Journal of Geotechnical and Geoenvironmental Engineering

It is well known within the dam engineering profession that the potential for earthquake-induced cracking is a key consideration in the design and construction of dams. Transverse cracking of embankment dams is a major concern of engineers involved in dam safety management and commonly represents a potential failure mode for such dams. If transverse cracking extends below the reservoir level, it may lead to leakage from the reservoir, internal erosion of the embankment soils, and dam failure. Considerable effort has been made in the engineering profession during the last few decades to study the nature and effects of dam earthquake cracking and to develop design measures to mitigate its potential occurrence. However, few studies exist on earthquake-induced cracking of embankment dams, and the profession’s ability to accurately predict its characteristics is limited. Thus, it seems useful to review key aspects of earthquake cracking of embankment dams and some of the lessons learned about cracking from the past performance of dams during earthquakes. This paper aims to fulfill this objective by providing an overview of the observed effects of earthquakes on embankment dams, discussing the potential for embankment dam failure by cracking, reviewing basic concepts about the mechanics of cracking in soils, discussing the analysis of embankment dam cracking in practice, and presenting two case histories of cracking of embankment dams. The case histories describe (1) the performance of Matahina dam during construction, reservoir filling, and the 1987 Edgecumbe earthquake in New Zealand and (2) the performance of Lenihan dam during the 1989 Loma Prieta earthquake in California and an analysis of the dam for that earthquake.

Natural soils often exhibit significant spatial variability due to their geological history of soil formation. As an attractive anchoring solution for floating offshore structures, this paper has investigated the pullout capacity of strip plate anchors in clay considering the inherent soil spatial variability. In this study, the soil properties were represented by random fields, and an analytical framework was developed to estimate the first two moments and the probability density function of the pullout capacity factor for shallowly and deeply embedded anchors. The analytical approach was validated by the random finite element method (RFEM) over a wide range of soil and anchor parameters. The results show that the coefficient of variation and correlation length of the soil significantly affect the prediction of anchor pullout capacity, providing evidence that sufficient site investigation is of great importance for cost-effective and reliable anchor design. Probabilistic charts were also developed to aid in the probabilistic analysis of anchor pullout capacity. Overall, the developed analytical framework can be used as a good approximation to the computationally intensive RFEM.

Plate anchors have been recognized as an attractive option for floating offshore structures due to their low cost and high effectiveness. The prediction of anchor pullout capacity often assumes a homogeneous sand with uniform properties over the entire soil mass. However, natural soils typically exhibit significant spatial variability due to their geological history of soil formation. To account for this inherent spatial variability, this paper has proposed an analytical approach to probabilistically estimate the pullout capacity of strip plate anchors in sand. The soil friction angle was represented by a random field, and the first two moments and the probability density function of the anchor breakout factor were determined analytically using a local average theory. The proposed analytical approach was validated by the random finite-element method (RFEM) over a wide range of soil and anchor parameters. The findings suggest that the anchor embedment depth ratio has a limited effect on the estimated anchor failure probability, whereas the coefficient of variation and correlation length of the soil friction angle have a significant influence. This confirms the importance of sufficient site investigation for cost-effective and reliable anchor design. Overall, the proposed analytical framework can be regarded as a reliable and practical alternative to the computationally intensive RFEM.

