The influence of different initial water contents, cold end temperatures and dry densities on sand water migration was studied using the self-developed water migration and frost heaving testing equipment. The influence of these three factors on frost heaving force and frost heaving capacity and the position of ice peak were determined. The results show that the initial water content and cold end temperature have obvious influence on the soil water migration and frost heaving effect. The water content increases from 0% to 10%, the peak water content increases by 5.00 times. The lateral frost heaving force and the frost heaving amount are increasing, and the ice front position moves up to 2.5 cm in height. The cold end temperature reduces from ?5 ℃ to ?15 ℃, the peak water content increases by 4.38 times. The lateral frost heaving force and the frost heaving amount are increasing, and the position of the ice peak moves up to 2.6 cm in height. Dry density has relatively insignificant influence on water migration and freezing characteristics of specimens. The results show an overall trend of slightly larger increases in specimen water content, lateral frost heaving force and frost heaving amount at smaller dry densities, and the position of the ice peak is concentrated in 2.2 ?2.5 cm in height. The prediction formulas of frost heaving force and frost heaving amount are put forward for different influencing factors, which can provide a reference for understanding the water migration law in sand under water vapor recharge and reasonably preventing frost heaving.

The characteristics and precursory law of rock burst in roadway are different when the coal-rock composite bearing structure is under different stress boundary conditions. The true triaxial loading and single face unloading test is performed on coal-rock combination bodies under high static and dynamic load, using high frequency vibration acquisition and borehole imaging triaxial static and dynamic load experimental system. The mechanical characteristics and strength conditions at the interface of coal and rock mass are analyzed. Failure modes, dynamic characteristics and evolution of acoustic emission signals of coal-rock combination bodies under different stress boundaries are investigated. The results show that: 1)the strength of sandstone at the interface is "weakened" due to the mutual restriction of coal and rock deformation. When the stress at the crack tip of the coal body at the interface is greater than the strength of the "weakened" sandstone, the crack will develop into the sandstone through the coal rock interface, and the sandstone presents the failure form of buckling spalling and splitting into plates. 2) Under high static loading, the deformation and failure characteristics of coal-rock combination bodies and the acoustic emission signal have obvious precursory law. Before the bearing failure of coal-rock combination bodies, the local particle ejection kinetic energy of coal body increases, the particle lumpiness decreases, and the acoustic emission signal changes from "high frequency and low energy" to "high frequency and high energy". The failure mode of coal-rock mass is mainly shear-tension composite failure. 3) Under the impact dynamic loading, the loading effect of sandstone on the top and bottom plates is weakened, the stress at the crack tip of the coal body can not be effectively accumulated, and the crack is blocked when it extends to the coal rock interface. The combination is dominated by the tensile failure of coal samples, and the acoustic emission signal shows the characteristics of "high frequency and high energy", but it is mostly concentrated after the impact failure, resulting in the dynamic failure of the combination, which is difficult to predict. 4) Compared with static loading alone, although the static load level is low under the superposition of dynamic and static loading test, the ejection quality and fractal dimension of ejection fragments of coal rock assemblage are large, and the average particle size of broken fragments is small. Dynamic load plays a positive role in the failure of coal-rock combination mass while static load provides stress and energy conditions for the dynamic failure of coal-rock combination mass.

In this paper, the digital image based triaxial shear, CT scanning and SEM tests were employed to investigate triaxial shear behavior and microstructure evolution of basalt fiber reinforced loess under drying-wetting action. The results show that the shear failure morphology of soil samples with relatively higher basalt fiber content changes from overall bulging to shear band failure with increasing drying-wetting cycles, while it exhibits the opposite variation with increasing fiber content at the early stage of drying-wetting process. Drying-wetting cycles and fiber content have no obvious effect on the type of stress-strain curves, which present strain hardening behavior. The deviator stress at failure decreases with the increasing number of drying-wetting cycles; however, the attenuation rate gradually decreases. The deviator stress at failure shows a parabolic variation with increasing fiber content and the optimal fiber content is 0.6%. A similar trend is observed between the ME value of CT scanning and the deviator stress at failure. Drying-wetting action induces cracking and loosening around the soil-fiber interface, thus weakening the fiber reinforcement effect. However, compared with unreinforced loess, fiber reinforced loess demonstrates strong stability in its microstructure. The macroscopic and microscopic damage variables reflecting the drying-wetting induced deterioration of loess samples were finally established, which shows consistent trend.

