Deep overburden deposits often contain low-permeability fine-grained interlayers, yet their role in liquefaction and post-seismic weakening under high in-situ stress remains insufficiently understood. In this study, two high-g centrifuge shaking table tests were conducted to investigate the seismic response of deep sandy deposits with and without a sandy-silt interlayer, representing prototype depths exceeding 80 m. The two models were designed to achieve comparable stress conditions and were subjected to similar seismic energy inputs. The results indicate that the presence of a fine-grained interlayer fundamentally reorganizes the spatiotemporal evolution of excess pore water pressure (EPWP) and the associated weakening pattern. Compared with the homogeneous deposit, the stratified configuration exhibits an earlier onset of nonlinear response in the shallow layer, stronger localization of cyclic deformation, and markedly intensified stiffness degradation. Hydraulically, the low permeability of the interlayer severely impedes upward drainage, creating a local maximum in hydraulic head and a localized reverse hydraulic gradient. This pore-pressure trapping effect leads to delayed dissipation and a prolonged reduction of effective stress in the deep zone. Based on the evolution of the EPWP ratio, both deposits developed a shallow liquefied zone, More importantly, rather than simply increasing the peak pore pressure, the interlayer generates a deep persistent weak zone characterized by extremely slow pore-pressure dissipation and depth-dependent delayed recovery.These findings demonstrate that the site response in deep stratified overburden is governed by stratigraphy-controlled hydro-mechanical coupling, where permeability contrast dictates pore-pressure redistribution and stiffness contrast controls deformation localization. This study provides new experimental evidence for understanding deeper and prolonged post-seismic weakening in complex stratigraphy.

Accurately describing slurry diffusion in fracture remains challenging due to the complexity of fracture roughness and the non-linear rheological properties of slurry. This study presents an analytical solution for the single-hole grouting in rough fractures considering the time-dependent viscosity of Bingham fluid. Fracture roughness is described by introducing two parameters, the fractal dimension D and the characteristic scale parameter G. The accuracy of the analytical solution is validated by comparing the slurry flow and diffusion radius from experimental results with predicted results. The corresponding slurry flow Q calculated from the analytical solution is used to delineate different areas. Variations in D and G shift the slurry flow resistance from rough (Q < 0.1 L) to transitional (0.1-0.8 L) and smooth (Q > 0.8 L) areas under constant other parameter conditions. Variations in yield stress and viscosity shift the slurry flow areas among low, medium, and high sensitivity areas. Additionally, numerical analysis of two-hole grouting in rough fractures is performed to determine optimal grouting hole spacing based on the percentage of the area covered by the slurry relative to the total fracture area. During two-hole grouting, mutual squeezing effect between slurry alternately promotes and impedes flow. The optimal grouting hole spacing of Bingham fluids with varying water-to-cement ratios decreases with fracture roughness and increases with grouting pressure. Bingham fluids with water-to-cement ratios of 1.0-2.0 exhibit greater sensitivity to grouting pressure in wide fractures due to complex flow characteristics, providing a reference for simplifying grouting process across varying geological conditions.

Accurately characterizing slurry behavior in discrete fracture network remains challenging due to the complexity of fracture geometries, the deformability of apertures, and the nonlinear rheology of slurry. This study develops a novel three-dimensional grouting model that integrates slurry diffusion, particle deposition, and permeability evolution in discrete fracture network. The evolving composition of the slurry is described by the volume fractions of deposited (S) and suspended (C) particles. A spatial-temporal viscosity field is incorporated to capture the time-dependent changes in slurry properties as it diffuses along the flow paths. Moreover, fracture apertures deform dynamically under the influence of slurry pressure and particle deposition, establishing a bidirectional feedback mechanism between the flow field and fracture geometry. The accuracy of the numerical model is validated by comparing the predicted and experimental results of permeability evolution in both single fracture and fracture network. The results show that permeability decreases hysteresis arises from delayed particle deposition, while increased fracture intersections accelerate the transformation from suspended to deposited particles. Sensitivity analyses reveal that slurry diffusion is primarily governed by slurry properties such as grouting pressure and water-cement ratio, whereas deposition and sealing behavior are strongly influenced by fracture aperture and network density. Finally, the model facilitates the determination of optimal grouting duration under various conditions, offering a reliable reference for designing effective grouting strategies in complex fracture networks.

