The stability analysis of high rock slopes is a classical research topic in geotechnical engineering. One analysis method based on a certain hypothesis is different from others. This paper is to discuss and compare the advantages and disadvantages of various existing rock slope stability analysis methods. In addition, on the basis of traditional slope stability analysis methods, such as the limited equilibrium method and the finite element method (FEM), we propose an idea using the FEM to directly solve stability of the slope. The stress field and the deformation field of a slope at all conditions can be well simulated with landslide units which have not yet slid after excavation and support, and thus slope safety factors can be calculated based on the real stress field. Furthermore, the stability analysis of rock slopes is gradually simulated during the step-step excavating process by combining on-site monitoring with computer simulation means. An example of analytical calculation shows that the rigid body limit equilibrium method is complemented and upgraded by this new method. We solve the challenge of determining the strength reduction ratio in the strength reduction method, and deal with the uncertain mechanical direction in the overloading method. The calculated results of the slope safety factor and the stability by using the new method show good agreement with actual field measurements. This paper provides a new thought and approach for the stability analysis of complex rock slope.

A thermo-hydro-mechanical coupled model was used to simulate hydraulic reaction of the PRACLAY in-situ heating experiment which was conducted in Belgium underground laboratory HADES. The 3D finite element simulation was performed for the sensitivity of parameters to the pore water pressure, effective stress and temperature of surrounding rock under thermo-hydro-mechanical coupled conditions by using a single factor analysis method. The effects of two coupling factors among temperature, seepage and stress were comprehensively investigated upon hydraulic reaction of rock around the repository. It is shown that the permeability, elastic module and thermal conductivity are the uppermost ingredients influencing the pore water pressure. The cohesion, friction angle and thermal expansion efficient have a negligible effect on the pore water pressure, but have a remarkable effect on the effective stress. The distribution of temperature field of surrounding rock is only affected by the thermal conductivity. The influence mechanism of various parameters on pore pressure, temperature, and effective stress is different. There exists a difference in the effects of two coupling factors among temperature, seepage and stress upon hydraulic reaction of surrounding rock. The strongest coupling is found from thermal to hydraulic and mechanical behaviour. The excessive pore pressure induced by heating greatly affects the stability of surrounding rock, so does the variation of effective stress caused by thermal expansion. The results of this study have a certain value for the determination of parameters and analysis of coupling mechanisms of clay tunnel for radioactive waste disposal.

To prepare samples with the same lithology, surface morphology and mechanical characteristics as those of natural rock structural planes, a new method was proposed to originally reproduce rock structural planes by integrating 3D scanning and 3D carving techniques. Marble structural plane samples with the same natural surface shape were made using this new method. Shear tests were carried out on the prepared samples at different normal stresses. In addition, the acoustic emission (AE) technique was used to monitor failure processes of marble structural planes. The results indicate that the reproduced samples are accurately similar to the original samples with natural structural plane. The curves of shear stress and shear displacement can be divided into sliding type and cutting type. The shear failure structural plane can be categorised into dilatancy zone and abrasion zone. A new formula of shear strength is put forward, and theoretical values are in good agreement with experimental results. The change laws of shear stress and AE energy rate, count and hit with time are consistent, and the location of AE events is similar to the asperity damage positions during the shear test.

Boom clay, as a study case for the potential geological disposal of high-level and long-lived radioactive waste in Belgium, can be considered as a transversely isotropic geomaterial. This paper presents a coupled thermo-hydro-mechanical (THM) elasto-plastic damage model which is based on the Drucker-Prager cap model. The model is able to reflect thermal effect on the strength, elastic modulus and permeability of Boom clay. The developed model was implemented in ABAQUS finite element code through subroutine USDFLD. Three dimensional numerical simulation analysis was conducted of the ATLAS III in-situ heating tests at the HADES underground research facility to validate the proposed constitutive model. The results of the numerical simulation are compared with in situ measurements in which the coupled THM properties of Boom clay were analyzed. It indicates that the model can reasonably depict the main features of coupled THM anisotropy behaviors of temperature and pore water pressure. Remarkably anisotropic characteristics were found on temperature and pore water pressure changes of Boom clay with thermal load. The pore pressure in the horizontal plan shows temporary decrease and then increase after increasing power, and temporary increase and then decrease after decreasing power. The pore pressure in the vertical plan shows immediate increase after increasing power and immediate decrease after decreasing power. This study shows that the anisotropic coupled THM elasto-plastic damage model can accurately reproduce the temperature and pore water pressure changes during the heating test. The results of this study can provide valuable information for the design and operation of similar engineering/in situ tests.

To study the influence of suction and degree of compaction on the deformation of unsaturated filling soils, one-dimensional confined compression test of unsaturated soil was carried out to control suction condition and degree of compaction. Results show that the gentle stage of the compressive curve extends with the increase of compaction and suction, indicating the synchronously improved yield strength of soil samples. The saturated soil sample has the largest difference of the specific volume v during the test, and the dehydration soil sample changes the smallest. The greater the suction is, the smaller the difference of specific volume v is. Compressibility of compacted soil decreases after dehydrating or suction growth. The suction compressibility coefficient and its empirical model are defined and established to characterize the influence of suction and compaction on compressive characteristics and measure the influence degree. It is found that the model exhibits exponential decay with increasing the vertical stress. The relationship between the parameters of the model and relevant parameters of compacted soil are analyzed and discussed. The average value of the parameters corresponding to the different suctions of the same compacted soil sample decreases linearly with the increase of compaction level. The parameters generally increase with the increase of suction. However, with the increase of compaction degree, the contribution of suction to the compressive strength decreases and the experimental value point is close to the linear approaching line. The compressibility of soil samples is adjusted dynamically with the change of compaction and suction. Based on the basic principle of the layered summation method, a modified calculation model for the compressive deformation of the high fill unsaturated filling soils (MS) is established. The application of this model should be established on the study of the relationship between soil and water of unsaturated compacted filling soils on the loading state.

