In order to study the dynamic response of ground fissure sites, based on a 13-layer shear model box, the shaking table test was used to simulate the dynamic response of the f4 ground fissure site in Xi’'an. The failure characteristics and dynamic response of the ground fissure site under different seismic waves were studied. The peak change laws of ground motion parameters were obtained. The test results showed that the main fissure gradually cracked and expanded on the surface, and the secondary fissures were developed at 45 degree joining the main fissure as the dynamic test continued. Moreover, the number of secondary fissures increased with the increase of input seismic intensity. The response of ground motion parameters at the ground fissure site had hanging-wall/ footwall effect. The ground motion peak of the hanging-wall reached the maximal value at the fissure on hanging-wall side and the peak gradually decreased from the main fissure to the two sides. Beyond that, the seismic intensity increasing weakened the site amplification effect. In addition, the changing frequency of acceleration at the hanging-wall site was faster than that at the footwall site. Meanwhile, there was a phase difference between the peak acceleration response on the hanging-wall and that on the footwall. However, the velocity and displacement time history curves of two sides were basically consistent. The research results are an important reference for seismic design of structure passing through the ground fissures.

Weathering is one of the most common diseases for earthen sites in northwest China. Surface sprinkling and hole grouting are two main methods to prevent weathering. Grouting experiments are performed on remolded soil samples with different dry densities to reveal the infiltration grouting diffusion laws of SH agent. An infiltration model of SH agent is established and the experimental phenomenon is reasonably explained. The volume of the soil reinforced by SH and the solid content of SH agent are obtained by calculation, and the solid content is proved to be sufficient to resist weathering for earthen sites. Through SEM images processed by Image J, pore distribution, pore fractal dimension and average Feret diameter under different dry densities are obtained. It is found that the permeability of soil decreases with the increase of dry density, because medium and large pores are compressed into micro and small pores. It is suggested that grouting experiment of SH agent should be applied to soil whose average Feret diameter is more than 4.98 ?m. Finally, by the method of multiple linear regression, it is found that grouting volume has the greatest effect on the infiltration radius, followed by average Feret diameter and aperture.

To study the damage of sandstone in different hydration environment conditions after repeated freeze-thaw cycles, saturated freezing and thawing experiments were carried out on sandstone samples under the freezing temperature of ?30 ℃ and melting temperature of 30 ℃. The acoustic emission (AE) signals were recorded during each freezing-thawing cycle, and the microscopic characteristics of sandstone were investigated by low field nuclear magnetic resonance (NMR) and an optical microscope after every three cycles. The uniaxial compression test was conducted at the end of cycles. The results show that since 3% NaCl solution leads to the internal microstructure destruction, with increasing cycle index, the spectrum shifts to the right, the total area of spectrum and the porosity increase. The AE absolute energy probability density of sandstone still meets the power law distribution after freeze-thaw cycle under uniaxial compression, but the critical exponent increases and the increment of 3% NaCl is greater than that of distilled water. In addition, the critical exponent of AE probability density for each freezing and thawing cycle increases first and then decreases with increasing cycle number, which is similar to existing ultrasonic testing results. The peak of critical exponent in thawing process is higher than that in freezing process, which means that thawing process has shorter breaking time and a lower damage degree. The freeze-thaw damage is mainly caused by the static pressure, osmotic pressure destruction and dissolution but also the erosion of rock by hydration medium. This study can provide some references for understanding failure mechanism and stability evaluation of rock engineering in cold regions.

Cushion is a key component of composite foundation, which ensures the deformation coordination and rationality of load distribution between soils and piles. However, the mesoscopic working mechanism of the cushion is not well investigated and the design parameters of cushion are mainly determined according to the experience. In order to reveal the coordination mechanism of the cushion, a test apparatus was designed to simulate the working of cushion in the composite foundation. Three visualized plane strain tests were carried out with different heights of the cushion. A series of 2D DEM simulations was conducted with an orthogonal array. Both the model tests and simulation results showed the three similar patterns of deformation which related to the ratio of the cushion height and the net spacing of the piles. Mesoscopic force chain distribution was also analyzed. In addition, a design suggestion of the cushion height was proposed by analyzing. Meanwhile, an index applied to assess the deformation coordination capability was presented. According to the variance analysis and range analysis, it was found that the height of the cushion and the net spacing of the piles were the significant factors which affected the deformation coordination capability. The upper limit value of the index was also explored and proposed to guide the design of cushion.

The aim of this study is to investigate the meso-damage and degradation mechanism of sandstone under the combined effect of water chemical corrosion and freeze-thaw cycles. Nuclear magnetic resonance (NMR) and uniaxial compression tests were conducted on the sandstone treated by acidic, neutral and alkaline solution immersion and freeze-thaw cycles, respectively. The variations of porosity and mechanical parameters of sandstone were analysed. With increasing immersion time in hydro-chemical solution, the strength and elastic modulus of sandstone decrease, and the maximum decrement occurs in specimens immersed by the acidic solution. With the increase of freeze-thaw cycles, the strength and elastic modulus significantly decrease, and the maximum decrement occurs in specimens immersed by the neutral solution. The porosity component of a small pore is trivially affected by water chemical corrosion, but significantly influenced by the freeze-thaw cycle. Moreover, the increase of small pore size porosity component in different solutions is close to each other. Both soaking and freeze-thaw cycle in hydrochemical solution greatly impact on the large pore-size porosity components of sandstone. Chemical corrosion has the greatest effect on the sandstone in acid solution, while the neutral solution largely affects the specimen under the interaction of chemical corrosion and freeze-thaw erosion. Under the condition of freeze-thaw cycle of hydrochemical solution, the micro-structure of small-pore sandstone is mainly damaged by frost heave, while that of large-pore sandstone is mainly damaged by chemical corrosion and frost heave. In terms of porosity, the superposition effect of frost heave and corrosion is manifested comprehensively. In the micro-structure of large-pore sandstone, the superposition effect of acid and alkali corrosion on the meso-structure of large-pore sandstone and frost heave is restrained to a certain extent. There exists a good correlation between initial fracture volume change and nuclear magnetic porosity change rate.