Ballasted railway tracks are increasingly required to withstand the loads of faster trains. Under high-speed train loads, the granular ballast layer can exhibit fluidization, leading to a rapid buildup of irreversible deformation and posing risks to travel safety. To understand the particle-scale mechanisms behind ballast fluidization, this study develops a coupled discrete element-finite element model to analyze the mechanical behavior of ballasted tracks under dynamic train loads. The sleeper is modeled with horizontally joined rectangular discrete elements, efficiently capturing its flexural behavior due to its high aspect ratio, offering a simpler and more efficient alternative to finite element modeling. The ballast is modeled with polygonal discrete elements, while the subballast and subgrade are modeled with finite elements. The results reveal that ballast fluidization is primarily caused by wave propagation in the track structure, resulting in asynchronous vibration in different regions of the ballast, weakening overall structural integrity and making it challenging to form a stable contact network. Intense vibration, particularly near free ends, hinders particle interlocking, giving the ballast fluid-like characteristics. Additionally, a lack of effective lateral support from the shoulder ballast results in outward movement of central ballast particles, causing rapid accumulation of permanent deformation in the trackbed. The findings offer important insights for the future design, construction, and maintenance of high-speed railway ballasted tracks.

In real-world geotechnical engineering applications such as multi-strutted deep excavations, the lack of prompt support for newly excavated faces could lead to horizontal soil arching phenomenon. However, very limited studies have been focused on the horizontal arching effect in clay. This study systematically investigates the development mechanism of horizontal soil arching in saturated clay and its effects on soil responses through centrifugal model test and numerical simulations. The findings reveal that with increasing displacement of the trap door, ground settlement exhibits a concave trough-shaped profile. Soil displacement is primarily horizontal, exhibiting a triangular distribution with peak values proximal to the trap door, whereas vertical displacement typically manifests as settlement above the trap door and heave below it. Negative excess pore water pressure develops in front of the trap door, while positive excess pore pressure arises near its edges. The presence of horizontal soil arching leads to a reduction in lateral earth pressure at the trap door and an increase in stress within the adjacent soil, resulting in the formation of distinct loose and soil arching zones. The study also finds that soil displacement serves as the primary mechanism initiating horizontal soil arching. Following the horizontal trap door’s movement, the soil behind is displaced toward the trap door, driven by horizontal stress forces. Concurrently, shear strains progressively develop along the upper and lower edges of the trap door. Within a distance of one trap door height in front of the trap door, the soil undergoes significant horizontal displacement, delineating a triangular region. This zone is characterized by pronounced stress relief, which subsequently propagates along its upper and lower triangular boundaries, thereby disrupting the local stress equilibrium. As a result, under vertical loading, the soil along the upper boundary experiences settlement, leading to the formation of a vertical soil arch. Conversely, under unloading conditions, the soil along the lower boundary exhibits rebound and heaving behavior, resulting in the development of a reverse vertical soil arch. This displacement-driven mechanism culminates in the formation of horizontal soil arching. These results contribute novel theoretical insights for the analysis of soil–structure interactions influenced by horizontal soil arching effects in clay.

The recirculation of concentrated leachate (CL) in municipal solid waste landfills has been shown to effectively reduce organic biochemical indicators in the effluent; however, it may also increase salinity within the landfill, raising concerns about its long-term sustainability. In this study, a laboratory-scale bioreactor was established to periodically inject nanofiltration-treated CL into waste with high and low organic matter content. Variations in waste composition, leachate characteristics, and landfill gas production were monitored. Results indicated that CL recirculation inhibited waste degradation and methanogenesis, leading to a 47.5% and 77.0% reduction in methane potential compared to the control groups. At the end of the experiment, 34.0% and 33.7% of the biodegradable material remained in the solid phase of the high and low organic waste, respectively. CL recirculation facilitated the release of NH3 and H2S, with total emissions being one to two times higher than those of the control group. It also promoted salinity accumulation; influenced by the temperature of the waste core, the final leachate salinity exceeded that of the recirculated CL, reaching 33,800-36,000 mg/L. Low organic matter waste was more vulnerable to the adverse effects of CL recirculation, with degradation and methanogenesis processes nearly halted. These observations were consistent with salinity-related inhibition and may also have been influenced by free ammonia toxicity and substrate competition involving sulfate-reducing bacteria. Considering methane's economic value, the microbial environment in the waste, and the emission of odorous gases, the sustainability of CL recirculation appears to be relatively low.

