The control of hidden hazards of large-scale high tailings ponds is an important task and key project to contain major accidents during the "13th Five-Year Plan" period. Key issues closely related to this research topic are the evolution of the meso-structure and macro-mechanical properties of tailings materials under high stress, the sedimentary regularity of large tailings dams, and the theory of degradation of high tailings dams under complex conditions. In order to reveal the meso-structure characteristics of the tailings, the tailings materials can be screened by optical microscope, electron microscope scanning, and CT scanning. In this paper, a three-dimensional reconstruction model of the tailings materials is established. In order to reveal the mechanical properties of tailings materials under high pressure, a high-stress triaxial experiment and high-stress conjoined consolidation permeability tests are carried out, and the strength, permeability and consolidation characteristics of the tailings materials under high stress are analyzed. To reveal the sedimentary characteristics, a large-scale model test of tailings materials is carried out. The spatial distribution of the slurry flow, siltation form, the physical and mechanical properties of the tailings, and the spatial distribution of the sandwich structure are obtained. To reveal the failure mechanism of high tailings dams, the strength criteria of tailings under high pressure are discussed through theoretical analysis and numerical models. The strength reduction stability analysis of high tailings dams is recommended. In order to reveal the clogging mechanism and service performance of tailings drainage facilities, tailings-geotextile permeability tests and microscopic observations are carried out. Fine-grained tailings are more prone to physical blockage, and the degree of chemical blockage is related to the type and concentration of solution ions. In order to promote the risk management and control technology of tailings dams, a major risk assessment and monitoring system platform for tailings dams are put forward, which can realize the extraction of basic data of tailings dams, real-time risk warning and other intelligent functions. Through the above-mentioned research activities, the level of dam building, disaster prevention and the control of large-scale high tailings ponds in China have been substantially improved.

The cyclic simple-shear behavior of soil-structure interfaces under different normal stresses is of great significance in theoretical analysis and engineering practices. A series of interface tests between gravel and steel was conducted under different normal stresses using a 3D large-scale simple-shear apparatus, and the influence of normal stress on the cyclic simple-shear behavior of the interface was explored in detail. The normal stress plays a crucial role in the magnitudes of the shear behavior of the interface, including deforming and sliding displacements, shear stress, reversible and irreversible normal displacement, and cyclic shear strength, while it has slight influence on the relationship pattern of shear behavior. Increased normal stress leads to increased deforming displacement at the first few shear cycles, accelerated reduction and decreased stabilized amplitude of deforming displacement. Enhanced normal stress also results in increased shear modulus, decreased shear modulus coefficient at the first few shear cycles, increased stabilized shear stress and cyclic shear strength of the interface. In addition, larger normal stress results in larger irreversible normal displacement, smaller peak reversible normal displacement, smaller transition tangential displacement and transition stress ratio of the interface. The cyclic shear strength behaves well in accordance with Mohr-Coulomb criteria, regardless of normal stress. Perfect consistency exists in the stress ratio versus tangential displacement response and irreversible normal displacement versus shear work density response, independent of normal stress, and can be described using hyperbolic models. These consistency characteristics will significantly simplify the constitutive modelling of soil-structure interfaces.

Based on the visual triaxial compression servo-control test system and the three-dimension digital image correlation technology (3D-DIC), the generalized stress relaxation tests under different rheological regions were carried out, and the evolution of strain in the rock surface was discussed. The experimental results show that if sandstone tends to fail during the rheological process, the axial and radial strain concentration area would gradually appear in strain fields. The axial strain concentration area concentrates from the layered discrete distribution to the position of the upcoming crack, while the radial strain concentration area is about to form with the centered crack. The differential evolution rates of axial and radial strains at different areas are positively correlated with the overall strain evolution rate. The development of axial and radial strain in cement near the crack is going through three stages: deceleration, constant velocity and acceleration. The strain in areas away from the crack may increase firstly and then decrease. And the isochronous curve of the strains at different positions indicates the area where the crack is about to form. As the rheological direction coefficient ? changes from 0.3, 0, ?3.0, ∞, and 3.0 in sequence, the mean and variance of the overall evolutions of the axial and radial strain fields increase during the rheological process, indicating that the evolution rate is accelerated, the strain concentration phenomenon is more significant, and the difference between the evolution of the strain field before and after the sandstone failure also increases.