In order to investigate the effect of confining pressure on deformation and failure characteristics of deep mixed ground, many specimens were made to mimic layered composite rock, considering the geological formation characteristics. Triaxial compression tests were carried out on these specimens to study the effect of confining pressure on deformation and failure characteristics of horizontal layered composite rock. The results showed that: with the increasing of confining pressure, the decreasing rate of partial stress-strain curves became slower and the degree of strain softening was gradually weakened; the damage form of horizontal layered composite rocks evolved with the increase of confining pressure and was translated from brittle to ductile. Under certain confining pressure, the constraint effect of confining pressure on soft rock expansion deformation was weaker than hard rock, which led to uncoordinated expansion deformation between the soft and hard layers. The relative dislocation occurred between the soft and hard layers under the influence of interlayer adhesion. These results can guide construction units to prevent and control the disaster of tunnel boring machine (TBM) in tunneling through deep mixed ground.

The Green functions of saturated soil derived from the Biot’s poroelastic dynamic equation are the results in u-w formulation, where u is a displacement of the solid phase medium, and w is an average displacement of the fluid phase relative to the solid phase. Decoupling fast and slow longitudinal waves with regard to Biot’s dynamic equation, the analytical solutions for the solid and fluid Green’s functions in u-P formulation, where P is the porous pressure of the fluid phase, subjected to an underground point source in a porous medium in frequency-domain are obtained respectively and readily in this paper. The effects of increasing roots of Green functions in the u-w formulation on BEM integration are eliminated. Through Hankel’s inversion and Somigliana’s representative integral over a half-space in numerical implementation, the displacement, porous pressure, water discharge and other corresponding results of dynamic response in the saturated soil are presented for discussing. It has important application value for the response calculation of subway and other traffic engineering, seismic engineering and structure-soil dynamic interaction (SSI).

The aim of this paper is to study the characteristics and generation mechanism of the low-frequency magnetic field generated during the damage of coal under static load, and to complete the monitoring and warning technology of dynamic disasters in coal mine. Laboratory and field experiments were conducted to analyse the characteristics in time and frequency domain of low-frequency magnetic signal. Laboratory experiment results show that the strength of the generated magnetic field is in the range of 19-156 nT. Both the maximum amplitude and magnetic field energy have a positive correlation with the intensity of coal and loading speed. It also shows that the magnetic field perpendicular to crack surface is the strongest, while the magnetic field parallel to the crack surface is the weakest. Then, by combining with the micro-seismic signal, generation mechanism of the low-frequency magnetic field was proposed. Results reflect that the magnetic signal and micro-seismic signal are synchronous in the time and frequency domain, and oscillation of crack surface with charge leads to low-frequency magnetic field. Field test results indicate that the magnetic signal includes the cluster pulse signal and small amplitude oscillation signal. The cluster pulse signal is caused by the crack surface vibration generated from blasting. The small amplitude oscillation signal is the result of lateral tensile failure generated during the formation process of the new stress equilibrium in coal wall, and migration, friction and rotation of coal particles with charge.

Bentonite, as buffer materials in disposal repository, will be under long term effect of decay heat which released from nuclear waste sealed in the canister. Whether the expansion self-healing ability of bentonite can be maintained after sustained high temperature has not been clearly concluded. The powdery bentonite was heated in muffle furnace at a constant temperature (105 ℃). But the heating periods were different, and the longest one was 90 d. A series of tests was carried out on the heated samples. The results show that the free swelling ratio and swelling pressure decrease dramatically with heated time. X-ray diffraction spectrums of montmorillonite demonstrate that the adjacent lattice distances is shortened, silica oxide cementation occurs after continuous heating for 90 d, which results in aggregation of montmorillonite particles. It can be inferred from infrared spectrum and thermogravimetric testing results that the hydration of interlayer ions of montmorillonite is removed by continuous heating, and some Na and Mg ions escape with the evaporation of water. The combination of the two causes the condensation of montmorillonite particles. This phenomenon is also found by SEM that montmorillonite particles closely linked to each other instead of distinctly stacked as original state. Moreover, this condensation or cluster behaviors are irreversible, which could be confirmed from particle size distribution curves of heated bentonite. Because the clusters of bentonite (heated 90 d) were not dispersed obviously, even after soaking 28 d.

The aim of this study is to investigate the stress and deformation characteristics of reinforced earth retaining wall under the action of traffic load. Large-scale model tests on reinforced earth retaining wall were conducted under static load and dynamic load to reveal the distribution of the vertical earth pressure, the deformation of the retaining wall-panel, the loading-plate settlement and the dynamic acceleration. The results showed that the horizontal influence of the reinforced earth retaining wall under the action of static and dynamic loads was about 0.16H-0.54H (where H is wall height). The earth pressure of the retaining wall was reduced in the horizontal direction from the center of the vibration source trend. At the same time, the main influence of the horizontal displacement of the retaining wall under the static load was about 0.55H. The internal structure of the prefabricated reinforced earth retaining wall was formed by gradual compaction. The settlement of the loading plate under the dynamic load was an inclining "step" shape with the increase of the load value at the critical value of the whole step-load. The reinforced structure had a significant dissipation effect on the acceleration response and decreased with the increase of the distance from the load.

Compared with the existing method combining soil arch shape and horizontal layer analysis, this study has no assumption of soil arch shape and shear stress is considered. The stress solution of any point in the sliding soil is obtained by solving equilibrium differential equations of the two-dimensional differential element, and the active earth pressure imposed on the rigid wall is further derived. Subsequently, based on the stress state of any point in the sliding soil, the analytic expression of the arch shape is established. Furthermore, the effects of internal friction angle, soil-wall friction angle on earth pressure and soil arching shape are discussed. Finally, the obtained results are compared with the results derived from existing theory and experimental results, and the conversion formula between the proposed active earth pressure coefficient and Coulomb theory is obtained. The results indicate that the proposed distribution of earth pressure is in good agreement with the experimental results, especially in the case of high wall. When the soil-wall friction angle is small, the calculated soil arch is similar to the arc, and the soil arch becomes steeper in terms of the increase of soil-wall friction angle, which is always under the parabolic arch. The practical formula is only related to , which is convenient to use in engineering.

