In this study, reactive magnesia (MgO) is used as a binder to carbonate and stabilise silty clay. The changes of unconfined compressive strength, soil pH, carbonation products, and microstructure of the silty clay carbonated and stabilized by reactive MgO are studied at different carbonation time and initial water contents through the experiments of unconfined compression, pH, X-ray diffraction, mercury intrusion porosimetry and scanning electron microscope, respectively. The micromodel of carbonation reaction of silty clay is proposed according to the intrinsic relationships between the strength of carbonated soil and contents of carbonation products as well as the cumulative pore volume. The results show that with the increase of the carbonation time or the decrease of the initial water content of carbonated samples, the carbonation products increase, and the strength of reactive MgO stabilized soil gradually increases with the decrease of cumulative pore volume. Moreover, the pH values of reactive MgO-stabilized soil decreased with the increase of the carbonation time, whereas the initial water contents show no obvious changes. Finally, the micromodel of carbonation reaction of silty clay is established, and the highest strength of reactive MgO-stabilized silty clay is determined after about 6 hours of carbonation.

Undisturbed Q3 loess is a typical unsaturated soil and has obvious structure which is closely related with its mechanical and deformation properties. The elastic-plastic constitutive model for undisturbed Q3 loess would consider structured evolution in the loading and collapsing process, which would meet the inherent characteristics of the mechanical deformation of undisturbed loess. Assuming that the yielding response of undisturbed loess is the coupling of remolded loess and structure of undisturbed loess, a new elastic-plastic constitutive model of undisturbed Q3 loess considering the influence of suction and structure is established based on meso-structured evolution in this paper. The new elastic-plastic constitutive model has two aspects including soil skeleton deformation and water changes, loading and loading-collapsing process are described, respectively. For soil skeleton deformation, the structured evolution equations obtained by CT triaxial tests of undisturbed Q3 loess in the process of loading and loading-collapsing are introduced into Barcelona unsaturated soil elastic-plastic modified model, respectively. The generalized soil-water characteristic curve is used to describe the aspect of water content considering influence of net mean stress and deviator stress. All of 22 material parameters of the new proposed elastic-plastic constitutive model can be determined by the unsaturated soil tests. The validity of elastic-plastic constitutive model is verified preliminarily by comparison of computation results and test data in this paper. The establishment of elastic-plastic constitutive model deepens the understanding of mechanical characteristics of unsaturated undisturbed Q3 loess, which will provide the gateway to reasonably analyze collapsible deformation coupling problems of loess foundation.

Referring to the original soft clay in the northern slope of South China Sea, the dynamic triaxial tests were carried out under different conditions of consolidation ratio and cyclic stress ratio CSR. The development rule of the residual dynamic strain, residual pore water pressure and the relationship between them were discussed. Considering the correlation between residual strain and residual pore pressure in the test process, the failure standard of dynamic triaxial test based on strain-pore pressure mode was proposed. The mechanism of mutual feed and its interaction mechanism were revealed by the scanning electron microscopy (SEM) technique. The research results indicate that under the condition of designated and different CSR, the development trends of the residual dynamic strain and cyclic vibration frequency curves show a significant difference, while the change laws of the residual dynamic pore pressure and cyclic vibration frequency curve are basically the same. With the increase of , the critical value of CSR gradually increases, while the final residual dynamic strain gradually decreases. The similar characteristics of changing law can also be found from the curve representing the relationship between residual dynamic strain and residual pore pressure. Furthermore, the conventional failure standard based on strain value is extended to the failure region controlled by the inflection point of the strain-pore pressure curve, which can effectively define the failure vibration frequency and completely describe the whole failure process of the specimen, even reveal the inherent mutual feedback mechanism among effective stress, strain and pore pressure. Research results from this study can provide reliable reference for the establishment of dynamic softening model in soft clay, the evaluation and prediction of marine geological hazards, and the foundation design of ocean engineering in the northern slope of South China Sea.

Coral sand is a type of marine biogenic granular material with extremely high calcium content. From the microscopic viewpoint, coral sand grains are characterized as highly angular, irregular in shapes and crushable. One-dimensional creep tests were conducted on coral sand under very high pressure to investigate the evolution of particle size distribution and change of particle shapes caused by particle breakage. Benefited from particle size analysis and shape evaluation with high speed dynamic image analysis apparatus, the evolution of particle shape along with stress was analyzed statistically, and shape factors including aspect ratio, sphericity and convexity increase with increasing pressure. Shape factors for particles with different sizes tend to reach a same value, which illustrated that the morphology of particle after breakage is scale independent and self-similar in this wide range of sizes. This paper analyzed the fractal feature after particle breakage, and found that the fractal dimension increases with increasing vertical stress, and eventually reaches the fractal ultimate breakage (D=2.5). This paper calculated the relative breakage by using Hardin’s and Einav’s methods, and found that relative breakage is increasing along with increasing pressure. The relative breakage is an exponent function with the pressure, which could be used to predict the relative breakage if the material and pressure are known. The increasing tendency with time is not obvious, because particle breakage is majorly caused by compression instead of creep.

As one of ‘the Belt and Road Initiative’ economic development program, the Cross-sea Bridge in Male-Airport island was constructed by China. The foundation of the bridge was located in coral reef which was similar to the geotechnical structure in South China Sea. Due to lack of design experience of rock-socketed pile in coral reef in China and overseas, it is necessary to obtain the load-bearing characteristics and relevant data of rock-socketed pile in coral reef by appropriate test method. Indoor experimental study was conducted on mechanical properties of reef limestone cores that came from pile foundation of the bridge in Male-Airport island. Basic physical and mechanical parameters such as density, relative density, saturate uniaxial compressive strength and triaxial shear strength of the reef limestone were tested. Combining the consequence with the results from model rock-socketed pile test on the reef limestone, the variation law of the bearing capacity with the pile tip displacement was obtained. The investigation suggested that the interface of the pile and rock experienced elastic shear, shear stress drop and fiction shear stage. In the elastic shear stage, the shear deformation was mainly elastic deformation, and the ultimate elastic displacement decreased with the increase of the elastic modulus. In the stage of shear stress drop, the stress rapidly softened, and transformed to the fiction shear stage. In the test, the shaft resistance of reef limestone was positively correlated with the saturate uniaxial compressive strength under low confining pressure. Meanwhile, shaft resistance was greatly influenced by the confining pressure when confining pressure rose.

