Bulletin of Engineering Geology and the Environment最新文献

Large reservoirs can intensify erosion and collapse of drawdown-zone soil banks, yet their evolution under contrasting geological conditions remains poorly understood. We combined field surveys along the Badong–Wushan reach of the Three Gorges Reservoir Area with four physical model tests on silty-clay and gravelly-soil banks. Wave motion, bank morphology, horizontal erosion distance, and horizontal earth pressure were monitored; erosion rates were derived from distance–time records. Erosion began near the still-water level within the run-up–backwash zone but followed contrasting pathways. The 60° silty-clay bank reached a horizontal erosion distance of 16.79 cm at 7 min, before overall collapse at approximately 8 min. The 16° silty-clay bank showed surface erosion and progressive retreat. The 34° gravelly-soil bank exhibited preferential fine-particle loss, exposure and rearrangement of embedded coarse particles, and progressive retreat with local collapse. The 12° gravelly-soil bank showed shallow erosion, surface coarsening, and 38.2 cm of horizontal retreat. Fitted continuous erosion rates decreased nonlinearly, while apparent interval rates showed peaks during rapid profile adjustment or detachment. Horizontal earth-pressure responses tracked morphological changes through local loading, progressive unloading, and rapid unloading after collapse. Field and model observations shared erosion notches, wave-cut notches, smooth erosional surfaces, and toe deposits. Under the tested conditions, slope angle was associated with wave-action extent and erosion propagation. Matrix strength and soil–rock composition were associated with post-erosion integrity, particle rearrangement, and failure mode.

The roughness of the structural surface serves as a pivotal factor that substantially influences the mechanical properties of the interface. This study identifies a physically interpretable critical normalized roughness governing interface shear mechanism transition and establishes a coupled framework linking friction mobilization with particle breakage evolution. Large-scale three-dimensional pile-soil interface shear tests were conducted on carbonate and silica sands using concrete and aluminum alloy piles with controlled surface roughness to investigate roughness effects on interface shear strength, friction coefficient, and particle breakage under circumferential pile-soil contact conditions. The results revealed that shear stress–displacement curves from all tests exhibited strain-softening. Compared with silica sand, carbonate sand demonstrated higher peak shear strength and more pronounced softening characteristics due to particle breakage and its irregular particle geometry. A critical normalized roughness equal to 0.25 was identified: below this value the friction coefficient increases with normalized roughness, while above it the friction coefficient tends to plateau. Using this critical normalized roughness as a boundary, the relationship between the friction coefficient and normalized roughness can be accurately described by a piecewise power function. Although an increase in surface roughness leads to enhanced particle breakage, the growth rate of breakage diminishes as roughness further increases. These findings provide a physically based reference for interpreting roughness-dependent interface behavior and offer guidance for surface roughness optimization in geotechnical interface design.

Lateritic earth dams were subjected to drying-wetting (D-W) cycles, which induced desiccation cracks and threatened its safety. This study examined the cracking behavior and micro-scale mechanisms of lateritic soil under varying compaction states and five D-W cycles. Desiccation cracks were quantified using image processing technique, while microstructure was examined by the mercury intrusion porosimetry and scanning electron microscopy tests. The results showed that the surface crack ratio R_sc decreased as compaction water content increased from the dry to the wet side, as reduced macro-pore volume and connectivity inhabited crack formation and penetration. With increasing dry density ρ_d, R_sc increased for the lateritic soil in present study but decreased for the pure clay soils reported previously, which was attributed to their distinct soil fabrics — a coarse–coarse contact dominated fabric for the lateritic soil and a fine–fine contact dominated fabric for the pure clay soils. Increasing ρ_d reduced permeability and intensified water content gradient between the specimen surface and bottom, thereby increasing tensile stress in both fabrics. Meanwhile, increasing ρ_d slightly increased tensile strength for the former fabric but significantly increased it for the latter, resulting in opposite cracking trends. With increasing D-W cycles, cracks underwent three stages: initiation (N = 0–1), propagation (N = 1–3), and stabilization (N = 3–5). These cracks acted as preferential pathways for deep soil evaporation and rainwater infiltration, threatening earth dam stability. These findings support stability assessment of lateritic earth dams under reservoir fluctuations.