In order to study the anisotropy of the water-softening damage characteristics and the micro-layered structure of deep-buried carbonaceous phyllite, rock mechanical tests of carbonaceous phyllite were conducted under different bedding directions and different water contents. Based on the energy damage evolution mechanism, the influences of bedding direction and water content on the energy storage and energy release mechanism of phyllite are analysed. The results suggest that the anisotropic characteristics and water-softening effects of deep-buried carbonaceous phyllite are inherent characteristics attributed to sedimentary beddings, layered lamellar minerals and high content of clay minerals inside the carbonaceous phyllite. The failure mechanism and the anisotropic strength of phyllite are dominated by the weak cementation between the bedding surfaces. This weak-surface effect increases with the decrease of confining pressure. With the increase of water content, the mechanical properties and brittleness of phyllite are weakened, and the macroscopic failure angle is increased, which results in the decrease of tensile failure and the increase of shear failure. Based on the energy evolution mechanism and the S-shaped curves between elastic energy and dissipated energy, the gradual failure process of rocks is divided into four stages. Based on the energy damage evolution mechanism, the energy mechanism and its damage evolution mechanism of rocks’ two typical failure modes, type I and type II, are studied. It is found that the bedding direction and water content both have great influence on the energy storage capacity of phyllite. The bedding direction has a small influence on the energy release mechanism and damage evolution mechanism of phyllite. The energy storage and energy release mechanism of phyllite are more sensitive to water content than to bedding direction.

To investigate the mechanisms of rockburst and to explain the process of rockburst occurrence, a two-body model of the ejection rockburst source was proposed by analysing the sources, the dissipation and the magnitude of the released energy. The rockburst is essentially a process of brittle crack growth and rapid energy release in rocks under the interaction between energy- release-body and rockburst-body. Based on the tensile fracture mechanism of rockburst-body, the influence of crack spacing and strain rate on the dynamic energy release rate was analysed by using a dynamic model of brittle cohesive crack-array growth. The role of energy-release-body for rockburst-body can be attributed to loading strain rate effect. It can be found that a higher energy release rate of energy-release-body results in a greater dynamic loading rate and energy release rate of rockburst-body. Meanwhile, a higher energy release rate of rockburst-body leads to a greater dynamic discharge rate and energy release rate of energy-release-body. Dynamic loading and unloading effect caused by the interaction between two bodies constitutes a positive feedback mechanism of rockburst energy release. Based on the above theory, it is concluded that the geometrical scale of the rockburst source is almost several to dozens of times larger than rockburst-body, which is consistent with the field monitoring results.

The investigation on the bearing behavior of pile foundation in coral-reef limestone formation is a hot issue in the geotechnical engineering. Based on the field static loading test of tip post-grouting pile in coral reef limestone formations in the project of a cross-sea bridge, this work analyzed the test results of the trial pile before and after grouting, and the influences of the preloading effect of post-grouting at the pile tip on the pile resistance. The mechanism of improvement for the grouted pile was further studied by the analysis of test results of post-grouting at the pile tip. The results show that the post-grouting technique can be applied to the coral-reef limestone formations, which can effectively improve the bearing capacity of pile foundation and decrease the settlement. Coring test reveals the distribution of pressurized grout, and verifies that the pore of coral-reef limestone is filled by the pressurized grout. The negative side friction is mobilized before application of vertical load because of the negative side friction arising from grout pressure after tip grouting. Thus, the side friction of grouted pile increases. Additionally, the preloading effect of post-grouting at the pile tip enhances the tip resistance and reduces the tip displacement. Consequently, the bearing behavior of pile foundation is improved by post-grouting at the pile tip.

Frictional characteristics of sliding planes are the critical issues in the kinetics of high-speed rockslides. The friction experiments of dry and wet limestone samples at different pressures and rolling velocities are conducted to investigate dynamic frictional characteristics (i.e., friction coefficient and wear rate) using the rolling wear test equipment. Experimental results show that for dry limestone, there is a negative correlation between the rolling friction coefficient and the rolling velocity, but it is a positive relationship for wet limestone. Whether the specimen is dry or wet, there is no apparent relationship between the rolling friction coefficient and the normal pressure. For dry limestone, the wear rate of rolling specimen (i.e., as sliding mass) and static specimen (i.e., as sliding bed) is positively and negatively correlated with the rolling velocity, respectively. However, for wet limestone, the wear rate is positively relevant to the rolling velocity. In addition, based on the Hertz Elastic Contact Theory, we deducted the maximum compressive and tensile stresses of the friction spot, and then proposed the friction-fragmentation mechanism of brittle limestone under corresponding conditions. This study can provide the design parameters for the disaster prevention of the high-speed limestone rockslides.

Particle morphology is an important factor affecting mechanical properties of granular sands (especially the shear strength, dilation and critical state behaviours under low stress conditions), as well as particle crushing behaviours under high stress conditions. Therefore, in order to study the morphological effects of sand particles, it is the prerequisite to accurately reconstruct and quantitatively characterize the three-dimensional (3D) morphological features of sand particles. In this study, the 3D morphological information of Leighton Buzzard sand and decomposed granite sand particles was obtained by means of high-precision CT scanning and a series of image processing techniques. The spherical harmonic (SH) function sequence was then used to reconstruct the accurate 3D particle morphology. Furthermore, the particle volume calculated by SH analysis was used to validate the efficiency of SH reconstruction of the 3D particle morphology. Based on the SH-reconstructed particle surface, we proposed practical methods to calculate the surface area, the surface curvature distribution and the 3D dimensions of the sand particles, and further calculate the 3D sphericity, roundness and elongation of the sand particles. The results show that the SH-reconstructed particle surface agrees well with the real sand particle including the general shape and surface texture when the order of the SH function reaches 15. In addition, the particle morphology of Leighton Buzzard sand is more regular and smooth due to geological action of water transportation and abrasion, while the particle morphology of granite residual sand is more complex and rough due to physical weathering and denudation. However, both two geologic actions have no significant effect on the 3D dimension ratio of the sand particles.