Based on the fault crossing situation of Xianglushan tunnel of water diversion project, we conducted systematic monitoring and analysis of key mechanical characteristics of tunnel such as surrounding rock rupture pattern, lining damage pattern and crack development, strain distribution characteristics during faults dislocation simulation through indoor physical model tests, and thoroughly studied the damage form and failure mechanism of tunnel with flexible joint crossing active faults under strike-slip fault dislocation. In terms of the mechanism of the design parameters of the flexible joint tunnel to resist faulting, the effects of factors such as liner section length, liner thickness, tunnel diameter, angle between tunnel axis and fault zone, tunnel section form, and mechanical properties of liner materials on the fracture resistance of the flexible joint tunnels were studied in detail. The results of the study show: 1) When the tunnel crossing the active fault is not articulated, the damage pattern presents a combination of shear and bending damage under the fault dislocation, and the lining damage is severe with a phenomenon of peeling off. The tunnel section shows elliptical deformation, and the overall collapse trend is obvious. The damage range of non-articulated tunnel reaches 4Wf (Wf is the width of the fault zone) in this scenario. 2) When the articulated design is adopted in the tunnel crossing the active fault, the deformation of the tunnel shows S-shape under the fault misalignment. The damage of the lining structure is in the form of inter-segment rotation and misalignment, while the lining segments are relatively intact and less damaged. The damage range of the flexible joint tunnel in this scenario is 2.14Wf, which is 48% less than that of the non-flexible joint tunnel, indicating that the articulated design can change the deformation and damage of the tunnel under the active fault dislocation and reduce the damage range of the tunnel structure. 3) Under the condition of articulated design, the maximum strain of the tunnel lining structure is mainly distributed in the fault zone, and the tunnel is prone to yield failure. Compared with the non-flexible joint tunnel, the maximum longitudinal tensile strain and compressive strain in the left and right side walls of the flexible joint tunnel are reduced by 56% and 68% respectively, which further indicates that the articulated design can effectively improve the tunnel’s resistance to fault dislocation. 4) In terms of the mechanism of the design parameters of the articulated tunnel, this paper concludes that the resistance performance of articulated tunnel can be enhanced by increasing the tunnel lining thickness, increasing the concrete strength level of the lining, reducing the length of the section and reducing the diameter of the tunnel. The optimal angle of the tunnel through the fault zone is orthogonal, and the circular section can improve the resistance of the flexible joint tunnel compared with the three-centered circular section. In summary, the research results can provide necessary theoretical reference and technical supports for the anti-faulting measures of cross-active fracture tunnel projects.

The shear deformation of rock joints is significant for the safety and stability of rock engineering. In order to study the constitutive relation of shear deformation in jointed rock mass under a normal stress, the direct shear tests under different normal stresses were carried out on the sandstone specimens with irregular joints using the RDS-200 rock direct shear test system. Based on the shear stress - deformation curve of the jointed rock, it can be divided into four stages including pre-peak compaction stage, linear stage, yield stage and post-peak softening stage. The post-peak softening stage can be further divided into three types including platform type, gradual decline type and drop type based on the decreasing magnitude and rate of shear stress at the post-peak period. The shear deformation constitutive model of the sandstone with irregular joints was established using piecewise function based on the shear deformation characteristics at different stages. Compared with the existing constitutive models, the new proposed shear deformation constitutive model of the jointed rock mass has a much higher fitting accuracy for the experimental results, which can better describe the deformation characteristics of the jointed rock in the whole shear process. The shear stress-shear displacement curve of irregular joints with different roughness coefficients under different normal stresses can be predicted after determining the relevant model parameters in the corresponding empirical formula after some direct shear tests. The research is practical for understanding the shear deformation of joints in rocks by numerical simulation and the safety evaluation of engineering.

Metal grillage foundations are increasingly used in infrastructure of transmission line in aeolian sand areas because of its convenient construction and relatively low cost. The full-model and half-model uplift tests under the influence of embedment ratio were carried out in foundation of medium dense aeolian sand, and the uplift load-displacement properties and uplift failure mode of the foundation under different embedment ratios were studied as well. The results show that the uplift load-displacement curve relates to the embedment ratio ?: when ? is 1.0, the curve varies gradually; while ? is in the range of 2.0?5.0, the curve shows clear strain softening with a peak load. As the ? increases, the uplift capacity increases accordingly. During the loading process, an approximately circular uplift area centered on the top of the foundation developed on the ground surface gradually, the uplift degree increased as the foundation was pulled up, and finally formed an overall failure, along with the sliding surface penetrating the ground. In the half-model tests, meridians of the sliding surfaces were observable, which can be approximately described by linear equation as the uplift angle decreased with the increase of the embedment ratio. According to the above results, based on Veesaert’s sliding surface friction strength theory and the limit equilibrium principle, an improved calculation method for uplift capacity of metal grillage foundation has been proposed. Compared with the earth cone method and the shear method, the improved method can obtain calculation results closer to the experimental values with smaller dispersion.

Constructing the underground gas storages in low-grade, multi-interbedded salt formations and using the pore space of the insoluble sediments at the bottom of the salt caverns excavated by solution mining to store natural gas are expected to greatly expand the storage capacity of the salt cavern. However, it is necessary to evaluate the feasibility and applicability of the insoluble sediments for gas storage. Firstly, large-size samples were prepared with insoluble sediments for computed tomography scanning, and the size and connectivity of the pores of the insoluble sediments were analyzed qualitatively based on the three-dimensional reconstruction models. Results demonstrated that the pores of the insoluble sediments have large size and adequate connectivity. Secondly, a self-developed experiment device of salt cavern insoluble sediments by solution mining was adopted to carry out the experiments of gas injection and brine displacement. The small insoluble particles and large-size interlayer blocks were mixed for preparing insoluble sediment samples, the porosities of which were measured before and after the experiments of gas injection and brine displacement, and the reasons for the change of porosity were analyzed. Finally, a prediction model of the pore volume of the insoluble sediments was established, and the theoretical equations of the mining space of salt cavern and the volume of the brine which can be discharged from the insoluble sediments were proposed. By considering the continuous accumulation of small insoluble particles and disordered accumulation of the large-size collapsed interlayer blocks, the mining space and drainable brine volume of the insoluble sediments were calculated, respectively, which were in good agreement with the field data of solution mining. This study can provide references for evaluating the volume of pore space of insoluble sediments for gas storage in low-grade, multi-interbedded salt formations.

