Current three-dimensional (3D) printing materials for rock masses remain inadequate in strength and brittleness to fully replicate real rock mechanical properties. This study presents a novel basalt powder-reinforced resin composite (BPRC) with tunable mechanical properties to address this gap. Uniaxial compression tests under ambient and cryogenic conditions revealed that: (1) BPRC achieved a peak uniaxial compressive strength (UCS) of 79 MPa at ambient temperature, surpassing that of existing rock-like materials; (2) Strength and brittleness increased linearly with basalt powder content; (3) Cryogenic treatment doubled strength while enhancing brittle failure characteristics, closely mimicking natural rock behavior. Additionally, a staged curing method for large-scale models significantly improved curing efficiency and enhanced ambient strength and brittleness, confirming its reliability and potential for rock engineering simulations. These findings provide a new material option for 3D printing highly brittle hard rock physical models to study their failure mechanisms. They also demonstrate the feasibility of printing large-scale, large-strength, and high-brittleness models, offering a process reference for the physical simulation of complex rock mass structures.
Understanding and predicting rock damage evolution under cyclic loading remains a central yet unresolved challenge, particularly when laboratory datasets are limited and commonly used damage indicators lack physical consistency. To overcome these limitations, we propose a physics-informed hybrid learning (PIHL) framework that integrates multi-source mechanical, energy, and acoustic-emission features. The framework embeds physics constraints distilled from consistent damage evolution patterns observed in the cyclic loading tests and employs an LSTM-RF hybrid architecture to fuse heterogeneous temporal features into a unified representation, thereby transforming empirical regularities into learnable supervisory signals and ensuring physically interpretable predictions. Bayesian optimization is further used to adaptively tune the model hyperparameters. The resulting fused damage variable yields physically consistent damage trajectories that align well with multi-source indicators and residual strain evolution, outperforms conventional AE-based indices, and consistently predicts the onset of residual strain acceleration well before macroscopic failure. Evaluation with four physical-consistency metrics confirms strong robustness across the tested confining pressures and loading frequencies. Overall, the PIHL framework effectively addresses the limitations of small-sample rock testing and provides a data-efficient and physically grounded approach under the tested conditions.
Shallow slope instability is frequently triggered by desiccation cracking, yet conventional chemical stabilizers often encounter environmental and economic constraints. This study evaluates the synergistic potential of biochar produced at 200 °C and 500 °C, each added at 9% by dry soil mass, along with tall fescue, to mitigate cracking and improve hydraulic properties. Through laboratory experiments and micro-scale analyses, the underlying mechanisms of this integrated treatment were elucidated. The results show that biochar-vegetation integration significantly suppressed desiccation cracking. Specifically, the 500 °C biochar–vegetation treatment (V-T5) yielded the most effective crack inhibition, reducing the surface crack ratio by 51.4% compared to bare soil (BS) after three dry-wet cycles. Regarding hydraulic performance, V-T5 extended the constant-rate evaporation stage and increased the cumulative retained water by approximately 35% over the tested drying period. Furthermore, the integrated treatment effectively limited the increase in saturated hydraulic conductivity (ks) typically induced by dry-wet cycles. Mechanistic insights indicate that biochar improves water retention through its micropores, strengthens soil aggregation via polyvalent cation exchange, and optimizes nutrient availability for enhanced vegetation growth. These physicochemical effects, together with mechanical root reinforcement and rhizosphere-induced biogenic cementation, form a sequential mechanism from rhizosphere improvement and enhanced plant growth to structural reinforcement. These findings demonstrate the potential of combined biochar and vegetation as an environmentally sustainable approach for stabilizing shallow slopes under alternating wetting and drying conditions.