Reliable characterization of rock fracture is essential in rock engineering. Rock fracture involves strong discontinuities and evolving crack topology. These characteristics challenge conventional mesh-based methods due to mesh distortion, remeshing requirements, and limited resolution of evolving mesoscale fracture processes. To address these issues, this study presents a meshfree framework that couples Total-Lagrangian Smoothed Particle Hydrodynamics (TL-SPH) with the phase-field method (PFM). Using a volumetric-deviatoric energy decomposition, a crack-driving formulation incorporating deviatoric dissipation is proposed to suppress nonphysical crack growth under compression-dominated states. Three representative brittle fracture benchmarks are used for validation, and the results capture key features of crack initiation, propagation, and coalescence, as well as crack paths and fracture topology. The proposed TL-SPH-PFM framework provides a meshfree approach for simulating brittle fracture processes in rock and rock-like materials.

To improve the measurement accuracy of scaled models in hypergravity environments underwater explosion tests, this research develops a flexible thin-film impact sensor based on polyvinylidene fluoride (PVDF), addressing limitations of traditional rigid sensors such as size effects and intrusive installation. The sensor features an integrated structural design with edge magnification, fabricated through a regional polarization edge depolarization process and protected by a composite thermal insulation layer, enabling reliable operation under extreme environments including hypergravity, strong impact, and transient high temperatures. Simulated hypergravity loading was achieved with prestress under normal gravity conditions, and calibration was conducted using a split Hopkinson pressure bar (SHPB) system to evaluate its force-electric response and stability. In normal gravity underwater explosion tests, the sensor demonstrated high signal fidelity and accurately captured pressure changes from both shock waves and bubble collapse under different prestress conditions, as validated by high-speed imaging. In hypergravity underwater explosion tests conditions, the sensor achieved accurate pressure measurements for the first time, though limited repeatability was observed due to residual stress accumulation in the polarized region. To address this, a gradient polarization encapsulation strategy is proposed to improve stability and reliability under hypergravity. This work provides a new technical pathway for highprecision pressure measurement in hypergravity underwater explosion environments.

With the increasing complexity of excavation projects, the demand for automated monitoring systems has grown due to their ability to provide continuous, real-time data capture. However, traditional methods of using automatic inclinometers are cost-prohibitive due to the dense configuration of sensors required for accurate data collection. This study introduces a novel recurrent neural network (RNN) framework that leverages sparse sensor data to predict excavation-induced horizontal displacements in automated monitoring. The predictive performance of three RNN variants-Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Bidirectional LSTM (BiLSTM)-was assessed and compared against traditional machine learning models, including Backpropagation Neural Network (BPNN) and Support Vector Regression (SVR), using field data from two deep excavation sites in Hangzhou, China. The BiLSTM model achieved the best performance, with a mean absolute error (MAE) of 1.90 mm and a coefficient of determination (R2) of 0.89, outperforming both other RNN variants and traditional ML models. The model's robustness was validated across different soil conditions and excavation strategies. The study also shows that optimized sensor deployment can reduce the number of sensors needed while maintaining high prediction accuracy, offering a cost-effective solution for real-time monitoring in construction projects. (c) 2026 Japanese Geotechnical Society. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/).

Per- and polyfluoroalkyl substances (PFAS) pose significant environmental and human health concerns due to their extreme persistence and bioaccumulation potential, and widespread contamination. Conventional clay-based barriers, such as geosynthetic clay liners, exhibit limited PFAS containment capabilities owing to their inherently low adsorption capacity. To overcome this challenge, this paper synthesized hexadecyltrimethylammonium-modified bentonite (HDTMA-NaB) as a high-performance adsorbent for engineered containment applications. Comprehensive batch adsorption experiments demonstrated that HDTMA-NaB achieves substantially greater PFAS adsorption capacity compared to pristine bentonite, with adsorption kinetics following a pseudo-second-order model and isotherms were well-described by both Langmuir and Freundlich models. Molecular dynamics simulations unveiled a distinctive “head-attracted, tail-parallel” adsorption configuration, where PFAS molecules align with HDTMA surfactants within montmorillonite interlayers. Chain-length-dependent adsorption patterns were observed, with short-chain PFAS preferentially adsorbing at clay particle edges while long-chain PFAS penetrated deeper into interlayer spaces. Potential of mean force analysis quantitatively verified stronger adsorption affinity for long-chain PFAS, consistent with adsorption experiment observations, through deeper and more numerous energy minima. Thermodynamic analysis indicated that PFAS adsorption was driven by both enthalpic and entropic contributions, with entropy playing the dominant role. The entropic contribution primarily arose from hydrophobic interactions between PFAS fluorinated carbon chains and HDTMA alkyl chains, providing a mechanistic explanation for the observed chain-length-dependent adsorption behavior in adsorption experiments. Meanwhile, the enthalpic contribution arose from electrostatic attraction between PFAS anionic headgroups and HDTMA cationic trimethylammonium groups. These molecular-scale insights provide a fundamental basis for designing high-performance clay-based containment systems for PFAS mitigation.