To verify the efficient rock breaking ability of high-energy laser technology using for tunnel boring machines (shorted for TBM) disc cutters (hereinafter referred to as disc cutters), taking the basic movement form of the hob in the process of rock breaking-rock intrusion as the research object, a reduced scale hob indenter is prepared with a ratio of 1:8 (the following is the hob indenter) referring to the 17 inch (432 mm) engineering hob, orthogonal tests of rock invasion by hob indenter L20 (4×5) with laser assist and comparative tests without laser assist were carried out with different hole spacing (2, 3, 4, 5 mm) and cutter-hole spacing (3, 4, 5, 6, 7 mm) as variables. The experimental results show that with the increase of indenter-hole distance, both volumes of rock breakage and rock fragmentation size increase; with the increasing laser hole distance, rock breakage volume, vertical forces and the difficulty of rock breaking decrease, while rock fragmentation size and specific energy increase. The optimal indenter-hole distance and laser hole distance are 6 mm and 3 mm respectively. Compared with the traditional rock-breaking method by disc cutters, laser-assisted rock cutting would promote the generation of rock tensile cracks, and improve the rock-breaking efficiency of indentation by the cutters. The research results also show the prospects for application of the proposed laser-assisted rock cutting model to TBM tunneling under the extremely hard rock grounds in the future.

To clarify the plastic law and hardening characteristics of deformation and failure of staggered zone under high in-situ stress unloading conditions with excavation in underground engineering, a series of triaxial tests under different confining pressures and different loading and unloading stress paths was conducted. The plastic deformation laws of the staggered zone under different loading and unloading stress paths were deeply investigated. Based on the results of the experiments, the dependency of stress path of internal variables such as the equivalent plastic work, the effective plastic strain and the plastic volumetric strain as hardening parameters were further explored in the stress space. Meanwhile, a modified formula for hardening parameters with the independency of stress path was proposed. The results showed that 1) The effect of stress path was significant on the deformation properties of staggered zone under high in-situ stress. The high initial confining pressure would inhibit the development of the circumferential plastic deformation of samples with the staggered zone under the same unloading stress path. And the plastic volumetric residual strain (6%) of samples with the staggered zone under the stress path of unloading axial pressure and confining pressure was significantly greater than that of other unloading stress paths (2%-4%) under the same initial confining pressure. The promoting effects of different stress paths on the plastic volumetric deformation of the staggered zone were as follows. The promoting effect of the stress path of unloading axial pressure and unloading confining pressure was the strongest, that of the stress path of constant axial pressure and unloading confining pressure was the second strong, and that of the stress path of loading axial pressure and unloading confining pressure was the weakest. 2) There were a certain of the dependency of stress path of its internal variables such as plastic volumetric strain, effective plastic strain and equivalent plastic work during the deformation and failure process of the staggered zone in different unloading stress paths. Therefore, it is not accurate to directly take any of the above state parameters as the hardening parameters and assume that it is independent of the stress paths in the elastic-plastic analysis of staggered zone. Therefore, a modified method of the equivalent plastic work was proposed. It was found that when the parameter ns reflecting the material properties of staggered zone was equal to ?0.4, the modified equivalent plastic work had the obvious independency of stress path, which was more appropriate for describing the unloading plastic strain hardening characteristics of the staggered zone under high in-situ stress unloading conditions. The plastic mechanical characteristics of the staggered zone revealed in this paper can further deepen the understanding of deformation and failure of the staggered zone under high in-situ stress unloading conditions, and provide a theoretical basis for the analysis of the failure mechanism and support control of the staggered zone in practical engineering.

Based on the existing three component local resonance structure, a four component local resonance pile structure composed of pipe pile, soft layer, pile core and matrix was proposed considering the flow plastic and elastic-plastic characteristics of geotechnical materials. By introducing the periodic theory, the effects of pipe pile density, elastic modulus and thickness on the band gap characteristics of four component structure were analyzed by using the dispersion curve. Compared with the three component structure without pipe pile, the four component structure with pipe pile is more prone to complete band gap, and the pipe pile only affects the upper boundary of band gap. The modulus of pipe pile has a qualitative and even greater effect on the band gap of quaternary structure comparing with the density of pipe pile. By modal analysis, the calculation formula of the band gap boundary for the quaternary structure was established. Through time domain analysis, it is found that compared with the three component structure, the vibration attenuation effect of the four component structure in the band gap can be improved by nearly 30%. The reason is that the pipe pile in the four component structure improves the overall stiffness of the matrix, so as to effectively prevent the occurrence of "breakdown" band and leakage wave of the matrix. This study provides a new insight into the design of new local resonance structure and the improvement of vibration isolation effect.