Several ground improvement techniques, which have been proven to be effective and economical solutions for increasing the lateral stiffness and strength of weak soils around piles, can often result in unnecessarily conservative large volumes of soil improvement. Additionally, there are no rigorous techniques available to analyze the seismic behavior of piles in improved soils that can be utilized in day-to-day engineering practice. In this study, a stand-alone one-dimensional finite element computer code called DYPAC (Dynamic Piles Analysis Code), which uses the beams on nonlinear Winkler foundation (BNWF) approach, was developed to analyze the seismic response of a single pile in both improved and unimproved soils. The computer code models the pile as a beam element and the nonlinear soil behavior as springs and viscous dashpots using a nonlinear $p\text{-}y$ element, where $y$ = displacement; and $p$ = soil reaction per unit length of the pile. This study proposes and validates a method for modifying the $p\text{-}y$ curves to consider the limited lateral extent of ground improvement. The $p\text{-}y$ curves were inputted into DYPAC to analyze a series of dynamic centrifuge tests of single piles in soft clay improved using cement deep soil mixing (CDSM). Free-field site response analyses were performed using the DEEPSOIL computer program and the soil displacement-time histories were inputted into the free-field ends of the nonlinear $p\text{-}y$ elements. The predictions made by DYPAC were validated using the centrifuge data and the results show that the DYPAC predictions are reasonable. Furthermore, the proposed method for modifying $p\text{-}y$ curves to characterize CDSM-improved soil appears to be appropriate and practical. Moreover, the benefits of nonlinear DEEPSOIL free-field site response analyses during large earthquakes, as compared to equivalent linear analyses, were also demonstrated.

End-of-life tire (ELT) rubber has been widely researched to replace aggregates in construction materials such as asphalt concrete and cementitious composites. Although most studies paid attention to its effect on engineering properties and the durability of the materials, very few considered chemical reactions with pore solution and the potential for environmental leachate, especially rubberized stabilized soil, which was proposed recently to increase the use of ELT rubber in civil engineering. This study proposes the use of rubberized stabilized soil (RSS) in which clayey soils (e.g., kaolin and bentonite) were stabilized by portland cement (PC) and end-of-life tire (ELT) rubber particles. The authors previously developed a methodology to extract zinc from the ELT rubber; therefore, this study explores the potential for RSS to immobilize the leachate from the ELT rubber before and after this treatment. Three main topics are addressed in this study: (1) the capability of clay to capture leached zinc under ambient and alkaline aqueous conditions; (2) engineering properties of RSS [i.e., unconfined compressive (UCS), flow] with 0%, 10%, 30%, and 50% ELT rubber added by clay volume; and (3) pore solution and leachability tests of RSS. A leaching experiment was employed for Topic 1, UCS and flow tests were conducted for Topic 2, and pore solution extraction and leaching tests were performed for Topic 3. The results showed that the clayey soils and ELT rubber are synergistic in terms of engineering properties and the capturability of zinc and total organic carbon (TOC). Although the ELT rubber and PC strengthen the clay structure, the clay absorbs leached zinc and TOC from the ELT rubber particles. Adding untreated ELT rubber into PC-stabilized clays significantly increased the RSS strength; however, this improvement was less significant for the treated ELT rubber. Ultimately, the results proved that the environmental and mechanical performance of RSS makes it a viable construction material.

An extensive database consisting of measurements of ground movements and pore pressures from 48 deep excavations is presented. The ground conditions for all cases are soft, normally consolidated clays, often underlain by a coarser layer of soil, i.e., confined aquifer, on top of bedrock. Traditionally wall and ground movements caused by deep excavations have been predicted without assessing effects of groundwater drainage or the influence of installation of tie-back anchors or foundation piles. This study clearly shows that the observed ground movements exceeded expected values based on induced shear movements and significant deformations occur at far distance from the excavation. The cause of the deformations is largely consolidation settlements due to pore pressure reduction, with a zone of influence recorded at up to 400 m distance from the excavation. The construction methods and opening of drainage paths to the confined aquifer is as important for pore pressure reduction, as exposing the confined aquifer itself. In addition, the groundwater drawdown is shown to be dependent on the ground conditions and mitigation measures. It is also shown that the pore pressure reduction and zone of influence may be very small if no drainage path is created, i.e., in case of strutted excavations and no piles or driven piles, or when the bedrock surface is not exposed. The drilling of tie-back anchors and piles are also seen to influence ground movements, especially at a distance of 2–4 times the excavation depth. Charts produced from the database enable estimation of expected groundwater drawdown including zone of influence. In addition, they can support decisions on construction methods, need for mitigating measures and extent of monitoring program in relation to deep excavations.