Deep underground excavations trigger repeated energy releases that often fracture conventional rock support systems. To address this issue, an enhanced multi-stage energy-absorbing cablebolt (MSCB) is proposed. The energy-absorbing system consists of an expandible octagonal tube (EOT), a contractable circular tube (CCT), and a constraining sleeve. Energy is absorbed sequentially through expansion and frictional deformation of the octagonal tube, plastic contraction of the circular tube, and tensile plastic deformation of the cablebolt. By controlling the yield thresholds and deformation paths of these components, the anchoring force increases stepwise to accommodate rock mass deformation. Based on frictional and plastic mechanics, mechanical models were established for each energy-absorption stage. Laboratory pull-out tests and numerical simulations show that the proposed cablebolt achieves a peak load capacity of 320.48 kN, a deformation capacity of 693.65 mm, and a total energy absorption of 166.37 kJ. The first stage provides an anchoring force of 153.42 kN with 73.32 mm deformation and 14.02 kJ energy absorption. The second stage maintains a stable load of 262.18 kN while accommodating 484.89 mm deformation and absorbing 140.57 kJ. In the third stage, the load increases to 320.48 kN before failure, with 43.88 mm deformation and 13.62 kJ energy absorption. Dynamic impact tests show that the cablebolt can withstand multiple consecutive 40 kJ impacts, producing an axial displacement of 744.01 mm and absorbing 147.67 kJ of energy. Good agreement among theoretical, numerical, and experimental results confirms the effectiveness of the proposed cablebolt for mitigating rockburst hazards in deep underground engineering under repeated energy-release conditions.
Monitoring tunnel displacements during construction is an essential element of the New Austrian Tunnelling Method (NATM) design philosophy, and it is crucial for ensuring safety and predicting the stability of the supported excavations. This paper utilizes Bayesian methods to update uncertain parameters affecting convergence in tunnels constructed in soft rock and subjected to time-dependent deformations, such as the thickness of concrete and the geomechanical parameters of the rock mass, i.e. Young's modulus, cohesion, and friction angle, and to account for existing uncertainties. The discrete element method (DEM) modeling approach combined with the rate process theory (RPT) is used to simulate tunnel excavation and performance; and the support vector regression (SVR) method is employed to establish an efficient surrogate model that maintains solution accuracy while achieving significant computational efficiency improvements. The Heshan Tunnel case in China is used to demonstrate the feasibility and effectiveness of the proposed method, and two Bayesian updating processes utilizing monitoring data at critical temporal points of the tunnel convergence curve demonstrate that the proposed approach significantly reduces the uncertainty of random variables, also improving the accuracy of subsequent numerical simulations.
Sandy soils often suffer from low geotechnical stability and require effective reinforcement. Microbially induced carbonate precipitation (MICP) has emerged as a promising method, yet spatial heterogeneity of CaCO3 precipitation restricts uniform reinforcement. To alleviate this issue, this study leveraged the bacterial-adsorption capacity of biochar (BC) by stratifying the soil matrix and incrementally increasing the BC dosage along the depth direction. Meanwhile, an existing transport-reactive model was refined by revising the urea hydrolysis kinetics equation to incorporate bacterial attachment, detachment, and decay, and was validated through solution tests and sand column experiments. Results show BC enhanced bacterial retention and CaCO3 precipitation, whereas uniform BC incorporation intensified the vertical heterogeneity of bacteria and CaCO3, leading to a typical “dense-top, loose-bottom” pore structure. In contrast, the gradient BC configuration promoted a more balanced vertical distribution of bacteria and CaCO3, thereby improving the coordination of porosity and permeability evolution. Among all tested conditions, the gradient BC configuration with a total dosage of 3 wt% (1.5, 3, and 4.5 wt% in the upper, middle, and lower layers, respectively) exhibited the most favorable overall performance, with the highest CaCO3 uniformity index of 0.71. Measured and simulated bacterial concentration, CaCO3 content, and permeability showed good agreement, supporting the applicability of the proposed model. The study demonstrates that gradient BC incorporation is an effective internal regulation strategy for enhancing the uniformity, while the revised transport-reaction model showed satisfactory predictive capability and can support MICP parameter selection and process design for sandy soil reinforcement engineering.