To study the nonlinear constitutive model of unsaturated intact loess under the condition of complex stress, by considering the effects of the intermediate principal stress and suction on the major principal strain under the condition of true triaxial test, a nonlinear constitutive model considering the effects of the intermediate principal stress for unsaturated intact loess is developed. The model contains 12 parameters, which can be determined by the true triaxial tests of unsaturated soil with constant water content. The effect of the intermediate principal stress on the deformation and strength of soil is studied, and the change rules of the model parameters are discussed. When the initial suction is constant, the cohesion of the soil enlarges with the increase of the intermediate principal stress, while the internal friction angle decreases linearly with the increase of the intermediate principal stress. When the suction is less than 162 kPa, the suction friction angle is equal to 21.8°. If the suction is greater than 162 kPa, the suction friction angle is equal to 5.5°. When the initial suction is the same, the bulk modulus of the soil under the condition of different intermediate principal stresses can be taken as a constant. The model is used to predict the variation of suction in the true triaxial isotropic consolidation test with constant moisture content. The theoretical results are found close to the experimental results. However, the prediction of suction of true triaxial shear test still needs to be further explored.

To study the microstructure change of coal and its damage and fracturing mechanism under temperature shock, cold impact and hot-cold impact tests were carried out on dry granular coal. The micro-morphology, fracture distribution, cracking and extension of coal samples before and after temperature impact were analyzed and compared by scanning electron microscopy (SEM). The cracking mechanism and propagation direction of micro-cracks in coal samples were analyzed based on fracture mechanics theory. The distributions of stress and displacement fields during micro-crack propagation were simulated by ANSYS finite element software to reveal the fracturing mechanism of coal samples. The results showed that the structure of coal was destroyed by two types of temperature shocks, and the thermal stress induced by temperature shocks eventually led to the expansion and extension of the original crack and the new crack. The main types of cracks induced by temperature impact were intergranular crack, transgranular crack, airfoil crack, cross crack, dendritic whisker crack and mesh crack. It is revealed that a bigger temperature difference results in a higher thermal stress. Hot-cold impact can generate more cracks and more extensive crack propagation than cold impact, leading to more serious damage of coal samples. Therefore, the hot-cold impact has better coal breaking effect than the cold impact.

soft dredger fill in Tianjin Binhai New Area of Lingang industrial zone is investigated by means of common vertical drainage board, improved vertical drainage board and improved transverse drainage board to take low vacuum pre-pressing test. The low vacuum pre-pressing of common vertical drainage board showed that in 28 days the strength of surface soil is up to 10 kPa. But the strength of soil in the middle part is still low. It is hard in both sides but soft in the middle. The low vacuum pre-pressing test of improved vertical drainage board showed that in the first 10 days soil strength increasing was similar to the common vertical drainage board. Later the soil strength increased significantly faster than the common drainage vertical board, in 28 days the strength of surface soil clamed to 12 kPa, the intensity of middle part was improved, but the overall uniformity was still poor. The low vacuum pre-pressing test of improved transverse drainage board showed that in 28 days the surface soil clamed to 11 kPa. The soil strength growth was slower than the improved vertical drainage board, but the strength of middle soil and uniformity were better than the improved vertical drainage board. The above test results are analyzed using the FEM simulation software, and compared with calculated settlement and pore water pressure, the improved vertical drainage board calculation result is found in good agreement with experimental data, improved transverse drainage board is a little poor, but it can basically reflect the low vacuum preloading soil settlement and pore water pressure in the process of value.

In this paper, the sandstones of the Huangmaqing Group (T2h) were taken as the study object. The compressive and tensile tests were carried out on the rock samples, considering the influencing factors of the slope along the Qinhuai East River in Nanjing city. Since water is the main factor that affects the stability of the excavated slope, both the natural and saturated soaking-drying cycles were taken into account during the experiments. Based on the experimental results of the sandstones, the void and deterioration characteristics and deteriorating mechanism were analyzed in detail. With the increasing number of the soaking-drying cycle, the compressive and tensile strengths, and the cohesion of the sand rock samples all gradually deteriorate. The more the cycle numbers, the more serious damage of the rock mass, which indicates that the cycles have cumulative damage on the samples. Under the natural water absorption conditions (no water entering the small open voids), the strength and modulus of the rock slowly decrease. However, under the saturated water absorption conditions (the small open voids are filled with water), the strength and modulus rapidly decrease, which indicates that the deterioration effect due to existence of the small open voids is obvious. The micro-structures indicate that the voids in the rock increase with the number of soaking-drying cycle. Combined with the shear strength characteristics results, the declination of the cohesive strength is the main factor of rock strength deterioration, and the frictional strength almost has no change. The cycles lead to the water-rock reaction, causing gradual loss of rock particle cementation and cracking of the edges, and the rock is degraded in macroscopic scale. Research results have an important reference for the study of stability of new excavation slope, and provide favorable evidences for the long-term stability analysis of river bank slopes.

For the aim to analyze the vibration characteristics of a single pile surrounded by radially inhomogeous soil, a simplified mechanical model for vertical vibration of a single pile embedded in a radially inhomogeneous viscoelastic soil is proposed by employing the viscous damping and three-dimensional axisymmetric continuum model with annular transferred complex stiffness. Firstly, the complex stiffness at the interfaces between soil and pile is derived using Laplace transform and complex stiffness transfer method. Secondly, an analytical solution for dynamic impedance at the pile head is obtained by using the compatibility condition of pile and radially inhomogeneous surrounding soil. Furthermore, the obtained analytical solution for dynamic impedance at the pile head is reduced to verify its validity by comparison with an existing solution. Extensive parametric analyses are performed to investigate the effects of the parameters on the vibration characteristics at pile head. The computational results show that the viscous damping coefficient and the damping factor only have significant influence on the amplitude of dynamic impedance at pile head, while the stiffness factor has significant influence on the amplitude and resonance frequency of dynamic impedance at pile head. The higher the degree of softening (hardening) of the surrounding soil, the larger (smaller) the amplitude of the dynamic impedance at pile head. The larger the range of softening (hardening) of the soil around the pile, the higher (lower) the level of the dynamic impedance at pile head. However, the influences of the degree of softening (hardening) of the soil around the pile and the range of softening (hardening) on the resonance frequency of the dynamic impedance at pile head can be neglected.