Stabilizer-treated soil possesses unique structures due to the cementation of soil particles and cement-hydrate. Compared with remolded normally-consolidated soil, treated soil usually holds a strong structure and overconsolidation ratio. Treated soil exhibits softening behavior due to the loss of cementation induced by the decay of bonded structure during the development of deformation. An evolution rule of cementation effect is proposed to consider the change of cementation strength with shear strain. The hardening parameter in UH model is modified and an elastoplastic constitutive model incorporating cementation effect for stabilizer-treated soils is proposed, in which non-associated flow rule is used. The comparison between predicted mechanical behaviors of cement-treated soil and lime-treated soil with triaxial compression test results indicate the validation and accuracy of the proposed model. Consideration of the contribution of cementation effect can describe the mechanical behaviors of stabilizer-treated soils. The soil behaves like overconsolidated soil due to the cementation strength, while the softening is faster with the decrease of cementation strength during deformation.

The soil water interaction in unsaturated soils can be divided into capillarity and adsorption. The current study of unsaturated soils is mostly limited to lower suction range where capillarity dominates. In practice, due to climatic changes, soils near the ground surface often experience cyclic wetting and drying, and the low water content and high suction conditions, where adsorption dominates. Using the vapor equilibrium technique on the saturated salt solution, the high suctions are controlled for non-expansive kaolin-sand soil under drying and wetting paths respectively. The strength and deformation behaviors under high suction are tested under four confining pressures of 0 (close to unconfined), 25, 50, 100 kPa. The results show that the soils under high suctions indicate strain softening failure and dilatancy. With confining pressure applied, the failure mode change from the longitudinal cracking to shear failure. It is proved that the modified Bishop's effective stress model is not suitable for describing the peak shear strength at high suction. The peak shear strength can be expressed by the net stress-dilatancy relationship. The peak shear strengths are directly related to specific surface area and depend on the development of aggregates and dilatation.

Through in-situ liquefaction tests under dynamic artificial loading, the pore pressure variation of saturated sand in level ground is measured, and an incremental model of predicting pore pressure is proposed. Using the results of in-situ testing in different cases, a new incremental model is established considering acceleration, buried depth, density of sand, common properties used to describe soil characteristics in-situ. The correlation and difference of response for pore pressure increase between in-situ test and laboratory test are also analyzed. Parameter analysis and validation through in-situ liquefaction tests indicate that the new model is reliable. The number-of-cycle effect by in-situ tests is different from the results obtained from laboratory tests. The main conclusions are as follows: (1) Under uniform cyclic loading conditions, the variation of pore pressure increment in saturated sand in field is not monotonically decreasing with increasing numbers of cycles, but increasing at the beginning followed by decreasing after a specific number of cycles, which can be defined as a threshold number; (2) The proposed incremental model can also be used to predict pore pressure buildup of saturated sand under irregular loadings.

In order to study the changeable characteristics of the hydrothermal field and the salt-expansion deformation laws of salty soil under natural climatic conditions in the northwest arid area of China, vertical apparatuses for monitoring deformation, and sensors for monitoring moisture content and temperature were buried in a test pit of 4.5 meters deep. With the seasonal changes, the different depths of soil temperature field, moisture and salt-expansion deformation of the pit were dynamically monitored for a year and analyzed. The results show that the response to climate change of soil layer within 0.6 meter from the ground is more positive than other soil layers, and the change of temperature is greater. The difference of temperature among the soil layers increases with the proceeding of cooling period. The moisture content of soil is mainly influenced by precipitation, evaporation and temperature gradient; the change of soil moisture above 0.4 meter is more significant than other soil layers, and the water migration along the altitude of the soil shows a zonal phenomenon. Salt-expansion is mainly affected by the temperature and the moisture transfer of soil layers; salt-expansion deformation mainly occurs in the depth of 1.0 meter soil layer from the surface of the earth, which mainly develops in the period from November to February next year.

Frost heaven force, related with tunnel axial length and time, has three dimensional space-time characteristics in cold region tunnels. In order to establish a frost heave model for studying the change rule of frost heaven force, three dimensional temperature field model of cold region tunnels was firstly established based on the field test results, and the changing law of surrounding rock frozen depth was obtained by Stephen formula. On this basis, a new temperature-frost heave model was established in combination with freeze-thaw circle theory and weathering layer theory. The results show that the frozen depth of surrounding rock influenced by environment temperature has a trend of decrease with the increase of tunnel axial length and changes periodically with time. According to this new model, when the frozen depth is smaller than the weathering layer depth, the frost heaven force is only induced by the weathering layer; when the frozen depth is bigger than the weathering layer depth, the frost heaven force in induced by both the weathering layer and freeze-thaw circle. The change rule of frost heaven force is similar with the change rule of surrounding rock frozen depth.

Extremely alternating extreme climatic events will aggravate weathering of rock, reduce rock strength, and affect the stability of geotechnical engineering structures. In this paper, the mechanical properties and microstructural changes of the weak interlayer were studied at room temperature and under dry and wet cycles (air drying at 60 ℃) by direct shear tests and scanning electron microscope. Significant variables of microstructure influencing the intensity were obtained through the correlation analysis and stepwise regression. Moreover, a regression equation was constructed to reflect the correlation between them. The results showed that the shear strength and cohesion of the weak sandwich in the red sandstone decreased drastically from the first to the fifth cycle, whereas the change was slower after five times. The cycle times had little influence on the change of the internal friction angle. As the number of dry and wet cycles increased, four microstructure parameters were decreased, including the average area, average diameter, average perimeter circumference and orientation probabilistic entropy, which showed that the positive correlations with the cohesion. However, three microscopic parameters, i.e., the number of particles, the average shape coefficient and the fractal dimension of the morphological distribution showed negative correlations with the cohesion. The results indicate that there are significant correlations between macroscopic mechanical properties and changes of microstructural parameters during wet and dry cycles. The weakening of weak sandwich is caused by the swelling and shrinkage, fragmentation and structural failure of red sandstone due to wet and dry cycles.