This research investigated dynamic vibration control in a monopile-supported wind turbine system through strategically placed resonators along its length. Four configurations were analyzed: Case 1 (conventional wind turbine model), Case 2 (resonators in the turbine tower), Case 3 (resonators in the monopile), and Case 4 (resonators in both tower and monopile). Utilizing spectral element formulations, the dynamic stiffness matrix was derived for the wind turbine system, embedded in cohesionless soil with frequency-dependent viscoelastic springs. Analytical responses at the tower top and monopile head under dynamic harmonic excitation were computed, and transmittance was determined as the logarithmic ratio of responses to input excitation. Experimental dynamic tests were conducted on a 3D-printed wind turbine scaled model, yielding displacement responses at the tower top and monopile head. Comparison with analytically determined transmittance values reveals consistent agreement across four configurations. Case 4 demonstrates substantial response reductions, highlighting the efficacy of resonators in both tower and monopile. Transmittance plots indicate notable response reduction near resonant frequencies, showcasing local resonance phenomena. Furthermore, the tower and monopile head responses are substantially reduced when the excitation frequency surpasses the resonator’s natural frequencies. This underscores the dynamic vibration control capabilities of periodically positioned resonators within the wind turbine scaled model. The research also explores the influence of parameters such as resonator mass ratio, soil shear modulus, and pile and tower slenderness ratio on transmittance plots.

Aging degradation of the geomembrane (GMB) significantly influences the dynamic shear characteristics of the composite liner interface, which comprises the GMB and the compacted clay liner (CCL), potentially jeopardizing the dynamic stability of landfills. In this study, cyclic shear tests were performed on two types of aging GMB/CCL interfaces, concurrently with shear tests on the nonaging GMB/CCL interface for comparison. The results suggest that the impact of aging on the dynamic shear characteristics of the GMB/CCL interface is essentially governed by the surface roughness and brittleness of the GMB, with the effect degree of brittleness influenced by the normal stress. Under low normal stress, aging increased the vertical displacement, dynamic shear strength, and shear stiffness of the GMB/CCL interface. However, under high normal stress, the dynamic shear strength and shear stiffness of the aging GMB/CCL interface were more likely to be lower than those of the nonaging interface. As the displacement amplitude increased, the influence of aging on the shear stiffness of the GMB/CCL interface gradually diminished. Aging also reduced the damping ratio of the GMB/CCL interface. The difference in vertical displacement between the exposed GMB/CCL interface and the in soil GMB/CCL interface caused by brittleness was not significant. In practical engineering, when the overlying load on the GMB/CCL composite liner is relatively small, aging makes the GMB more susceptible to tearing under seismic loads, whereas with larger overlying loads, aging is more likely to increase the shear displacement, thereby increasing the likelihood of instability in landfill. Finally, based on classic models of soil, fitting models for the normalized shear stiffness and damping ratio of the GMB/CCL interface were established and validated. This study can provide reference for analyzing the dynamic stability of landfills during long-term use.

Traditional free vibration-based forward models generate theoretical dispersion curves under the assumption of planar waves, neglecting the influence of the actual source-receiver configuration. The 2D/3D discretization-based numerical wavefield modeling approaches can mimic real field scenarios considering active source-receiver information. However, numerical methods are computationally inefficient. This study introduces an active sourced semianalytical wavefield modeling approach for laterally homogeneous horizontally stratified media, incorporating source-receiver data acquisition layouts. The method considers a cylindrically spreading wavefield described by the Hankel function instead of the planar wave assumption. The approach considers both propagating waves characterized by real wavenumbers and decaying waves with complex wavenumbers, allowing for the calculation of surface displacements in both the far and near fields. The proposed model captures the complete wavefield, including source-offset effects and leaky waves, while maintaining computational efficiency comparable to any free vibration-based approaches. The method entails solving the eigenvalue problem constructed through the higher-order thin-layer method. Subsequently, it calculates the frequency domain vertical and radial surface responses at any desired location in space generated by a vertically positioned active source. The proposed method’s overall performance is investigated on diverse subsurface profiles, including regularly dispersive media, low-velocity layer models, and a field cross-hole model. The Rayleigh wave’s vertical and radial component dispersion images obtained from the proposed method are validated against the numerical approach. The proposed method is at least two orders of magnitude faster than the numerical approach. Notably, it addresses mode misidentification issues arising from modal osculation at low frequencies and effectively captures the smooth transition of modal energy from fundamental modes to higher modes. Note that the present method is limited to laterally homogeneous media. However, this method provides a valuable tool for advancing the accuracy and efficiency of active surface wave methods in various engineering applications.