Rubber-sand mixture (RSM) has long been recognized as a light weight and energy absorbing material with widely usage in civil engineering. Shear strength of RSM usually decreases with inclusion of rubber particles. As geogrid is usually used to improve the shear strength of granular materials, it is of great interesting to find out how geogrid affect the strength of RSM when RSM is reinforced by geogrid. Triaxial tests were carried out on dry RSMs to study the influence of reinforcing geogrid layers on the strength parameters, such as peak deviatoric stress, pseudo cohesive strength and internal friction angle. Five rubber contents (0%, 10%, 20%, 30% and 40%) and three confining pressures (50 kPa, 100 kPa, 200 kPa) were taken into accounted in tests. Test results indicate that peak deviatoric stress, pseudo cohesive strength and internal friction angle of RSMs reinforced by geogrid are significantly improved comparing to those of RSMs without geogrid. These parameters increase with the increase of the number of geogrid reinforcement layers and with the decrease of confining pressure. The reinforcing effect on strength of RSM by geogrid reaches a peak when the rubber content is 20%. The strength recovery coefficients of geogrid reinforced RSMs exhibit linear relations with the reinforcement density of geogrid, and can be estimated well by the empirical curves of this study.

Fluid flow tests were conducted on artificial fracture networks with different numbers of intersections and subjected to various boundary load conditions. For all cases, the inlet hydraulic pressures were ranged from 0 to 0.6 MPa, and the lateral pressure ratios were increased from 1.0 to 5.0. The test results show that the Forchheimer’s law provides an excellent description of the nonlinear fluid flow in fracture networks. Both the linear and nonlinear coefficients in the Forchheimer’s law generally increase as the lateral pressure ratio increases but decrease as the number of intersections increases. During the fluid flow process, relationships between nonlinear effect factor E and hydraulic gradient J can be well described using a power function. The nonlinear effect factor E shows an increase with the hydraulic gradient. As the lateral pressure ratio increases, the critical hydraulic gradient shows an increasing trend. For all numbers of intersections (1-12), in the range of lateral pressure ratio from 1.0 to 5.0, the critical hydraulic gradient shows an increase from 0.63-12.13 to 6.01-81.55. A mathematical model of for decreased normalized transmissivity against the hydraulic gradient was established. An increase in the lateral pressure ratio shifts the fitted curves upward. The coefficient generally presents an increasing trend with the lateral pressure ratio. Equivalent permeability of the fracture networks decreases with the lateral pressure ratio.

In order to understand the self-healing property of the damaged salt rock in excavation disturbed zone (EDZ) during the utilization of salt cavern, self-healing experiments of damaged salt rock under the effect of the brine soaking and drying were carried out. The results are as follows. The mechanical properties of the damaged salt rock can be recovered after being soaked in brine and dried at different temperatures. The recovery effect increases with the increase of temperature in the test temperature range 35-80℃. The acoustic emission characteristics of the damaged salt rock specimen during the secondary loading process after healing show that the Felicity ratio is less than 0.1. It can be concluded that the Kaiser effect is not established, which indicates that part of the damage of the salt rock can heal at the mesoscopic scale during the self-healing process. Based on the mesoscopic morphology characteristics of the salt rock specimens after healing, three different physical mechanisms of damage healing in rock salt are summarized: healing of microcracks based on diffusion, filling of microcracks based on crystallization of NaCl crystal, healing of the boundaries of crystal in the crushing zone. During the process of salt rock damage healing, the brine filled in the cracks not only provides a channel for the material transfer but also provides a number of base material for crystalline, which lays the foundation for damage healing. The increasing temperature provides more energy for the healing and increases the recrystallization density, improving the speed and the degree of damage healing.

Wille Geotechnik ring shear apparatus was applied to conduct large shear strain ring shear tests on three samples of uniform carol sand with different sizes. This paper investigates the effects of particle breakage on peak strength and residual strength. The experimental results indicate that particle breakage increases with increasing the shear distance, and also increases with increasing particle size. Particle breakage has subtle or negligible effect on the peak strength and residual strength. However, particle breakage remarkably influences the volumetric strain, where the volumetric strain increases with increasing the particle size because of large particles suffered greater particle breakage. An equation for fractal dimension considering length-width ratio, sophericity and camber is established. As the self-similar and no scale of particle along with ranges of sizes, the fractal dimension calculated through the cumulative number of particles with consideration of particle shape is the same with that of particles without considering of particle shape.

Debris flow destroying piers is a common destructive form of bridge under the impact of debris flow. To investigate the magnitude of debris flow impact forces on bridge piers, we adjusted the contents of clay, sand, gravel, and water to generate debris flows with different rheological properties and densities. Two types of pier scale models with circular and square cross section were impacted in debris flow trough by using the above debris flows to comprehensively investigate the relationship between rheological characteristics, flow velocity, pier shape and impact force. The results show that the obtained debris flow materials have distinct rheological properties which can be easily measured through a rotational viscometer and represented by Newtonian fluid or Bingham fluid. The velocity of debris flow can be calculated by the Manning equation, and the roughness coefficient and the viscosity of debris flow in the equation satisfy a power function relationship. In the same cases, the impact forces on a round pier and on a square pier are significantly different. Generally, the drag coefficient of impact force on a round pier is much larger than that on a square pier. Because using non-Newtonian fluid Reynolds number (Re) can comprehensively represent the debris flow’s rheological properties and velocities, the drag coefficient of the round pier be expressed as a function of Re. However, there exists this function for the square pier. For the convenient application in engineering, the drag coefficient of a round pier can be selected as 2.3 and 0.9 for viscous debris flow and sub-viscous debris flow, respectively. For a round pier, the drag coefficients are 2.6 and 1.9 for viscous debris flow and sub-viscous debris flow, respectively.