To study the fracture characteristic behaviour of the defective cemented backfill, three-point bending tests were carried out on the cemented backfill specimens with the offset ratios of 0, 0.25, 0.50, 0.75 and with the seam height ratios of 0.10, 0.25, 0.50, respectively. A high-speed camera was employed to capture the full-range crack propagation, and two-dimensional particle flow software PFC2D was used to analyse the crack propagation process, the fracture mode and the fracture mechanism of the cemented backfill. The results showed that the fracture peak load increased with the increase of the crack offset ratio at the same seam height ratio. When the offset ratio was constant, the peak load was smaller with the increase of the seam height ratio. When the crack offset ratio was at 0, 0.25 and 0.50, the crack propagated from the offset, and the deflection angle increased with the increase of the offset ratio. When the crack offset ratio was at 0.75, the crack propagated from the centre. The crack propagation process was divided into three stages, and it was serrated-shaped. The fragmented particles were generated and shed during the crack propagation process. Two-dimensional particle flow model was employed to simulate the force chain network, velocity field and fracture mode of specimens. The obtained results were compared with experimental results of macro mechanics to explore the mesoscopic fracture mechanism. The peak loading of the fracture was not more than 3.8% compared with the experimental value.

To analyze the settlement and deformation characteristics and the failure modes of embankments on soft ground reinforced by plain concrete piles, three centrifugal test models of the embankments on soft ground reinforced by plain concrete piles with different spaces between piles and their numerical models have been performed. The results show that changing pile spacing has a remarkable effect on the settlement and deformation, the pile shaft strain, the failure modes of piles and reinforced cushion of the models, under embankment self-weight load, and track structure and vehicle load. When the pile spacing is not greater than 4 times of the pile diameter, the reinforced cushion can basically maintain integrity, the settlement of the models gradually tends to stabilize. Once the pile spacing increases to 6 times of the pile diameter, the reinforced cushion is impaled, the function of deformation coordination and load transfer would not be available, resulting in a continual increase of the settlement of the model embankments. While the pile spacing reaches 4 times of the pile diameter, an abrupt decrease in the maximal strain value of the piles occurs with increasing upper load, and the plain concrete piles near the embankment toe fail firstly in bending failure mode rather than shear failure mode. The plain concrete piles closer to the embankment center fail gradually in bending failure mode as the pile spacing increases to 6 times of the pile diameter.

This study aims to solve the problems on the nonlinear behaviour of rock mass or mortar against the pressure side of the bolt and the dilatancy effect of the structural surface. Theoretical formulas for the axial force and axial deformation, transverse shear force and transverse deformation of anchor bolts were derived, and the formula for the shear resistance of anchor bolts was also established based on the classical beam theory. The validity of the theoretical solutions was verified by the direct shear test on bolted structures. Furthermore, this study performed the effects of joint dilatancy coefficient, joint friction angle, surrounding rock strength and bolt installation angle on the deformation and shear resistance of anchor rods. The results showed that the theoretical calculation results were in good agreement with the experimental results. The higher dilatancy coefficient led to the higher shear resistance of the bolt, but less significant of plastic behaviour. Thus, the shear resistance of bolted joint can be improved by the inherent shear strength of the structural surface. When the strength of surrounding rock decreased, the anchor bolt had to undergo specific deformation to exert a large shearing effect. The higher strength of rock resulted in the yield failure of the bolt within a shorter time, and the damage of the anchor bolt was gradually transformed from the axial tensile failure into tensile shear failure. The optimal installation angle of the bolt increased with the increase of the internal friction angle of the structure. The optimal installation angle of the bolt was estimated within the angle ranging from 30° to 68° in working joint conditions.

The resilient modulus tests of polymer rockfill materials were performed using a middle-scale triaxial instrument. Mechanical properties of polymer rockfill materials and their evolution laws of resilient modulus were analysed under unloading and reloading conditions. The results showed that the volume-contraction was influenced by the confining pressure and stress level significantly. The specimen gradually changed from the initial state of shear dilation to shear contraction with the increase of confining pressure. Under high confining pressure, the unloaded volume-contraction increased nonlinearly with the increase of the stress level. At the confining pressure of 100 kPa, the specimen presented a volume-dilation during unloading, and the values of unloaded volume-contraction were similar and smaller at different stress levels. The ratio of the average resilient modulus to initial Young’s modulus was approximately between 3.9 and 4.2 under different confining pressures. The corresponding resilient modulus at the stress level of 0.7 was similar to the average resilient modulus. The Duncan-Chang model was applied to simulate the resilient modulus of the polymer rockfill materials, and the resilient modulus coefficient was approximately 4.0 to 4.2 times of the initial Young’s modulus coefficient in the numerical calculation process.

This paper presents a model developed for one-dimensional consolidation of unsaturated soil. This model uses the piece-linear finite difference approach and considers soil nonlinearity and large strain problems. Fortran programming package is used to implement the computation of the model. This model is verified by analytical solutions and experimental results. The numerical solution of this model agrees with the existing solutions in the case of instantaneous loading. In process of loading, the approximation of consolidation settlement obtained by this model is in good agreement with the test results. Then this model is implemented for a large strain consolidation. The case study compares the settlement occurred at the instantaneous deforming stage and the subsequent consolidation stage. The case study also investigates the effect of the pore air permeability to the hydraulic conductivity ratio on soil layer settlement, saturation degree, and consolidation degree of different strain. It also analyzes the difference of pore water (air) pressures and soil layer settlement on large and small deformation consolidation theory of unsaturated soil.

The movement of wall plays an important role in the calculation of lateral earth pressure and the design of retaining structure. Regarding the process of backfill approach active or passive pressure state as the shearing process of soil sample in simple shear test or direct shear test, the backfill process reaches active or passive earth pressure state when the soil deformation equals to the ultimate value (shear strain in the simple test or shear displacement per unit length in the direct shear test). Based on the geometric relationship between the soil shear deformation and wall displacement, the theoretical calculation method of wall displacement required to reach active or passive earth pressure is provided, where the soil stress-strain behavior and initial stress state are considered. The analysis indicates that the magnitude of needed wall displacement to reach active or passive earth pressure is controlled by soil ultimate shear deformation, the range of active or passive zone, and the initial earth pressure state. The first factor is the most important among them, which contributes to variation of wall displacement among different soils. The wall displacement in passive state is greater than that in active state, because the range of passive zone is larger than that of active zone. The theoretically calculated wall displacement attaining active state is about 0.5‰~13.2‰ H (where H is the height of the wall), of which the non-cohesive soil is larger than the cohesive soil. As to the case of passive state, the wall displacement is ?0.4%~?5.2% H. The theories are concordant with the model test results from relevant literatures.