The cross joints, which are ubiquitous in real rock mass, have significant influences on the fracture characteristics of rock mass but it has not been thoroughly studied. In this paper, the 3D printed samples with cross joints are made. The mechanical fracture characteristics of rock samples with cross joints under uniaxial compression are studied by means of acoustic emission, CT scanning technology and DIC image analysis. The results show that the relationships between fracture fractal dimension (DB) and uniaxial strength, elastic modulus and Poisson’s ratio have similar trends when the dip angle of primary joint is small, which has a better accordance than that of a much larger primary joint dip angle. There is a much better regularity of fracture mode of the former than that of the latter. When the dip angle of the primary joint is large, the clustered microfracture AE events occur under the tension between the joint and the block in the compaction and elastic stages, while this phenomenon will not occur when the dip angle of the primary joint is small. In DIC strain nephogram, the micro strain on the fracture propagation path is prior to the macro fracture. The fracture mode can be predicted by calculating the maximum area of micro strain under compression. The primary joint plays the main role in the fracture initiation angle and fracture propagation mode, and the acoustic emission and fractal characteristics are unified. When the secondary joint is parallel to the maximum principal stress, the total rock damage is mainly caused by the propagation of the primary joint, and the accumulative AE events (AAEE) and DB are the largest, implying that the fracture mode is more complex when the AAEE and DB is much larger. The three-dimensional fracture mode is mainly characterized as 2-3 tension wing fracture surfaces caused by the expansion of joints ends. On the internal three-dimensional macro fracture surface, there are discontinuous structures affected by incomplete tension failure.

The construction waste generated by underground excavation in coastal cities has the characteristics of high fine content, high water content and looseness. Landfilling is the most common method for disposal of such construction waste. As the waste production exceeds the capacity of disposal, many of the construction waste dumps have the problems of fast filling, lack of drainage facilities, overload of landfill capacity and over height, which are likely to cause safety accidents such as landslide. These irregulated operations reflect the poor understanding of the strength growth law of the high saturation degree waste during rapid landfilling. In this paper, completely decomposed granite (CDG) waste is taken from the site of Shenzhen Hong’ao landslide, and a series of triaxial tests with the drain and exhaust valves closed is performed on the samples with different initial saturations. The experimental results reveal that the undrained shear strength of CDG nonlinearly increases with the increase of confining pressure, and the growth rate is closely related to the initial saturation of the sample. The growth rate of the undrained shear strength significantly decreases as the saturation exceeds 0.7. Inspired by the principle of undrained strength analysis, an estimation method for undrained shear strength is proposed, in which Hilf’s equation and the modified Cam-clay model are employed. The proposed method is verified to be capable for predicting the undrained shear strength for CDG with an initial saturation exceeding 0.6 by comparing the estimation results with the measured values in the triaxial tests. The undrained shear strength cu and normal stress determined by the estimation method can be used in the stability analysis of waste dump with high filling rate.

As a key structural element of fracture networks, fracture intersections play an important role in water infiltration in unsaturated fractured rocks. With the decrease in flow rate, the flow mode within intersecting fracture changes from continuous and stable rivulet flow to intermittent and unstable droplet flow. Most studies have focused on the rivulet flow, while less attention has been paid to the droplet flow. In this paper, theoretical and experimental work on the characteristics of droplets flowing through unsaturated fracture intersections is performed. It is found that there are two regimes of droplet splitting: capillary dominated splitting regime and capillary-gravitational splitting regime. Based on force balance analysis, we propose a quasi-static model and a semi-analytical solution approach to predict the dynamic splitting behavior of droplets at fracture intersections. Additionally, by investigating the detailed processes of different splitting regimes, we clarify the microscopic mechanism about the combined influence of gravity, capillary force and viscous force on the dynamic splitting behavior. We elucidate the mechanisms based on the variation of interface velocities and droplet splitting volume ratio under different initial droplet lengths and inclination angles. This work provides a theoretical and experimental basis for prediction of seepage in unsaturated fractured rocks under low flow rate and low saturation conditions.

The mechanical behaviors of rockfills exhibit strain-softening and dilative features, which is also influenced by the pressure, loading paths and density. In this paper, a new fractional sub-loading surface model for rockfills is proposed based on the critical state mechanism, which is a typical double yield surfaces model. In the p-q plane, the current and reference stress points are located on the sub-loading surface and the reference yield surface, respectively. The relative position between the sub-loading surface and reference yield surface is controlled by the void ratio difference ?. Compared with the state parameter ?, ? is able to account for the effect of loading paths additionally. The proposed model is able to account for the extent between the plastic flow direction and the loading direction without introducing a plastic potential function. A fractional plastic flow rule is developed by applying the Caputo fractional stress operator on the yield function, which provides a unified description for the associated and non-associated plastic flow rules. The formula of the proposed model is quite simple and the included seven model parameters can be determined through laboratory tests conveniently. The model predictions were compared with the drained triaxial test results of Tacheng rockfill under the conditions of various initial void ratios and confining pressures. A good agreement between the model predictions and experimental results is obtained, which indicates that the proposed model is capable of describing the stress-strain curves and deformation features of Tacheng rockfill. In addition, the effect of initial void ratios on the critical state line in the e-lnp plane is also captured by the proposed model.