As a basic portion of frozen soils, ice has a significant impact on the accelerated creep stage of frozen soils. Temperature gives rise to hardening and weakening of frozen soil structure by affecting the freeze-thaw process of ice in frozen soils, as well as its viscoplasticity flow, thus becoming one of the key factors to determine the behavior of frozen soils. Meanwhile, the external stress also causes hardening and weakening, affecting the creep of frozen soils. Both hardening variable and damage variable are introduced here to consider hardening and weakening of the frozen soils resulted from variation of temperature and stress. Among them, the hardening factor H represents the magnitude of hardening effect in the process of creep, while damage factor D represents the reduction ratio of related parameters of frozen soils caused by the weakening effect, and then an improved Nishihara model is developed, the expression of improved Nishihara model under complicated stress state is derived as well. The comparisons between model predictions and available experimental results show that the improved model can not only describe the initial creep stage and the stable creep stage well, but also agrees well with the strain law growing exponentially with time in the accelerated creep stage, which demonstrates its accuracy and usefulness.

The damage of rock caused by freezing and thawing is closely related to its bedding structure, and the freeze-thaw damage is a comprehensive performance. A simple consideration of the change of a single defect inside rock mass or a single external factor under freezing and thawing cycles is insufficient to objectively reflect the damage. In this paper, freeze-thaw tests subjected to 0, 20, 40 freeze-thaw cycles were carried out on the vertical and parallel layered samples. The damage development characteristics of the layered sandstone during the freezing and thawing process were explored. Based on the ?100 mm SHPB test device, impact tests at five types of speed were carried out, and the distribution law of impact crushed degrees of layered sandstone under the condition of freeze-thaw cycles was analysed. Based on the correlation between the accumulation of freeze-thaw damage and the degree of crushing, a freeze-thaw degradation model of layered sandstone was established by the deterioration of mechanical properties. It is found that under the action of freezing and thawing, the longitudinal wave velocity, pore development and impact crushing morphology of these two types of layered sandstone samples are significantly different. There are three correlations between the degree of fragmentation and the number of freeze-thaw cycles, i.e. positive correlation, negative correlation and fluctuation. The calculated values of the intensity and peak strain obtained from the negative correlation region data, as well as the proportional and overlapping area data, are very close to the measured values. The calculated values of the intensity based on the positive correlation data are also in good agreement with the measured values.

Since the Jiweishan landslide in Chongqing in the front edge of the “key block” is of certain concealment, it is of significance for the early warning of landslide to study the failure property the rock bridge in the front locking section of “key block”. Based on the geological condition of Jiweishan, after prefabricating different length edge cracks, infilling soft materials and forming the joints of soft interlayer at the ends of the sample, direct shear tests were conducted on the prepared samples under different confining pressures. The propagation rules of crack, the characteristic information of the block spalling and the fracture condition of the rock bridge were analysed and the k value of changed rate of shear stress was proposed in a would-be sliding stage. The length of joint of the soft interlayer has a certain effect on the failure modes and the extent of block spalling, and the longer the end of rock bridge is, the smaller the k value of the growth rate of the shear stress is. The failure modes of the rock bridge at the end include shear failure, tensile-shear failure and tensile failure, however the extent of block spalling of the same joint is different under varied coalescent modes. The ratio of shear stress of spalling point to that of failure point is in the range of 79.5%-92.2%. The short joint is slower than the middle joint, but the long joint is faster than the middle joint in the time of would-be sliding stage. The three types of failure modes of the end rock bridge meet certain fracture conditions and present three stages and the crack growth stably. The conducted direct shear tests reveal the failure and characteristic information of rock bridge at the end and the fracture condition, and thus this study can provide a theoretical basis for the instability and failure estimate of the rock slope containing the front edge of the locking section which is called “key block”.

To investigate the meso-damage evolution characteristics of rock under different drying-wetting cycles, a series of drying-wetting cycle experiments was conduct on saturated argillaceous sandstone from a slope in Three Gorges Reservoir. Low field nuclear magnetic resonance(NMR) was used to test the damage behaviors of the sandstone subjected to drying-wetting cycle. The amplitude of NMR T2 spectrum significantly changed under the influence of the drying-wetting cycle. With increasing drying-wetting cycles, the peak stress of argillaceous sandstone decreased exponentially, and the T2 spectrum area and porosity showed exponential increase. The internal pore of rock increased gradually, and the pore size increased, finally small cracks coalesced and extended gradually. The final cracks distribution of MRI of specimen were similar to the SEM ones under different wetting-drying cycles. The gray value distribution of the pixels in MRI coincided with logarithmic Gauss distribution, and the expectation of the distributions was gradually saturated with drying-wetting cycles. As the number of cycles increased, the mechanical parameters gradually decreased, and the decrease of peak stress was positively correlated with the increase of porosity.

Based on the geological structure of columnar jointed rock mass, artificial columnar jointed rock mass specimens with different column dip angles and transverse joints were made of cement mortar. Uniaxial compression tests were carried out to investigate the effect of dip angles and transverse joints on mechanical behaviour and failure modes. The results showed that the shape of various curves of deformation modulus and peak strength with dip angles were similar to the letter “U” which indicated significant anisotropy. With the dip angles changed, the failure modes of columnar jointed rock mass under uniaxial compression would be different. Four typical failure modes were summarised, i.e., the splitting failure orthogonal to column axis, shear-sliding failure, comprehensive failure of splitting and shear-sliding, splitting failure along the column axis. Since the transverse joints cut off columns, the integrity of rock mass was reduced along the axial direction, which influenced the distribution and coalescence of splitting fracture. Thus, the bearing capacity of columnar jointed rock mass was largely affected by the transverse joints.

Considering the effects of the pile installation on the properties of the soils surrounding the piles, the load transfer models for pile shaft and pile base are established using the exponential function type load transfer curve. Then, according to the deformation mode of the soil around the piles, the stress transfer method is adopted to model the nonlinear behaviour at the pile-soil interface and the shear displacement method is used to evaluate the pile-pile interaction effects. A hybrid calculation method is proposed for assessing the nonlinear load-settlement behaviour of pile groups. The proposed method is verified by the centrifuge model test. Investigations are conducted on the effects of pile installation and nonlinear behaviour on bearing performance of pile groups. The results show that, the strength and stiffness of the soil around the piles are significantly enhanced by the pile installation due to the squeeze effects, which greatly improves the bearing performance of the pile group. The stiffness of pile-soil interface decreases with the increase of the settlement, and hence the pile group exhibits apparent nonlinear behaviour during the loading process.

