To explore the anisotropy mechanism of joint shear strength from its damage characteristics, shear strength and failure characteristics were quantitatively analyzed. The study found that the shear strength and failure characteristics of joint show similar anisotropy with its morphology, and their anisotropy is weakened with the increase of normal stress. In addition, the influence of the inclination and height characteristics of the sawtooth joints on the shearing behavior was investigated. The analysis shows the angle and height characteristics of asperities have a positive correlation with shear strength, and the angle characteristics also determine the locality of asperities contact state during the shear process. The rough joint surface is composed of many microscopic asperities. The essential reason for the anisotropy of shear strength is the different contact area on the joint surface in different shear directions and the difference of asperities height and dip angle in the contact area. With the increase of normal stress, the failure of more and more micro convex bodies on the joint surface evolves from sliding wear to shear failure, and the difference of failure volume in different shear directions gradually decreases, which leads to the gradual consistency of the energy required for shear, which is the main reason why the anisotropic characteristics of shear strength are weakened by normal stress.

To investigate the description and evolution of the damage state of rock-like brittle materials, the physical meaning of each elastic modulus method parameter based on the strain equivalence hypothesis and the limitations of the model application are discussed. The modulus change during the triaxial cyclic loading and unloading test of limestone is studied. Moreover, the defects of the unloading modulus substitution method and the statistical damage evolution model in damage evolution analysis are discussed. The results show the existing elastic modulus method can only be used to reflect the damage evolution process of rock under uniaxial compression, and the unloading modulus substitution method cannot correctly describe the damage state and its evolution law. In addition, the statistical damage constitutive model can only be regarded as a theoretical self-consistent solution under the numerical range [0, 1] of the statistical damage evolution model. Based on the above research, a damage characterization variable and its evolution model considering the effects of damage strain threshold are proposed. Additionally, the constitutive model below the damage strain threshold and the damage constitutive model above the damage strain threshold are established, respectively. The sensitivity of model parameters is also analyzed in this study. The final results show that the proposed model and method can not only reasonably explain the damage mechanism of rocks under triaxial compression, but also accurately simulate the full stress-strain process, which is rationable and feasible.

As the preferred riser system for deep-sea oil and gas resource extraction, steel catenary riser (SCR) has a large impact on the depth of burial and fatigue life of the pipeline in the touchdown zone due to its interaction with the seabed soil body. According to the nonlinear soil resistance-intrusion depth p-y curve of pipe soil interaction, it is simplified as a three segment linear elastic soil stiffness attenuation model, and the analytical solution of initial penetration static equilibrium of catenary riser on seabed is obtained based on the nonlinear Pasternak foundation model. Compared with three-dimensional finite element and five model test cases, it is found that conventional Winkler foundation model overestimates the pipeline vertical deformation and bending moment of soil, which verifies the rationality and applicability of the analytical solution based on nonlinear Pasternak foundation model; the increase of seabed shear strength Su0 significantly improves the stiffness of three-stage linear elastic soil and reduces the pipeline vertical deformation. In addition, according to the changes of water depth H, riser laying angle ?, pipeline outer diameter D, pipeline elastic modulus E and material density ?, the physical and mechanical properties of pipeline penetrating soil are compared and analyzed. The results show that with the increase of vertical pipe laying angle ?, the pipeline vertical deformation will be reduced, while the bending moment and shear force will be increased. When the angle exceeds 82o, it is prone to yield failure. With the increase of pipe outer diameter D, the soil vertical resistance, vertical deformation, bending moment and shear force in touchdown zone will be increased simultaneously. When the outer diameter exceeds 0.4 m, it is easy to yield. The larger the elastic modulus E is, the smaller the pipeline vertical deformation is, the greater the bending moment and shear force are, when the elastic modulus exceeds 275 GPa, it is easy to yield. When the density of pipeline material is higher, the pipeline vertical deformation is larger, the bending moment does not change obviously, but the shear force increases, and when the density exceeds 14 850 kg/m3, it is easy to yield. The above conclusions can provide a theoretical basis for the preliminary design of catenary riser of offshore pipeline.