In this work, the effect of gas jets used in the deep vertical vibratory compaction technique are studied. Gas jets play a vital role in treating structured loess foundations by the pneumatic-vibratory probe compaction method. Utilizing the geotechnical particle finite-element method numerically, we estimate the limit gas injection pressure and delineate the injection-induced damage and plastic zones. The behavior of structured soil is described using an elastoplastic constitutive model considering its structure evolution. The analysis of structured loess under gas injection is based on the cavity expansion approach. Experimentally, we performed a scale model test of gas injection to investigate the mechanism of the gas jets on the surrounding soil and compared relevant results with numerical results. Numerical results show that the limit gas injection pressure for structured loess beyond a depth of 8.0 m ranges from 1,409.7 to 1,467.2 kPa, increasing with the increase of overburden depth while the current cavity expansion radius decreases. The radius of the plastic zone induced by cavity expansion is 2.0 to 3.0 times the current cavity radius within this depth range; for the damage zone, however, it ranges from 0.1 to 0.4 times. The horizontal pressure recorded during the model test is observed to be lower compared with the numerical simulation results. This discrepancy can be attributed to factors such as the neglect of gas leakage within the soil and the utilization of a uniform parameter. The gas jets expand soil in cyclic shear form. It goes through a process from destruction of soil structure to compression in the horizontal direction; then, its pressure gradually drops to zero in the expansion direction of the dominant channel in soil.

The presence of bound water in clay has a significant impact on the physical and chemical properties of clay, particularly its strength, permeability, and creep behavior. In this paper, the bound water in clay has been studied from the perspective of water potential. Initially, the adsorption isotherms of powders and consolidated samples for Na- and Ca-bentonite and illite were measured, and the bound water content was determined by subtracting the capillary water in the isotherms, with the capillary water being calculated by mercury intrusion porosimetry tests. The results indicated that bound water is independent of the void ratio and pore structure of clay, which is consistent with previous studies. Then, metadynamics was conducted to determine the adsorption free energy landscapes of the three clays, and the lowest suctions of the three clay minerals were determined to be $-1.7$, $-5.4$, and $-2.1\mathrm{GPa}$, respectively. Finally, through a comparison of the simulations and experiments in this study and in the literature, three important conclusions were drawn. Firstly, the lowest water potential for montmorillonite exhibits a linear relationship with the hydration free energy of exchangeable cations. Secondly, the critical water potential for tightly bound water is determined as the first inflection point in the relationship between water potential and water content, and the critical value of montmorillonite is found to be correlated to the valence of exchangeable cations. Lastly, the boundary between loosely bound water and capillary water is determined as the starting point of capillary water formation. Overall, this research highlights the importance of considering water potential as a key factor in understanding the behavior of bound water in clay.

Land reclamations often demand large quantities of fill materials. However, many countries commonly engaged in land reclamation are facing an escalating scarcity of natural fill sources. To address this issue, this study proposes a sustainable land reclamation approach that involves dredged marine sediments as fill materials. The dredged marine sediments are improved by a combined method that incorporates the use of horizontal drains with vacuum preloading for preliminary soil treatment and vertical drains with vacuum preloading for further improvement. A field trial has been designed and conducted to verify the performance of the proposed preliminary treatment of marine sediment slurry using prefabricated horizontal drains (PHDs) and vacuum preloading under a “membrane-free” condition. Field measurements, including settlement, pore pressure, effective earth pressure, vacuum pressure, water content, and undrained shear strength, were systematically recorded during the field trial period. After a four-month treatment, the average water content of the filled soil was slightly below 50%, which is one-third of the initial water content. Owing to the combined effect of the vacuum preloading and drying process, a crust layer with the undrained shear strength of 40 kPa was formed on the top of the treated soil. Below the crust layer, the average undrained shear strength of the treated marine deposits was approximately 32 kPa. Overall, the notable effect of using PHDs and the “membrane-free” vacuum preloading method to rapidly consolidate marine deposits slurry has been successfully demonstrated with field monitoring under the site trial condition.