Groundwater seepage is one of the primary drivers for the migration of contaminants at a site. Currently, the long-term direction of groundwater seepage is critical to the service life of vertical pollution-barrier walls in engineering design. Based on the convection–diffusion equation, we created a numerical model in three dimensions for this investigation. The site’s horizontal hydraulic gradient was 0.02 and the barrier’s insertion depth was fixed at 12 m. This was done to examine how the angle (α) between the suspended vertical barrier’s length direction and seepage direction affected the barrier’s resistance to breakthrough from flow around the bottom for contamination containment. The numerical model is then validated against experimental and analytical results in order to demonstrate its accuracy. According to simulation results, changing α alters the distribution of maximum bypass flow velocities and resulting in an uneven flow-field distribution inside the barrier-controlled area. The position of the breakthrough failure gradually moves from the middle of the downstream bottom of the barrier to the end of the barrier when the value of α falls between 45° and 90°. The service life against breakthrough can be increased by up to 10.7

The waste sector is the third-largest anthropogenic source of methane emissions, significantly contributing to the greenhouse effect. Accurate quantification of landfill methane emissions is essential for effective waste management and emission reduction policies. However, current regional inventory-based assessments often overlook site-specific heterogeneity, while emerging satellite observations remain temporally sparse. This study provides a nationwide, site-level characterization and quantification of methane emissions from MSW landfills in China by integrating bottom-up inventory estimates with hyperspectral satellite observations. We first developed a comprehensive database containing site-specific information for more than 300 major MSW landfills to analyze their spatial and temporal distribution patterns. Methane emissions from individual landfills were then estimated using the IPCC first-order decay method, revealing an increase from 1.015 Mt in 2005 to a peak of 2.161 Mt around 2015, followed by a decline to 1.98 Mt in 2023. We further compared these inventory results with hyperspectral satellite observations (EMIT, PRISMA) for three typical landfill sites. The comparison revealed that satellite-detected instantaneous emissions consistently exceed inventory-based averages, quantifying a systematic bias in current IPCC models which tend to underestimate fugitive leaks. Uncertainty in the results primarily arises from parameter assumptions in inventory models and wind-related variability in satellite-based flux inversion. Finally, a scenario analysis projects that the full implementation of China's "Zero Waste" policy could reduce landfill methane emissions by approximately 59 % by 2030. Our findings highlight the importance of combining bottom-up inventories with top-down satellite monitoring for improved landfill management and provide valuable insights for national carbon mitigation strategies.

Erosion-induced removal of fine plastic particles, concurrent with the retention of low-plasticity fine particles, increases the susceptibility of completely decomposed granite (CDG) soil to static liquefaction under undrained shear conditions in laboratory settings. The impact of this phenomenon on slope behavior remains unclear. This study aims to elucidate this issue through centrifuge modeling techniques. Two centrifuge models, featuring identical slope geometries but incorporating CDG with varying fine particle types, were tested using a newly developed inflight angle adjustment apparatus. In the CDG model with low plasticity fines, a rapid accumulation of excess pore water pressure and sudden slope failure were observed when the model angle was adjusted to 13.8 degrees. Static liquefaction was found to initiate at the toe of the slope and propagate upward toward the crest, affecting increasingly deeper soil layers. Conversely, the CDG model containing plastic fines exhibited neither significant excess pore water pressure buildup nor slope failure during the angle adjustment process. The centrifuge observations underscore the critical role of fine particle migration in enhancing the fluidization potential of CDG landfill slopes. Moreover, the study highlights the effects of fluid level variations due to tilting methods on the onset of static liquefaction and the impact of centrifuge densification on the accuracy of instrumentation measurements.