Rockburst mechanism of surrounding rock masses in underground cavern group under high stress and strong unloading is a frontier subject, and needs to be studied and solved urgently at present. The rockburst process of the granite from the Shuangjiangkou underground caverns under various intermediate principal stresses was simulated through three-directions and five-sides true triaxial tests. The fracture surface of the specimens after rockburst was observed by scanning electron microscope and energy dispersive spectrometer. The strength and deformation characteristics, macroscopic failure modes, rockburst process and meso-fracture modes of main minerals of the specimens under various intermediate principal stresses were analyzed. The experimental results show that: 1) With the increasing intermediate principal stress, the macroscopic failure of the specimens after rockburst changes from tensile failure to shear failure. 2) With the increasing intermediate principal stress, the specimen rockburst types change from time-delayed rockburst to immediate rockburst, and the rockburst intensity first increases and then decreases. 3) Under each intermediate principal stress, the fracture pits of the specimen after rockburst are primarily shear fracture, while the fractures of the ejection debris are tensile fracture. 4) The meso-fracture mode of main minerals in the studied granite is independence on the intermediate principal stress: the probability of intergranular and transgranular fracture of quartz and K-feldspar is almost the same, while plagioclase and mica are more prone to transgranular fracture. The research results can provide technical support for rockburst prevention during subsequent excavations.

Geosynthetic clay liners (GCLs) are newly used in combination with the vertical barrier wall in waste landfills. One of the critical points of this application is to control the preferential flow in the overlapping area of adjacent GCLs. However, there is still a lack of reliable data. This paper aimed to investigate the influence of the overburden pressure on the equivalent hydraulic conductivity (ko) of the overlapping area with bentonite pastes, to the reference base case in which no bentonite paste was presented. A penetration apparatus with internal dimensions of 1 200 mm×700 mm×700 mm(length×width×height) was presented. The 500 mm long GCL overlap was tested under a combination of the hydraulic head being 1 m and the overburden stress being 10, 25, 50, 100 and 150 kPa, respectively. It was found that: 1) The preferential flow around the overlapping area of GCLs occurred at the initial stage of penetration. The ko of the overlapping zone under an overburden pressure of 10 kPa was 5.6 times greater than that of the GCL due to the existing preferential flow. 2) The ko of the overlapping area decreased with an increase in the overburden pressure. The ko of the overlapping area subjected to a low overburden pressure of 10 kPa was equal to 2.27×10?8 cm/s, while it decreased to 5.93×10?9 cm/s when the overburden pressure reached to 100 kPa. 3) The permeability was significant reduced by the presence of bentonite pastes in the overlapping area. The ko of the GCL overlapping area was lower than that without the bentonite paste, and the value of ko reduced to 5.15×10?10 cm/s when the overburden pressure was as high as 150 kPa.

The unsaturated permeability curve is governed by the pore-size distribution curve, which can be used to predict the permeability curve. In order to investigate the applicability of this method for compacted loess, three groups of compacted loess samples with different dry densities were prepared. The pore-size distribution (PSD) curves of soil samples were measured using mercury injection porosimeter test. The unsaturated permeability curves of soil samples were measured by small soil column equipment designed by our research group. Then the PSD curves were used to predict the permeability curves and compared with the measured data from soil column test. The results show that the permeability curves of compacted loess can be divided into a low matric suction stage dominated by capillary water, and a high matric suction stage dominated by adsorbed water. In the low suction stage, the permeability curves of the three soil samples differ greatly. While in the high suction section, the permeability curves of the three soil samples coincide, indicating that the permeability of the high suction section has nothing to do with the density of soil. In addition, the predicted results of three groups of soil samples are in good agreement with the measured data in the low suction section, but the predicted results in the high suction section are smaller than the measured results. It can be seen from the principle of the prediction method that this method is suitable for capillary water but not for adsorbed water. Therefore, a modified method is proposed for the prediction of permeability curve in the high suction section, and the modified method can describe the permeability curve in the whole matric suction range.