Blasting-induced rock fragmentation in sandstone is commonplace in tunneling and mining engineering. This paper presents an experimental and numerical investigation into the disintegration characteristics and energy consumption of sandstone under blast loading. Twelve lab-scale blast tests using sandstone cylinders are conducted with three stemming configurations (no-, partial-, and full-stemming) and four charge densities. Blast-produced fragment size distributions are first obtained by sieving and weighing. The shape of fragments is analyzed by image processing, and the energy consumption is determined using the Griffith fracture criterion. Additionally, LS-DYNA simulates pressure evolution, crack development, sandstone disintegration, and energy dissipation. The results quantitatively demonstrate that rock disintegration becomes significantly finer in a sequence of no-stemming, partial-stemming, and full-stemming, with the average fragment size decreasing by approximately 18.4% and 27.9% for partial- and full-stemming configurations, respectively, compared to unstemmed blasts. The aspect ratios of fragments are primarily falling within the range of 0.6-0.8, showing no statistically significant dependence on stemming configuration (p > 0.5, rs = -0.062). The proportions of explosive energy utilized for rock fragmentation under unstemmed, partially stemmed, and fully stemmed blasting are quantified as 5.15%–8.54% (mean 6.57%), 5.90%–11.90% (mean 9.36%), and 8.28%–14.00% (mean 10.40%), respectively, with the explosive energy utilization enhanced by approximately 42.47% (partial-stemming) and 58.30% (full-stemming) compared to the unstemmed case. Numerical modelling confirms that stemming enhances confinement, prolongs pressure duration, and improves the energy transfer from explosive detonation to internal and kinetic modes, thereby leading to greater efficiency in both rock fragmentation and energy conversion.
This study evaluates the effectiveness of microbially-induced calcite precipitation (MICP) as a bio-cementation strategy for soil stabilization in shallow soils without direct access to the soil surface (e.g. beneath pavements or other surface-impermeable infrastructure), with emphasis on injection method and treatment scale-up. The objective was to assess how different delivery strategies and solution formulations influence calcium carbonate distribution and mechanical improvement in treated soils. Experimental testing was conducted at two scales: small-scale (15-cm diameter) columns to investigate injection layout and media chemistries, and an intermediate-scale (30-cm-by-30-cm) specimen to explore spatial uniformity and scale-up effects. Two injection methods, top-down (TD) and injection well (IW), were applied using both laboratory-grade and commercially sourced nutrient recipes. Strength was assessed through California bearing ratio (CBR) testing, and calcium carbonate precipitation was quantified via acid digestion across multiple depths. The CBR test was selected as a practical and minimally invasive method to evaluate strength improvements across multiple depths and locations, and it correlates well with dynamic cone penetration (DCP) test results for assessing near-surface strength in confined systems. Results demonstrated that the IW method improved deeper treatment while TD method enhanced surficial bonding. It was observed that commercial yeast-extract-based media produced comparable or greater strength improvements than traditional lab-grade solutions. CBR values increased from 11% to as high as 188%, indicating the potential of MICP for in situ strengthening of soil. While findings support the viability of MICP for scalable ground improvement, limitations related to sample replication and uniform fluid distribution in larger specimens warrant consideration.