In this study, microfluidic experiments and theoretical analysis are conducted to elucidate how gravitational, viscous, and capillary forces compete to control drainage dynamics and pore-scale interfacial behavior in porous media. A tilted microfluidic platform with precise angle adjustment to modulate the magnitude and direction of gravitational acceleration is used to perform drainage experiments on water–oil and gas–oil displacement pairs under seven gravity conditions and six viscosity ratios. The results reveal that in the capillary-fingering regime, gravity aligned with the flow promotes backward/transverse meniscus motion that stabilizes the front, while counter-flow gravity accelerates longitudinal interfaces, triggering a transition to gravitational fingering. In the crossover zone, gravity aligned with the flow reduces the frontal velocity and viscous resistance, producing a compact tip-splitting morphology. In the viscous-fingering regime, gravity influences pattern evolution without regime transitions, with stabilization dominating below a critical velocity. Quantitative analysis of Haines jumps and the associated energy conversion and dissipation in the capillary–gravitational forces-dominated regime indicates that gravity alters the entry capillary pressure threshold. Increasing gravity in the flow direction shortens the buildup stage preceding Haines jumps, thereby increasing the frequency of these jumps. This induces a more compact displacement morphology, with the net efficiency, denoted as the ratio of external work converted to surface energy, decreasing from 63% to 46% for gas displacing oil and from 50% to 45% for water flooding as gravity increases from 0.5g to 1g. These findings bridge pore-scale interface dynamics and reservoir-scale flows to optimize subsurface processes.

A novel stress-based hybrid phase-field model is developed to describe mixed-mode cracking in frictional rock-like materials, with particular attention to gravitational stress-gradient effects. To capture tensile-shear competition, the formulation introduces two physically motivated driving forces: the tensile stress-related elastic strain energy governing tensile cracking, and an equivalent shear stress-related energy driving shear cracking while incorporating internal friction. Their interaction is embedded in a mixed damage evolution criterion. In addition, a damage-hardening law is proposed to describe the evolution of shear fracture toughness with accumulated local compressive strain, enabling transition between tensile and shear dominated cracking modes. Implemented within a finite element framework, the model is first validated against four representative benchmarks. It is then applied to plate specimens with multiple pre-existing flaws under uniaxial compression subjected to both normal gravity and hypergravity conditions. The results indicate that, under normal gravity, cracks may initiate at different locations and a transition from shear to tensile dominated cracking can occur, whereas under hypergravity cracks preferentially initiate in the lower region of the specimen and propagate upward as tensile wing cracks. Moreover, the crack initiation angle and peak strength are influenced by the gravitational stress gradient. Hypergravity exhibits a clear suppressing effect on shear cracking.

Geological carbon sequestration (GCS) mitigates climate change by storing anthropogenic carbon dioxide (CO2) in geological formations. CO2 undergoes complex physical and chemical transformations in deep geological formations, governed by various interacting trapping mechanisms. Because the trapping mechanisms operate over a wide range of different timescales, their long-term interplay remains unclear. We develop an integrated numerical modeling framework to analyze and track the plume footprint and phase transition processes that occur throughout the entire cycle of the injected CO2 in saline aquifers. The key novelty of the modeling framework lies in its capability to describe multiple hydrodynamic processes and their interactions, including injection, dissolution-driven convection, reactive transport, and gravity-induced Ostwald ripening. The results suggest that dissolution reduces the lateral migration of free-state CO2, while geochemical reactions generate preferential pathways for CO2-rich flow. For the scenarios we analyze, after 500 years of mass transfer, dissolved CO2 accounts for 42.80 % of total trapped CO2 mass, while reactive CO2 contributes less than 1 %. The results also illustrate that low vertical permeability is unfavorable for the long-term transition of CO2 from physical trapping to dissolution trapping. When the permeability anisotropy index γ increases from 0.5 to 10, the total dissolution storage amount within the domain is reduced to one-third over the 500-year simulation period. This integrated modeling framework provides critical insights into the long-term evolution of CO2 plume migration and phase transition behavior, thereby offering a practical tool to quantitatively assess the long-term fate of the injected CO2 in saline aquifers.

With intensifying global climate change, the frequency and intensity of extreme precipitation events continue to rise, posing a significant threat to railway system safety and stability. This study integrates historical records with systematically collected internet-based disaster reports to establish a comprehensive and long-term database of precipitation-induced railway disasters (PIRDs) in China. A systematic analysis is conducted to examine the spatiotemporal distribution patterns of these disasters and their impact on railway systems. The results indicate that PIRDs exhibit strong seasonality, with 83.8