Magnesium oxychloride cement (MOC) is innovatively introduced into sludge solidification. The influence of initial water content, molar ratio of MgO/MgCl2, curing age and MgO activity on sludge solidification efficiency and the corresponding driving mechanisms are systematically studied by unconfined compressive strength, water content of solidified matrix, scanning electron microscopy (SEM) and energy dispersive spectrum (EDS). The results show that the higher the initial water content, the lower the strength of solidified samples, and the needle-rod shaped crystals (phase 3 or phase 5) can be clearly observed in MOC solidified sludge with medium water content. The increase in molar ratio of MgO/MgCl2 leads to an improvement in the compressive strength of samples, accompanied by the transformation of hydration products from amorphous gel to phase 3, phase 5 and Mg(OH)2 crystals. The strength of solidified sludge shows a increase trend with the prolongation of curing age, and the strength prior to 28 d increases relatively faster, followed by a stable trend after 28 d. Especially, the surface of specimens with high molar ratio of MgO/MgCl2 is prone to have frost at a longer curing stage. An enhancement in MgO activity produces more reactive components and leads to higher compressive strength of MOC solidified sludge, whereas the activity degree of MgO has no significant effect on the evolution of hydration products. The obtained results can provide a theoretical support for the development of green and low-carbon MOC-based cementitious material and its application in soil reinforcement fields such as sludge solidification.

Freezing and thawing processes can change the structure of frozen soil and reduce the mechanical properties, thus affecting the stability of engineering infrastructures built in frozen soil. Due to the different settings of freeze-thaw cycles, a large number of experimental results cannot be effectively compared and analyzed. In addition, estimating the soil mechanical characteristics under freeze-thaw action has also become a research difficulty. In this paper, based on the frozen soil genetic creep theory, a spherical indenter is developed, and a freeze-thaw cycle–physical time analogy method is proposed. Using the number of cycles (numbers of freeze-thaw cycles) and the duration time (physical time–minutes), we can obtain the long-term strength curve family of the frozen soil and mapping the curve family into the same stress space to achieve the normalized strength curve, and thus we can predict the long-term deformation and shear strength. Finally, two soil samples are selected for testing, and the relevant equations for predicting the long-term shear strength are obtained. This method has important theoretical significance for the comparative study on the mechanical behavior of frozen soils under freeze-thaw actions. It also has value for engineering practice and stability analysis of infrastructures built in cold regions.

Based on the wave propagation theory of saturated porous medium proposed by Biot and wave propagation theory of inviscid compressible ideal fluid, considering the fluid-mechanical coupling of underwater saturated soil, an analytical solution of the scattering problem of the incident plane P1-wave around the undersea cavity is presented using the Hankel function integral transformation method (HFITM). Compared with the “large arc hypothesis” in traditional research, the HFITM can better deal with the surface boundary of the half space. Using the analytical solution, the effects of permeability conditions of the cavity, incident angle, incident frequency, seawater depth and the porosity of saturated soil on the displacement (horizontal and vertical displacement) of the interface between water and saturated soil and the stress (hydrodynamic pressure and total circumferential stress) on the surface of the cavity. The results show that the permeability conditions on the surface of the cavity have a small effect on the displacement of the interface between water and saturated soil; as the oblique incident angle increases, the vertical displacement on the interface between the ideal water and saturated soil decreases; the horizontal displacement of the interface between water and saturated soil increases as the incident frequency increases; when the depth of seawater is 2.5 times the wavelength of the SV-wave, the maximum horizontal displacement on the interface between water and saturated soil and hydrodynamic pressure on the surface of the cavity are the largest; the displacement of the interface between water and saturated soil and the total circumferential stress on the surface of the cavity decrease with the increasing porosity, while the hydrodynamic pressure on the surface of the cavity increases.

Laterite is prone to form as aggregate due to its high clay content, leading to a difficultly in uniformly mixing itwith lime. Meanwhile, is the weak acid property of laterite promotes the interaction between the lime and laterite, which shows a significant influence on the final treatment effects. In this paper, the causes of the weak acidity of laterite are firstly studied through the charges distribution characteristics of water molecules and clay minerals. Then, the interaction between lime and laterite is simulated through soaking the lime-treated laterite in the acid and alkaline solution respectively. Based on the results, the method to inhibit the lime-laterite interaction by metakaolin is proposed. Laterite with two groups size, mixed with different dose of metakaolin and lime are adopted to conduct the unconfined compressive strength tests. The results show that compared with the laterite treated with lime alone, neither excessive nor insufficient metakaolin has contribution to its strength improvement, and the best treatment effect can be obtained with the dose of 5.0% metakaolin, which verifies the feasibility of metakaolin to inhibit the lime-laterite interaction . Reactions between metakaolin and limecan form not only the silicon and aluminate cementation bonded by ionic bond but also the reticular cementation bonded by covalent bond. Since the latter has a significant ability toresist erosion by acid, the internal reason of lime-laterite interaction undermined by metakaolin is interpreted.