The classical Richards infiltration control equation belongs to partial differential equation and is strongly nonlinear, that it is difficult to obtain analytical solution. In this study, taking infiltration time as the minimum action, the time functional of infiltration path was established based on Richards equation to transform the vertical infiltration problem of unsaturated soil considering gravity into functional extremum problem, which was solved through the constructed equivalent Euler-Lagrange equation. The calculation results reveal a functional relationship of the diffusion coefficient D (? ) with the distance of the generalized wetting front. Three solutions can be found when the form of diffusion coefficient D (? ) is known: the moisture content at the wetting front under the optimal path, the minimum moisture content at the distant wetting front and the maximum entropy distribution of soil moisture content. Meanwhile, the accuracy of solving Richards equation by Boltzmann transformation and linear transformation is tested by examples. In the solving process, there is no new variable introduced to simplify the Richards equation, and the structure of the original equation remains, so that the solution is universal and can be used as a supplement to the mechanical calculation of unsaturated soil.

The deformation of underground structure during earthquake is mainly controlled by the deformation of surrounding soil. Based on this idea, a simplified seismic analytical method, namely the boundary displacement method, was proposed for underground structure by applying soil deformation to the lateral boundary of soil-structure finite element model, while there is a lack of relevant experimental research. To explore the feasibility of the test method of applying pushover displacement to soil lateral boundary, a large-scale pushover test of soil-underground structure system was carried out taking the self-developed geotechnical comprehensive test model box as the test platform and the 1:10 scale Dakai station section tunnel as the research object. Response characteristics of underground structure and surrounding soil in the process of test were revealed based on the analyses of the strain, displacement, and stress. The results show that due to the strong nonlinear characteristics of soil materials, the inverted triangle deformation applied on the side of soil will attenuate in the transmission process, and the underground structure is subject to the coupling effect of shear deformation and extrusion deformation. The junction between central column and bottom plate is the seismic weak region in the whole structure. Horizontal coefficient of subgrade reaction is related to the level of soil displacement and the failure stage of structural side wall. The integral stiffness of structure is greater than that of the equivalent soil, and the ratio of lateral deformation between structure and soil is less than 1, which gradually increased with the increase of the pushing level. The soil-structure interaction could be quantified through deformation characteristic, which effectively fills the gap in the experimental study of soil structure interaction coefficient. The test method and conclusions have important guiding significance for seismic analysis of underground structure and feasibility study of pushover test.

In order to describe the anisotropy characteristics of rock strength, a modified Hoek-Brown yield criterion based on the microstructure tensor is proposed, in this process, the microstructure tensor of the material is introduced, considering the change of uniaxial compressive strength with the bedding dip ?. Conventional triaxial compression tests are conducted on shale samples with different bedding orientations (0°, 15°, 30°, 45°, 60°, 75° and 90°) under different confining pressures (0, 5, 10, 15 and 20 MPa). The results show that the peak strength of shale and the bedding dip ? have a "U"-shaped curve relationship. When the confining pressure is low, it shows a "Shoulder"-shaped curve relationship. Elastic modulus Et and Poisson’s ratio ν are not sensitive to confining pressure, and the relationship between their average value and the bedding dip angle ? is inverted "U" shape and "shoulder" shape, respectively. The micro and macro failure modes of shale are further discussed, and the results show that its failure mode is controlled by both confining pressure and bedding dip. The comparison between experimental data and fitting results indicates that the modified Hoek-Brown yield criterion can simulate the anisotropy of shale strength satisfactorily. Meanwhile, the experiment data of uniaxial and triaxial compression tests on Koteshwar phyllite, Koteshwar slate and chlorite schist are conducted to further verify the application of the proposed modified Hoek-Brown yield criterion in the simulation of most transversely isotropic rock with bedding structure. The simulation effect of the peak strength of the yield criterion is evaluated in the case of more and less test data.

Starting from the calculation model of the two-dimensional soil-cantilever rigid wall under horizontal simple harmonic resonance action, and based on the theory of wave mechanics, the dynamic response characteristics of two-dimensional soil-cantilever rigid wall were studied considering the vertical stress and vertical displacement of soil layer. Firstly, the vibration equation of the soil layer was transformed to obtain the equation of the volumetric strain ? , which was solved by the method of separation of variables. Then, the inhomogeneous equation of the displacement motion equation was obtained by substituting the solution back to the vibration equation. The definite solution of the vibration equation was obtained by combining the interaction conditions between the wall and the soil layer as well as the far-field boundary conditions. Then more strict analytical solutions of the earth pressure on the wall, the shear force at the bottom of the wall and the bending moment of the rigid underground wall were obtained. Compared with the vertical stress-neglected solution and the vertical displacement-neglected solution, it is shown that the obtained solution can reflect one more resonance frequency, and when the Poisson’s ratio of soil is greater than 0.45, the vertical displacement-neglected solution is meaningless. Through parameter analysis, it is shown that the excitation frequency and soil damping factor have great influence on the dynamic response of the wall, whereas the vibration mode order considered has less impact.