The vector sum method (VSM) based on real stress state has been applied recently for analysing the slope stability, considering its clear physical meaning. However, the factor of safety (FoS) defined by VSM could be negative in the case where total normal force vector projection is negative, and its absolute value is larger than total shear force vector projection. Moreover, deduced by Pan’s Principle, the direction of shear resistance stress vector of every point on potential slip surface is currently defined to be opposite to that of the projection of the potential slip direction of the possible slip mass on the tangential surface through this point. However, according to the definition of VSM safety factor and Pan’s Principle, the direction of the projection of the potential slip direction on the tangential surface through this point is opposite to that of the total resistance stresses including shear and normal ones. In this study, a vector-sum-based slip surface stress method for analysing slip mass stability was proposed to overcome these two disadvantages above of VSM. The slip surface was treated as a thin slip band, and forces undertaken by an arbitrary element of the slip band was analysed. The direction of the limit resisting shear stress of an arbitrary element was defined as the opposite direction of the real shear stress, and the possible slip direction was defined as the opposite direction of the resultant limit resisting shear force. Finally, FoS was defined as the ratio of the projection of the limit resisting shear force vector sum in the opposite possible slip direction to that of real dynamic shear force vector sum in the possible slip direction. Compared with the limit equilibrium method, the proposed method satisfied the force and moment equilibrium equations and compatibility conditions without any assumption. Compared with the shear strength reduction method, the proposed method defined that the FoS was explicit form avoiding iteration. The classic two-dimension static slope case was used to verify the proposed method with comparisons among main methods for slope stability analysis, which indicated that the proposed method was feasible. Additionally, the proposed method was applied to a three-dimensional engineering case to study its stability during the excavation process, which showed the proposed method could be used for practical application.

To simulate the failure of loess under undrained condition in the actual engineering, a series of isotropic consolidation and shear tests with different intermediate principal stress ratios b under constant water content is performed on intact loess with various initial suctions by using unsaturated soil true tri-axial apparatus. The mechanical characteristics of unsaturated intact loess are studied. The results show that the isotropic compression yield stress increases with initial suction, suction decreases with the increase of net mean stress, the extent of suction impact of net mean stress increases with increasing of the initial suction. The damage state lines of generalized shear and net average stress are parallel to the linear relationship and the test points of generalized shear and effective net mean spherical stress can be normalized to drainage shear failure state line of saturated soil under different the ratios b. The suction of shear failure increases linearly with the ratio b; void ratio is approximately linear parallel to logarithm of net average stress critical state lines, whose slopes are larger than those of saturated loess and larger than those of isotropic compression before yielding; both void ratio and logarithm of effective net mean spherical stress critical state lines of unsaturated soil are approximately linear parallel to the critical state line of saturated soil. The test points that ratios of damage state void ratios between unsaturated loess and saturated loess and gas saturations are distributed in a linear zone of 1-1.2 under different net confining pressures, but those points can be normalized to a nonlinear curve at the same net effective stress.

In this study, the model test with the geometric scale of 1:20 was performed to study the behaviour of the slope reinforced by prestressed anchor cable under earthquake conditions. Inputting seismic waves and excitation amplitudes were adjusted to investigate the acceleration and displacement responses of the reinforced slope and the dynamic force characteristics of the retaining structure according to the similitude law. The results showed that the rock at the top of the slope was shattered, and denudation and rockfall at the lower part of frame beam occurred, but the overall stability of the anchored slope remained well. During the propagation process of seismic wave, there was an energy accumulation effect in the weak interlayer. With the increase of the amplitude of inputting seismic wave, the diversity of centrobaric acceleration amplification coefficient decreased under different earthquake waves. The mechanism of pre-stressed anchor cables showed the characteristics of collaboratively at different altitudes. The prestress loss occurred at the top first and then gradually transited to the lower-middle part. In the seismic design of the anchor cable frame system, it is significant to consider the acceleration amplification effect of the upper part of the slope and the prestress loss of the cable for the overall stability of the slope. The experimental results provide useful references for the seismic design of the prestressed anchor cable frame system.

The vertical vibration of a single pipe pile in saturated soil is studied with 3D wave model. The soil around pile and in the pile core are regarded as two-phase porous medium, and the pipe pile is regarded as uniform circular tube unit. A dynamic model of saturated soil-pipe pile with coupled vertical vibration is developed with soil 3D wave effect model. Considering the boundary conditions of soil, the vertical vibrations of the saturated soil around the pile and inside pile core are solved by using potential functions and method of separated variables. The vertical vibration of a single pipe in saturated soil is solved and the vertical complex stiffness at pile head is obtained by using the orthogonality of trigonometric functions considering the boundary conditions at pile end. The results of soil 3D wave effect model and simplified model neglecting the radial displacement of soil are compared and analyzed by numerical examples. The influences of the main parameters of pile and soil on the vertical vibration of pipe pile in saturated soil are investigated. The results indicate that the radial displacement shouldn’t be neglected when the pipe pile wall is thin and frequency is low. The liquid phase of the soil shouldn’t be ignored at the peaks and valleys of the curves of dynamic stiffness factor and the equivalent damping varying with frequency. The pipe pile wall should not be too thin. The thickness of pipe pile wall, the ratio of length to diameter, density ratio and shear modulus ratio between the pile core saturated soil and saturated soil around pile, pile-soil modulus ratio have great effect on the vertical vibration of a single pipe pile in saturated soil. The relevant parameters should be considered synthetically in the design of pipe pile.