Dispersive clay is a type of water sensitive soils, which has the characteristics of dispersing and material losing when encountering water. The influence factors and modification mechanism of dispersive clay modified by calcium lignosulfonate were studied using dispersity tests, mechanical property tests, chemical property tests, microstructure tests and simulated rainfall scouring tests. The results indicated that with the increase of calcium lignosulfonate content, the dispersity, disintegration and erosion resistance of the modified soil were gradually improved, and it had a good modification effect when the content reached 3.0%; the unconfined compressive strength first increased and then decreased, and the compression coefficient first decreased and then increased, and both achieved extreme values at 0.5% content. With the increase of curing age, the dispersity and disintegration of the modified soil decreased gradually, the compression coefficient decreased significantly, and the compressive strength increased gradually. At the age of 28 days, the unconfined compressive strengths of the modified soil at 0.5% and 3.0% contents were increased by 50% and 20% respectively, compared with that of dispersive clay. The engineering properties of dispersive clay were improved by the addition of calcium lignosulfonate mainly through reduction of thickness for electric double layer, cation bridging, particle cementation and hydrophobic effect of hydrophobic group. However, when the content of calcium lignosulfonate was too large, the calcium lignosulfonate would preferentially combine with itself and weaken the attraction between soil particles, which would increase the porosity and decrease the mechanical properties of soil. The research concludes that calcium lignosulfonate has a good modifying effect on the dispersive clay, which can significantly improve the water sensitivity, water stability and erosion resistance of clay.

The special joint network of columnar jointed rock mass makes its anisotropy significant, and accurately understanding its deformation and strength characteristics are crucial for engineering safety. Based on the natural structure of columnar jointed rock mass, quadrangular, pentagonal and hexagonal prism columnar jointed rock mass specimens with different dip directions and dip angles were manufactured by using gypsum and other materials. Uniaxial compression tests were carried out to study the influence of cross-sectional shape, dip direction and dip angle on the anisotropic characteristics. The typical failure modes and the mechanisms of columnar jointed rock mass were summarised in accordance with the final appearances of specimens. The empirical equations were adopted to estimate the strength and deformation of the columnar jointed rock mass. The results show that the cross-sectional shape mainly affects the deformation and strength parameters of the specimens with different dip directions, but has little effect on the anisotropy of the specimens with different dip angles. The anisotropy of quadrangular, pentagonal and hexagonal prism columnar jointed rock mass are orthotropic, orthotropic and quasi-transverse isotropic, respectively. The final appearances of specimens present four typical failure modes, and the cross-sectional shape mainly affects the failure mode of the specimens with different dip directions. Furthermore, the calculated results of the proposed empirical equations are in good agreement with the existing results, and have high engineering value.

Expansive soil combined with expanded polystyrene(EPS) can absorb expansive energy through deformation characteristics of EPS, improving swelling and shrinkage of expansive soil. The combination can also have excellent effect of heat preservation and vibration reduction. The dynamic stress-strain, dynamic elastic modulus and damping characteristics of EPS modified expansive soil under different conditions are studied by means of GDS dynamic triaxial instrument under different confining pressures and frequencies. The test results show that: 1)With the increase of dynamic stress, the dynamic strain and energy loss of EPS modified expansive soil increase non-linearly, while its dynamic elastic modulus decreases. 2) With the increase of frequency, the dynamic strain of EPS modified expansive soil decreases while the dynamic elastic modulus increases. 3) With the increase of confining pressure, the dynamic elastic modulus of EPS modified expansive soil decreases first and then increases. The damping ratio and dynamic stress curves of EPS modified expansive soil at different frequencies have obvious intersection points, which deviate to the left with the increase of confining pressure. When the amplitude of dynamic stress is small, applying high-frequency load will produce a larger damping ratio. In the later development period of dynamic stress, a larger damping ratio corresponds to a lower-frequency load. Under the condition of same frequency, the damping ratio of improved soil decreases with the increase of confining pressure.

In this paper, a large number of one-dimensional tests, including constant water content compression and soaking under constant stress, are conducted. The microstructure evolution and deformation mechanism of the compacted loess under loading and wetting conditions are investigated with mercury intrusion porosimetry (MIP) and scanning electron microscope (SEM) analysis. Experimental results show that, as the saturation of compacted loess increases at a constant moisture content, it will develop into a saturated consolidation process under further compression. At the microscopic level, the compression of the unsaturated compacted loess results from the collapse reduction of its macrospores, while the distribution of microspores is unaffected in compression. During increasing wetting under the constant vertical stress, the wetting deformation of compacted loess shows a trend of increasing and then decreasing with the increase of vertical stress, and the maximum wetting strain occurs near the compaction stress. Under wetting conditions, the bonds between particles and aggregations are weakened, and the particles and agglomerates collapse and slip, resulting in the reduction of macrospores and the increase of microspores. Also, the soil structure tends to be more uniform and stable after wetting. The creep of compacted loess is caused by the further slippage of particles under constant load and further compression of macrospores. In addition, the settlement law of compacted loess fill is summarized from the construction and post-construction period according to testing results.