At present, the improvement of the horizontal bearing capacity of the piles by pre-consolidation of the soft soil foundation has been well recognized by practising engineers. However, how to estimate the increment of horizontal bearing capacity of piles during the pre-engineering process is still difficult. In this article, a practical calculation method for estimating the increment of horizontal bearing capacity of piles is established based on the Bowles[1] method and by considering the impact of pre-drainage and pre-consolidation treatment of the layered soft soil foundation. This method provides an effective way to calculate the shear strength index and pre-consolidation treatment time based on the shear strength of undisturbed soft soil by laboratory test. Meanwhile, the elastoplastic solution of the horizontally loaded pile and the calculation formula of the plastic zone depth of layered soft soil foundation are analytically derived, based on the influence of elastoplastic yielding of soils surrounding the pile. In addition, the source code for computing the horizontal displacement of the pile top and the maximum bending moment of the pile body are given. Finally, the horizontal displacement, bearing capacity and the maximum bending moment of piles in the sluice pile foundation engineering case in Zhejiang Province are calculated according to the proposed method. The results of the field tests before and after the pre-consolidation treatments are compared. It is found that the estimated results are close to the test results, which may provide a good reference for similar engineering designs.

The study on the seepage in wide-grading soils, the main material of colluvial landslides, is the premise and foundation of the research on the mechanism of rainfall-induced colluvial landslides. The seepage in wide-grading soils is a complex process including the transport of water and fine particles. However, the migration of fine particles is usually ignored. Therefore, three soil samples with different D15 /d85 (D15 is the particle size which 15% of the coarse-grained group by weight is finer than; d85 is the particle size which 85% of the fine-grained group by weight is finer than) were tested by self-made large-scale permeameter to study the transport of water and fine particles in wide-grading soils. The results show that the D15 /d85 has a vital influence on the permeability coefficient and the migration of fine particles. The smaller permeability coefficient is, the more difficult migration of fine particles occur in soils of small D15 /d85; the larger permeability coefficient is with the violent change, and significant fine particles movement observed in soils of large D15 /d85. The change of permeability coefficient reflects the movement of fine particles in the soil, and three modes of particle migration are proposed for the wide-grading soils. The research improves the understanding of the permeability characteristics of wide-grading soils and provides a new mechanism of rainfall-induced colluvial landslides.

Soils are commonly existing in unsaturated condition on the earth’s surface. Compared with the classic Biot consolidation theory for saturated soils, the consolidation theory of unsaturated soils is still largely unexplored. Based on the Fredlund’s double stress variables consolidation theory of unsaturated soils, this paper derives the governing equations of axisymmetric fully-coupled consolidation of unsaturated soils under varying loadings, by removing the assumption that the total stress remains unchanged during the consolidation. With the aid of Laplace-Hankel transform, the governing equations are converted into ordinary differential equations by eliminate the variables r and t. The extended precise integration method is further applied to solve these equations and obtain solutions of layered unsaturated soils in the transformed domain. The actual solutions are acquired by the inverse Laplace-Hankel transform technique. The feasibility of the presented results is further verified by comparing with those of degenerated saturated soils in the existing literature. At last, three numerical examples are provided to investigate the influence of varying ramp time T0, varying coefficients of the water volume change due to the net normal stress and the stratification on the consolidation characteristics of unsaturated soils.

The underground pipelines are connected by socket joints, and the continuity of the pipeline depends on joint rigidity. However, previous studies mainly simplified pipelines as continuous structures, and the effect of joint rotation on pipeline deformation mechanisms was ignored. By conducting centrifuge model tests and two-stage numerical parametric studies, tunneling-induced deformation mechanisms of pipelines with different joints stiffness are explored in this study. Because of the existence of flexible joints, overall flexural stiffness of jointed pipelines is much smaller than that of continuous ones. Moreover, tunneling-induced joint rotation results in jointed pipelines to have a better ability to deform with surrounding soils. Thus, tunneling-induced settlement in jointed pipelines is much larger than that of continuous pipelines. By simplifying jointed pipelines as continuous structures, tunneling-induced pipeline settlement is grossly underestimated. A linear relationship between the joint rotation angle and volume loss is observed. Based on parametric studies, a dimensionless group of relative pipeline-soil stiffness is established to differentiate relatively rigid and flexible pipelines. Calculation charts between the relative values of the length of pipe segment, pipe-soil stiffness and joint rotations are established for relatively rigid and flexible pipelines, respectively. Centrifuge test results are used to verify the proposed calculation charts. The proposed method can effectively predict the joint rotation angle due to tunnel excavation. Research finding from this study can be used in underground pipeline network reconstruction projects in urban cities.