Abstract

Soil–steel composite structures are commonly considered competitive alternatives to conventional small road and railway bridges. The structure is made from corrugated steel plates and comes in various profiles and shapes, up to 32 m at most. These structures are also found in seismic areas. Despite this, knowledge of their seismic behavior is limited. The paper analyzes the seismic behavior of an existing soil–steel composite bridge in Poland, where full-scale tests have been conducted. The analyzed bridge has a total height of 6.05 m and a span length of 17.67 m and was built with $140\times 380\mathrm{mm}$ corrugation and 7-mm thick steel plates. Numerical analyses are performed on the bridge using the finite element (FE) program DIANA FEA with seismic records from the 1940 El Centro earthquake as a reference. Expanded polystyrene (EPS) geofoam is applied under the structure and studied in the numerical models to reduce the impact of seismic waves on the bridge. The first numerical model was created without geofoam for calibration purposes, and the latter five with EPS given different stiffness properties (densities and Young’s modulus). The results suggest that EPS has advantageous features (especially with low stiffness) in reducing the maximum deformations and stresses on the steel shell by absorbing (damping) the energy of the seismic waves and rearranging the stresses away from the structure and to the soil. This trend was particularly evident when comparing the five models built with EPS to the model without EPS, where stresses in the structure were significantly higher, demonstrating the material’s ability to rearrange the stresses away from the structure.

Practical Applications

Soil–steel composite bridges are commonly considered competitive alternatives to conventional small road and railway bridges. The shell structure is made from corrugated steel plates backfilled by granular soil and comes in various profiles and shapes, up to a maximum of 32 m. These structures are also found in seismic areas. Despite this, knowledge of their seismic behavior is limited. This paper analyzes the seismic behavior of an existing soil–steel composite bridge. The EPS geofoams were applied under the structure and studied in the numerical models to reduce the impact of seismic waves on the bridge. The results suggest that EPS has advantageous features in reducing the maximum deformations and stresses on the steel shell by absorbing the energy of the seismic waves and rearranging the stresses away from the structure and the soil. The obtained results may contribute to the dissemination of these bridge structures in seismic areas.

In order to improve the understanding on static liquefaction phenomenon, this paper presents an insight into the instability mechanism of a silty sand originating from the Sakarya region (Turkey). Monotonic undrained simple shear tests were performed in either strain-controlled or stress-controlled modes. The interpretation of the experimental results allowed us to draw three interesting conclusions on instability conditions: (1) Hill’s stability postulate can be conveniently used to identify the onset of instability for strain-controlled tests, whereas in stress-controlled conditions, an alternative new parameter ${\eta }^{*}$ (modified apparent viscosity) was introduced. Depending on the strain rate, ${\eta }^{*}$ allows to identify stability conditions more clearly than second-order work; (2) the value of the pore pressure ratio at the onset of instability ($r_{u,\mathrm{is}}$), ranging between 0.30 and 0.50, is linearly linked to the state parameter $\psi$; and (3) at the onset of instability, the dependence of both the major and minor principal stress directions (${\alpha }^{\mathrm{is}}$ and ${\beta }^{\mathrm{is}}$) and Lode’s angle (${\theta }^{\mathrm{is}}$) on $\psi$ was also observed. These findings are expected to be of some help for a more consistent approach to the constitutive modeling of the static instability mechanism.