To investigate the deformation behavior of loess in the typical “Y” shape long and narrow gully landforms, a series of one-dimensional consolidation tests are carried out indoor under high-pressure conditions. Test results show that the Gunary Model is better than the hyperbolic model or the exponential function to represent the relation between vertical compressive stress and vertical compressive strain of compacted Malan loess. Using secant modulus, the modified formulas of deformation of foundation with compacted Malan loess are constructed to calculate the deformation considering the variation of loading, water content and compactness. The relation of secant modulus and the reciprocal of water content is in the form of quadratic polynomial. The widely used linear relation formula of secant modulus and compactness K is not consistent with the actual situation. Research results have an important reference value to the design and construction of gully type high loess-content landfill projects.

In this study, the uniaxial compression test and Brazilian splitting test were carried out on Silurian silty slates in northwest Hubei province to investigate their mechanical and anisotropic properties. The anisotropic mechanical properties and deformation failure modes of specimens were analysed under different loading conditions, and the corresponding mechanisms of different failure modes were revealed. Meanwhile, the numerical analysis was used to study failure modes and mechanical mechanisms of slate slopes with different angles of bedding planes. The results showed that bedding planes in silty slates were weak surfaces that affected mechanical properties of rock mass, resulting in obvious anisotropic characteristics of the silty slates. Under the uniaxial compression condition, the deformation of silty slates were easier in the direction of the vertical bedding plane than that of the parallel plane, and their deformation was more significant. The failure mode of the specimen was the vertical splitting tensile failure when the direction of the parallel plane was loaded, while the failure mode was the splitting shear failure cutting the bedding planes when the direction of the vertical plane was loaded. The anisotropy of measured mechanical parameters was relatively apparent. Under the action of splitting load, the failure modes of silty slates mainly included the tensile splitting failure and shear failure along the bedding plane. The obtained tensile strength was the largest in parallel bedding plane direction and was the smallest in the vertical bedding plane direction. The tensile strengths in both directions were lower than the compressive strengths. As the tensile strength between bedding planes was extremely low, tension-split failure or tension-shear failure was easily caused by the splitting load that intersected with the bedding plane at a small angle. Hence, the tension failure or tension-shear failure along bedding planes should be avoided as far as possible in practical engineering. The dip directions and angles of slates have great influences on failure modes and mechanics of slates slopes, which should be considered in the protection and treatment of rock slopes. Therefore, the results provide significant references for the protection and treatment of rock slopes, design and construction of other rock engineering projects in the distribution area of silty slates.

To accurately describe the mechanical behavior of stress relaxation of expansive soils, the software element is constructed by applying the theory of fractional calculus to replace the ideal viscous element in the traditional element model. The damage degradation of viscosity coefficient is considered. A new three-element model is established to better reflect the nonlinear stress relaxation behaviors of Nanning expansive soils. The stress relaxation of Nanning expansive soil under different water contents is studied by a series of indoor tests. The relevant model parameters and stress relaxation equations are determined according to the analysis of the tested results. By comparing the proposed model with the traditional integer order models (such as the traditional St.Venant model and Nishihara model), the results show that the proposed model can more effectively simulate the whole process of the stress relaxation of the expansive soils. Furthermore, the proposed model has advantages on simple structure, few calculation parameters, high accuracy, easy to practical application. The research results provide a theoretical basis for the analysis of the stress relaxation of expansive soils under long term load.

The dynamic stress-strain and dynamic elastic modulus-strain of cemented soil mixed with different amount of polypropylene fiber or the basalt fiber were studied by using the dynamic triaxial apparatus. The results show that the dynamic strength and dynamic elastic modulus of cement soil is in connection with confining pressure, fiber types and fiber contents. With the increase of fiber contents, the dynamic strength and dynamic modulus of cemented soil increase, and the dynamic deformation decreases. The hysteresis curve and the microstructure of cemented soil show that the cement soil containing basalt fiber presents the best dynamic performance . The maximum dynamic elastic modulus increases with the increase of confining pressure and fiber content. The normalized curves of cemented soil mixed with basalt fiber under different confining pressures gather with each other, so the dependency on the confining pressure is weakened.

The combination of cutoff wall and composite geomembrane is usually used as the imperious system recent years. The uncoordinated deformation between the cutoff wall and the fillings will cause the composite geomembrane cracks. The connecting form between cutoff wall and composite geomembrane is studied through geotechnical centrifuge tests. Four model tests considering the cutoff wall, wall mud, the backfilled materials and the connecting form are performed to investigate the variation of the strain of the composite membrane. When the geomembrane is laid horizontally on the top level of the cutoff wall, it is seriously strengthened at the connection near the cutoff wall, and the geomembrane is broken down when the differential settlement increases to a certain value. When the geomembrane is first laid vertically above the cutoff wall and then horizontally extended outward, the tensile strain of geomembrane is less at the connection. The longer the distance to the connection point, the less the strain. The existence of wall mud can effectively alleviate the local strain of the geomembrane.

A new type of similar material for the surrounding rock-support system was developed to control the permeability coefficient, based on the Shenzhen East Transit Expressway Link Project adjacent to Shenzhen reservoir. A self-made seepage model test device was used to analyse the effects of dynamic and hydrostatic heads with different heights on water pressure, water discharge and seepage field. The results showed that when the groundwater level was relatively high, the inner walls of grouting ring underwent several procedures such as heavy wetting, the occurrence of saturation line, thoroughly wetting, overhanging water drops above the arch and intermittent water flow. With the decrease of the water head, the seepage velocity slowed down and the seepage time increased greatly. The water pressure at each feature point reduced approximately linearly with decreasing the height of the hydrostatic head. Besides, when the hydrostatic head became lower, the water pressure values at the same position of secondary lining and grouting ring were closer, and the dispersion of tunnel water discharge test was also smaller. Under the action of the dynamic head, the water pressure presented an obvious time lag, and the time effect was not conducive to the stability of the structure. The shallow buried areas were more easily affected by the decrease of the dynamic head, which indicated that it was necessary to strengthen the capacity and anti-seepage design of the tunnel roof. The water discharge of the tunnel exhibited a decreased tendency with the decrease of the dynamic head, and it took a longer time to reduce the same water-head difference.