When shield is driving along a curve alignment or during deviation correction, the asymmetrical thrust will generate an additional bending moment at the head of the tunnel, which will cause construction problems such as longitudinal deformation of the tunnel and dislocation between rings. Based on the theory of elastic foundation beam, the shield tunnel is simplified as a Timoshenko beam in Winkler foundation, an analytical model for evaluating the additional response of the shield tunnel induced the asymmetric thrust is established, and the analytical solutions of longitudinal deformation and internal force of shield tunnel are deduced. Then the correctness and applicability of the analytical solutions are verified based on the finite element method, and the sensitivity of key parameters in the analytical model is further analyzed. Finally, the influence range of additional bending moment and the second order effect of shield thrust are discussed. The research results show that the proposed analytical model is reliable and has good applicability for evaluating the additional response of shield tunnel under asymmetric thrust. The influences of foundation stiffness and tunnel stiffness on the longitudinal deformation of the tunnel are more significant than the internal force. The influence range of the additional bending moment is more sensitive to the changes of the foundation stiffness, and an exponential attenuation relationship is found to exist between the two. The shield thrust improves the longitudinal bending stiffness of the tunnel, and the second-order effect produced is relatively low. Under the condition of lower bending stiffness of tunnel and foundation stiffness, the second-order effect of axial force is enhanced.

A roof-rib pillar support system is often formed in the metal mines that applied to the backfill mining method. In order to ensure the safe mining, it is of great significance to explore the failure mechanism of the roof-rib pillar support system under backfill mining. On the basis of considering the side pressure effect of the backfill on the rib pillar, a mechanical model of the roof-rib pillar support system under backfill mining was established. Furthermore, the catastrophe theory was used to explore the failure mechanism of the supporting system under the action of the filling body, and the influence of the structural parameters of the support system before and after filling on the stability of the stope was analyzed. The research results show that under the condition of certain mechanical properties of the rock mass, the stability of the roof-rib pillar support system under backfill mining is controlled by the stope structure parameters (roof thickness, stope span, rib pillar width, rib pillar height), the overburden load and the side pressure of the backfill. The addition of the filling body will reduce the stiffness ratio of the support system, thereby improving the stability of the stope. When the stope is at an unfilled state, the optimization sequence of the structural parameters of the support system should be roof thickness, stope span, rib pillar width and rib pillar height. When the stope is at the state of filling, the optimization sequence should be roof thickness, rib pillar width, stope span, and rib pillar height. The combination of theoretical derivation and numerical simulation is applied in the supporting project to verify the correctness of theoretical derivation.

X-section cast-in-place concrete pile (X-section pile) is a kind of new special-shaped pile. Considering the distribution of dynamic friction resistance on concave arc section, convex arc section and soil-pile coupled vibration, treating the soil as axisymmetric homogeneous viscoelastic medium, the analytical solution in frequency domain is obtained by using Laplace transformation technique and coordinate transformation. The solution is compared with that of solid circular pile to verify the rationality. Moreover, by analyzing the effect of pile length, open arc spacing and open arc angle on complex dynamic stiffness and velocity admittance, the influence of dimension parameters on the vertical vibration characteristics of X-section pile is investigated. It shows that the complex dynamic stiffness at the top of pile increases, the oscillation amplitudes and resonant frequencies of the velocity admittance curves significantly decrease with the increase of pile length. The open arc spacing and open arc angle have little influence on the velocity admittance at low frequencies. At high frequencies, the velocity admittance decreases with the increase of open arc spacing and increases with the increase of open arc angle. However, oscillation amplitudes of the velocity admittance curves increase with the increase of open arc spacing and decrease with the increase of open arc angle.