Calcareous sand is widespread in tropical marine regions (including the South China Sea). It is one type of special geomaterial with unique hydro-physical properties different from terrigenous sands. Through the conventional constant head permeability tests, the effect of particle size distribution with varied uniformity coefficients and curvature coefficients on the permeability of the calcareous sand has been investigated. In light of the unique shape of calcareous sand particle, the ratio, , of sphericity to circularity X was introduced to quantitatively describe the particle shape based on the SEM images. To explore the effect of particle shape on the permeability of the calcareous sand, an investigation was performed by comparing with Fujian standard sands and glass beads with the same particle size distribution and relative compaction. The test results indicate that the permeability of the calcareous sand increases with the growth of uniformity coefficients and curvature coefficients, being consistent with the effect of particle size distribution on common sands. Compared to the other two kinds of granular materials, the particle shape of the calcareous sand has significantly structural and nonuniform properties, which can significantly reduce the permeability. The results of the present study will provide a guide for the designs of the island building and platform foundations.

The type of anchorage slurry and its compatibility with blot and slurry are the research focus in the field of earthen sites’ anchorage. Glutinous rice soup was used as major cementing material modified by mixing with clay and fly ash. The experimental program examined three anchoring systems, i.e., wood, glass fiber reinforced plastic, and steel rebar, cemented by modified glutinous rice mortar. The test monitored the ultimate load, load-displacement characteristics, stress-strain on the blot-slurry interface, and the response extent of interfacial monitoring point strain on load step. The preliminary analysis of the mechanism of the anchorage concludes that the anchorage performance the system highly depends on the mechanical compatibility of blot-slurry, which is further determined by its force mechanism, deformation and strength characteristics ultimately. The research results provide a basis and reference for the application of modified mortar mainly made of glutinous rice soup in the field of earthen sites’ anchorage.

The mechanical properties of rockfill materials are not only depend on stress state, but also relate to degree of compaction or material state of rockfill materials. Large-scale triaxial tests under isotropic consolidation and drained shear were carried out to investigate the effect of dry density on mechanical properties of rockfill materials. Test results show that the forms of stress-strain relationships depend on the combined effect of density and confining pressure. Before failure state, the strength and deformation of rockfill materials are determined by density and confining pressure. When the strain is large enough, as in the asymptotic state or critical state, the effect of initial dry density on stress and volumetric strain of rockfill materials decreases until it disappears. For stress softening curve, the stress of phase transformation state is less than the stress of asymptotic state, and the stress of asymptotic state is less than the stress of failure state. Regardless of the value of confining pressure, the hardening of the stress-strain curve gradually decreases and the softening is gradually enhanced as the initial dry density increases.

To observe and analyze the motion characteristics of the drag embedment anchor (DEA) in soil, the dragging experimental apparatus comprised of glass flume, dragging system and tension sensor is established. Transparent soil is prepared using silica gel and water. Based on the ratio of the effective unit weight of plate anchor to the undrained shear strength of soil, the material of the plate anchor is chosen. At the reduced scale of 1:50 in this paper, the dragging experiment of the fixed fluke DEA with the dragging setup is performed. The effects of the padeye offset ratio and fluke angle on the travel trajectory of the fixed fluke DEA are experimentally investigated. The results show that the penetration of the anchor into the soil can be directly observed by the test setup in this paper. The tangential eccentricity and normal eccentricity of padeye are not independent. The padeye offset ratio instead of the padeye eccentricities can describe the effects of padeye offset on the travel trajectory. The fluke angle is an important parameter of influencing the travel trajectory of the fixed fluke DEA in soil. With the fluke angle increasing, the ultimate penetration depth of the fixed fluke DEA firstly increases and then decreases. With the padeye offset ratio increasing, the ultimate penetration depth of the fixed fluke DEA firstly increases and then decreases. The travel trajectory of the fixed fluke DEA in soil approximately follows a negative exponential form.

Diffusion is the dominant solute transport process in high slump backfill with low hydraulic conductivity. However, the method to measure the diffusion coefficient is not available. Based on the theory of dynamic leaching test, a dialysis test method was adapted to measure the diffusion coefficient of NaCl diffusing from flowing-state backfills with different bentonite contents at different initial NaCl concentrations. The solution for solute transport in a finite cylindrical medium was used to calculate the diffusion coefficient. Test procedure and problems in data analysis were discussed. The results show that, as more data are used, the effective diffusion coefficient of NaCl decreases and the relative regressing error increases. Diffusion in backfill is obviously faster than that in compacted specimen. The total data sets are truncated until a relative error of less than 0.5% is achieved. This relative error quantitatively describes the dominant role of diffusion in solute transport process, and assures the validity of the data fitting using the pure-diffusion analytical solution. The measured diffusion coefficient decreases as the bentonite contents and initial NaCl concentrations increase. However, variations of the values are roughly small in the test range. The dialysis test is a feasible method for quick measurement of solute diffusion coefficient in cutoff wall backfill.

Two centrifuge model tests on foundation pit engineering with cantilever retaining wall in sand and silty clay ground are carried out respectively. Test procedures and some feasible solutions to the key problems in the tests are introduced, and results of the two tests are listed and compared in this paper. It can be summarized that the preparation of unsaturated soil ground is difficult because it is hard to control parameters of remolded soil. Therefore, stratified compaction method needs to be improved. In contrast, sand ground prepared by sand-rain method is easier to control parameters.Evolution of soil pressure, foundation deformation, and bending moment in the sand ground test is better than that in the silty clay ground test. Therefore, geotechnical centrifuge model tests in unsaturated soil ground can be replaced by the tests in sand ground appropriately. In foundation pit engineering with cantilever supporting structures, the induced surface settlement curve is an exponential pattern in sand ground but a linear pattern in silty clay ground according to test results, and the induced ground deformation range is larger in silty ground than that in sand ground. The induced bending moment of the retaining wall is larger in sand ground than that in silty ground, and the positions of the maximum bending moment for both grounds move down during the excavations. In sand ground, soil excavation causes decrease of the earth pressure behind the retaining wall, while in silty clay ground, there is an increase at the bottom of the wall. The complete test procedures and data processing methods presented in this study provide technical premises and guidance for further tests.