As an important part of the subgrade, the filling material quality of subgrade bed has a significant influence on the dynamic response of the subgrade, track, and train. In practical engineering, although filling material of group A, B can be directly used according to the design code in China, improved soils and even unimproved soils also may be used if the economic factors and mechanical properties are considered to meet the requirements. In order to study the influence of different filling materials, such as red mudstone, filling material of group A, B, and improved red mudstone, on dynamic response of subgrade, field vibration tests are carried out. According to the comparative analysis of the tested data, the following conclusions about dynamic stress, acceleration, and settlement have been obtained. When the red mudstone is used, the dynamic stress and settlement can meet the design requirements during the loading period. Compared with the responses when using red mudstone, the distributions of dynamic stress and acceleration along the cross-section of the subgrade surface are more uniform when filling material is group A, B and lime-modified red mudstone, so the overall bearing capacities of these subgrades are better at this time. Besides, the attenuations of the dynamic stress and acceleration along the depth of the subgrade are more obvious than that in the red mudstone; the settlement of the subgrade is smaller, especially when the improved red mudstone is used.

A bounding surface plasticity model for describing the stress-strain behavior of saturated sand subjected to drained cyclic loading is presented within a critical-state framework. A hardening rule depending on incremental deviatoric strain is adopted. During the initial loading, the stress state always locates on the bounding surface. During the unloading and reloading processes, the bounding surface is the historical maximum yielding surface. This hardening rule can describe the softening phenomena of dense sands and remember the stress history by the bounding surface. The shape of the bounding surface is a modified ellipse, which enables the model to describe the plastic strain during loading with constant stress ratio. This model incorporates state-dependent dilatancy and adopts the non-associated flow rule, so it can reasonably describe the volume change behavior of sandy soil. A mapping rule passing through stress reversal points is adopted. A single set of 10 model constants calibrated by conventional triaxial tests is needed for one type of sand under different initial void ratios and different confining pressures. The predicted results by the model for the monotonic and cyclic triaxial tests on Hostun sand, Nevada sand, Toyoura sand and Fuji River sand demonstrate that the model can reasonably describe the stress-strain characteristics of saturated sand.

To investigate the evolution of gas permeability of granite in triaxial cyclic loading/unloading tests, the cyclic loading/ unloading tests and gas permeability tests were carried out on granite specimens under different confining pressures by using a rock T-H-M-C coupled testing system. The test results show that the plastic hysteresis loops are formed by the loading curve and unloading curve of the former cycle. The lower limit of axial strain in each cycle increases with increasing number of cycles. The volume compression of specimens occurs at the first several cycles, and the main deformation is along the axial direction. Afterwards, the specimen volume changes from compression to dilatancy when deviatoric stress increases to a certain value. All specimens are divided into several parts by macroscopic cracks when shear fracture generates, and the orientation of crack is inclined to axial direction. Gas permeability curves are divided into three stages: a stable decline, a slow rise and a sharp rise. The gas permeability increases by 2-3 orders of magnitude after the brittle failure happens. The inflexion point of volume strain curve and the lateral strain curve can be used as an important reference to predict the change of permeability.

Adhesion between deep-sea sediment and metal surface were studied using the deep-sea sediment collected from the Pacific C-C Zone (Clarion-Clipperton Zone), under different void ratios, normal pressures and loading speeds. A formula for calculating top soil adhesion index between deep-sea sediment and metal surface is developed based on the calculation of separating force of parallel disc. This research selects four commonly used marine metals (aluminium alloy 5052, titanium TA2, titanium alloy TC4 and titanium alloy STi80). INSTRON 5943 universal testing machine was used for the adhesion tests. The results show that the adhesion force increases with a high growth rate, as the void ratio of soil sample increases. As the normal pressure increases, the topsoil adhesion index increases exponentially. As the loading speed increases, the topsoil adhesion index decreases gradually. And the adhesion between deep-sea sediment and 5052 or STi80 are weaker. The results provide an important theoretical basis for the safe of deep-sea mining in China.

Accurate prediction of embankment settlement is critical for risk mitigation and cost reduction in embankment projects. Traditionally, the prediction only using data from site investigation usually deviates from the monitored settlement. In this article, it is proposed to integrate multi-source information based on Bayesian theory to predict embankment settlement. The finite element method is adopted to simulate the consolidation process of multiple soil layers, and the posterior high-dimensional distributions of soil parameters are obtained by efficient Markov Chain Monte Carlo simulation. The proposed method is validated by site data from a trial embankment constructed at Ballina, New South Wales, Australia. The results indicate that the proposed multi-sources information integration method based on Bayesian theory can effectively integrate date from site investigation and field monitoring, based on which the embankment settlement can be accurately predicted. For the trial embankment at Ballina, the accuracy of prediction is improved in terms of the overall trend as more monitored data is incorporated into Bayesian updating. The accuracy of surface prediction can be satisfied based on data from 0-116 d, while the data of 0-496 d can be used to monitor settlement for all the monitoring points. For the Ballina embankment, the prior information affects slightly on the posterior prediction, while the measurement error barely affects the prediction.

In general, the properties and parameters of fractured rock mass are changed with the increase of rock mass size. But when the size increases to a critical state, the properties and parameters of fractured rock mass remain unchanged. The critical size is called the representative elementary volume (REV) of fractured rock mass. In this paper, based on an underground oil project in China and the method to determine REV size in radial flow configuration, the REV size for permeability and the permeability coefficient of fractured rock mass in the area of underground oil depot cavern are investigated, and the REV size is 37.5 m for radial flow configuration, which is 1.3 times of the trace length. The influences of trace length, spacing and gap length on the REV size and permeability coefficient are investigated as well. The REV size and permeability coefficient are influenced by the trace length and spacing, and the REV size is not influenced by the gap length. REV sizes and permeability coefficients under the conditions of radial and unidirectional flow configurations are calculated, and the REV size and permeability coefficient obtained in radial flow configuration are closer to the actual project and the calculation method is more convenient.