Considering that the measured interface properties between soils and geosynthetics are prone to being influenced by experimental apparatus, authors applied large-scale laminated shear apparatus to conduct the shear tests on interface between geotextile and sand instead of common direct shear apparatus. The comparative analysis of the results from laminated shear tests shows that the displacements of laminated rings of sand with geotextile are smaller than those in shear tests of sand. The horizontal displacement of each ring is closely related to the dilatancy of sand. The geotextile limits the movement of soil particles in lower shear box and reduces the dilatancy ratio of sand at the peak interface shear strength. It can be inferred from the experimental results that the shear band caused by the interface shear between geotextile and soil is much narrower than its influencing range. The effect of geotextile in the interface shear test cannot be reflected only by the size of shear band, but also by the shielding function of geotextile from shear and its influence on the development of shear band in soil.

Theoretical and field experiments were carried out to reveal the relationship between the hydraulic fracturing method and the Brazilian splitting method for measuring the tensile strength of rock. Based on the classical hydraulic fracturing theory, borehole rupture pressure and tensile strength of rock were deduced under different confining pressures. The fracture mechanics theory was applied to establish the relationship between these two methods. Meanwhile, the crack initiation pressure of the hydraulic fracture was studied by using the prefabricated crack method to simulate natural crack. The results show that when the maximum principal stress was three times the minimum principal stress, its measured tensile strength was higher than that without confining pressure. The relationship between these two methods was related to three variables, including stress field, crack length and fracture toughness. The prefabricated crack experiment was performed at Wangtaipu Mine in Jincheng mine area. It was proved that the tensile strength formula established by fracture mechanics was more suitable for practical applications. The research results can provide references for the hydraulic fracturing design of prefabricated cracks in the hard and stable roof.

In the process of CO2 geological storage, the mechanical stability of rock can be influenced by changes in the composition and pressure of pore fluid due to CO2 injection into deep rock stratum. In addition, adverse consequences may further arise, such as reservoir caprock rupture, surface uplifting, and small and medium scale earthquakes. The triaxial compression tests of rock were conducted to compare the effects of pore fluid media of supercritical CO2, N2 and H2O under single-phase fluid. The fluid temperature and pore fluid pressure were controlled and adjusted to analyse the strength, elastic modulus and Poisson's ratio under single-phase fluid coupling. It was found that the peak strength and elastic modulus of dry sandstone were reduced at various degrees by high-pressure pore fluid. However, Poisson’s ratio increased obviously, and the impacts of pore fluid were H2O, CO2 and N2 in the descending order. The pore fluid decreased the brittleness of dried yellow sandstone and enhanced plastic deformation to a certain degree. Besides, the water-bearing sandstone had the strongest plastic property. The effect of pore fluid on mechanical properties of sandstone depended on the strength of interaction between the fluid and mineral composition of rock. The selective absorption of CO2, N2 and H2O by different mineral compositions of sandstone resulted in a significant difference in the effect of pore fluid on sandstone strength. The action effects ranked as H2O, CO2 and N2 in the ascending order.

As the earliest concrete in the world, the modified ginger nut has been proved to be proper restoration material for restoring Chinese ancient buildings and cultural relics. In order to study mechanical properties of modified ginger nut, we carry out conventional triaxial compression tests on modified ginger nut. During triaxial test, the failure surfaces are shear failure under low confining pressure (2 MPa). When the confining pressure increases to 6 MPa, there are no obvious failure surfaces but only volume expansion. Mohr circle envelopes are drawn to calculate shear strength parameters. The triaxial test results indicate that the peak strength, peak strain and elastic modulus linearly increased with the increase of confining pressure. The test results indicate that the modified ginger nut has strong deformation capacity as well as mechanical strength. Based on XRD diffraction results and SEM images, we find that quartz particles are covered by hydration and carbonation product and this special structure may be the likely reason to explain why the mechanical strength and deformation capacity are high. The presented results are also expected to provide useful references to the restoration of cultural relics engineering.

The geometry is one of the key factors which can influence the mechanical and hydrological properties of coarse granular soil. In this article, the laboratory test and analysis method are adopted to study the particle appearance characteristic. Two types of coarse granular soil particles with different appearance characteristics are selected in the particle diameters of 10－20 mm, 20－40 mm, 40－60 mm separately. The wax-coated method is used to measure the volume, and the particle surface area is calculated by measuring the thickness of the wax coating. By building the mathematical fitting between the particle volume and the surface area, the correlations between volume and surface area of the two coarse granular soil particles are founded. With analogy, the two fitting equations proposed in this paper are found to be in excellent agreement with that of normative sphere, thus the concept of sphericity-similarity is raised and the coefficient of sphericity-similarity is calculated, which providing a method to describe the appearance characteristic of particle quantitatively.

This paper studied the seismic stability of the 1# surge chamber in the Baihetan hydropower plant under the influence of interlayer shear weakness zone (ISWZ) C2 regarded as the large dominating geological discontinuity. Based on the dependent behaviour of normal stress, a nonlinear continuous yielding (CY) model was adopted to describe the complex mechanical properties of the ISWZ C2 under static and seismic dynamics. In this model, the deformation characteristics of the discontinuity surface were expressed in terms of a power function, and the progressive destruction of the strength during shear failure was also considered. Besides, 3DEC software was applied to verify the CY model. Then the applicability of the CY model was proved by comparing experimental results with theoretical solutions. Three ground motion waveforms were utilised to conduct the seismic analysis of the #1 surge chamber after the special response spectrum matching process. The seismic analysis confirmed the control effect of ISWZ C2 on the seismic stability of the cavern. The seismic displacement of the cavern was mainly the elastic body movement and was supplemented by the plastic deformation. Furthermore, most of the deformations were caused by the contact deformation of C2. For the contact deformation of C2, the magnitude of permanent shear deformation was larger than that of the normal deformation. The magnitude of permanent shear deformation was more obvious along the strike direction of C2, and the permanent normal displacement of C2 mainly occurred along the dip direction of C2. Finally, the seismic stability of the cavern was determined by the overload method. The measured seismic safety factor of the cavern was approximately 2~3. The findings in this study may provide helpful references for the seismic design of the underground caverns.