Transient fluctuations in groundwater and surface water levels caused by extreme climate events increasingly threaten the hydro-mechanical stability of transportation embankments. This study investigates the coupled seepage-deformation behavior of geosynthetically reinforced embankments subjected to cyclic water-level rises and rapid drawdowns through large-scale 1g physical modeling. Four configurations with unreinforced, geotextile-, geogrid-, and geocomposite-reinforced embankments were tested under two representative hydraulic boundary conditions, namely basal-permeable (combined internal-external infiltration) and basal-impermeable (external-only infiltration). Results show that boundary permeability primarily controls pore-water pressure redistribution and transient seepage gradients. High-transmissivity geotextiles accelerated excess pore-pressure dissipation, reducing drawdown stabilization time by about 40%. The geocomposite provided the most effective pressure relief, mainly through internal in-plane drainage within the core layer, creating a preferential flow pathway and lowering adverse hydraulic gradients. Mechanically, reinforcement greatly impacted deformation patterns and failure modes. Maximum settlement decreased by 28%–46% compared to the unreinforced case, with the geocomposite showing the least cumulative settlement (around 5.4 mm) due to combined tensile confinement and hydraulic dissipation. The geogrid provided superior control of lateral displacement through enhanced interlocking and increased stiffness. Conversely, the unreinforced embankment experienced significant tensile cracking and traction-type instability during rapid drawdown, linked to delayed pore-pressure dissipation and outward seepage forces. These findings highlight that embankment resilience under cyclic hydraulic loading depends on the interaction between drainage capacity and tensile stiffness rather than on reinforcement strength alone, providing a mechanistic basis for performance-based geosynthetic selection in climate-resilient embankment design.
This study presents a novel self-healing system that uses epoxy resin microcapsules to encapsulate Sporosarcina pasteurii spores for autonomously sealing fractures in rock masses. The microcapsules, known for their strong mechanical properties and high encapsulation efficiency, are designed to rupture when cracking occurs, thereby initiating microbial-induced calcium carbonate precipitation (MICP). An extensive experimental program, including Brazilian splitting tests, industrial computed tomography (CT) scanning, and triaxial permeability tests, was conducted to assess the system's effectiveness. Results showed that crack width was the main factor influencing healing efficiency, with up to 89.3% healing observed in 0.3–0.4 mm cracks after 28 d. An optimal mix, consisting of 2% content and 150 μm particle size, decreased the permeability coefficient by over 70% on average, roughly one order of magnitude. The relationship between permeability and confining pressure followed a power law, and notably, the sensitivity of permeability to stress changes decreased significantly in well-healed specimens. Numerical simulations supported these results, revealing that MICP healing greatly improved the hydraulic barrier function of rock masses by lowering seepage velocity and raising the pressure gradient needed for fluid flow. Overall, this work quantitatively clarifies the relationship between healing extent, permeability reduction, and in situ stress, providing a scientific foundation for designing engineered microbial barriers for geotechnical applications such as groundwater isolation and seepage control.
The pronounced strain-softening behavior of cement-stabilized clay, particularly its evolution with the incorporation of supplementary cementitious binders, remains poorly understood. This study investigated dredged clay stabilized with ordinary Portland cement (OPC), red mud (RM), and phosphogypsum (PG) through a series of isotropically consolidated undrained (CIU) triaxial tests. The strain-softening response of stabilized clay was evaluated under varying water-to-cement ratios, industrial waste binder (IWB) proportions, and consolidation pressures. Scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) tests were conducted to explore the microstructural evolution. The results show that the brittleness index increases with decreasing water-to-cement ratio and increasing RM content. The enhancement is primarily attributed to more complete pozzolanic reactions and a more pronounced cementation structure. As consolidation pressure increases, brittleness initially drops sharply before reaching a plateau at higher pressures. The coexistence of cementation damage and secondary cementation formation is demonstrated in the stabilized clay under compression. Moreover, binary medium theory was applied to interpret the underlying mechanism. The analysis suggests that the evolution of the strain-softening response depends on the synergistic interaction between bonded elements and frictional elements within the OPC-RM-PG stabilized clay. The addition of RM promotes C-S-H formation and increases the proportion of bonded elements. In contrast, PG promotes the formation of ettringite, which acts primarily as a pore-filling material and increases the proportion of frictional elements, resulting in a transition toward strain-hardening behavior.