Resonance of multi-layer unsaturated railway ground induced by moving surface load is studied using a proposed semi-analytical approach. Influences of the soil saturation on ground vibrations are studied. The half space of railway ground is modelled as a layered unsaturated poroviscoelastic medium, the multiple Fourier transform method, the interface stress and displacement conditions for each soil layer are used to derive the global dynamic stiffness matrix of multi-layer unsaturated ground in the frequency-velocity domain based on the governing equations of unsaturated porous medium. A semi-analytical model for coupled track-ground is established by assembling a track model including a rail, rail pads, sleeper and ballast. The dynamic response of the multi-layer unsaturated ground caused by a moving harmonic load on the track is analysed. Considering the influences of the soil saturation on the soil shear modulus, the different dynamic response of the ground with various degrees of soil saturation for the surface layer and interlayer are discussed. The shading representation dispersion curves are used to study resonant modes for the unsaturated ground in the spectral domain. This research found that the influences of the soil saturation on ground displacements depend on soil layer properties, load speeds and frequencies. Besides, the ground saturation reduction can increase the appearance frequency of the ground resonant mode as well as the ground critical speed.

Widely-graded gravelly soil is composed of gravelly and clayey material in a certain proportion. It has the characteristics of low compressibility and high shear strength. At present, it is widely used as the core material for embankment dams and subgrade filling. Due to the low permeability of well-graded gravelly soil, the consolidation process is slow in conventional consolidated- drained (CD) triaxial tests, resulting in a long test period. In order to improve the efficiency of consolidation process, a quick triaxial CD test method with cylindrical sand core in the center of specimen is proposed in this paper. A comprehensive study on the influence of various factors on the quick triaxial CD test of well-graded gravelly soils is performed by changing types of sand cores, diameters of core sand cylinder and gravel contents. Results show that the proposed quick triaxial CD test method can effectively accelerate the consolidation process of the specimen, thus speeding up the whole test. Various factors including type of sand core, diameter of core sand cylinderand gravel content show different levels of influence on the speed of consolidation process and the stress-strain curve in the subsequent shearing stage. The smaller the diameter of the sand core is, the closer the test results are to those of specimens without sand cores.

When there is invariant/uninfluenced pumping drawdown and waterproof curtain at a distance r =rb from the pumping well, the aquifer system is laterally finite in extend (termed finite aquifer) and rb is termed ‘finite radius’. A practical calculation model of transient flow to a partially penetrating well in a typical aquitard-confined aquifer of finite extent is developed. Also, the model considers the effect of well radius and wellbore storage. The drawdown solution is obtained by the Laplace transform coupled with separation of variables, and inverted into the time-domain solution utilizing the Stehfest method. Moreover, the obtained solutions can reduce to some available solutions and are further validated by comparison with other solutions. Finally, the effect of the lateral boundary and well configuration on pumped aquifer drawdown behaviors are analyzed based on the proposed analytical solution. The results show that the influence of the lateral boundary is obvious during the late pumping time, the drawdown of the finite aquifer system is larger than that for Case 1 (constant head boundary at r = rb) and much smaller than that for Case 2 (no-flux boundary at r = rb). A smaller finite radius rb results in a larger error between the finite aquifer system (Case 1/Case 2) and the infinite aquifer system. The impact of the well configurations occurs at the whole pumping stage for different cases, and the drawdown at the top of the pumped aquifer becomes smaller with the increase of the length and location of the well screen.

The pile bottom sediments cannot be cleaned up completely during the construction process of drilling with pre-stressed concrete pile cased (DPC) piles, which may lead to the loss of bearing capacity of the pile. Model tests for DPC piles with consideration of the pile bottom sediment effect are carried out in this paper. The load-settlement curves of DPC piles with enlarged cement-soil pile end show slow-varying, while other tests of DPC piles with pile end sediments show sharply down. The influences of the sediment thickness and enlarged pile end on the vertical bearing characteristics are studied. Experimental results show that the pile end sediment has a certain impact on the vertical bearing capacity of the DPC pile, and its bearing capacity can be improved by around 22% after removing the sediments. Compared with the bored cast-in-place pile with a certain thickness of sediment, the sediment at the bottom of DPC piles has less effect on reducing the bearing capacity. Besides, DPC piles with expanded cement-soil pile end can increase the bearing capacity by 37% in approximation. The external load of the DPC piles with pile bottom sediments is generally borne by the pile side frictions for more than 90%, and the pile axial forces near the pile end become smaller with thicker sediments. In dense sand strata, the pile end resistance ratios of DPC piles are less than 15%. Both in-situ (pile length 15.5 m, pile length-diameter ratio 15.50) and indoor model test (pile length 1.0 m, pile length-diameter ratio 15.87) results show that DPC piles with pile bottom sediments are the type of end bearing friction piles. These research findings may provide a reference for further understanding the bearing characteristics of DPC piles.