The large diameter and super-long pile is the development trend of the pile foundation engineering of high standard buildings. But there is few research about its horizontal bearing characteristics. In this paper, the horizontal deformation and stress distribution of the large diameter and super-long pile were obtained by pile test in field of Jingjiang cultural center construction project. The piezocone penetration test (CPTU) p-y curve method was used to construct the numerical model of the large diameter and super-long pile under horizontal loading. After comparison and verification with the results of pile test, the numerical model was used to further study the influence of size effect, embedding types, inclination of pile and vertical loads on the horizontal bearing performance. Finally, the quantitative evaluation about the parameter sensitivity of the large diameter and super-long pile was achieved by calculation of different sensitivity values. The research results show that in-situ CPTU test is a very good application in analysis of horizontal bearing capacity of the large diameter and super-long pile, and the order of sensitivity values is: embedding type, size effect, vertical load, inclination.

In terms of support form for foundation pit, the most common calculation method of internal force and displacement of retaining piles (or enclosure wall) plus internal bracings is incremental method, which is based on the theory of beam on elastic foundation. Reasonable calculation of load increment is the key to implement this method. At present, the load increment of this method mainly consists of earth pressure increment and reaction increment released by the excavated soil elastic resistance. Based on the basic principle of the method of beam on elastic foundation and incremental method, combined with excavation construction process, the following conclusions were drawn through the comprehensive analysis. In addition to the above two parts, the load increment also included reaction increment by soil elastic resistance released, which was partially caused by the decrease of the excavation face and the reduction of the horizontal resistance coefficient. The reaction increment of random point of the retaining piles (or enclosure wall) in this part that below the excavation face was equal to the product of the lateral displacement of this point in the previous working condition and the proportional coefficient of the soil horizontal resistance coefficient (m) , and the excavated soil thickness in the current working condition. The comparative analysis of theoretical calculation results and measured results of engineering examples show that: compared with the traditional incremental method, utilizing the load increment mentioned in this paper to do load increment calculation (the improved incremental method) yields a value of lateral displacement of retaining piles (or enclosure wall) that is much closer to the measured value. The calculation result is more in line with the engineering practice.

The upper bound numerical method for reliability analysis of soil slopes is proposed, by combining the theory of upper bound limit analysis, finite element discretization, stochastic programming theory and Monte Carlo method. Firstly, three-node finite element is used to discrete the soil slope, and then shear strength parameters of soil are set as random variable. According to the upper bound theorem, kinematically admissible velocity fields are established to satisfy the plastic flow constraint conditions of triangular finite elements, the plastic flow constraint conditions of velocity discontinuities and the velocity boundary conditions. The objective function is established based on the internal and the energy-work balance equation. Then the upper bound limit method stochastic programming model of soil slope reliability analysis are established. The Monte Carlo method is used to solve the upper bound stochastic programming model. At the same time, an estimation method for coefficient of slope failure risk based on velocity field of upper bound method is proposed, it is especially suitable for slope risk analysis with multiple failure modes. Finally, two examples of soil slope are selected, and the results prove the correctness of the method.

Desaturation is a method to improve the liquefaction resistance of ground by decreasing the degree of saturation of saturated sandy soils. A simplified single-phase fluid numerical method considering water-air flow in desaturated sand as a saturated pore fluid, a modified single-phase fluid numerical method is proposed to investigate the static liquefaction behavior considering the change of pore fluid bulk modulus. The numerical simulations of the undrained triaxial test are carried out in this study. By comparing the numerical results with laboratory tests, it indicates that the modified method can predict static liquefaction performance of saturated sand accurately. Moreover, some conclusions are obtained from numerical results of different methods, namely, simplified single-phase fluid method, modified single-phase fluid method, and two-phase flow method. First, the initial value of pore fluid bulk modulus will have a significant impact on the prediction of the undrained behavior of desaturated sand in simplified single-phase fluid numerical method due to the constant pore fluid bulk modulus. When the initial value of pore fluid bulk modulus increases from 100 kPa to 200 kPa, the peak deviator stress can decrease by 30% and the pore pressure can increase by 40%. Second, the stress-strain response and volumetric strain curve of modified method agree with that of two-phase flow method and the error is less than 5%. Overall, the modified numerical method is a simple and reasonable method for describing the static liquefaction of desaturated sand.

Since many earth dams were built and are to be built on overburden, it is essential to describe the dynamic interaction behaviour between earth dams and the overburden reasonably for the seismic safety evaluation of dams. The seismic wave input method is based on the equivalent load and artificial boundary. This method can well reflect the dynamic soil-structure interaction, and thus it has been widely used in dam engineering and underground engineering. Moreover, this method is easy to be realised in the homogeneous linear elastic foundation but difficult in the overburden foundation with the soil nonlinearity. In this study, a simplified finite element model for the free field is established according to the deformation mode of the free field at the lateral boundary. This simplified model can conveniently and precisely obtain the nonlinear dynamic response of homogeneous or layered overburden under the action of normal incidence of the multi-direction earthquake. Then, a seismic wave input method is developed for nonlinear dynamic analysis of earth dams built on overburden. Besides, the established method is combined with the dynamic real-time nonlinear artificial boundary of ground material parameters. The numerical examples show that this method can reduce the amount of calculation grid remarkably, meanwhile can reflect the effect of overburden on the spectrum of earthquake response with high accuracy.

With the purpose of solving the problem of model sharing among stages of shield tunnel construction, an IFC-based data model of shield tunnel was built in previous research. Based on the IFC-based data model, a BIM-based shield tunnel structure information modelling method is introduced in this paper. According to the extended IFC description of shield tunnel information model, the modelling method of single segment and flow and process of tunnel axis are proposed. Based on this, a shield tunnel assembling method is built, as well as a flow chart and a parametric modelling method of shield tunnel. By this way, the BIM-based shield tunnel structure modelling method is formed. The modelling method is examined by applying it to a metro shield tunnel case. The research shows that, by introducing the concept and technology of BIM, it is suitable for the communication and sharing of the information of shield tunnel structure that builds the shield tunnel structure information model based on the uniform IFC standard. The validity of the IFC-based data model of shield tunnel is verified further. The result in this paper can provide the initial model for numerical calculation and analysis, and provide foundation for realizing seamless connection of visualization model and analytical model.