To investigate dynamic behaviour of shield tunnels and surrounding soil, a physical model test was conducted. An electromagnetic shaker located at the bottom of the shield tunnel was used to apply sweep excitation and train vibration load. The data of accelerometers are applied to calculate the frequency response function (FRF) and the maximum acceleration of the tunnel and soil. It is found that FRF is insensitive to the excitation amplitude, sweep direction and period, which represents dynamic characteristics of tunnel lining structure and surrounding soil. The results also show that the high-frequency response is greater than the low-frequency response at the tunnel lining. The attenuation of dynamic response along the longitudinal direction of the tunnel is obviously faster at the tunnel invert comparing to at the tunnel apex. For surrounding soil, a variety of dynamic response with depth is observed. A clear degradation of soil response along the longitudinal direction of the tunnel is found at all depths. The soil response increases with the increase of excitation frequency at the first measurement layer above the tunnel lining. However, at the second and third measurement layer, soil response increases linearly at the frequency of 30-90 Hz. At higher frequency range, soil response does not show a clear increasing trend with frequency. The dynamic response under train-vibration load is consistent with sweep excitation load. Both tunnel and soil responses decrease in the longitudinal direction. Tunnel response at the tunnel invert is larger than that at the tunnel apex. With the increase of the train speed, tunnel and soil responses are significantly amplified. It is also found that the soil responses at the free surface are more significant than the soil responses inside the soil layer from train induced vibration.

Due to the abundant negative charge existing on the surface of clayey soil particles, pore solution has significant influence on the physical and mechanical properties of clayey soils. The demanding geotechnical applications necessitates an effective numerical analysis of the detailed problems of chemo-mechanical coupling. Therefore, it is very important to develop a constitutive chemo-mechanical model on clayey soil saturated by salt solution. Based on the modified Cam clay model, a simple chemo-mechanical constitutive model is proposed. In the proposed model, the osmotic suction is adopted to describe the chemical state of pore solution in clayey soil. Formulas are presented to relate the pre-consolidated stress, the slope of critical state line M and elastic modulus to the osmotic suction. The comparison between simulated results and experimental data demonstrates that the proposed model is able to capture the isotropic compression, compression behavior for clayey soil saturated bysalt solution. The proposed model also can reproduce the response of clayey soil sample subjected to chemical and mechanical alternative loading. The simulated result for triaxial compression test also implies that the proposed model can capture the fundamental features of triaxial compression response of clayey soil.

In permafrost regions, canal is usually subjected to the horizontal frost heaving force, which impairs service life of canal. Based on the field test of U-shape subgrade canal in permafrost regions of Qinghai-Tibet plateau and the results of numerical calculation, the influences of canal depth, water content and width of surrounding soil replacement on temperature field and horizontal frost heaving characters were analyzed. The results indicate that U-shape canal depth, water content of soils as well as the width of surrounding soil replacement only have influence on temperature values of surrounding soils, but not on the shape of temperature contour of soils. The horizontal frost heaving force firstly increases and then gradually decreases along the depth of the structure, exhibiting varying values. The maximum frost heaving force occurs at about 1/2-1/3 depth of canal, which has the highest probability to have frost force-induced failure. The horizontal frost heaving force against the trapezoidal canal is about 13%-15% more than against the U-shape canal with the same depth. However, the horizontal frost heaving forces against two different canals have almost the same contour characteristics. The U-shape canal is superior to the trapezoidal structure in the perspective of reducing the horizontal frost heaving force. The more the width of soil replacing, the less the maximum horizontal frost heaving force. For the U-shape canal with 1.7 m in depth, the horizontal frost heaving force changes obviously when the width of soil replacement is from 0 to 2.8 m. However, the change is hardly seen when the width is beyond 2.8 m.

The thermal-moisture dynamics of the active layer directly affects the stability of permafrost and engineering projects in cold regions. Previous studies mainly focus on the thermal stability of permafrost based on the adherent layer theory. The migration process of liquid water and water vapor and its effects on the active layer are still lack of consideration. Considering the physical process and mechanism of liquid water-vapor transfer in unsaturated soils, a new heat and mass transfer model in saturated-unsaturated partially frozen soil was developed, in which the moisture migration in both vapor and liquid phases and heat transfer by means of conduction, convection and phase change were accommodated. The established water-vapor-heat transport model was used to analyze the water-vapor-heat transport in the shallow unsaturated zone of active layer under the in-situ meteorological conditions. The results show that the liquid water and water vapor are driven by the temperature gradient flow downward during daytime but upward at night, the liquid water and water vapor are driven by the temperature gradient flow upward in warm season but downward in cold season. Vapor water accounts for more than 15% of the water flux in the active layer and the water vapor driven by the pressure head can be neglected throughout the year. The moisture transport is mainly controlled by temperature gradient in sunny days. During and after rainfall events, rainfall infiltration is significantly enhanced and liquid water and water vapor mainly infiltrate downward. Rainfall can significantly decrease surface soil heat flux, heat conduction and soil temperature, which mitigate the permafrost degradation process.

To study different degrees of initial damage, different drying-wetting cycles and the coupled effects on the development and evolution of pore and fissure of expansive soil, drying-wetting cycle tests and triaxial shear tests (suction is 50 kPa, net confining pressure is 100, 200, 400 kPa, respectively) are conducted on Hefei expansive soil specimens for 3 groups with different initial degrees of damage (the diameter of the cylindrical hole is 0, 2.5 and 5 mm, respectively). If there is no drying-wetting cycle, a certain extent of initial damage can increase soil strength. If drying-wetting cycle is 1-2 times, the initial damage contributes to outcome of crack direction and penetration mode, forming intact block structure that has higher strength than that of the soil without initial damage. If drying-wetting cycle is 3 times, crack is able to expand completely and the strength of the soil with different degrees of initial damage becomes consistent with the soil without initial damage. The evolution mechanism of soil pore-fissure is that: the different strength and shrinkage deformation induced by hydraulic gradient when subjected to drying-wetting process will produce tension and compression stress in micro-crack tips and tear the tips up, which makes the tips combine with the initial hole damage, causing intersection and transfixion of the damage. Furthermore, the fracture is able to develop due to the increased water vapor exchange interface area in the fracture to further balance the compressive stress. The research provides a new insight for the study of expansive soil engineering with initial damage.

