Journal of Geotechnical and Geoenvironmental Engineering

Most of the existing design methods for laterally loaded monopiles were developed for uniform soil profiles. The application of such design methods in cases involving layered sand profiles can lead to incorrect capacity estimations. In order to investigate the impact of the soil layering on the response of monopiles to lateral loads and to develop a design method that is applicable to layered soil profiles, a series of three-dimensional (3D) finite-element (FE) analyses were performed on laterally loaded monopiles embedded in layered sand profiles using an advanced two-surface-plasticity constitutive model. The analyses take into consideration a wide range of pile geometries (pile diameter, slenderness ratio, and wall thickness), load eccentricities, sand types and relative densities, and layered soil profiles. The study examines the change of the lateral capacity of a monopile in a two-layer sand profile as the top layer thickness varies, as well as the impact of the presence of a thin loose-sand layer in a dense-sand profile on the pile lateral capacity. Based on the results of these analyses, a set of design equations is proposed to estimate the lateral capacity and load-rotation response of monopiles in layered sand profiles. The proposed method accounts for the 3D pile–soil interaction and provides estimates of the lateral capacity that are in close agreement with those obtained from the 3D FE analyses.

Two centrifuge model tests were used to examine the effect of soil interlayering on the measured cone-penetration resistance in a layered soil model. The first centrifuge model included four soil profiles: two uniform loose and dense soil profiles and two two-layer sand profiles of loose over dense and dense over loose sands. The other centrifuge model involved a layer of loose or dense sand with varying thickness between overlying and underlying layers of low-plasticity clayey silt. Multiple cone-penetration soundings were performed using cone penetrometers with diameters of 6 and 10 mm. The results showed that the measured tip resistance in layered soils as well as the sensing and development distances depend on sand layer thickness and stiffness contrast between the subsequent soil layers. The results of these experiments were compared with the results obtained from analytical and numerical solutions. The results provide insights on the effect of thin-layer presence on cone tip measurements and an archived data set for evaluating design procedures and numerical analysis methods.

Repeated rapid drawdown (RDD) and rapid rise in water level during extreme events lead to the progressive development of plastic strain zones within the earth embankments with subtle, rather than obvious, visible signs of distress. The traditional approach within the framework of limit equilibrium does not account for accumulated permanent deformation with repeated hydraulic loading. Work presented herein is focused on quantifying the level of deviatoric strain, in terms of key surface deformation and distress level of earth embankment slopes, with repeated hydraulic loading. A simple linear relationship between the deviatoric strain and surface deformation at the toe of the slip surface is proposed as a function of the geometry of the slope for rotational sliding. This relationship is applied using the stress-strain data obtained from conventional triaxial testing and provides a simple means to estimate the ultimate performance limit state that corresponds to the onset of embankment slope instability. Results from a parametric study show good agreement between the numerical results and proposed analytical criterion. The proposed criterion is also compared with data from field cases reported in literature by others, and reasonably good agreement with onset of failure is obtained. Results from applying the proposed model to the case studies indicate the applicability of the proposed approach as a framework for various loading conditions, slope geometries, and material properties.

An earthquake of moment magnitude 5.4 occurred in Pohang, South Korea on November 15, 2017. This earthquake of moderate magnitude induced failures of structures and buildings and led to ground displacements. Approximately 600 sand boils were observed in the vicinity of the epicenter. We conduct 26 multichannel analysis of surface waves (MASW) tests and obtain three shear wave velocity ($V_S$) profiles from suspension Compression (P) and Shear (S) wave logging at liquefied and nonliquefied sites near the epicenter. The phase velocities at frequencies lower than 10 Hz at the liquefied sites were smaller than those at the nonliquefied sites. The cyclic stress ratios (CSRs), cyclic resistance ratios (CRRs), and factors of safety were calculated. The CRR curves for these approaches predicted sand boils in Pohang with an accuracy of 79.5%. The liquefaction potential index (LPI) was calculated at 29 sites and is geospatially interpolated to create maps near the epicenter. The LPI maps had a maximum accuracy of 77% when LPI thresholds were 8 and 2 for two approaches from the literature. The high accuracies from both local and regional liquefaction evaluations demonstrate the applicability of the existing approaches for moderate earthquakes in South Korea.