To unify the source power of water migration, a pressure-suction element model of film water is constructed on the basis of the film water hydraulic driving force model and the surface adsorption force model. Model analysis shows that under the dual action of net suction and actual liquid pressure (or theoretical suction and actual ice pressure), the surface adsorption force can be generated, which drives the tangential migration of water along the surface of the substrate. In view of the fact that the surface adsorption force has nothing to do with the boundary conditions, it is suitable for any form of unfrozen water, and it is the unified source power of water migration. Based on this, the pressure-suction element model is introduced into the frozen fringe theory, and it is found that the actual ice pressure determines the temperature and position of the segregated ice formation, the theoretical suction determines the direction of water migration, and the surface adsorption force determines the velocity of water migration. Finally, substituting the main parameters from the Konrad (1980) test into the surface adsorption force equation, it is found that even if the temperature gradient increases from 0.1 ℃/cm to 0.67 ℃/cm, the sample height increases from 6.4 cm to 28 cm, as long as the segregation freezing temperature and the overburden pressure keep unchanged, the surface adsorption force is always constant at ?23 kPa, which verifies the correctness of the surface adsorption force equation. In short, the development of this model has important theoretical value and practical significance for improving the existing frost heave theory and guiding engineering practice.

The variation of relative humidity induces the migration of vapor, which has an important effect on the strength and deformation of earth. The sorption-desorption and migration of vapor inside the earth are of great significance for revealing the disease breeding mechanism of earth structure. Aiming at two different earth materials named STR and CRA chosen from a few existing construction sites located in Lyon region in the southeast of France, the sorption-desorption tests were conducted through two different methods, saturated salt solution method and dynamic vapor sorption (DVS) separately, and the experimental results were compared accordingly. In addition, an innovative apparatus with high-precision strain measuring system was invented and used for monitoring the vapor migration, the variation of relative humidity and gas pressure with time at the top and the bottom of earth specimens was explored, and the deformation characteristics of earth specimens during vapor migration were also analyzed. The experimental results indicate that the sorption curve obtained through DVS is in accordance with that from saturated salt solution method. The sorption curve can be divided into three parts: monolayer adsorption, multilayer adsorption and capillary condensation. Meanwhile, the retention characteristics of earth increases with the rise of clay activity. The retention characteristics have important effect on the relative humidity at the outlet and relative gas pressure at the bottom of specimens. The dilatancy phenomenon occurs during the vapor migration process.

The constitutive relationship between suction and degree of saturation is of great significance for estimating the shear strength and deformation behavior of unsaturated soils. A unimodal/bimodal soil-water retention curves (SWRC) is proposed considering the effects of various pore structures and the effects of void ratio on capillarity and adsorption. The different mechanisms of water retention through capillarity and adsorption are explicitly distinguished in the proposed model. The relationship between the capillary degree of saturation and suction is described as a specific function related to the characteristics of pore-size distribution, while the adsorptive degree of saturation is modeled considering the effect of capillary condensation explicitly. Subsequently, the decoupling formula of capillary and adsorptive saturation is further put forward. The formula lays a foundation for the model to account for the dependence of capillary part of SWRC on void ratio, which is consistent with the results from micro-scale tests. Eventually, an approach to estimate the shear strength of unsaturated soils with different initial void ratios has been proposed based on the improved SWRC model. The model is verified using data from water retention and direct shear tests reported for various types of soils in the literature.

Gassy soils are widely distributed in offshore sediments around Yangtze River and Pearl River in China. The pore gas usually exists in the form of discrete bubbles and is sensitive to the variation of the temperature and the pressure, which in turn leads to the changes of soil properties such as compressibility and permeability. Thus, to investigate the influence of thermal expansion of occluded gas phase and creep deformation of soil structures on the consolidation process of gassy soil in the offshore region, a study of thermos-hydro-mechanical coupling (THM) behavior of gassy soil is conducted based on the viscoelastic model. In the governing equations, the change of gas phase volume subjected to thermal and mechanical loadings is introduced to take the thermal expansion into consideration. The fractional derivative Merchant model is also adopted to describe the process of rock creep. In addition, these problems are solved by precise integration method (PIM), which proves to be efficient and several orders more precise than conventional numerical methods, combining with integral transform. For verification, the results obtained by this newly proposed method are compared with the analytical solution for THM problems of saturated soils and numerical prediction by FEM for consolidation problems. Typical examples with different saturations and viscosity coefficients are performed to investigate the effects of viscoelasticity and thermal expansion of gas phase on the time-dependent behavior of gassy soils. It is found that the viscosity has a significant impact on the thermal consolidation process. The creep characteristic shows a greater influence on deformation than the excess pore pressure. When the viscosity coefficient is greater than 1×1012 MPa·s, the excess pore pressure is similar to that of elastic soil foundation, and meantime the deformation process is obviously delayed. The occluded gas phase increases the peak values of excess pore pressure and surface deformation, which may reduce the stability of buried pipelines with high temperature. For Gibson soils, using the means of shear modulus to predict the excess pore pressure may result in non-ignorable deviations but the predicted surface deformation is with a deviation of less than 5%.