This paper describes the dynamic properties and Poisson's ratio of silty clay, aiming at the nonlinear characteristics of Poisson's ratio under low amplitude strain. A series of resonant column tests was conducted in torsional mode and flexural mode to determine the shear modulus G, elastic modulus E and Poisson's ratio ν. The results show that both E and G decrease with increasing strain, but not synchronously. The reduction of G precedes that of E with increasing strain. And Poisson's ratio ν decreases with the increase of confining pressure, and increases with increasing shear strain. In addition, an empirical model is proposed to describe the linear relationship between ν and in the semi-logarithmic scale, and to predict Poisson’s ratio. Theoretical analysis proves the inevitability of this linear relationship between ν and in small strain level. And with the increase of shear strain (or decreasing ), the ν- curve develops from linear phase to nonlinear regime, and finally approaches to a constant value. Moreover, the relationship between the fitting parameters and the constants of Hardin-Drnevich model is established. The fitting parameters obtained from linear phase can be used for the estimation of the Poisson's ratio in nonlinear phase through this relationship. The case study demonstrates that this method provides good results and has practical value.

Mechanical parameters of gypsum rock are influenced by the relative humidity in the mine atmosphere. Based on failure characteristics of the roof bed in gypsum mines, we proposed a simplified three-hinged arch analysis model under the uniform load q. Then a cusp catastrophe model was established to investigate instability mechanisms of the roof bed system influenced by the relative humidity. Experimental results show that mechanical parameters of gypsum rock are greatly reduced due to the influence of relative humidity. Analysis and calculation results show that the increase of the relative humidity around the roof bed leads to the instability of support system and the appearance bifurcation set. Therefore, it is more likely that the roof bed undergoes a catastrophic event. The algebraic value of control parameter (a) of the three-dimensional space gradually increases with the increase of relative humidity in the mine atmosphere (i.e., absolute value decreases and changes from negative to zero), and the control parameter (b) gradually increases as well. The path of the control parameter (a, b) of system instability shows the northeast direction, which is the probable path for the system to lose stability.

Based on the Young-Laplace equation, the relationship between pore size distribution and soil-water characteristic curve is estimated. The lagging model of soil-water characteristic curve is established by referring to the statistical model of unsaturated seepage. The capillary water corresponding to the upper limit of the capillary radius under a certain suction of the substrate has a blocking effect on the capillary water of larger pore radius. The probability of obstruction is directly related to the pore distribution function. The pore size distribution function reflects the heterogeneity of spatial distribution of soil pores. The model shows that the desorption curve is almost identical to the absorption curve when the substrate is in high or low matric suction. The water content in the middle section of desorption is higher than that of the absorption curve, and there is a clear peak. The results show that the model works well for medium and fine granular soil, but the sandy soil porosity is beyond the model hypothesis and resulting large errors. For sandy soils, the introduction of blocking rate correction coefficient shows that the best correction factor and SWCC half log under the SWCC maximum slope inversely proportional relationship.

Geological disasters occurred frequently in the reservoir area since the Three Gorges reservoir was built up, which highlights the importance to study the stability and failure mode of reservoir bank under the condition of reservoir water level fluctuation. A method to evaluate the slope stability by centrifugal test is introduced with Liangshuijing landslide as a testing case. Soil slope models considering geological conditions of the Three Gorges are also constructed. Through the techniques of digital photography, laser displacement sensor, pore water and soil pressure sensor, controlling of drawdown and charging of reservoir is implemented in the test, and a series of important test data such as pore water pressure, displacement, pore water pressure and soil pressure is obtained. Results indicate that the model displacement is mainly in the vertical direction from the beginning to the end of the water adding stage, and the horizontal displacement is much larger than that in water falling stage. With the rising and plunging of the reservoir water level during the experiment, crack formation appears at the leading edge first, then gradually extends at the middle and trailing edge. In this process, large deformation is generated at the middle edge and localized slip appears. In general, failure pattern of the model shows the trend of retrogressive mode. At the end of the experiment, volumetric water content gradually increases from the trailing edge. During the rise of water level, model deformation is relatively small. The main deformation occurs during the decline of water level. Therefore, the landslide is a dynamic-water-pressure type. The deformation and failure are mainly due to the dynamic water pressure effect. The failure mode and the mechanism of deformation instability under the condition of reservoir water level fluctuation are revealed by the experiment, which provides an important basis for the prevention and control of landslide in the reservoir area.

To reflect the different mechanical characteristics of sand under high and low pressures, the mechanical test results of sand, i.e., the strength, isotropic compression and critical state characteristic parameters, under a large range of confining pressure are analyzed to build a unified model. Correlation factor of stress path is used to adjust the plastic deformation related to stress path, and the hardening parameter can be used to describe the dilatancy behavior of relatively dense sand under low pressure. Based on the critical state behavior of sand, potential state surface is proposed to reflect the sand inherent state. With the dynamic relationship between the yield surface and potential state surface, the current compactness parameter and potential strength of sand are defined. The hardening parameter can reflect shear softening and shear contraction hardening properties of relatively dense sand under low and high pressures. Evolution law of yield surface and potential state surface is analyzed, and model prediction under different pressures is carried out. Results comparison validates the proposed model with the capability to reflect shear dilatancy softening and shear contraction hardening properties of sand under different pressures.

To simulate the rockburst process of deep circular cavern under three-dimensional (3D) stress conditions, the granite material with extremely strong rockburst tendency was used for testing. True-triaxial compression tests were conducted on the cubic granite sample (100 mm×100 mm×100 mm cube granite sample with a diameter of 50 mm through circular hole) by using a self-developed TRW-3000 true-triaxial rock electro-hydraulic servo mutation testing machine. In the experiment, an initial stress environment of 1 000 m depth was simulated. Then, the stress in the vertical direction was increased gradually, and the hole wall was monitored with a real-time micro-camera and acoustic emission (AE) device. The results show that failure process can be divided into four stages, i.e., calm stage, pellet ejection stage, rock exfoliation stage and rockburst stage. The middle parts of the circular hole have the highest stress concentration factor and the damage is initiated. Then cracks develop along the radial direction to the deep part of the tunnel, and ultimately form two symmetrical V-shaped notches. Three typical loading points reflecting the failure states of the hole wall are in line with the statistical relationship of the far field stress state and failure mode of deep circular roadway without support. Four kinds of rockburst criteria are also applied to analyse failure states of the hole wall under different loading conditions, and the results are consistent with the failure states recorded in the testing process. The failure of the wall is accompanied by an increase of AE count and AE energy. The more severe the rockburst is, the more active is the AE.