Reservoir slope stability is affected by the spatial variability of soil parameters and time-varying environmental loads, including precipitation and reservoir water-level fluctuations. Traditional numerical methods for time-dependent probabilistic stability analysis are computationally demanding, while existing surrogate models often overlook temporal correlations or spatial variability. This study proposes a deep learning-based surrogate framework that combines a residual convolutional neural network (ResNet), a convolutional block attention module (CBAM), and a bidirectional long short-term memory network (BiLSTM), termed the ResNet-BiLSTM-CBAM network (RBLA-Net). ResNet extracts spatial features from random-field representations of heterogeneous soil parameters, while CBAM adaptively recalibrates these features. In parallel, BiLSTM learns the temporal evolution of environmental loads. The fused features enable RBLA-Net to predict the time-dependent factor of safety (F_s) and its probability distribution. The Baijiabao slope in the Three Gorges Reservoir Area, China, is used to evaluate the framework. RBLA-Net outperforms two baseline models, achieving a coefficient of determination (R²) of 0.975, compared with 0.922 and 0.946, respectively. It also achieves a mean absolute error (MAE) of 0.035 and a root mean square error (RMSE) of 0.044, compared with 0.050 and 0.053 for MAE and 0.064 and 0.077 for RMSE. Moreover, once trained, RBLA-Net reduced the computation time for a full-year analysis from 24,230 s to 15 s. These results indicate that RBLA-Net provides an efficient surrogate for time-dependent probabilistic stability analysis of reservoir slope.

To investigate the instability process of sandy loess slopes under mining disturbance-rainfall infiltration effects, a two-dimensional physical model test was conducted on the sandy loess slope in Dabaibao Village, Shenmu City, Shaanxi Province. Displacement monitoring, crack statistical analysis, and pore structure characterization were employed to focus on the deformation response and failure mechanism of sandy loess slopes under the influence of mining disturbances-rainfall infiltration. The results indicate that slope failure exhibits a progressive evolution characterized by “tension-shear-sliding”, which can be divided into four stages: mining-induced pre-damage, tensile crack propagation, slope toe localized slip, and overall sliding instability. During the test, the maximum settlement displacement was 2.1 cm, equivalent to 2.1 m at the prototype scale (1:100), demonstrating significant localized deformation characteristics. The development of cracks intensified notably during the rainfall phase, with the rate of change in the slope’s fractal dimension being higher than that during the mining phase. These observations support the interpretation that mining disturbance predisposed the slope to subsequent instability through initial structural damage, whereas the subsequent rainfall phase was associated with accelerated crack propagation, deformation, and eventual sliding. Furthermore, pore structure coarsening and connectivity enhancement were most pronounced in the slope surface area, followed by the toe, and were relatively weaker at the top, indicating a strong correlation between microstructural degradation and the progression of macroscopic sliding. These research findings provide experimental evidence for the stability assessment and disaster prevention of sandy loess slopes in mining areas.

Understanding coupled heat and moisture transport in soils is essential for evaluating the performance of near-surface geotechnical and geoenvironmental systems subjected to atmospheric thermal forcing. This study investigates the depth-dependent thermo-hydraulic behavior of near-surface unsaturated silty soil through controlled laboratory soil-column experiments, complemented by field observations from a landfill cover system. Two complementary laboratory experiments were conducted. The 8-h continuous-heating tests varied initial gravimetric moisture content (6%, 12%, and 18%) and dry density (1.5, 1.6, and 1.7 g cm^− 3) to evaluate their effects on thermal propagation and moisture redistribution. A separate 30-day diurnal cycling test was performed at a dry density of 1.6 g cm^− 3 and an initial gravimetric moisture content of 18% to quantify depth-dependent attenuation and phase lag under repeated thermal forcing. The continuous-heating tests showed that initial gravimetric moisture content primarily controlled the magnitude of the thermal response, with dry-density effects becoming more apparent under wetter conditions. The cyclic test demonstrated persistent attenuation and delayed thermal response with depth, while deeper layers remained comparatively buffered. Field monitoring showed a similar depth-dependent response pattern under natural atmospheric forcing. The findings of this study provide a scientific basis for the design and optimization of cover barrier systems and offer a valuable dataset for the development and calibration of coupled thermo-hydraulic analytical and numerical models.

