Journal of Geotechnical and Geoenvironmental Engineering

Suffusion refers to the loss of small particles in a soil matrix without volume changes, caused by hydrodynamic forces and alternation of pore fluid chemistry. This study investigated the ionic concentration (IC)-induced suffusion of sand–clay mixtures under saturated and unsaturated conditions through laboratory soil-column experiments. The impacts of clay mineralogy and sand-grain size on suffusion were investigated based on the breakthrough curves (BTCs) obtained from the soil-column experiments. In addition, the degree of saturation during the experiments was measured to assess the impact of the saturation state on suffusion in the sand–clay mixtures. The observed BTCs in this study demonstrated that suffusion was more significant under saturated conditions than unsaturated conditions for the sand–kaolinite and sand–illite mixtures. In contrast, the sand–bentonite mixtures exhibited higher susceptibility to suffusion under unsaturated conditions. Moreover, for the sand–kaolinite mixtures, a higher susceptibility to suffusion was observed with smaller sand-grain sizes under saturated conditions, whereas the opposite trend was observed under unsaturated conditions. This study provides insights into the suffusion behavior of sand–clay mixtures, considering influential factors such as clay type, sand-grain size, and IC gradient under both saturated and unsaturated conditions, which highlight the importance of considering the saturation state for a comprehensive understanding of suffusion in the sand–clay mixtures.

Due to time, budget, and/or technical constraints, geotechnical site investigation data from a specific site are often limited and sparse, leading to a long-lasting challenge in characterization of spatially varying geotechnical properties. During preliminary stages of site characterization, geotechnical data from neighboring sites or sites with similar geological conditions are often collected and used as valuable prior knowledge in geotechnical engineering practice. Nevertheless, existing methods for spatial variability characterization often rely solely on site-specific data and cannot effectively incorporate prior knowledge or existing databases. To address this issue, this study proposes a novel machine learning method that systematically combines sparsely measured data at a specific site with existing data from neighboring sites or sites with similar geological settings for characterization of property spatial variability in a data-driven manner. The proposed method starts with the construction of a dictionary that draws the dominant spatially varying patterns from a property measured at sites with similar geology under a dictionary learning framework. Leveraging the developed dictionary, the spatial variability of a property is interpreted from sparse site-specific measurements using Bayesian learning. The effectiveness of the proposed method is demonstrated using real data, and improved performance over existing methods is observed.

The contraction behavior of monotonically expanded cavities is intriguing as it offers insights into certain geotechnical scenarios, especially for pressuremeter tests, where the unloading data is equally informative as the loading data. Despite many solutions for cavity expansion, attempts for the analyses of cavity contraction from an expanded state were rarely made. To extend previous solutions to include contraction, this paper presents a novel semianalytical solution for drained contraction of spherical and cylindrical cavities from an initially expanded state in soils characterized by a unified state parameter model for clay and sand (CASM). Given the nonself-similar nature of the contraction after expansion problems, the hybrid Eulerian-Lagrangian (HEL) approach is employed to derive distributions and evolutions of stresses and strains around the cavities during the unloading process. Combined with the previous expansion solution, the complete loading-unloading cavity pressure curves and stress paths at the cavity wall are presented and verified against numerical simulations. Following validation through comparisons with calibration chamber pressuremeter tests conducted in Stockton Beach sand, a new method for the interpretation of pressuremeter testing data is developed based on the proposed solution. This method demonstrates its capability in the back-calculation of the effective horizontal stresses and state parameters for four distinct types of sands.