A numerical model, called HT2, is presented for one-dimensional heat transfer in saturated incompressible layered soil with steady fluid flow, effective porosity, and possible local thermal nonequilibrium (LTNE) between solid and fluid phases. The model uses a series-parallel approach and accounts for advection, conduction, linear and nonlinear thermal dispersion, and interstitial heat transfer. The key to HT2 is the definition of separate columns for the solid matrix (SM) and mobile pore fluid (MPF). The SM column includes the solid phase and immobile pore fluid and consists of fixed elements. The MPF column consists of moving elements and uses Lagrangian element-tracking to follow the fluid motion, which reduces numerical dispersion and simplifies heat transfer within the column to that of dispersive flux between contiguous elements. The model uses an interstitial heat transfer coefficient to calculate kinetic heat transfer between SM and MPF elements. The development of HT2 is first presented, followed by verification checks and a limited parametric study. Numerical simulations indicate that larger particle size and higher fluid discharge velocity yields greater local temperature disequilibrium between solid and fluid phases during heat transfer through saturated soil.

This study developed a new method to simultaneously measure gas dispersion coefficient and gas coefficient of permeability (GCP) of an unsaturated soils with due theoretical consideration of steady-state gas transport mechanisms. Correspondingly, a soil column apparatus was developed. Based on the proposed method, gas mixture containing air (with a volumetric concentration $>90\%$) and tracer gas were applied to the column base. The gas dispersion coefficient was estimated based on a measured steady-state tracer gas concentration gradient and a dispersive flux of the tracer gas, the latter of which was determined based on mass conservation of the tracer gas. On the other hand, the GCP was determined based on the total flux of gas mixture and gas pressure gradient following Darcy’s law. The new method has the following advantages over existing methods: (1) has an ability to determine gas dispersion coefficient and GCP at any soil depth rather than an average value of a test region; (2) measures the two transport properties at varying soil moisture conditions rather than being limited by uniform moisture distribution; and (3) requires only a simple hand-calculation method to estimate the gas dispersion coefficient without the need to apply nonlinear regression on a gas breakthrough curve. The proposed method to determine GCP and gas dispersion coefficient of a silty sand was verified by independent element tests.

For the design of ground-source heat-pump systems, the local subsoil is an invariant factor. To improve the evaluation of the local heat exchange capability, significant efforts recently have been devoted to identifying the ground thermal conductivity vertical profile. In recent years, an innovative method using hybrid optic fiber cables inserted into the ground has been developed. The technique relies on copper wires that thermally stimulate the ground. Optical fibers measure the temperature variation over time all along the cable at a high spatial and temporal resolution. In this work, the hybrid cable was grouted into a 125-m well located in the Po Plain in Northern Italy. The provided core defined the geological environment as a continuous succession of unconsolidated alluvial deposits of very limited thickness, grouped in 15 different granulometric units. Three enhanced thermal response test (ETRT) data sets were acquired in different seasons; for 5 days of heating followed by 5 days of recovery, the soil temperature was recorded continuously along the well, with a spatial resolution of 1 m. A new approach using a multiple linear regression is proposed to analyze the data sets to distinguish the thermal conductivity of each individual granulometric unit. The obtained thermal conductivity values were compared and discussed considering the standard thermal response test outputs and the thermal conductivity data obtained from direct measurements performed on the cores. The analytical method’s reliability stands due to the high repeatability of the obtained results, despite the increased complexity of the treated geological setting.

Water absorbing polymers (WAPs) are gaining attention in the field of geotechnical engineering, with application to bioengineered slopes, vegetation growth on landfill covers, and maintenance of green infrastructure in arid climate regions. WAP-amended soil is often exposed to alternate drying–wetting cycles, which influence its water retention characteristics. Determining the hysteresis in water retention characteristics of WAP-amended soils is crucial to evaluate their hydraulic performance with time. The present study proposes a methodology to measure the hysteresis in soil–water retention curve (SWRC) of WAP-amended soils subjected to alternate drying–wetting cycles. For this purpose, the drying–wetting SWRC of three different textured soils (sand, silt loam, and clay loam) were determined for varying WAP concentrations (0.1%, 0.2%, and 0.4% on weight-by-weight basis). Fourier transform infrared (FTIR) spectroscopy was used to confirm degradation in the polymer chain of the WAPs with the progressive wetting–drying cycles, which significantly affected the hysteresis phenomenon. A predictive model was proposed to estimate the wetting SWRC from the drying SWRC for WAP-amended soils, which is valid until the complete degradation of the WAPs. The proposed model was validated for two consecutive drying–wetting cycles for the selected soils with varying WAP concentrations, which showed its effectiveness for predicting hysteresis behavior in SWRC of WAP-amended soils. Further, the degree of hysteresis in SWRC of WAP-amended soil was quantified after each drying–wetting cycle. The experimental results revealed that the presence of WAP and its degradation with time significantly increased the degree of hysteresis in SWRC.