Temperature gradient, moisture gradient and pressure gradient significantly affect hydrothermal migration in unsaturated soils. The unsaturated loess sample under 0.10 MPa of high-temperature water vapor is tested to study the hydrothermal migration under the gradients of steam pressure, temperature and moisture. The results show that the high rate of temperature rise, and water vapor humidification observed within the range of water vapor migration and the steam humidification are mainly driven by steam heat transfer and pressure gradient. In the region beyond water vapor migration, the heat conduction caused by temperature gradient and water migration driven by moisture gradient coupled with temperature gradient are the main factors. Under the influence of soil particle obstruction and steam pressure dissipation, the rate of soil heating, humidification and temperature diffusion decrease with the increase of radial distance, and the rate of humidification is less than temperature diffusion. The maximum moisture content of water vapor humidified soil is close to the optimal moisture content, and the humidifying effect is good, which can effectively improve the compaction performance of soil within a certain range. Based on the boundary conditions of the model test, a set of algebraic explicit special solutions to the two-dimensional heat-wet migration equation in a moist porous medium is determined. The measured and calculated values of temperature and water content are compared and analyzed at 20 cm measurement points. The results can provide theoretical support for the research on the transmission of water-vapor-thermal in unsaturated loess and the new technology of water vapor humidification.

Rapid and accurate measurement of permeability coefficient for the graded macadam used in heavy-haul railway subgrade bed surface layer is of great significance to guarantee the construction quality and period. First, a large-scale constant-head permeameter suitable for graded macadam was developed based on Darcy’s Law, and a series of penetration tests considering wall effects was carried out using the aforementioned device. These initial tests were conducted to explore the influence degree and properties of pore ratio e, unevenness coefficient Cu, and characteristic particle size parameters on the permeability coefficient K20 of a specimen. Besides, based on the Terzaghi model, an improved calculation model that took into account both pore and gradation characteristic parameters was established with reference to test data, and the accuracy of the model was verified through experiments. The results show that the graded macadam permeability coefficient K20 has a linear relationship with e2 and the square of mean particle size d502, and has a negative exponential relationship with Cu; d50 has the greatest influence weight on K20, followed by Cu, and e has the smallest influence. The proposed calculation model demonstrates a high degree of accuracy and can provide a reference for rapid determination of permeability coefficient for the fillers used in newly heavy haul railway subgrade bed surface layer under construction. Using this calculation model, the permeability stability of operating subgrade beds can be evaluated.

When solving the water infiltration problem in unsaturated soils, the hydraulic function is a function of water content or suction, which makes the equation governed by hydraulic function exhibits strong nonlinear characteristics resulting in the difficulty of solving the horizontal infiltration problem. Based on the assumption that the water flow in soil medium follows the path of time - consuming extreme value, a time functional was introduced, and the horizontal infiltration problem was transformed into a functional extreme value problem based on a variational principle. By solving Euler-Lagrange equation and combining with the boundary conditions, the explicit analytical solution of the nonlinear transient horizontal infiltration problem was obtained. Combined with the Brooks-Corey hydraulic function, the distribution of volume water content of this type of soils in the unsaturated state was explicitly solved. By calculating the water horizontal infiltration laws of four different types of soil samples through theoretical and numerical methods, the results obtained by the solution matched well with the existing results and numerical results, verifying the effectiveness of the method. The results show that the distribution of volume water content has a power function against location distance and wetting peak distance ratio, and the power exponent depends on the shape parameter of soil water characteristic curve. Initial conditions and boundary conditions had different effects on the distribution of volumetric water content.

According to the deformation characteristics of step-like landslides in the Three Gorges Reservoir area, a new method for predicting the landslide displacement is proposed. The monitoring displacements of points ZG118 and XD-01 in Baishuihe landslide are taken as example analysis. By using the empirical mode decomposition with soft screening stop criteria (SSSC-EMD), the cumulative displacement-time curves and the influencing factor time series are adaptively decomposed into multiple intrinsic mode functions (IMF). The K-Means clustering method is adopted to cluster and accumulate IMFs. The displacement components (including the trend, periodic and stochastic displacements) and the influence factor components (including high-frequency and low-frequency factors) which contain physical meanings are obtained. The trend displacements are fitted by the least square method. The periodic and stochastic displacements are predicted by combating fruit fly optimization and least squares support vector machines (FOA-LSSVM) model. Finally, the cumulative prediction displacement is found to be the addition of the three component prediction values. The results show that the proposed (SSSC-EMD)-K-Means-(FOA-LSSVM) model has the capability of predicting the displacement variation of step-like landslides. The prediction accuracy of this model is higher than those of traditional SVR and LSSVM models. Furthermore, the single factor analysis is performed by changing the length of the training, and it is positively correlated with the prediction accuracy.