Coal, as a typical sedimentary rock, has a naturally anisotropic feature. For simplicity, coal is considered as an isotropic material in the study of coal permeability, and the corresponding isotropic permeability models have been proposed. However, the actual situation of gas-solid coupling in the field and laboratory tests cannot be well reflected by these models. In this paper, coal is treated as transversely isotropic, and an anisotropic permeability model is developed by using different directional modulus reduction ratios as the key parameters. The developed model is further implemented in COMSOL multiphysics software to comprehensively investigate the effect of coal anisotropy on gas diffusion and penetration. Theoretical and numerical results show that different directional modulus reduction ratios ( ) reflect the degree of anisotropy of the coal structure. When is different, the coal permeability is also not the same. Coal permeability is mainly governed by mechanical effects and desorption effects, and meanwhile, these two effects on each direction of the permeability of coal are controlled by boundary conditions. is a reflection of these two effects. Under the uniaxial strain or displacement control boundary condition, the horizontal modulus reduction ratio ( ) has more significant effect on the amount of permeability change in the vertical direction ( ) than on that in the horizontal direction ( ). However, the vertical modulus reduction ratio ( ) has less effect on than on .

Triaxial compression tests were conducted on salt rock specimens with an interlayer to investigate its failure modes and failure mechanisms. Based on the force balance of an interlayer element and the transfer mechanism of shear stress on the interface, we established a differential equation of the interlayer's radial displacement. The theoretical formulas of shear stress on the interface, and stresses and strains of the interlayer were obtained by solving the differential equation. Numerical simulation was also performed to verify the accuracy of these stress formulas. Moreover, failure modes and failure mechanisms of salt rock with an interlayer are analysed by various failure criteria. It is found that shear stress on the interface and stress of the interlayer are not uniformly distributed, and their maximum values are typically in the edge and the centre. With regards to vertical splitting of the interlayer, calculations suggest that tensile stress or tensile strain in the horizontal direction occurs in the interlayer. A ring-shaped white crack is observed in the edge of the interface of some specimens, which is caused by the maximum shear stress on the interface. Thus, the sliding failure of the interface is prone to happen. The induced failures of the interface and the interlayer lead to a decrease in tightness of salt rock with an interlayer.

Based on the principle of thermodynamics and the 2D computational model of liquid bridge, the relationship between matric suction and volumetric water content is derived with three unequal diameters of soil particles. The influences of particle size, gradation and the contact angle of soil particles on the soil-water characteristic curve are studied by simplifying soil particles as the constrained particle size, median particle size and effective particle size. The computational results show that the larger the size of the soil particles has the smaller corresponding matric suction under the same volumetric water content. Under the same matric suction, the volumetric water content of the soil particles is lower when the coefficient of non-uniformity ( ) is larger or the soil particle is more uneven. The smaller the contact angle is, the better the hydrophilic property of the soil particles is. The volumetric water content under the same matric suction is also greater when the contact angle is smaller. The air-enter value of the soil-water characteristic curve decreases with the decrease of the contact angle.

In this paper, materials such as plaster, pumice and diatomite are used to simulate water-bearing porous rock as similar material specimen. The experiment adopts four-factor (cement/plaster/diatomite ratio, the standard ratio of pumice sand, bone-glue ratio and barite powder ratio), five-level orthogonal experimental design to measure the density, uniaxial compressive strength, porosity, elastic modulus and softening factor. The measurements are conducted to determine whether the simulation materials can be used to simulate water-bearing rocks. With the use of direct analysis method, we determine how various factors affect physical and mechanical nature of the specimens. Analysis results show that: Due to different material ratios, the strength of the specimens is widely distributed. Comparing with the parameters of the porous layer, under certain similar conditions, we find that the specimens are with similar properties to the porous layer rock. Diatomite ratio is negatively correlated with density, compressive strength, elastic modulus and porosity, and it is positively correlated with softening coefficient. Pumice ratio is negatively correlated with density, compressive strength, elastic modulus and softening coefficient, and it is positively correlated with porosity. Therefore, addition of the two materials can simulate the porous nature of the rock sample well. Within the allowable range, the established empirical equation to simulate porous water-bearing rocks can be used in engineering practice.

The geological structures of the left bank slope at Baihetan hydropower station are extremely complicated. Many steep faults and interlayer staggered zones are the most significant risks which influence the stability of the left bank slope. Based on the analysis of the survey and field monitoring data, a numerical software UDEC was then adopted to build the numerical model for deformation analysis of the left bank slope. Under the condition of unsupported excavation, the stress and displacement changes of the key points of F17, LS331 and LS3319 at hanging wall and footwall rock masses were analysed to identify the potential instability failure areas. Meanwhile, the behaviours of potentially unstable block reinforced using pre-stressed anchorage cables were simulated at different construction procedures. The deformation characteristics, failure mechanisms and overall stability of the left bank slope were comparatively analysed using different models. Numerical modelling and field observation both show that the slope stability is strongly related to rock mass structural planes. Moreover, the deformation and damage of the left bank slope are mainly induced by the shear failure of interlayer staggered zones and tensile fracture of faults under unloading conditions. Pre-stressed anchorage cables can efficiently improve the stability and reduce the deformation of the rock slope. To efficiently control the shear deformation of rock slope, the excavation and support should be carried out simultaneously. The obtained results have particular significance to the design of reinforcement and construction process.

To consider the impact of displacement factor on the progressive development of landslide, a modified Janbu slices method considering the relationship between shear stress and shear displacement of sliding face was presented to evaluate the stability of a landslide. The modified Janbu slices method is improved in the following two aspects: firstly, combined the constitutive shear stress-strain model and shear displacement model along the slide face to simulate the progressive failure process of landslide with deformation development; secondly, the overall safety factor was redefined using concept of partial safety factor with clearer physical meaning. The formula of the simplified method was derived and implemented by Matlab code. An example of simulation shows that the stability factor for slope with strain-softening behavior depends on the strength parameters, and also on the shear stress-shear displacement relationship of slip zone soil. It is unsafe for the limit equilibrium method to be applied to practical engineering, This method can also effectively simulate the development process from the beginning towards the entire sliding face failure of slope. The simulation realizes the quantitative correspondence relationship between displacement and safety factor, and can be used to predict the long-term stability of slope.

