to proceed without any wall effects distorting the resulting fabric. By following the parallel processing paradigm of the LAMMPS code, the number of particles simulated was limited only by the hardware on which the simulation was run. The results of the simulation showed remarkable consistency with previous published studies over all the void ratios explored and the transition from overconsolidated to normally consolidated behavior was captured. Fabric analysis showed that reorientation occurred during unloading and subsequent loading and followed a nonlinear path in virgin compression. Additionally, the median spacing between particles was determined, and particle contact was found to coincide with a stiffness change during consolidation.

Cement stabilization is a well-established technique involving the addition of cement to increase shear strength of soft soils. Studies on the prediction of cement-treated slurry clay generally do not cover a comprehensive suite of experiments to develop a more in-depth strength-development model, and applicability is often limited to their specific database. This paper presents a large database of unconfined compressive strength tests of cement-treated marine clay over a wide range of mix ratios and curing ages. An improved predictive strength-development model is proposed that considers the effect of mix proportions and curing duration. The actual and predicted strength values are found to agree with a correlation coefficient of 0.997. The model is also shown to be applicable to several clay and cement types; the good fit to experimental data is indicated by correlation coefficients exceeding 0.95, which is superior to existing classic prediction models.

The conventional practice of calculating consolidation settlement in one-dimensional conditions involves dividing soft ground into multiple layers to enable an accurate estimation. However, the effect of layering on consolidation settlement has not been clearly verified. The authors propose an analytical solution to the case where the ground is divided infinitely for normally consolidated soils. The effect of error caused by layering in conventional practice is presented as well. Based on the suggested analytical solution, it is found that the conventional settlement calculation method always underestimates consolidation settlement. In addition, a representative depth for the conventional method in which the soft ground is not divided is suggested for calculating the same consolidation settlement as that obtained by the analytical solution.

Sands age and their engineering properties often improve significantly during performance life cycles and are of practical interest to engineers. Although various hypothesized aging mechanisms have been proposed, experimental evidences identifying the underlying mechanisms have not yet been provided. The experiment of this study, using a true triaxial apparatus equipped with the bender element and tactile pressure sensor, is therefore designed to complement this missing part. The experimental results show a distinct and continuous increase in the measured small-strain shear modulus of a dry, predominately quartz sand sample, subjected to a constant confining pressure for 30 days of aging, during which the measured volumetric strain is negligible. In addition, the measured contact normal forces (by the tactile pressure sensor) continuously redistribute and ultimately become uniform over time, i.e., the process of contact force homogenization. Sand particles in the weak force network gradually share the forces that are redistributed from the strong force network. These findings indicate that the weak contact force network in sand is gradually strengthened, thus reinforcing the soil structure and giving rise to a higher small-strain shear modulus during aging.