Layered distribution of natural soft soils and the threshold gradient of the flow in soils have been gradually recognized. However, with consideration of the threshold hydraulic gradient and the nonlinear compressibility and permeability, the theory of large-strain consolidation of double-layered soil has rarely been reported in the literature so far. In this study, a model for large-strain nonlinear consolidation of double-layered soil is developed in Lagrangian coordinate by employing the excess pore water pressure as a variable. Meanwhile the corresponding finite difference solution for this model are provided. The numerical solution for the proposed model is verified by comparing with the numerical solution of large-strain nonlinear consolidation of a single-layered soil with threshold gradient. Finally, the influences of dimensionless parameters R1, R2 on consolidation behaviors are analyzed. The differences of the dissipation of excess pore water pressure, and the differences of the consolidation settlement under large- and small-strain conditions are investigated. The results show that the influences of R1 on consolidation behavior of double-layered soil are more evident than that of R2; the moving speed of seepage flow front of double-layer soiled under large-strain assumption is faster than that under small-strain assumption; the dissipation rate of excess pore pressure under large-strain assumption is faster than that under small-strain assumption, and the settlements of double-layered soil with threshold gradient under large-strain assumption is larger than that under small-strain assumption.

In this paper, the extended precise integration method is used to solve the plane strain problem of transversely isotropic multilayered soils. Extended precise integration method is of high accuracy and efficiency, which is an effective method to solve differential equations. Compared with analytical methods, it can save a lot of time for theoretical derivation. Starting with the governing equations of elasticity in Cartesian coordinates, the matrix differential equation in the Fourier transform domain is derived. Then, the establishment and combination of adjacent layer elements are introduced, and the extended precise integration solution of multilayered soils subjected to internal loads can be obtained further. The comparison with existing results verifies the accuracy and correctness of this method, and numerical examples are presented to elucidate the influence of transverse isotropy, stratification and load position. It turns out that the vertical displacement decreases with the increase of modulus ratio n, and increases with the increase of modulus ratio m. Besides, soils above the loading point is more sensitive to change of load position, while the modulus of upper soils is more important to the results of vertical displacement, and stratification of the soils has more significant influence on vertical displacement compared with vertical stress.

A mobile jack-up rig is an important facility in offshore oil and gas development and often works in close proximity to a permanent jacket platform. Penetration of the spudcan could displace a large volume of soil and induce considerable loadings to the pile foundation of a nearby jacket, which may jeopardize the designated performance of the jacket platform. The aforementioned effect is more significant in cohesive soils where deeper penetration is expected. However, the currently available industry guideline and approaches have obvious limitations on this important issue. Therefore, a numerical approach is developed in this paper which has two steps to study the effects of the spudcan penetration on an adjacent jacket pile in cohesive soils: (1) computation of the lateral soil displacement induced by the spudcan penetration in a free field assuming the jacket is absent through the arbitrary Lagrangian Eulerian (ALE) numerical method, and (2) computation of the pile lateral loading induced by the aforementioned soil displacement based on the P-Y curves by the conventional beam-column method. Feasibility and reasonability of the approach are verified by calculated results against centrifugal experiment results published in the past. The paper reveals the deformation mode and flow mechanism of the soil in the free field, indicates the variation of additional loadings of the adjacent pile induced by spudcan penetration, and especially finds that the induced soil displacement is heavily dependent on soil deformation characteristics such as , and the induced pile loading is found to be not sensitive to the variation of within a reasonable range. Finally, the study discusses how to define the boundary condition at the pile head where the pile is connected to the upper structure, and whether the linear superposition law is appropriate for loadings induced by the spudcan penetration and the jacket.

Using Laplace transform, the calculation on lateral vibration impedance of pile-soil interaction embedded in layered soil is investigated. Vibration differential equation of pile-head in layered-soil with axial loading is transformed into algebraic equations by means of transfer matrix method and Laplace transform to solve the parameters of piles for the lateral vibration response. Combined with dynamic Winkler model in the frequency domain, the lateral vibration impedance of a pile is derived. Based on the obtained solution, the pile-soil-pile lateral dynamic interaction factors is deduced. Through theoretical analyses of practical problems, the present method is verified to be applicable and efficient. The method is simple to use, and the calculated results agree well with the existing results. Moreover, the present method can give a continuous solution and applied to the analysis of pile-soil interaction.

The fully grouted anchor is a commonly used support method for surrounding rock in underground caverns. It is important to investigate the stress distribution along the fully grouted anchor for evaluating the stability of the combined structure of the anchor and the surrounding rock. By utilising the element embedding formulation, an equivalent method was established to simulate the anchor in the combined structure. The internal deformation of the anchor was simulated by using the axisymmetric quadrilateral element. Accordingly, the internal force and tangent stiffness of the element, the inherent internal force and algorithmic tangent stiffness of the anchor model were derived. Finally, formulations of the additional internal force and tangent stiffness of the anchor model were deduced by using the transformation between the global and local coordinates. The nonlinearity of the anchor was mainly caused by the interface failure and anchor yielding. This method can easily apply the implicit finite element formulation to perform an integrated analysis for the nonlinearity of the surrounding rock and the anchor. Therefore, this method can not only simulate the supporting effect of anchor for the surrounding rock, but also reflect the stress distribution along the anchor.

A self-made hollow cylinder torsional apparatus for rock was developed to control four loading parameters independently, including axial force, torque, internal and external confining pressures. This study presented several stress paths and loading methods which were easy to implement in practical engineering projects and meet their requirements through mathematics and mechanics analysis. The tensile strength of rock can be obtained in the premise that the axial force and internal confining pressure satisfy specific relationships. By controlling the internal and external confining pressures and axial force, conventional triaxial tests and true triaxial tests can be carried out by using the hollow cylinder torsional apparatus for rock. Thus, the existing disadvantages of the complicated true-triaxial test apparatus and the large loaded surface friction were overcome. The mean stress p and the coefficient of intermediate principal stress b can be kept constant when the axial force and the internal and external confining pressures meet certain relationships, respectively. Those stress paths can be used to study the effect of the stress rotation on rock mechanical properties. Therefore, the realisation of the above stress paths is of great significance to the study of rock mechanical properties and the development of laboratory tests.