The parallel small clear-distance tunnels have obvious dynamic amplification effect under earthquake loading. Small neighbourhood shield tunnels in the strong earthquake region were selected to investigate the dynamic response of shallow-buried parallel small clear-distance tunnels. A series of dynamic stress solutions was deducted for shallow-buried composite lined tunnels subjected to vertically propagating incident P waves using the Fourier-Bessel expansion method. Moreover, we discussed the influence of the distance between two tunnels and the grouting reinforcement zone on dynamic stress of tunnels. The results show that the distance between tunnels plays a vital role on the dynamic stress response of tunnels, and dynamic stress concentration factors of linings increase obviously as decreasing the distance of tunnels. The critical distance between the centres of two circular tunnels is proposed as 2.5 times diameter. The aseismic and damping design should also be suggested when the distance is below the critical value. The dynamic stress of linings decreases with grouting reinforcement ring zone set outside around tunnels, and the dynamic stress of linings decreases with increasing the stiffness of the grouting zone. The optimal ratio of shear wave velocity of the grouting zone to that of surrounding rock is suggested as 1.2. The recommended ratio of the thickness of the grouting zone to the tunnel radius is 0.5.

The increased length of mining face is one of the significant factors to increase the underground pressure in longwall (LW) face. The physical model tests showed the fractured shape of overlying stratum exhibited as trapezoid platform structure (TPS) after roof caving. Its forming mechanisms were investigated in detail and the calculation principle of geometric parameters of TPS. The effect of mining face length on the fracturing behaviour of overlying strata was studied according to the TPS, and then the relationship between the TPS and key strata theory was also given. The thin plate theory was applied to calculate roof weighting interval. Then, the influence of mining face length on TPS’s parameters was analysed as well as underground pressure. The results show that due to TPS-shaped fracturing characteristics of overlying strata, the thickness of loading layers and roof weighting intensity decrease with the decrease of mining face length. Field measurements were conducted in LW1305 with the face length of 85 m and in LW1302 with the face length of 180 m in mine No.2 of Zhaozhuang. The measured data show that the resistance of the supports should not be less than 4 738 kN and 7 623 kN in LW1305 and LW1302, respectively, and the roof is maintained well with the resistance of 5 500 kN in LW1305 and 7 800 kN in LW1302.

To address the reliability analysis problems of earth slopes system, a framework was developed consisting of three major components: generation of random sample by variance-reduction techniques, minimization of slope safety factors by global optimization algorithms, and determination of system reliability of slope by Monte-Carlo method. A relatively simple method for system reliability analysis of slope stability of high earth and rockfill dams was established within the proposed framework. The sample values for input random variables were produced using the Latin hypercube sampling technique. The minimum safety factors for slip surfaces were gained using the commercial slope stability analysis software STAB. Finally, the system reliability of dam slope was calculated by reliability-index method or Monte-Carlo method. The engineering example demonstrates that variability in shear strength parameters of dam construction materials has significant effect on the location of critical slip surfaces. The system reliability of slope is less than the minimum slope reliability of single slip surface. The proposed approach can be applied to system reliability analysis of actual dam slopes.

It is essential to assess the stability of coal pillar before mining for the evaluation of its rockburst risk and determination of mining schemes. Based on the geological conditions of a deep coal mine in Shandong province, China, these methods including in-situ case study, theoretical analysis, numerical simulation and engineering practice were adopted to evaluate the stability of the coal pillar with thick conglomerate upper strata. The deformation of conglomerate strata and coal pillar was analysed on the basis of a simplified mechanical model. We investigated the primary source of stress that caused coal pillar deformation and its deformation form, and obtained the stress-strain curve of coal pillar. Then, by considering the stress in coal pillar, surrounding rock stability and deformation, a mechanical criterion of coal pillar failure was proposed. It indicates that the location and size of a coal pillar, as well as the deformation of conglomerate upper strata, have significant impacts on the failure of coal pillar. The deformation of upper strata results in dynamic loading on the coal pillar. The deformation of coal pillar mainly includes collaborative deflection compression deformation induced by the concentration load F and the gravity settlement deformation caused by the concentration load G. The coal pillar failures when the supporting stress (p) is greater than its ultimate strength (Rc). It is found that the coal pillar with 50 m width has a high risk of rockburst, the risk decreased by optimising the mining scheme and achieved good effects. The results of this study are of significance for the analysis of the stability of coal pillar with similar conditions.

To improve the in-situ stress calculation method applied in hollow inclusion measurement, the formulas to calculate in-situ stress components are deduced by the least squares fitting method of the linear parameter by considering the installation mode of the hollow inclusion strain cell. We obtained a modified calculation method of six in-situ stress components with standard errors. In Gongchangling underground mine, measurements were performed by using the complete temperature compensation technique, and the improved hollow inclusion strain devices are applied at three points with the levels of -160 m, -220 m and -280 m, respectively. Three sets of borehole wall strain data are measured. The results show that the standard error of in-situ stress components calculated by the modified calculation method is less than that calculated by the traditional method. Therefore, the reliability of calculated results is improved by using the modified calculation method. Based on six obtained in-situ stress components calculated by the modified method, the stress states of Gongchangling underground mine are analyzed. The results show that the in-situ stress is dominated by the horizontal tectonic stress field in the mine area. The maximum horizontal principal stress is oriented in SEE-NWW direction. The maximum horizontal principal stress, the minor horizontal principal stress and the vertical principal stress are increased with the depth.

As one type of green energy, offshore wind power has been paid more and more attention. There are extensive research focused on local scour of the monopile foundation. In this paper, a model of offshore wind turbine system with different scour depths is established. The static and dynamic analysis of the model is carried out. Results are in good agreement with the previous scouring model test. As the increase of scour depth, the maximum horizontal displacement, the maximum horizontal stress and the maximum bending moment values of the support system increase gradually. When the scour depth is twice the pile diameter, the maximum horizontal displacement, stress and bending moment are increased by 3.61%, 12.7% and 10.3% respectively, comparing to those under the condition without local scour. The natural frequency of the system decreases with increase of scour depth, and frequencies of high order are more susceptible. Scour around pile foundation has a significant impact on dynamic characteristics of system. The horizontal displacement at the top of the tower constantly increases with scour depth. The influence of scour on the bearing performance of the supporting system must be considered in the design of offshore wind turbines.