In this paper, based on the combined point estimation method and finite element method (i.e. PEM/FEM) for the spatial variability of geotechnical mechanical parameters, the application of PEM/FEM was extended to study the spatial variability of joint distribution through the geometric models and meshing characteristics of joints. A pumped-storage hydropower station in China was adopted as a case to verify the accuracy and rationality of the expanded PEM/FEM by analyzing the influence of spatial variability of joint distribution on the deformation and plastic zones of the surrounding rock mass in the underground caverns. A finite element analysis model for spatial variability of joint distribution was established through the probabilistic statistics of more than 1,400 joints in the underground caverns. The effects of spatial variability of joint distribution on the stability of the surrounding rock mass was comprehensively studied through the expanded PEM/FEM probability analysis method. Results show that: (1) The comparative analysis of the deformation probability distribution of the surrounding rock mass obtained by the expanded PEM/FEM and on-site monitoring shows that after removing some abnormal values, most of the monitoring deformation values are within the range of the displacement probability distribution, indicating that the spatial variability of joint distribution is the main factor that cause the fluctuation of the monitoring deformation. (2) The probabilistic distribution of the standard deviation of the deformation can be adopted to identify the incidences of the spatial variability of joint distribution on the deformation. For the studied case, the influence of the spatial variability of joint distribution on the deformation is the most significant in places near the floorsand least near the top arches. (3) The probabilistic zoning of the plastic zones of surrounding rock mass can reasonably determine the areas and ranges that are greatly affected by the joints during the excavation of the underground caverns, which provides a support for the design of engineering construction.

The transverse settlement induced by the construction of a parallel shield tunnel alongside the building has raised considerable attention, whereas few studies focus on the longitudinal settlement. Therefore, the spatial deformation of ground developed from this tunneling form is investigated. In this study, some field measurements from the shield tunnel section of Tianjin Metro Line 6 parallel to four similar masonry buildings in vicinity are analyzed first, and the deformation pattern is established. Then, a hardening soil model calibrated against field measurement, considering small strain stiffness, is implemented in a three-dimension finite element simulation to evaluate the longitudinal deflection of the buildings, the ground deformation, and the soil stress distribution. Additionally, the effect of building aspect ratio is discussed. The simulation results show that tunneling-induced sagging deformation develops along the longitudinal direction of the building, and the settlement at the middle of a longitudinal wall is twice of that at the corners. Therefore, the study of tunneling parallel to buildings cannot be simplified to a plane strain problem. The building construction and tunneling activity result in the soil above the tunnel crown experiencing a complicated stress history, which can be divided into six stages. In longitudinal direction, compared with the part below the building foundation corners, the soil in the middle initially behaves larger compressive deformation due to building construction, followed by greater unloading deformation caused by tunnel excavation. In addition, the longitudinal sagging is significantly reduced for the buildings with aspect ratio less than 2.

Based on the reinforced retaining wall of Chengdu-Kunming railway, the field tests were carried out. The comparison of the reinforced retaining wall with full-height rigid panels, the reinforced retaining wall with modular panel embedded return package structure and the reinforced retaining wall with modular panel within 9 months after the completion of construction were analyzed to study the structural properties. During the test period, two earthquakes occurred and changes of the deformation of retaining wall and the earth pressure were monitored. The results show that the geogrid strain rate, wall compression and wall horizontal displacement can be well controlled by the reinforced retaining wall with full-height rigid panel, showing the best stability and seismic performance. The earth pressure on the back of the reinforced retaining wall with full-height rigid panel and reinforced retaining wall with modular panel are distributed following with a "M" type along the wall height. The distribution of earth pressure on the back of the reinforced retaining wall with modular panel embedded return package structure is similar to "S" along the wall height. The lateral earth pressure coefficients of the reinforced retaining wall with full-height rigid panel change little along the wall height, which are less than the FHWA calculated values. The lateral earth pressure coefficients of the reinforced retaining wall with modular panel embedded return package structure have an increased trend along the wall height. The values of the middle and lower part of the retaining wall are less than the FHWA calculated values, and the values of the upper part are close to or higher than the static earth pressure coefficients. Most of the lateral earth pressure coefficients of the reinforced retaining wall with modular panel are ranging from the FHWA calculation values to the static earth pressure coefficients. The geogrid strains of three retaining walls change nonlinearly along the wall height. The retaining wall compressions increase gradually with the increased time, especially in the first 50 days. Then the increase rates gradually decrease until the coming of the rainy season, the hydraulic effect promotes the increase rates to increase again. After the earthquakes, the earth pressure on the back decreases, the geogrid strain increases, the wall compressions increase and the wall panels have an obvious trend to move out.