To address the significant carbon footprint challenge of existing ground treatment methods, a novel magnesium oxide-based steel pipe carbonation composite pile (MSCP) technology was proposed. Field tests, microstructural characterization, and carbon footprint assessments were conducted to evaluate its solidification characteristics and carbon reduction efficacy. Results indicate a significant positive correlation between the maximum temperature increment (ΔTmax) of the MSCP outer pile and its 14-d unconfined compressive strength (qu), providing a real-time strength control method based on critical and safe ΔTmax. Higher gas injection flow rates are found to significantly enhance carbonation uniformity and sufficiency. Microscopic observations reveal that the hydrated magnesium carbonates (HMCs) reshape the soil microstructure through densification, filling, interlocking, and cementation. The carbon emissions of MSCP single-pile are reduced by approximately 30% compared to traditional solutions, primarily stemming from the use of low-carbon binders and the utilization of industrial CO2 tail gas. MSCP technology not only offers a new ground treatment method for soft ground with controllable performance, but also opens up a new pathway for carbon sequestration and emission reduction in geotechnical engineering.
The geological conditions of deep engineering are complex and are impacted by high stress and excavation disruption, which can lead to engineering disasters like severe surrounding rock deformation and roof collapse. The real-time, accurate measurement of rock mechanical parameters is a necessary condition for analysing the stability of deep surrounding rock and preventing disasters. In this paper, drilling tests are conducted with an intelligent detection anchoring system for underground engineering. An investigation is conducted into the correlation and change law between drilling parameters and the elastic modulus E, and a testing model for E while drilling, is proposed. According to the test results, the average difference in E is 0.43 GPa, and the average error rate is 5.49%. On this basis, an in situ zoning approach for surrounding E while drilling is established, and field drilling experiments are carried out for deep surrounding rock. A three-dimensional (3D) zoning and mapping relationship model is established for the equivalent elastic modulus Eeq of the surrounding rock, enabling in situ testing of the Eeq of the surrounding rock and effective identification of rock interfaces and fracture zones. While constructing the boreholes for rock bolts and anchor cables, the drilling testing method established can constantly and quickly determine the E of rock in real-time. This work provides theoretical and technical support for the on-site, real-time acquisition of rock mechanical parameters required for preventing and controlling failures of surrounding rock in underground engineering.
Slurry-infiltrated granular composites offer a sustainable strategy for underground backfilling, in which a pre-formed aggregate skeleton is bonded by an infiltrating slurry to reduce binder use. Their mechanical integrity is governed by slurry retention within intergranular voids rather than uniform mixing. However, a quantitative link between fresh-state rheology and both the resulting bonded configuration and hardened mechanical performance remains insufficiently established. This study develops an experimental–theoretical framework linking slurry rheology, retention behavior, and uniaxial compressive response. Slurries with varying rheology were prepared by adjusting polycarboxylate ether dosage, and infiltration tests were performed using coarse gangue aggregates of different sizes. The retained slurry mass () was measured and used to define a structural state descriptor, the initial bonded volume fraction (), to characterize the initial bonded configuration. A constitutive model is formulated by decomposing total strain into elastic and compaction components, while progressive degradation of bonded regions is described using a statistical damage approach. Model parameters are identified independently from distinct segments of the stress–strain curves to ensure identifiability. Validation using independent mixtures demonstrates that the proposed model captures early-stage compaction, elastic stiffness, peak strength, and post-peak softening, with relative errors generally within 15%. Within the present uniaxial material-point scope, slurry retention provides a measurable link between fresh-state rheology and hardened mechanical performance, supporting rapid comparison and preliminary assessment of load-bearing behavior in slurry-infiltrated granular systems.