It is of important theoretical value and practical significance in geotechnical engineering to describe the lateral earth pressure coefficient at rest in overconsolidation stratum. In this study, based on the theoretical analysis evolution of lateral pressure at rest during overconsolidation, the excess lateral stress release characteristics of overconsolidated silty clay were investigated through confined rheological tests. The relationship between the residual excess lateral earth pressure and the overconsolidation ratio (OCR) was established, and a new method for calculating the at rest lateral earth pressure coefficient in overconsolidation stratum was proposed. The proposed method was validated by the in-situ KSB test and laboratory test, respectively. Moreover, the proposed method possessed clear physical meanings and achieved unification with empirical formulas under medium-high OCR levels. The results revealed that excess lateral earth pressure went through rebound process and relaxation process, and finally stabilized at residual excess lateral earth pressure . For silty clay, the release of excess lateral earth pressure mainly occurred during the rebound process. The ratio of residual value and initial value of excess lateral earth pressure increased with increased OCR, and no longer changed when OCR>OCRr. The conclusions could provide reference and basis for the calculation and numerical analysis of earth pressure of retaining structures in overconsolidation stratum.

Jiangsu offshore belongs to the South Yellow Sea, which is the most concentrated area of offshore wind farms in China. At present, it accounts for about 70%?75% of the total installed capacity of China. The overburden layer thickness in this area is very large, soils along the pile length depth (generally 40?60 m) are mainly silty sand, silt and silty clay, especially the easily liquefied silty sand silt layer at the depth of 20 m below the mud surface. The South Yellow Sea is also an earthquake activity area in China. Therefore, the seismic liquefaction characteristics of the horizontal site of offshore wind farm in this area are studied. Firstly, based on the statistical characteristics of drilling data of 50 sites in an offshore wind farm in the South Yellow Sea, the generalized formation model is established. The soil dynamic parameters are calibrated based on dynamic triaxial test and resonance column test, and three ground motions are inversed (EL-Centro, Northridge and Kobe), and the peak ground acceleration (PGA) near the surface is adjusted to 0.05g, 0.10g, 0.20g and 0.40g respectively, and the site liquefaction is researched. The characteristics of excess pore pressure ratio, total settlement and layered settlement under earthquakes are analyzed. It is found that the strata in this area are liquefiable soils, the excess pore pressure ratio of each layer is less than 1.0 when input PGA is 0.05g, and the total settlement is about 1 cm. When input PGA is 0.10g and 0.20g, only the surface layer (within 12 m) is completely liquefied, and the total settlement is 10 and 17 cm respectively. When input PGA is 0.40g, the 20 m strata below the mud surface are completely liquefied, and the total settlement is about 30 cm. Under different earthquakes, the settlement of surface soil (12 m) accounts for the largest proportion (more than 95% when PGA is 0.10g and 0.20g). Therefore, the liquefaction characteristics of offshore wind farm under earthquakes should be considered for turbine pile foundation.

Block collapse or sliding is one of the main failure modes of rock slope engineering. Namely, block stability analysis plays a key role in rock slope engineering. Taking the Shenxianju rock slope in Xianju County, Zhejiang Province as the research background, this paper mainly conducts the stability analysis and visualization of rock blocks. A new method for fitting structural planes and free faces is proposed based on the linear regression method and the non-uniform rational B-spline method. Then, based on the coordinate projection method, the method for calculating the stability coefficient of the single-sided sliding surface and double-sided sliding surface blocks is proposed. Finally, the unmanned aerial vehicle (UAV) measurement technology combined with the coordinate projection method is used to develop a CPG program using Matlab, which can be adopted in the stability analysis of planar polyhedron blocks and curved blocks in rock slope engineering. This program enables the spatial representation and visualization of structural planes an free faces and unstable blocks. Engineering practice shows that the new proposed method is effectively applicable to engineering geological disasters, such as rockfall and collapse. The results of program calculation are basically consistent with those of the coordinate projection block theory, demonstrating that this method is reliable and the developed CPG program is feasible. This method is of vital significance in practical engineering since it can greatly improve the efficiency of block stability analysis.

The construction of urban subway tunnel inevitably produces disturbance to surrounding rock and causes ground surface settlement. Dynamic prediction of ground surface settlement caused by tunnel excavation is an important method to ensure the safety of above-ground buildings and tunnel construction. In view of the difficulty of accurate dynamic prediction of ground surface settlement during tunnel construction, based on the definition of longitudinal excavation coefficient ? , a dynamic prediction model of lateral ground surface settlement is established. The model can accurately describe the variation of the settlement of the same monitoring location with the advancement of the tunnel face, and then realize the dynamic prediction of the ground surface settlement at the construction site. The results show that under certain constraints, this model can be degenerated into Peck model and stochastic medium theory prediction model. The accuracy and applicability of the dynamic prediction model are verified by on-site construction. The tunnel can be divided into three affected segments longitudinally (i.e., intense influence, moderate influence, and mild influence) based on the obtained ? , which well reflected the influence degree of the excavated tunnel face on the same monitoring section at different positions. Through the analysis of the influence of the buildings and isolation piles on the ground surface settlement curve, it can be found that the building and its adjacent ground surface present the characteristics of cooperative deformation and joint bearing. Moreover, installing geological drill isolation piles on the side of the tunnel can reduce the ground surface settlement of that side up to 71.9%. The research results have a certain guiding and reference significance for the on-site construction of the Central Yunnan Water Diversion Project and similar projects.