Generic models for estimating the earthquake-induced displacement of Earth slopes are developed based on a numerical approach. A number of 14,112 slope models with different configurations of slope geometry and soil property parameters are developed to represent generic Earth slopes. Thousands of slope dynamic analyses are then conducted in FLAC to estimate the coseismic slope displacements. Based on the displacements calculated, 18 ground-motion intensity measures (IMs) and eight slope variables are considered as candidate predictor variables to develop predictive displacement models using the light gradient boosting machine (LightGBM). Comparative results indicate that yield acceleration ($K_y$) and Arias intensity (AI) are the most efficient scalar variables in regressing the displacements. Based on the efficiency, sufficiency, and computability criteria, the vector IMs of (AI, peak ground velocity) and (AI, peak ground acceleration), together with $K_y$ and initial shear modulus, are regarded as the preferable predictor variables, respectively. Two sets of predictive displacement models are thus proposed using the preferable variables via the LightGBM- and polynomial-based approaches, respectively. The aleatory variability in predicting the slope displacement for the polynomial models is approximately 15%–30% larger than that of the LightGBM models, indicating that the predictive performance of the LightGBM models is superior to the polynomial models.

Accurate prediction of excess pore water pressure (EPWP) generation in saturated sandy soils remains one of the most challenging issues in sandy site responses to strong earthquakes and extreme marine environments. This paper presents experimental results of undrained and drained multidirectional cyclic hollow cylinder (MCHC) tests on saturated coral sandy soils under various cyclic loadings. The results show that threshold generalized shear strain ${\gamma }_{\mathrm{ga},\mathrm{th}}$, below which EPWP and volumetric strain can be neglected, is an inherent property depending only on the soil type and initial state. Furthermore, there exists a virtually unique form of relationships between the generalized shear strain amplitude (${\gamma }_{\mathrm{ga}}$) and the cumulative dissipated energy per unit volume of soil ($W_{\mathrm{c}}$) at different relative density ($D_{\mathrm{r}}$), irrespective of drainage conditions and cyclic loading conditions. These findings highlight the fundamental mechanism for cyclic deformation behavior and the uniqueness of correlations among $r_{u\mathrm{p}}$ (peak EPWP ratio), ${\epsilon }_{\mathrm{vp}}$ (peak volumetric strain), and ${\gamma }_{\mathrm{ga}}$ of saturated sandy soil at the similar $D_{\mathrm{r}}$, regardless of cyclic loading conditions. Based on these findings, a novel unified model of ${\gamma }_{\mathrm{ga}}$-based cyclic shear–volume coupling and EPWP generation is established, which is independent of cyclic loading conditions over a wide loading frequency range. Then the applicability of the proposed model is validated by the experimental data of the same tested coral sandy soil and siliceous Ottawa sand, as well as the data of siliceous fine sands in previous work. It is found that the proposed model surpasses the existing strain- and stress-based models in accurately predicting EPWP generation under complex cyclic loadings, which can offer new insights into the mechanisms of the EPWP generation in saturated sandy soils.

Deep horizontal vibrocompaction is an efficient method of compacting granular soils that has been used and optimized over decades. The state of research indicates that the best possible compaction results are achieved when the vibrator operates at the natural frequency of the vibrator-soil interaction system. However, this approach proved to be unsustainable in practical application. This paper presents a concept for determining the soil response by means of the soil contact force and its phase angle and proposes that an optimized compaction is possible at the natural frequency of the soil response. The considerations are explained using a single-degree-of-freedom (SDOF) model. The admissibility of the approach is demonstrated using measured data as an example. The soil contact force and the phase angle of the soil response are evaluated for selected compaction tests. The evaluation shows that reducing the excitation frequency allows compaction close to the natural frequency of the soil response, requiring less electrical energy, shorter compaction time, and less jetted water. Approaching the natural frequency of the soil response promises to be a valuable criterion for optimizing the deep vibrocompaction process.