Nearly 30% of surface soils in Australia can be classified as expansive. Structures built on these soils face additional stresses from ground movement associated with moisture changes. Lightweight structures founded at shallow depths, such as residential footings, pipelines, and road pavements, are especially vulnerable. Ground movement in expansive soil is often controlled by interactions at the soil–vegetation–atmosphere (SVA) boundary. During wet seasons, moisture content increases due to precipitation, and in dry seasons, evaporation/evapotranspiration occurs. This moisture cycle can create shrinkage cracks extending to significant depths in certain soil types. Such cracks are likely to affect moisture movement through the soil and, in turn, ground movement. This study investigates the effect of cracks on SVA boundary interactions and the expected changes in the behaviour of geotechnical structures due to their presence. Depth and width of cracks were monitored across different seasons at a research site near Adelaide, Australia. The SVA boundary interactions with soil cracks were simulated to better understand their effect on moisture dynamics. The effect of such changes was evaluated for a hypothetical residential footing design in a semi-arid climate zone, highlighting the importance of accounting for local soil features such as cracks in the current design practices.
Probabilistic analysis of rock structures is often impractical due to insufficient input data, limiting reliable probabilistic modelling. Although non-probabilistic uncertainty models such as interval/fuzzy and convex approaches offer alternatives, their effectiveness in capturing spatial variability with limited data remains inadequate. Existing interval/fuzzy field models rely only on bounds/bounds with a most plausible value, leading to loss of available information and inadequate treatment of input correlations and spatial dependency. To address these limitations, this study proposes a non-probabilistic, optimized, bounded-field reliability framework that explicitly accounts for the exact level of available information. A minimum-volume super-ellipsoid convex model is first constructed to optimally bound the input data. This optimized model is adopted for its ability to capture inter-parameter correlations, thereby reducing the over-conservatism associated with conventional hyper-box and ellipsoidal models. Spatial variability is represented using reduced eigenvectors and uncorrelated coefficients through a non-probabilistic series expansion (NPSE). The generated bounded field is then implemented in a FLAC-2D (fast Lagrangian analysis and continua) numerical model while preserving spatial dependency. The key novelty of this work lies in integrating an optimized convex uncertainty representation with spatially correlated bounded-field generation, and in quantitatively benchmarking against established methods. The methodology is demonstrated through a rock slope case study, in which the simulated spatial data remain bounded within the super-ellipsoid, validating the proposed framework and yielding factor of safety (FOS) bounds [2.17, 7.52], thereby confirming slope stability. Comparative analyses using interval and random field approaches further show that the proposed method avoids overestimating response bounds while eliminating the need for probabilistic distributions under limited data conditions.
Open-pit mines possess abundant water resources, and the influence of groundwater on the strength of ore rock intensifies with increasing mining depth. The blastability of water-saturated ore rock differs from that in its natural state. To investigate the effects of air–water coupled media on the blasting damage and fragmentation characteristics of iron ore under both natural and saturated conditions, hematite was examined. X-ray computed tomography (CT) scanning and three-dimensional (3D) reconstruction, combined with fractal theory, were utilized to analyze the 3D fractal dimensions and damage degrees of blast-induced fracture fields with different charge structures. Furthermore, the sieving method was employed to determine the fragmentation fractal dimension D, which is a quantitative indicator characterizing the complexity of particle distribution following blasting. A higher D value signifies more thorough fragmentation of the ore rock and a more uniform particle size distribution. Results show that, under identical charge conditions, water-saturated iron ore experiences less blasting damage and fragmentation than natural ore, with fragmentation fractal dimensions of 1.59 and 1.7, respectively. Under saturated conditions, air-coupled charges perform poorly, whereas water-coupled charges enhance effectiveness. When the explosive is placed at the bottom of the blast hole in a water-coupled structure, cracks propagate horizontally, leading to a fragmentation fractal dimension of 1.8. Conversely, when a water column is present at the bottom, cracks propagate obliquely and vertically, yielding a fragmentation fractal dimension of 1.73 but decreased toe formation. This study provides a strong foundation for designing blasting parameters in water-rich metal mines.