In order to study the deformation and failure laws and stability control of surrounding rock under the influence of repeated mining in double roadway layout working face, numerical simulation, theoretical analysis and field observation were conducted to analyze the influence of double roadway spacing on the spatial distribution characteristics and deviatoric stress of plastic zone under repeated mining. The induced expansion mechanism of secondary mining on plastic zone of retaining roadway was revealed, and the second expansion inhibition method of rock plastic zone was proposed. The results show that the plastic zone of surrounding rock in repeated mining roadway can be divided into six areas in the axial direction of roadway, the larger the distance between two roadways is, the smaller the damage depth of plastic zone is, and its shape changes from asymmetric distribution to symmetric distribution. The main reason is that the stress rotation degree and the peak value of deviator stress caused by the increase of distance between two roadways are low. One mining hysteresis affects the longest maintenance distances and cycles in the stability zone and it is the basis of superimposed expansion of plastic zone in secondary mining, and it is the key control area of surrounding rock stability in repeated mining roadway. On the basis of this, the purpose of improving the stress environment and adjusting the failure state of surrounding rock of roadway can be reached by adjusting the spatial position of roadway. Besides, the control method and control technology system of regional reinforcement and support were established. The field measurement results show that the technical system can achieve the goal of roadway safety and stability, and the effect is satisfactory.

Domestic geotechnical engineers paid more attention to the drained stability of high filled ground; however, the undrained stability brought by rapid constructions was more of a problem for ground with a high groundwater level and poor drainage performances. The relevant standards existing in current specifications and the common methods about undrained stability analysis of slopes were reviewed. By analyzing the relationship between total stress friction angles and stress paths and by calculating the typical total stress paths of original ground in high-fill projects, an explanation was given why the undrained stability of high filled ground was overestimated when strength indexes of total stress from consolidated-undrained (CU) triaxial tests were applied to undrained analysis, which brought theoretical defects and engineering hidden dangers. Taking a simple hypothetical slope model and a real case of a proposed high-fill airport as example, the factors of safety (FOS) of the undrained stability were calculated through five undrained analysis parameters or models using simplified Bishop method respectively. The results show that the undrained stability of high filled ground is overestimated by the CU total stress method all the time; the FOS of undrained stability by other four methods are close, which verifies the relative applicability of these methods.

The piezocone penetration test (CPTU) is widely applied in soil layer division and soil classification. However, its various soil classification methods have certain applicability and limitations. Five common classification charts are analyzed based on two groups of practical engineering cases. The results show that the existing classification chart can well distinguish cohesive soil and non-cohesive soil, but there are certain errors in the discrimination between silty clay, silty and silty sand. Besides, most classification charts adopt the transitional zone method, which will underestimate the classification accuracy. In this study, according to the actual test data classification results, the soil behavior type index Ic is modified through the soil behavior type index classification chart by Robertson. As a result, a new method of CPTU based modified soil behavior type index is proposed by modifying the concentric circle model to a parallel line model. The application results show that the proposed classification method can well satisfy the classification of practical engineering soil and the discrimination between silty clay, silty soil and silty sand, respectively.

To study the process and mechanism of rock tensile failure, the advantage of weighted inversion method in solving moment tensor of Brazilian splitting test is first verified by simulation calculation. Acoustic emission monitoring, event location and moment tensor inversion are performed during the Brazilian splitting test on granite, and the difference of moment tensor results between standard inversion method and weighted inversion method is analyzed. Then the moment tensors inverted by the weighted inversion method are analyzed by the K-means clustering algorithm. The results show that the weighted inversion method can reduce the moment tensor inversion error of Brazilian splitting test. In the laboratory granite Brazilian splitting test, the weighted inversion method greatly optimizes the distribution of the double couple component and pressure/tension axis of the acoustic emission event, which makes the inversion result more reasonable. For the weighted inversion result, the events selected for moment tensor inversion can be divided into three clusters, and the source type, stress state, focal mechanism, radiation pattern of events in the same cluster are similar. There is no significant difference in the occurrence time and position of events in different clusters, but in the later stage of the test, events with obvious tension characteristics in the east-west direction increase sharply and gather near the center of the disc and then dominate the whole rupture, which eventually leads to the split of the specimen. The results reasonably explain the failure process and mechanism of the Brazilian splitting test on granite, and can provide further guidance for the study of rock mechanical behavior.