The liquefaction of saturated fine sand is a crucial difficulty hindering the development of long auger cement fly-ash gravel (CFG) pile construction technology. However, the relationships between liquefaction and two construction-related parameters of CFG piles, namely the pile spacing and the pipe rotation speed, during the piling process have not been well investigated. Thus, in this paper, the influences of these two parameters on the liquefaction of saturated fine sand are analyzed, and the relationships between the two parameters and liquefaction are established. Besides the theoretical studies, an experiment platform is also established to reproduce the construction process of a long auger CFG pile in the saturated fine sand. The results of laboratory tests show that the increase of pipe rotation speed and the decrease of pile spacing promote the liquefaction of saturated fine sand, shorten the pile length, and increase the pile diameter. In order to relieve the liquefaction of saturated sand, the pipe rotation speed should be reduced and the pile spacing should be increased.

During similar simulation experiments in geotechnical engineering, cracks and speckles falling off appear on the surface of the equivalent material model, and invalid areas will form at the excavation zones and model surface penetrations. However, the traditional digital image correlation (DIC) method had some limitations in these cases. To improve the applicability of the DIC method in similarity simulation experiments, a novel method was proposed based on the image feature matching algorithm. In the proposed method, firstly, reliable subsets were selected by feature matching of images before and after deformation; secondly, deformation parameters of reliable subsets were obtained by correlation operation; finally, deformation parameters of monitored points were determined by the least absolute deviation fitting of displacements of reliable subset center points in certain areas around the monitored points. To validate the proposed method, deformations of the similar simulation experiment of mining-induced fault slip were measured using the proposed method, and the results were compared with those of the traditional method. Results show that cracks and invalid areas can be accurately identified by the proposed method, and displacements measured by the proposed method are more complete than those by the traditional DIC method in the vicinity of these areas. The proposed method has strong stability for speckles falling off on the surface of the model.

High-frequency impact drilling speed-up theories and its tools have rapid developed in recent year, and effectively solve the problem of the stick-slip, rebound and vibration of PDC (polycrystalline diamond compact) bit in complex wellbore. However, the failure of PDC teeth under impacting load remains serious. For this reason, a new type of drill bit combining of separated impact and cutting is proposed, this drill bit can effectively solve the adverse effect of impact load on the service life of PDC teeth. The rock breaking of the drill bit is totally different from that of traditional percussion tools, and the mechanism of the rock breaking is unclear, it is thus urgent to perform relevant study. Based on the new type of drill bit combining of separated impact and cutting, the rock breaking experiments under different combinations of unit tooth impact, cutting and their different combinations are carried out, and the corresponding numerical analysis is carried out. The results show that the rock with significant impact pit is damaged by the shock wave after the impact tooth loading, consequently, reducing the strength of the rock around impact pit. As a result, it also effectively reduces the cutting force of the cutting teeth during cutting, and increases the crushing volume. At the same time, based on the orthogonal experiment, the parameter sensitivity analysis is carried out considering the impact speed, the impact angle and the offset distance between the impact pit and the cutting tooth. The results show that the impact speed has the greatest influence on the cutting force of the cutting tooth. The above research reveals the mechanism of rock breaking under combining of separated impact and cutting and the best combination of impact tooth and cutting tooth. The present work has important reference value for the design and research of impact drilling tools.