Loess-mudstone interfacial landslides represent a major geological hazard on China’s Loess Plateau. The existence of interfaces makes the properties of heterogeneous soil structures different from those of single homogeneous soil, and brings new disaster effects, most of which are related to water seepage. However, the microstructural controls on the stagnant water effect at the loess-mudstone interface remain insufficiently understood. To address this issue, loess-mudstone samples from Baota District, Yan'an City, China, were investigated using nuclear magnetic resonance (NMR), scanning electron microscopy (SEM), and laser particle size analysis. The test results show that the loess-mudstone mainly undergoes a coupling effect of wetting collapse and swelling softening. Under water supply condition, the high initial water content of the mudstone limits its swelling capacity, resulting in an increase in microcracks at the interface during the seepage, weakening the stagnant water effect, promoting the connection with the top fractures, and ultimately triggering the loess-mudstone cut-through strata landslide. In contrast, extreme rainfall triggers a squeezing-pore clogging cycle, where the continuous swelling of the mudstone and the disintegration of aggregates block the pore throats, enhancing the stagnant water effect, and directly triggering sliding along the interface. These findings are helpful for further developing and improving the research on the mechanism of loess-mudstone landslides, and provide a scientific basis for the prevention and mitigation of such landslides.
Coarse-grained red-stratum mudstone fillers are extensively employed in subgrade construction in eastern Tibet. A detailed understanding of their hydro-mechanical behavior is required to anticipate and mitigate the geotechnical hazards associated with them. This study investigated the impact of water content and dry density on this behavior using nuclear magnetic resonance, scanning electron microscopy, soil-water retention tests, and triaxial tests. A dilatancy equation applicable to red-stratum mudstone fill was developed by incorporating energy dissipation associated with particle breakage. Water immersion induces the development of porous, honeycomb-like microstructures in the mudstone, characterized by unevenly distributed pores, increased microporosity, and microfractures, ultimately reducing its structural integrity. The dry density and initial water content substantially influence the water-retention properties and dual-pore structure of the fillers. Furthermore, the van Genuchten model proves effective in predicting the soil-water retention curves under various initial states. Red-stratum mudstone fillers display distinct hydro-mechanical behavior, with peak shear strength positively related to initial dry density and negatively related to water content. The critical stress ratio varies with confining pressure instead of remaining constant. In the e–ln p plane, the critical-state lines at different initial dry densities remain parallel, while their slopes decrease as the initial water content increases. Using the test data, a novel dilatancy equation incorporating energy dissipation from particle breakage was derived. Comparison with the test data confirmed that the developed dilatancy equation accurately captures the dilatancy characteristics of red-stratum mudstone fillers.
As a sudden geological hazard chain, landslide-induced river blocking involves coupled interactions among landslide mass, water, and air. This study proposes a scenario simulation framework for landslide-induced river blocking based on the multiphase flow finite volume method (FVM). In this framework, unmanned aerial vehicle (UAV) imagery and point-cloud data are integrated with non-uniform rational B-splines (NURBS) surface fitting to construct a three-dimensional real-scene model for finite volume simulation. The rheological model and volume of fluid (VOF) interface-tracking method are then used to couple the three phases and reproduce the process from slope failure to river blockage. Finally, simulation results are integrated with the real-scene model in ParaView for dynamic visualisation. The 2023 Xiongshan Village landslide in Songpan, Sichuan Province, China, is used as a case study. The simulated accumulation morphology agrees well with UAV photogrammetry, with an error of 8.93%, confirming the engineering applicability of the proposed framework. Based on material behaviour and energy evolution, the disaster process is divided into three stages: initial rapid acceleration, intermediate gradual deceleration, and terminal water-entry retardation. Potential energy is converted into kinetic energy and frictional dissipation, explaining the velocity attenuation mechanism. Phase-fraction evolution shows intense water-air interface deformation and local three-phase mixing during water entry. A post-disaster scheme combining “monitoring and early warning - emergency traffic assurance - permanent treatment” is proposed to support risk mitigation and engineering decision-making.