As a kind of rolling stone with large volume and mass, collapse, instability and long-distance movement with high-speed and high-energy for a rock boulder often lead to destructive disasters of the buildings and traffic lines along the runout pathway. The K4580 typical landslide along the G318 national road in Tibet Autonomous Region was taken as the engineering background, the characteristics and phenomena of the entire process of the collapse, instability and movement of the rock boulder were studied using 3D discontinuous deformation analysis (3D-DDA) method. The 3D-DDA numerical models of the rock boulder collapse were built for three slope shapes: slope shape without landslide, slope shape with shallow landslide, and slop shape after deep landslide. The accuracy of the 3D-DDA simulation results for the boulder movement was then verified based on an empirical model of transverse offset for the block movement. In addition, the instability mechanism of the boulder collapse was investigated; the movement trajectory and kinetic energy evolution were analyzed for the three slope shapes after failure. The results show that the 3D-DDA has the capacity to effectively simulate the whole dynamic process of the boulder collapse and instability, movement development, violent impact, collision, and terminate stages. The boulder collapse presents an instability mode that including the mode transformation of sliding→toppling-sliding→toppling→overturning-falling. The rock boulder movement is manifested by various movement forms such as collision, bouncing, flying, rolling and sliding, as well as 3D spatial movement characteristics such as transverse offset and lateral deflection. The boulder disaster may be induced once the rock boulder passing through the road and colliding with the viaduct. Under different slope geometric shapes from no landslide, to shallow landslide, and to deep landslide, the main factors all decrease as the slope shape changes such as the movement deviation, bouncing height, colliding time to the slope bottom, and final stability time of the boulder. Through the 3D-DDA analysis of the collapse and movement of the boulder, the whole movement process, influence range, impact energy, and terminate location of the boulder can be then predicted, which can provide a reference for the boulder disaster prevention and mitigation and measurement policy.

To study the influence of the movement mode of retaining wall and the width of fill on the active earth pressure of cohesionless finite soil, the discrete element simulation of translation (T) mode, rotating about the base (RB) mode and rotating about the top (RT) mode of retaining wall were carried out under different widths of fill. According to the discrete element simulation results, the active earth pressure, the failure mode of the soil behind the wall and the stress state were analyzed. The results indicate that the failure mode and stress state of soil are transferred by the change of retaining wall movement mode and the widths of fill, which lead to the variation of active earth pressure distribution. The values of mobilized internal friction angle in the sliding wedge of T and RB modes will increase relative to the initial value, and small principal stress arches will appear in the sliding soil wedges of T mode. The RT mode is more special, when the width of fill is small, the stress state is similar to the T mode; when the width of fill is large, in the upper part of the sliding wedge, there will be a region in which the internal friction angle is reduced relative to the initial value, and a large principal stress arch will appear.

Generally, reliability analyscs of landslide stability under rainfall infiltration only consider the spatial variability of calculated parameters of soil materials, while the influence of non-uniform distribution of initial water is ignored. To solve this problem, the governing equation for Green-Ampt infiltration model with arbitrary distribution of initial water content under different rainfall conditions is derived, and a variable-step compound Simpson method is used to solve the governing equation for exploring the functional relation between wetting front depth and rainfall duration with different spatial distribution of initial water content. Then, the time-varying probability of landslide during the process of rainfall infiltration is calculated using multidimensional normal cumulative distribution function (Mvncdf). Finally, an actual case is used for verification through comparative analysis. The results show that different spatial distribution of initial water content has a significant impact on the time-varying reliability of landslide stability under rainfall conditions. Compared with the uniform and trapezoidal distribution of initial water content, the exponential distribution has the smallest decrease rate of time-varying reliability for landslide stability under the same rainfall condition. The improved model is feasible for arbitrary distribution of initial water content of soil, which is conducive to the universal application of Green-Ampt model in landslide stability evaluation.

The impact of rockfall on the shed cave structure is complicated and lacks a unified expression of rockfall impact force. First of all, the falling rock is simplified into a rigid sphere, the theoretical expression of the rockfall impact force is derived using half-sine algorithm based on the Hertz contact theory. Considering the inelastic characteristics of rockfall impact on shed cave, the method of function fitting is used to calculate the impact force of rockfall under normal impact according to the characteristics of the rockfall acceleration curve during the collision between the rock and the material, and then a numerical calculation model of the rockfall impact on shed cave is established to study the dynamic characteristics of the rockfall impact on shed cave at different impact speeds based on ANSYS/LS-DYNA software. Finally, compared with the existing common methods, some conclusions are drawn. The impact force of rockfall obtained by Hertz half sine method is much larger than that obtained by function fitting method and numerical method, and the time-history relationship of rockfall impact force obtained by function fitting method in this paper is close to that obtained by numerical method. It indicats that function fitting method can better reflect the dynamic relationship of the collision between the falling rock and the shed cave. Compared with other calculation methods, it can be obtained that the Hertz algorithm is suitable for analyzing the elastic collision problem without energy loss, and the Logistic algorithm is suitable for the situation of large plastic deformation of the material. There are differences between the elastoplastic contact theory results and the dynamic finite element results, and the maximum impact force and the impact time of the rockfall obtained by the calculation method derived from the function fitting are closer to the dynamic finite element method, and it can better reflect the dynamic response characteristics of the rockfall impact on the shed. The calculation method of rockfall impact force deduced in this paper can provide theoretical reference for the design of protection measures in practical engineering.