Columnar jointed basalts are a unique geological phenomenon in the volcanic region. Its engineering characteristics and excavation responding characteristics are very different from those of common rocks because of the unique geological origin. In order to evaluate the variation characteristics of relaxation depth of tunnel columnar jointed rock mass under different borehole properties, an in-situ test method, which combined digital borehole camera with acoustic velocity measurement, was proposed. The comprehensive evaluation method of relaxation depth was proposed as well. The unloading relaxation properties of columnar jointed basalts after excavation under different angles and diameters were obtained. Results showed that the diameter of borehole almost had no effect on the relaxation depth in the jointed rock mass, while the relaxation depths in borehole with different angles were significantly different. In addition, the borehole deviation angle, the angle between borehole axis and the jointed columnar axis had a substantial influence on the evaluation of relaxation test results of relaxation depth. Further discussion on the mechanism of relaxation depth variation under different borehole conditions was made. This study provides important references to the stability analysis of surrounding rock mass and support scheme optimization of tunnel project.

Slope stability evaluation is a classic problem for soil mechanics. In order to study the stability of heterogeneous slopes, this paper uses the upper limit analysis method to establish the failure mechanism of segmented logarithmic spiral arcs for heterogeneous slope. Then fitting polynomial is used to approximately describe the logarithmic spiral arcs. External force power and internal energy dissipation power of the failure mechanism can be determined. The calculation process is applied to repeat the strength reduction until the slope gradually transition to the ultimate balance and the slope stability factor can be obtained. Taking the south slope in an open-pit mine as the background, the slope stability factor is determined by the method of repeated strength reduction. The stability of the calculation process is verified by the method of changing initial value, and the accuracy of the calculation result is verified by the method of strict Morgenstern-Price. The results show that the process is stable and is not affected by the initial value. Compared with the strict Morgenstern-Price method, the difference of stability factor calculation results between the two methods is less than 5%. The stability factor calculation results can fully satisfy the requirements of engineering practice. At the same time, the most dangerous slip surface formed by upper limit analysis method can satisfy the speed separation relationship.

In this study, Chengchao western iron mine is taken as a case to systematically analyze the variation of groundwater and the sudden collapse caused by underground mining by considering the lithology of rock mass and the influence of underground mining. The surface collapse is mainly controlled by underground mining. The relationship between goaf height and the mining depth is linear when the surface subsidence occurs under certain geological and hydrological conditions. The magnitude and extension range of surface deformation are closely related to lithology of strata around the goaf. Meanwhile, continuous rainfall may induce a sudden surface collapse. The surface collapse could be divided into 4 stages: intermittent upward caving stage, cracks connectivity and expansion stage, drainage collapse stage and formation stage of surface collapse pits group. Moreover, the collapse of rock mass above goaf is not sufficient. The accumulation of rock mass cracks with time and the dynamic change of rock mass collapse show intermittent and leaping upward development. The caving height of the overlying rock mass is related to lithology of strata. When rock mass strength decreases and rock mass integrity becomes poor, the single caving height becomes smaller while the total caving height becomes larger. A funnel-like water level drop gradually forms with the orebody exploitation. The decrease of groundwater level is not only related to the relative position of the goaf, but is controlled by the lithology and structure of surrounding rock. Furthermore, the formation and development of cracks in the rock mass provide favorable channels for the downward flow of groundwater. Soluble rocks are eroded continuously, clay and debris are washed away by groundwater, consequently, a hidden space is easily formed underground. The deformation and failure of the whole gypsum may cause the groundwater inrush which generates an instantaneous high negative pressure in the subsurface area, inducing the surface collapse.

The coupling analysis model of the surface runoff and water flow in fissures is established to study the effect of surface runoff and groundwater seepage on the stability of slope under heavy rainfall condition. The Navier–Stokes is used to describe the runoff and subsurface runoff, and the Brinkman-extended Darcy equation is used to describe the seepage in the soil and rock. Under the boundary conditions of the equal velocity and continuous shear stress at the interface of different media, we could inquire into the flow distribution of internal seepage and subsurface runoff. The drag force generated by water flow on the slope is obtained by applying the Newtonian friction law, and the drag force is embedded into the rigid body equilibrium theory to analyze the slope stability. The calculation results of the actual slope show that the safety coefficient of the slope is 1.164 when the effect of drag force is not considered. And the safety coefficient of the slope is 1.089 when the drag effect exerted by the water flow on the slope soil is taken into consideration, and the safety factor of the slope is decreased by 6.44%. It shows that the drag force effect has a great negative influence on the stability of the slope, and the drag force can play a decisive role in the critical steady state. Finally, the relationships between the safety coefficient of slope and runoff depth, soil thickness and slope inclination are discussed respectively considering drag force or not. The analysis shows that under similar conditions, the steeper the slope, the more obvious the decrease of the safety factor of the slope soil. The safety factor of the slope decreases with the increase of the slope inclination and the thickness of the soil layer.

A microseismic event often contains some abnormal channel signals, strong noise interference signals and weak signals. The picking accuracy of these channel signals’ P-wave phase arrivals is often low, and even those P-picks are false. Currently, the invalid P-picks are not screened out before automatic source location in microseismic monitoring system, and the manual intervention need to be carried out. To solve this difficult problem, a modified Akaike information criterion (AIC) picker was proposed firstly. Secondly, the quantitative relationship between the P-picks accuracy and the source location accuracy was analyzed based on a microseismic monitoring case of a deeply buried tunnel, and then a new concept of P-pick admissible error was introduced. Furthermore, all kinds of microseismic signals’ P-wave phases were picked up by the modified AIC picker, and the corresponding waveform parameters were calculated. A great amount of statistic work has revealed the relationship between P-pick accuracy and waveform parameters. Finally, the P-picks were classified into valid group (tagged 1) and invalid group (tagged -1) based on the admissible error. And support vector machine (SVM) classifier was applied to build a forecasting model for identifying valid and invalid P-picks. Practical application shows that the invalid P-picks are identified correctly and effectively by this quality control method. The accuracy of microseismic source location is improved greatly after eliminating the invalid pickups.