Faults, joints and fractures are widely distributed in rock masses, and these discontinuities of existence and development have a significant effect on the whole strength, deformation and stability of rock mass. Thus, studying the evolvement of initiation, propagation and coalescence of primary flaws is of significance in theory and practice for estimating the safety and reliability of rock engineering. The extended finite element method (XFEM) is a numerical method for modelling discontinuities within the classical finite element framework. Since the computation mesh in XFEM is independent of the discontinuities, re-meshing for moving discontinuities can be overcome. Owing to the unique advantage for fracture analysis, XFEM has been employed to simulate hydraulic fracture and crack propagation of rock mass. The basic theory of XFEM and its application in simulating crack propagation are studied in detail. The numerical model of solving the frictional contact problem and hydraulic fracture propagation is developed. The calculation model is used to solve some geotechnical problems, such as deformation and failure of jointed rock slopes and modelling fracture propagation of gravity dam foundations.

Fracture seepage can result in a change of temperature fields of rock around cracks, which is more apparent in particular under low temperature. In addition, heat exchange between fracture water and the cold rock medium may induce water/ice phase transition in fractures. The generation of solid ice prevents water from flowing in fractures, which leads to a change of the seepage field of fractured rock mass. Thus, this hydro-thermal coupling action of fractured rock mass under low temperature is extreme. By considering water/ice phase transition and fracture seepage, a coupled hydro-thermal model is developed for fractured rock mass under low temperature. To illustrate the influence of fracture flow on the freezing process, an example of the artificial freezing method is investigated. The results show that the rock medium far from fractures is frozen earlier for fracture flow and the completed freezing time of seepage fractures are more than that of rock medium. Both the fracture width and delivery head of fracture water affect completed freezing time. The completed freezing time increases with the increase of fracture width and delivery head of fracture water. The seepage velocity in fracture gradually decreases along with freezing time, and the fracture seepage stops after fracture water is frozen. Finally, by building a stochastic fracture network model, the impact of seepage in fracture network on the freezing process is studied using the proposed coupled hydro-thermal model. The calculated results indicate the significance of considering fracture seepage during artificial ground freezing.

Constitutive model of tailings sand plays an important role in the numerical simulation and security evaluation of tailings dam. However, current researches focus on nonlinear elastic model such as Duncan-Chang model, but less on elastoplastic model. A modified generalized plasticity model considering the stress-strain property of tailings sand is proposed to describe the mechanical characteristics of tailings sand. The constitutive model is developed in ABAQUS with user defined material subroutine UMAT and explicit stress integration of Runge-Kutta method. Finally, the modified model is verified in ABAQUS by triaxial test simulation. The deviatoric stress curve shows that the results of finite element method can reflect the influence of confining pressure on stress-strain relationship. And the shear strength gradually increases with strain increases, followed by strain softening after reaching the peak strength. Volumetric strain calculation results show that the generalized plasticity model can well describe the volumetric strain development, which is in good agreement with the experimental curve. Meanwhile, finite element computation data match well with test data, and the theoretical value. The research results can be further used in the stress-strain analysis and security evaluation of tailings dam.

Calculation of buoyancy plays an important role in the anti-floating design of underground structures. However, the determination of the buoyancy in weakly permeable clay is controversial. An apparatus of model test is designed to avoid the effect of the friction in the determination of the buoyancy of underground structure in soft clay. The weight unloading is adopted in a simple and explicit mechanical model. The buoyancy is equal to the surplus weight at floating moment. And the critical state of floating is determined by the real-time monitoring of pore pressure and displacement. Results of experiment show that the varying pore pressure in soft clay is influenced by seepage, temperature, unloading rebound, and soil water interaction, and the quantity is often less than the theoretical water pressure. Measured buoyancy is 0.93 times theoretical value, which presents a small amount of reduction. The real-time curves of water pressure and displacement indicate that the floating of structure is a continuous and dynamic process till to suddenly occur.

The optical fibre sensing technology has been widely used to measure the internal strain of similar material. However, the coupling between the optical fibre and similar materials has not yet been resolved. In previous studies, theoretical results of strain transfer cannot be validated in the experiments. A one-dimensional optical fibre Bragg grating (FBG) similar material was designed and realized using three layers of fibre-adhesive-similar material. The sensor provides a method to solve the problem of coupling between the fibre and similar material. The FGB sensor without coating layer was used as the sensing element, and the axial centre strain was measured by using the fibre embedded in the similar-material strain testing pieces. We investigated the behaviour of strain transfer using theoretical analysis, numerical simulation and a newly designed calibration. The results of the average strain transfer rate of the sensor obtained by different methods were very close, which demonstrated good coupling among fibre, adhesive and similar material. Moreover, the accuracy and reliability of theoretical analysis and numerical simulation were verified by a new calibration experiment. The method provides a feasible solution for the three-dimensional strain measurement of similar material in the model.

There are a large number of long deep-buried tunnels either being designed and under construction in Western China. The mechanisms of rockburst hazards and their monitoring and early warning in those deep-buried tunnels are particularly important. Microseismic (MS) monitoring, as a spatial monitoring technology for microcrack initiation and propagation, plays a vital role in predicting and warning rockburst in underground caverns. However, the application of evolution of MS source parameters to assess rockburst hazards has not been efficiently applied in practical projects due to the problems of discreteness (spatial), fluctuation (temporal) and deviation of measured MS data. To improve and solve these problems properly, this paper establishes an EMS method to evaluate and predict rockburst hazards, based on three commonly used source parameters (i.e., seismic energy, the seismic moment and apparent stress) which can be easily obtained and closely interrelated. This method fully defines the concepts of evaluation parameters, such as seismic energy, seismic moment and apparent stress, and thus the relationship among these three parameters can be obtained. Then, these three parameters are applied to predict rockburst and to evaluate the grade of potential rockburst. Meanwhile, the method defines concepts of MS path and the space of MS source parameters. Finally, the EMS method is adopted to analyse a rockburst case in the Micangshan tunnel of Bazhong-Shaanxi highway, Southwestern China. The progressive fractures and rockburst accompanied by MS events are detected and analysed, which achieves an efficient way for daily analyses of MS data and rockburst prediction.