Buried pipelines subjected to permanent ground deformations (through, e.g., earthquake-induced liquefaction or fault rupture) often experience widespread damage. Regardless of the direction of ground movement, pipelines tend to respond and experience damage axially due to their directional stiffness characteristics. In addition, case studies and previous testing have shown that damage is concentrated at the pipe joints due to their lower strength compared with a pipe barrel. Previous testing has also shown that axial forces increase significantly when pipe connections have jointing mechanisms, such as coupling restraints, with larger diameters than the pipe barrel alone. These enlarged joints act as anchors along the pipe, increasing the soil resistance at these locations. Current methods for predicting the axial force along a pipe underpredict the force demands and oversimplify the mechanics of soil resistance on the joint face. This study conducts a series of 12 pipe-pull tests in a centrifuge, varying joint diameter and burial depth, to quantify the axial forces developed. A strong linear correlation was observed between the soil resistance on a joint face and the joint surface area and burial depth. The study also proposes an analytical solution based on pullout capacity design equations for vertical anchor plates as a function of soil and pipe joint properties. The proposed solution to calculate joint resistance is in good agreement with the centrifuge tests performed for this study and previous full- and model-scale experiments. The proposed prediction equation is anticipated to have future applications to other buried structures because it is based on mechanisms of passive resistance commonly encountered in underground structures and lifelines.

Bentonite, used as a barrier material in landfill systems and the geological disposal of radioactive waste, undergoes various processes such as the cation exchange with landfill leachate and repository groundwater, desiccation shrinkage due to heat released from solid waste or nuclide decay. In the context of these engineering applications, it is crucial to consider the effect of interlayer cation composition on the desiccation-shrinkage behavior of bentonite for assessing the long-term performance of barrier systems. This study investigates this impact by conducting desiccation tests on four different cation-type bentonite samples (i.e., K-, Na-, Ca-, and Mg-bentonite) and elucidating the underlying mechanisms through mercury intrusion porosimetry test, X-ray diffraction analysis, and suction measurements. The results indicate that K-bentonite exhibits the highest water evaporation rate and the lowest shrinkage potential, with no shrinkage cracks detected, and that Na-bentonite evaporates water at a relatively fast rate but presents the most pronounced volumetric shrinkage. In contrast, Ca- and Mg-bentonite have the highest water-retention capacity, and their shrinkage potentials are less than Na-bentonite. An improved shrinkage model is proposed, which is shown to describe the measured shrinkage curves for four types of homoionic bentonite samples. Additionally, the mechanism behind the effect of different interlayer cations was interpreted in terms of the interlayer–cation–water interaction, pore structure evolution, and soil suction variation.

The applicability of particle-scale modeling using the discrete-element method (DEM) is typically evaluated by comparing simulation results with stress–strain responses observed in elementary tests. This validation at the global level may not guarantee that the simulation can capture realistic particle-level motion. Thus, this study investigated the applicability and limitation of two types of DEM models, through the comparison with experimental results of biaxial shearing tests on bidisperse granular assemblies comprising circular (round) and hexagonal (angular) particles under various confining pressures. Experimental data wherein particle rotations were identified by novel image analysis technique were used to evaluate whether the DEM models could accurately reproduce macroscopic stress–strain relationships and microscopic particle responses. Experimental findings suggested that particle rotations play a crucial role in granular deformation and are influenced by the particle shape. A detailed DEM model with precise particle shapes effectively replicated both macroscopic stress–strain relationships and microscopic responses, including particle rotation and interlocking at global and local levels. Conversely, a simpler ad hoc DEM model, which incorporates rolling resistance for circular particles, could imitate the stress–strain relationships of hexagonal particles but fell short in replicating microscopic responses accurately.

A full scale load test on the apron is a very effective method for validating the design of a new operative quay wall. The availability of the monitoring data from the load test on a sheet pile wall of a major Italian port and the well-known geotechnical context of the site made it possible to explore the potential of three-dimensional (3D) numerical modeling to understand the response of the work to the applied load. In order to achieve a good match between real observations and numerical results, the limited impact of the working load on a robust geotechnical structure, the use of advanced constitutive models for soils and careful modeling of the construction phases and details of the work were taken into consideration. For this reason, the procedure adopted to calibrate the numerical model based on geotechnical test results is presented after the description of the load test monitoring data. Thanks to these efforts, the 3D numerical model, validated on the monitored data, allowed us to gain further insights on the structural behavior of the quay wall during construction and loading phases, underlining the relevant role played by some apparently marginal and often ignored details.