The effect of rainfall on the failure of slope can be analysed through experimental testing, field investigation, and numerical simulations. In numerical simulation for rainfall-triggered slope failure, how accurate boundary conditions of infiltration are initialized and initial pore water pressure is obtained in a dynamic flowing boundary condition are the most important aspects. To solve these problems, firstly, disadvantages of traditional numerical boundary in infiltration are found through numerical simulation and experimental results. Thus, new boundary conditions of infiltration are presented, in which the rainfall infiltration is processed through air units, as a dynamic boundary condition, and then into unsaturated soils. Considering these new infiltration conditions, a variety of numerical simulation is carried out in a soil column test for infiltration. It is shown that the results from simulation are consistent with those from experimental tests and field monitoring by others. The proposed method is proved to be effective and accurate for the rainfall-triggered slope failure. Meanwhile, the tensile strength for water at nodes is set to express suction in unsaturated soils. Because the relative permeability coefficient and suction are the function of saturation, the changes of saturation in each node can update the relative permeability coefficient and suction in each step of simulation. Then the seepage is developed using FISH language in the FLAC software. Moreover, a dynamic flowing boundary condition is proposed to consider the presence of suction in unsaturated soils. In this condition, the natural rainfall infiltration is simulated through the change of groundwater, and the initial pore water is precisely obtained after the development of seepage modules in the FLAC software.

It’s well recognized that the soil response is generally linear and elastic at very small strain level (i.e. below 10-6). Previous numerous experiments showed that the soil elastic properties mainly depended on the effective confining pressure, void ratio, soil grading and sample preparation method. In this paper, the physical experiments on regular packing of steel spheres by Duffy and Mindlin were simulated by three dimensional discrete element method in order to reveal the fundamental mechanism related to soil elastic properties at meso-scale. The simulations showed that the elastic stiffness of the soil depended on the coordination number, the normal contact force and its distribution, which correspond to the macroscopic parameters of void ratio and effective confining pressure. It was found that the stress exponent n was higher than 1/3 as predicted by the Hertz-Mindlin contact model and it was mainly due to the increase of coordination number and uniformity of contact force distribution during the increase of confining pressure. Meanwhile, Poisson's ratio decreased with increasing coordination number instead of a constant as commonly assumed, which was consistent with experimental results.

To predict the development trend of ground settlement around foundation pit accurately, a prediction model for maximum ground settlement of foundation pit was proposed based on shuffled frog leaping algorithm and generalized regression neural network model(SFLA-GRNN model). Firstly, through the settlement mechanism analysis and the initial selection of the input variable set, grey correlation analysis was used to quantify the correlation between model input and output variables. Some of input variables that are significantly less correlated with output variables were eliminated. Secondly, the smoothing factor of the generalized regression neural network model (GRNN) was optimized by using the shuffled frog algorithm (SFLA), so as to reduce the adverse effects of human factors on the accuracy and generalization ability of the model. Finally, a generalized regression neural network model for predicting the maximum settlement of the foundation pit was established by using the selected input variables set. Example application and comparative analysis show that input variables selection based on gray correlation degree and smoothing factor optimization based on shuffled frog leaping algorithm all can effectively improve the accuracy and generalization ability of GRNN model. The above conclusions can provide reference for similar deformation prediction.

1. School of Water Conservancy and Environmental Engineering, Zhejiang University of Water Resources and Electric Power, Hangzhou, Zhejiang 310018, China; 2. School of Naval Architecture, Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, China; 3. Reconnaissance and Design Institute, Qiantang River Administration of Zhejiang Province, Hangzhou, Zhejiang 310016, China

To understand the restraint of surrounding rock, vibration characteristics and evaluation standard of underground powerhouse structure, field tests and finite element analyses were conducted on a large pumped storage power station by using an electric testing method, to achieve the modal and dynamic response of underground powerhouse structure. Four types of finite element model are established according to different constraints between the upstream and downstream sidewalls and surrounding rocks. By comparing the results of finite element modal calculation and field test, it is found that when calculating the finite element modal of the underground powerhouse structure of the pumped storage power station, the reasonable calculating condition is that the boundary node of the powerhouse sidewall contacts with the surrounding rock in the normal direction. The deformation modulus of surrounding rock is selected according to the third type of surrounding rock, and the deformation modulus and Poisson's ratio are 10 GPa and 0.25, respectively. The elastic modulus of concrete should be calculated according to the dynamic elastic modulus. Through the dynamic response test of underground powerhouse structure, the maximum displacement is observed when the power generation is started or shut, which indicates that the start-up and shutdown of power generation are the most disadvantageous vibration conditions of underground powerhouse structure in normal operation of the pumped storage power station. The vibration caused by the start-up and shutdown of the power plant does not affect the stability of surrounding rock of the underground powerhouse, and the surrounding rock of the underground powerhouse is in a safe state. The frequency of dynamic response is mainly low-frequency components such as 0.5, 0.75 and 8.25 Hz that may be from the fluctuating pressure in the draft tube or spiral case and the basic frequency or frequency doubling of the unit, which means the vibration response of the underground powerhouse structure is mainly caused by the flow fluctuation and the unit frequency. Based on the safety assessment and analysis of underground powerhouse structure, it is suggested that 0.2 and 0.8 mm should be used as the evaluation criteria for the vibration control of underground powerhouse floor under steady and transient conditions, respectively. The anti-vibration performance of underground powerhouse structure basically satisfies the safety requirements. The analyses of finite element calculation and vibration tests of modal and dynamic response not only make up for the deficiency of the vibration test, but also provide a reference for anti-vibration design and vibration safety evaluation of underground powerhouse structure for pumped storage power stations.

Earth pressure is an important test parameter in centrifugal model experiment, it can be affected by many factors such as the performance of the earth pressure cell, the stability of the centrifuge data acquisition system and the external environment, so accurate measurement of earth pressure is hard to obtain. As a component of earth pressure measurement, the performance of earth pressure cell can directly affect the accuracy of earth pressure measurement. To get more accurate test data, two common resistance strain type earth pressure cells are selected to test using the centrifuge data acquisition system. The calibration test shows that two cells’ calibration coefficients of sand are smaller than the factory calibration coefficients. The calibration coefficients of sand for the type I are 64.75% lower than the factory calibration coefficients, and that of type II are 18.77% lower. The coincidence degree between type II earth pressure cell and factory data is better than that of type I earth pressure cell. In a centrifugal model test for measuring the lateral earth pressure distribution of counterfort retaining wall with wall height of 10-30 m, type I earth pressure cell has some shortcomings, such as data distortion measured by factory calibration coefficients, poor stability and low sensitivity, compared with the type II cell. The results of calibration test and centrifugal test show that type II cell with self established data acquisition system has better performance than type I with static strain data acquisition system.