Widely-graded soil is a good anti-seepage filling material for core of earth dams and embankment, and its impermeability is determined by its hydraulic properties. For the same soil, compaction is the primary factor that controls its hydraulic properties. In this study, the effect of compaction on hydraulic properties of a widely-graded soil is studied by laboratory tests, and through data analysis, the hydraulic property model is established, which is based on the effect of compaction. The results demonstrate that the higher the degree of compaction, the less macropores between particles of the soil, the higher the air-entry-value of the soil, the lower the saturated permeability coefficient. A prediction model of osmotic suction is proposed and it can accurately describe the changing principle of osmotic suction for unsaturated soil upon drying. The air-entry-value of the soil has linear relationship with the relative compaction on semi-log coordinate, and saturated permeability coefficient has linear relationship with the dry density on semi-log coordinate. An experiential model among air-entry-value, saturated permeability coefficient, unsaturated permeability coefficient and relative compaction is established, also an experiential calculation formula for soil-water characteristic curve (SWCC) of widely-graded soil is established, which has general applicability.

Columnar jointed rock mass with unique joint network structure is a special type of jointed rock mass, which is a binary structure composed of high strength "basalt block" and specific "dominant joint". Relaxation, opening and slippage of columnar jointed surfaces and disintegration of columns occur easily during excavation under high ground stress, which eventually lead to disastrous collapses in columnar jointed rock mass. The construction safety of underground engineering under high stress is severed restricted by it. By combining acoustic wave, borehole camera and other integrated in-situ testing technology and numerical simulation, the mechanism of stress-structural collapse in columnar jointed rock mass under high stress is studied based on multiple columnar jointed rock mass collapses at left bank tailwater connection pipe of Baihetan hydropower station. The proposed controlling measures of excavation and support are also provided. The key of columnar joined rock mass collapse is the redistribution of stress in surrounding rock mass after the excavation and the strong unloading relaxation of columnar joined basalt, which causes the opening of its internal joint surfaces and structural deterioration, result in the disintegration of basalt columns. Thus, the chain catastrophic process of continuous unloading relaxation and progressive collapse of columnar jointed rock mass is induced. The research can provide reference for the prediction and control of deformation and failure of jointed rock masses in underground engineering under high geo-stress.

Based on dynamic testing technology applying (earth pressure cells, EPCs) and (linear variable differential transformers, LVDTs), combining with the field vibration compaction tests, the internal stress and deformation characteristics and energy input mechanism of rockfill were studied. In the process of compaction, the maximum dry density of rockfill was determined. The results showed that: 1) The additional peak stress of rockfill in the depth range was between 0.3?1.4 MPa. The dynamic peak stress was between 0.22?0.82 MPa, which decayed exponentially along with the increasing of the depth of rockfill. 2) The measured dynamic strain significantly lagged behind the dynamic stress. With the increase of rolling times, the hysteretic cycle became steeper and smaller, while the soil stiffness increased and the damping decreased. 3) Dynamic stress, static stress and deformation were obtained through spectrum analysis and filtering methods, based on which the relationship between the input energy of the vibratory roller and the measured dry density was established. Then, the maximum dry density of rockfill and its relative density under different rolling times were calculated. This method can provide a basis for quality control of rockfill dam using the relative density index.

In this study, a block instability risk assessment method based on catastrophe theory is proposed to tackle the difficulty of block failure risk assessment of underground caverns under complex geological conditions. Firstly, the finite element mesh modeling technology and the method of element reconstruction-aggregation technique are used to identify the rock blocks around the complex caverns. Then the influencing factors of block instability and failure are analyzed, starting from two aspects of internal and external factors of block with 6 elements, including the development trend, spatial location, structural block property, external load, structural surface characteristics and environmental factors. Eleven factors are selected as the underlying evaluation index, on which the mutation model of block instability risk assessment is constructed. After that, the normalized formula is used to seek the solution of the mutation state variables layer by layer, and the membership degree of the obtained total mutation series is converted to improve its clustering and finally the risk state of the block is determined. The analytical results of engineering example show that the method is simple in calculation and high in reliability. It allows comprehensive consideration for the influence of various factors on block stability and avoids the limitations of rigid body limit equilibrium method. The law revealed from the calculation results is basically consistent with the known law, which better reflects the real occurrence state of cavern block. It provides a new way for risk assessment of cavern block under complex geological conditions.