Numerical simulation is a normal method for slope stability analysis in open-pit mines. Slope model construction is the basis of numerical simulation, and the accuracy of topographic measurement data has an important influence on the accuracy of slope stability analysis. In recent years, low-altitude and low-speed small unmanned aerial vehicles (UAVs) have been widely used in geological surveys, which provide a feasible method for rapid terrain measurement under complex conditions. Considering the situation of the rapid change of slope shape of the open-pit mine, the complex terrain, and the difficulty of manual measurement, a single-lens small unmanned aerial vehicle (UAV) was used to perform low-altitude photogrammetry on an open-pit mine slope. Through UAV flight tests, the influence of the rotation angle of the UVA mounted lens on the slope measurement accuracy was compared and studied. Finally, a high-precision slope digital elevation model (DEM) was obtained, which was converted into a three-dimensional FLAC3D model using several softwares such as Surfer and Rhino. A rapid establishment was realized from the UAV measurement to the 3D numerical model. The following research conclusions can be obtained for reference: when the target slope is less than 45°, the rotation angle of UAV mounted lens has a large influence on the slope measurement accuracy. When the lens rotation angle is between 70° and 90°, relatively higher measurement accuracy can be obtained. Compared with manual measurement, the slope shape obtained by the UAV photogrammetry is more in line with the actual slope shape. After a series transformation of the aerial photos of UAV using the PhotoScan, Surfer, Rhino, etc., a 3D numerical model can be quickly built, and better numerical simulation results can be obtain to meet the needs of slope stability analysis.

In order to effectively simulate the process of random propagation of hydraulic fractures in fractured shale reservoirs, a new method of random propagation of hydraulic fractures based on the finite element mesh embedded with zero-thickness cohesive elements is proposed. This new method is based on the topological data structure of element nodes and the splitting mode of mesh nodes. The accuracy and effectiveness of the random propagation method are verified by comparing with the analytical solution of KGD model and two kinds of laboratory experiments. Meanwhile, the influences of horizontal in-situ stress difference and reservoir heterogeneity on the process of random propagation of hydraulic fractures are evaluated by running numerical examples. The results show that: (1) the new method makes up for the deficiency that the cohesive element built-in ABAQUS platform can not effectively simulate the random propagation of hydraulic fractures; (2) under a higher horizontal in-situ stress difference condition, the stronger the heterogeneity of a shale reservoir is, the easier it is to reopen a high-angle natural fracture intersecting with hydraulic fractures. The proposed method can accurately describe the random propagation behavior of complex hydraulic fractures, and thus provide a novel means for numerical simulation of naturally fractured shale reservoirs.

Common geological disasters such as landslides, mudslides, rock slides, etc. usually involve the movement of particles of different shapes, and most of these particles have different sizes and contents. Based on a typical particle column collapse test, the parameters required for the discrete element simulation were first determined according to the test method, and then the random polyhedron method was used to generate large particles with a controllable slenderness ratio. After that, the discrete element method was used to determine the contents of different large particles. The effect of morphological changes on the collapse characteristics of the binary particle column was then studied. The results of the study showed that: (1) The discrete element method could better reproduce the collapse process of the binary particle system composed of small spheres and polyhedrons in laboratory experiments; (2) In a binary particle column system composed of irregular large particles and small balls with different slenderness ratios, when the content of large particles was higher than the critical content of 20%, the duration of the collapse of the binary particle column varied with the non-spherical large particles; (3) In a binary particle column composed of irregular large particles and small balls with different slenderness ratios, when the content of large particles was higher than the critical content value of 20%, under the same percentage of large particle content, the increase of the slenderness ratio of the large particles increased the average coordination number of the large particles and reduced the movement ability of the particles. A stronger interlocking effect was formed between the large particles, which reduced the overall fluidity of the particle column and made the column finally pile up to a higher height, a shorter maximum distance and a smaller normalized kinetic energy peak; (4) In the binary particle column composed of irregular large particles and small balls with different slenderness ratios, small particles could significantly reduce the friction and interlocking effect between large particles, increase fluidity, and reduce the impact of large aggregate shape on the collapse process.

A water-heat-vapor coupling model of unsaturated frozen soil is established based on energy and mass conservation equations by considering ice-water and water-vapor phase changes, and influences of heat transfer by water vapor and temperature potential on water vapor migration are also considered. The smoothed particle hydrodynamics (SPH) can be used to calculate their evolution process conveniently. In the solution, ice content and vapor content are solved at first, and then followed by solving of unfrozen water content and temperature field. With two steps as such, coupling of temperature field between water and vapor field is implemented. Distributions of unsteady temperature field, volumetric moisture content and water vapor flux in semi-infinite medium with thermal boundary conditions of the first kind are simulated. The results are compared with the analytical solutions without coupling. The water and vapor transfer shows a greater impact on the thermal field. Under the seasonal periodic temperature boundary, the distributions of thermal field and volume moisture content of embankment section are predicted in high compatibility with the actual data, which can well reveal the characteristics of water-thermal-vapor transfer and its phase transformation in unsaturated frozen soil.