The original stress equilibrium state of the soil is destroyed after the excavation of the double-line shield tunnel, resulting in the settlement of the stratum above the tunnel and bringing adverse effects on the surface buildings and subsurface pipelines near the tunnel. The reasonable prediction of subsurface settlement is significant for protecting adjacent underground structures. The wave propagation model is introduced to describe the settlement mechanism of the stratum above the tunnel based on random medium theory, and the subsurface settlement model of single-line tunnel is derived through the Fourier transform. The disturbance influence factor of double-line tunnel is defined, and the corresponding settlement model is established. The three-dimensional subsurface settlement models of single-line and double-line shield tunnels are proposed by introducing the surface displacement release rate of palm surface and maximum slope of surface vertical settlement. The models are verified using monitoring data, and the effects of the parameters on the model are analyzed.

Abstract

This paper presents an analysis of an underwater retrogressive slope failure caused by the concurrent construction of a wharf access causeway and dredging near the end of the causeway for the wharf structure cofferdams. A cross section was developed along the causeway and analyzed to simulate the retrogressive slope failure using limit equilibrium analyses. The compound failure surface for the inverse stability analysis of each of the five retrogressive failure masses agrees with observations before, during, and after the failure. Soil stratigraphy is discussed and the mobilized undrained strength of the seabed clay underlying the causeway fill was estimated using an inverse analysis of each of the five failure masses. The inverse analysis shows that the concurrent dredging and causeway construction reduced the factor of safety (FoS) of the causeway underwater slope, which eventually initiated the retrogressive failure. The stability analyses show that the causeway construction contributed to the reduction of the FoS but dredging triggered the first slope failure, which started the retrogressive failure. Nevertheless, had dredging not occurred, a slope failure would still have occurred if the causeway construction had extended another 5 m from where the first slope failure occurred.

Practical Applications

There are several practical lessons from the slope failure discussed in this paper. First, in any slope modification, the existing soils in the area of construction should be sampled and tested. Engineers and owners should be cautious of construction methods that make implicit assumptions about soil strength and/or other engineering characteristics. Even though gathering underwater geotechnical data is difficult, it is recommended to perform field tests to evaluate the shear strength of soil layers to perform slope stability analyses. Further, planned modifications to a slope that could destabilize it should be thoroughly analyzed, and such analyses should not be restricted to a circular failure surface. If possible, the construction sequence should be modified to reduce the extent to which the slope is destabilized. In this case, that would mean constructing the wharf before the causeway. Finally, construction should be monitored to verify that the proposed methods are working as intended. In this case, it was assumed that the rockfill would displace the marine clay when constructing the causeway. This should have been verified as filling progressed. Photographs of the dredged marine indicated a medium to medium stiff clay, which indicated that the clay was not as soft as anticipated.

Wetting–drying cycles have a significant impact on the gas diffusion coefficient ($D_p$) and gas permeability ($k_a$) of unsaturated soils. The soil volume change during wetting–drying cycles limits the application of the traditional rigid-wall device for measuring $k_a$ and $D_p$, due to the gas preferential flow along the interface between soil and the rigid-wall container. Although flexible-wall devices for measuring $k_a$ are available, no such device exists for measuring $D_p$. Thus, the effects of wetting–drying cycles on $D_p$ remain unclear, particularly for fine-grained soil. The present study developed a flexible-wall device to investigate the effects of a wetting–drying cycle on the $D_p$ and $k_a$ of unsaturated soils. Both the flexible- and rigid-wall devices were adopted to measure $k_a$ and $D_p$ of three soil types, including fine sand, silt and kaolin. The rigid-wall device could overestimate $k_a$ by up to approximately one order of magnitude, whereas it overestimated $D_p$ by approximately 2–3 times. Regardless of $k_a$ and $D_p$, the difference in measurements between the rigid- and flexible-wall devices became more significant at a lower water content and along the drying path because of the gas preferential flow caused by soil shrinkage in the rigid-wall device. Accordingly, the kaolin exhibited the largest difference in $k_a$ and $D_p$ as measured by the flexible- and rigid-wall devices because it had the largest clay minerals and the finest particle size, resulting in the largest volume shrinkage. The $k_a$ and $D_p$ measured by the flexible-wall device along the drying path were generally larger than those along the wetting path, probably because of entrapped gas in the soil caused by water spray during wetting.