The performance of compacted soils that are used for the construction of railway embankments are impacted by weather-driven suction and water content fluctuations. Hence, the accurate measurement of the water retention properties of soil during cyclic loading is essential to evaluate the soil cyclic behavior under repeated train-induced loads. This paper presents a suction monitoring setup for soil cyclic triaxial testing and its application to evaluate soil water retention and volumetric response. The setup uses a high-capacity tensiometer to measure suction and on-sample displacement transducers to measure volume changes mounted at the midheight of the soil sample. Compacted soil samples were subjected to cyclic loads under different testing conditions: (1) testing a saturated sample under a free-to-drain condition showed the accumulation and dissipation of excess pore-water pressure during cyclic loading, (2) testing unsaturated samples under constant water content conditions demonstrated that the accumulated volumetric strain was smaller than the saturated sample and suction decreased during cyclic loading due to an increase in the degree of saturation, and (3) applying successive packets of cyclic loads and wetting to unsaturated samples showed a progressive increase in the volumetric strain and degree of saturation, leading to loss of suction. The accumulation of the volumetric strain measured for the tested soil was dependent on the magnitude of the confining stress, cyclic deviator stress, and suction. The soil water retention response during cyclic loading was governed predominantly by the suction level and the distance between the current soil water retention state and the main water retention curves. The accurate measurements of the soil water retention properties allowed the evaluation of the evolution of the stress path during cyclic loading in terms of Bishop’s stress and the dynamic nature of water retention properties of the soil. The void ratio of the tested soil measured in the resilient state was found to be dependent on the applied stress levels and degree of saturation, and changes in the void ratio also affected the water retention behavior under cyclic loading.

A generic predictive model of earthquake-induced slope displacement subjected to shallow crustal earthquake events is developed using displacements computed from finite-element (FE) analysis. The maximum displacement on the slope surface at the end of shaking was computed by nonlinear FE simulations for 49 slope models each subjected to 1051 earthquake motions. A unified predictive model of seismic displacement is developed that characterizes the slope in terms of its yield acceleration ($k_y$), the depth of the slip surface relative to the height of the slope ($H_{\mathit{ratio}}$), and the natural period of the full slope height ($T_{\mathit{slope}}$). Across five intensity measures and 10 combinations of intensity measures, peak ground velocity (PGV) is found to be the most efficient parameter for the displacement prediction, leading to significantly smaller aleatory variability. The displacement variability is partitioned into two components: between-slope variability, which represents the variability associated with different slope models, and within-slope variability, which represents the variability due to different input ground motions. The developed generic predictive model can be applied to the probabilistic seismic hazard analysis of slope movements and used for deterministic earthquake scenarios in the design/analysis process.

With the aim of testing an alternative sealing system for noise protection embankments containing environmentally hazardous substances, a fill of compacted clay with organic and elevated arsenic content was tipped. The fill was extensively monitored to measure the water balance and seepage. Due to the low permeability of the soil, a simple capping, consisting of a geosynthetic drainage mat and a clayey top layer, was used. Measurements over a 10-year period demonstrate that the chosen capping system is very efficient in minimizing seepage from the core. The success is based on a high water retention capacity of the topsoil in combination with the capillary-breaking effect of the drainage mat. Numerical simulations of the water balance of the test fill indicated that the calculation results are very sensitive to the fineness of the finite-element mesh, the choice of the time steps, and the hydraulic properties of the drainage mat. After calibrating the hydraulic properties of the drainage mat using the measured data of the first 3 years, the measured water balance of the 10-year period could be reasonably predicted by the model. Differences between the measured and the calculated discharges from the drainage mat and the core material are attributed to heterogeneities of the materials in the field and to the hysteresis of the soil water retention curve not being considered in the model.

Apart from spatial variability of soil properties, a geotechnical system can have many other sources of uncertainties. To efficiently analyze such a system in a probabilistic manner, many strategies have been proposed in the literature. This paper presents a deep learning technique for an efficient geotechnical reliability analysis with multiple uncertainties. The proposed method involves using convolutional neural networks (CNNs) as metamodels of the physics-based simulation model of a geotechnical system. In the present study, the spatially variable soil properties and the external loads are simultaneously considered in the analysis of a geotechnical system. The proposed neural network method configures these uncertainties to form a multi-channel “image.” CNNs can then simultaneously learn high-level features that contain information about the multiple uncertainties. With an appropriate architecture and adequate training, the trained CNNs can replace the computationally demanding physics-based simulation model for Monte Carlo simulations. Application of the neural network method is illustrated using a synthetic geotechnical example. The results reveal that the proposed neural network method effectively handles multiple uncertainties and efficiently predicts a failure probability value that is in good agreement with the benchmark result obtained using direct Monte Carlo simulations.