Given the current lack of systematic research on the parameters of the hardening soil model with small-strain stiffness (HSS) for alluvial-proluvial sand in eastern Beijing, this study addresses this gap by relying on the Zone 02B project of the Beijing Sub-center Comprehensive Transportation Hub. Through comprehensive laboratory tests, key parameters of the HSS model, particularly the stiffness and small-strain characteristics, were systematically determined. By integrating these experimental results with empirical values, a complete HSS parameter set for the regional sand was established. The research reveals the unique mechanical characteristics of the sand in this area: the reference threshold shear strain (({gamma _{0.7}})) is concentrated around (4.0 times {10^{ - 4}}) (approximately twice that of international standard sand), and the failure ratio (({R_{text{f}}})) is approximately 0.72, which is significantly lower than the conventional default value of 0.9. To facilitate practical engineering applications, a conversion model based on conventional site investigation indices was established, and a stiffness proportional relationship referencing the compression modulus (({E_{text{S}}})) was proposed: ({E_{text{S}}}):(E_{{50}}^{{{text{ref}}}}):(E_{{{text{oed}}}}^{{{text{ref}}}}):(E_{{{text{ur}}}}^{{{text{ref}}}}):(G_{0}^{{{text{ref}}}})≈1:1.19:1.17:4.88:7.32. Finally, a three-dimensional slice model of a representative cross-section was used to verify the complex “pit-in-pit” excavation in Zone 02B. The results demonstrate that the HSS model utilizing the proposed parameters can control the deviation between the calculated and the measured maximum lateral displacement of the retaining structure to within 2 mm, significantly outperforming both the Hardening Soil (HS) and Mohr-Coulomb (MC) models. This study provides a reliable basis for determining HSS model parameters for sand strata in this region.

Directional drilling technology is widely used worldwide to construct underground pipelines and utility corridors. However, the process generates large volumes of drilling fluid and excavated solids that present substantial waste-management challenges. Disposal is often constrained less by the absence of treatment technologies than by their site-specific cost, practicality, and regulatory acceptance. Solid-liquid separation can substantially reduce the volume requiring disposal; the separated water may be treated, while the solids may be reused, solidified, or landfilled. In many projects, local treatment capacity is limited, so construction units contract specialist companies to handle the waste mud, which increases project cost. Inadequate management can therefore create construction, financial, and environmental risks. This review synthesizes international literature and reported practice. It examines waste-mud generation and composition; compares regulatory and management approaches across North America, Europe, Oceania, and other jurisdictions; evaluates source reduction, gravity separation, physical-chemical conditioning, solidification, landfill, and land application; and reviews resource-utilization pathways after solid-liquid separation. A prioritized selection framework is proposed to help engineers compare treatment options by mud properties, site constraints, environmental risk, and relative cost.

Enzyme-induced carbonate precipitation (EICP) is an innovative soil stabilization method, though its performance can be limited by insufficient nucleation sites and uneven calcite distribution. This study investigates the use of coal-derived char as an additive in EICP-treated sodium bentonite-sand mixtures to improve their mechanical and physicochemical properties. Soybean-derived urease was used to catalyze carbonate precipitation in mixtures containing 0 to 20% coal char by dry weight of the soil. Performance was evaluated using bulk density, pH, thermal conductivity, calcium carbonate content (C_cc), unconfined compressive strength (UCS), California bearing ratio (CBR) with soaked swelling tests, unconsolidated undrained triaxial compression, XRD, and SEM analyses. EICP treatment with coal char improved soil performance. UCS increased by up to 35%, with 10% char yielding optimal strength across curing periods. Calcium carbonate content (C_cc ) reached 2.79% at 28 days with 10% char, compared to 0.18% in controls. Bulk density increased by 15–22%, swelling decreased, and CBR values improved. Thermal conductivity declined to 0.35 W/mK at 20% char. SEM observations are consistent with calcite precipitation on and within char pore surfaces, enhancing nucleation and matrix densification. The combined EICP-coal char system offers a promising method for stabilizing expansive soils while valorizing coal by-products.

Land subsidence induced by thermal stress degradation of rocks surrounding coal seams during the underground coal gasification (UCG) and underground coal-mine fire (UCF) constitutes a significant engineering geological hazard. The Jharia basin of India represents a critical case, as it is both a potential site for syngas production using the underground coal gasification techniques and is extensively affected by underground coal fire. To better understand the response of Jharia sandstone in extreme thermal environments, this research examines the physico-mechanical behaviour and fracture dynamics at various treatment temperatures between 25 °C and 800 °C. The research examines how grain size affects the strength and acoustic emission properties of sandstone at elevated temperatures. Crack initiation and damage thresholds were determined from linear-to-exponential trends in cumulative acoustic emission (AE) counts that marks the onset of stable microcrack nucleation. In contrast, crack damage stress threshold ratio indicates accelerated crack coalescence, evidenced by sharp increases in acoustic emission counts and energy. Microcracking initiated at 400 °C, and the strength parameters peaked when transgranular cracks were dominant. However, the strength parameters decreased significantly when the intergranular cracks were dominant at 600 °C. The grain-size and acoustic emission analyses revealed that the cumulative count and cumulative energy increased with increasing grain size. In contrast, the crack damage stress threshold showed an inverse relationship with the grain size.