In order to study the synthesis and decomposition characteristics of natural gas hydrates, the exploitation performance of the reservoir, and the sand production and control during the mining process of natural gas hydrates, a set of multi-dimensional natural gas hydrate production simulation test system has been developed. The test system is mainly composed of gas injection system, constant pressure liquid supply system, steam or hot water injection system, 1D,2D&3D model, back pressure control system, outlet metering system, data acquisition and control processing system and corresponding auxiliary systems. The main innovations of the experimental system are as follows: 1) The one-dimensional reactor wall made of high-pressure glass and the high-definition camera are combined to achieve visualization for monitoring the synthesis and decomposition of hydrate and the sand migration in the production process under medium and low pressure; 2) The self-developed double-cylinder constant speed and pressure pump is used to control the back pressure valve, so that the pressure in the reactor can be uniformly reduced at a decrement of higher precision during the entire depressurization production process. Consequently, the gas hydrate reformation is greatly reduced, and the gas production rate becomes more uniform. Using quartz sand of different particle sizes, deionized water and methane gas with a purity of 99.9% as raw materials，the one-dimensional phase equilibrium test and two-dimensional depressurization production test are conducted. The reliability of the test system is verified based on the primary application results from these tests.

To improve the liquefaction resistance of calcareous sand foundations, microbially induced calcium carbonate precipitation (MICP) technology combined with fiber reinforcement technology was proposed to treat the calcareous sand in the South China Sea. Based on dynamic triaxial tests, the dynamic behaviors of MICP and fiber-treated calcareous sand were studied. The dynamic strain, dynamic pore pressure, cyclic stress-strain response and dynamic elastic modulus were analyzed. Then, the strengthening mechanism of MICP and fiber on the mechanical properties of the treated calcareous sand was explored from the microscopic point of view, based on the scanning electron microscope (SEM) test results. The results show that: (i) MICP could improve the deformation resistance and liquefaction resistance of calcareous sands. Compared with the untreated calcareous sand samples, the dynamic strain and dynamic pore pressure of calcareous sand treated by MICP decreased by 95.74% and 92.46%, respectively. (ii) The addition of fibers further improved the reinforcement effect of MICP. Compared to the MICP-treated samples, the dynamic strain and dynamic pore pressure of MICP and fiber-treated samples decreased by 74.32% and 74.18%, respectively. (iii) MICP and fiber reinforcement technologies improved the deformation resistance and liquefaction resistance of calcareous sand subjected to cyclic loading by reducing the cyclic activity strength and energy dissipation, increasing the dynamic elastic modulus and reducing the decay rate of the dynamic elastic modulus. (iv) The results of the SEM test showed that MICP and fiber reinforcement had a synergistic effect on the improvement of the mechanical properties of calcareous sands. The incorporation of fibers provided more spots for bacterial adhesion and promoted the formation of calcium carbonate crystals, which not only increased the bonding strength between sand particles, but also enhanced the restraint of fiber nets by fixing fibers and sand particles together.

To further explore the fracture initiation mechanism of fracture grouting in typical sandy mudstone from Huainan and Huaibei mining areas in China, a conventional triaxial fracture grouting test device was developed, and the model test of fracture initiation pressure of slurry fracturing in similar material of sandy mudstone was carried out. Based on the test results, the influences of rock strength and stress state on grouting fracture initiation pressure and fracture propagation pattern were analyzed, and the fracture initiation mechanism of fracture grouting in sandy mudstone was revealed. The results show that there is a positive correlation between the initiation pressure and the compressive strength of rock; the larger the compressive strength of the rock is, the more complex the fracturing path is. The sensitivity of fracture initiation pressure to confining pressure is much greater than that of axial pressure; the larger the stress difference Δσ = σ _V −σ _H is, the more regular the fracture shape is. Under the triaxial condition of pore pressure, the rock tensile strength determined by slurry fracturing method in sealed open hole section is approximately 2.5 times the uniaxial tensile strength. The research results can provide a reference for the design and construction of fracture grouting in similar rock strata in the future.

 A series of triaxial tests was carried out under coupled cyclic-seepage loads for Tianjin coastal soft clay, and the variation of permeability in different loading stages was analyzed systematically. Combined with SEM and MIP, the evolution mechanism of permeability was elaborated. The results showed that under the coupling cyclic-seepage loads, the permeability of soft clay present three variation stages. With seepage loads increasing, the permeability increased initially and then decreased, but just reverse for increase of critical dynamic stress ratio. The evolution of permeability was caused by the adjustment of micro-features of pore shape, size, and distribution: for initial vibration, large compression of super large pores (D>2.5 μm) between particles occurred and pore shape varied little, leading to the linear reduction of permeability; at the medium loading stage, the large particles were broken, and the newly formed small particles were densely filled between the large particles, inducing a large number of “compact and zigzag” small pores (0.05 μm <D <0.1 μm), which led to the slow decrease of permeability; at later loading stage, the structure was compact and stable, and permeability was basically unchanged. For the unstable deformation, the “strips” like macropores were re-formed between the particle clusters (aggregates), and the permeability increased. The results can provide a theoretical basis for the determination of the permeability in the fluid-solid coupling analysis.