Due to the complexity and obvious random distribution features of joints and fractures in the rock slope, accurately simulating the random fracture network is crucial for slope stability evaluation. However, the current methods cannot effectively make full use of the measured data of structural outcrops. This paper adopts a Bayesian updating approach to optimize the probability distributions of geometric and shear strength parameters of structural planes and correct the random fracture network of rock mass by using field measured data. The field measured data of structural planes are characterized as sample distributions. Based on these, a non-intrusive stochastic finite element method is employed to conduct reliability analysis of jointed rock slopes considering the uncertainties of geometric parameters (e.g., dip and trace length) and shear strength parameters (e.g., friction angle and cohesion) of structural planes at the same time. Finally, a simplified slope model selected from the left bank of Xiaowan hydropower station is adopted to validate the effectiveness of the proposed method. The results indicate that the probability distributions of structural plane parameters inferred from the Bayesian updating approach agree well with the corresponding analytical solutions. The Bayesian updating approach can effectively reduce the estimation of uncertainties and optimize the probability distributions of the structural plane parameters by incorporating the field measured data. Furthermore, more realistic fracture network models and reliability analysis results of the slope can be obtained. When the posterior information of structural plane parameters is used to generate a fracture network model and conduct slope reliability analysis, the estimated posterior probability of slope failure will be greatly smaller than the prior probability of slope failure.

The steeply dipped fault zone is the weak link of the seismic stability of underground caverns. Aimed at the complex dynamic interaction characteristics between surrounding rock and faults, based on the Ladanyi’ shear strength model, a seismic deterioration coefficient was introduced, and a shear strength model considering nonlinear mechanical properties and seismic deterioration effect was established. Given the discontinuous deformation characteristics between surrounding rock and faults, a three-dimensional dynamic contact force method was proposed, which considered both the complex shear strength and multiple contact states. The method was applied to Jinchuan underground powerhouse to study its seismic damage characteristics under the influence of the steeply dipped fault F31. The results indicate that after considering the interfaces and seismic deterioration effect, the seismic response of the cavern increases, the dislocation between the surrounding rock and faults is more evident, and a certain depth of separation and sliding failure zones occurs. The steeply dipped fault cuts the high sidewalls of the main powerhouse, and forms a weak zone where the surrounding rock thinness is thin, resulting in larger deformation and damage. The flexural toppling deformation and slipping deformation into the cavern tend to occur on the upstream and downstream sidewalls, respectively. The distribution of the sliding and separation failure zone of the interfaces between surrounding rock and faults changes dynamically with the seismic process and extends to the depth. Among them, the separation zone is relatively large at the arch abutment and rock anchoring beam. The numerical results reveal the dynamic failure mechanism of surrounding rock of underground cavern intersected by a steeply dipped fault, which can provide reference for seismic design.

Agricultural irrigation in loess platform has led to a continuous rise of the groundwater level and triggered a series of loess landslides. To gain a better understanding of the infiltration behavior of irrigation water in loess, a field infiltration test with a diameter of 20 m was conducted on the South Jingyang tableland, Shaanxi province. The spatiotemporal characteristics of moisture content and matric suction were monitored. Then the infiltration laws under the conditions of soaking and intermittent irrigation were simulated. Results show that the soak infiltration can be divided into three stages, namely, uniform infiltration, preferential flow infiltration, and stable infiltration, respectively. Preferential flow associated with the vertical cracks in the Malan loess was observed when the width of the crack was more than 2 mm and the soil above the crack was saturated. The hydraulic conductivity of the lower part of paleosol layer (S1) was relatively weak compared to the upper part of S1, and a transient perched water was observed above the lower part of S1. The three-dimensional infiltration numerical model was established to reappear the soak infiltration process. Due to the multiple irrigation events, the downward percolation of water was significantly promoted by the superposition effect of wetting front. The infiltration capacity decreased with the increase of depth. The velocity of infiltration below 5.6 m depth was less than the hydraulic conductivity of the soil, the infiltration was driven solely by the gravity, and the moisture content and matric suction were almost unchanged. This finding explains the phenomenon that water infiltration in deep loess is difficult to be monitored in the previous studies.