Tunnel roof collapse is a progressive failure process. In order to study the progressive collapse characteristics of deep tunnel roof with weak surrounding rock, we establish the three-dimensional progressive collapse mechanism of deep tunnel based on the limit analysis upper bound theorem and nonlinear Hoek-Brown failure criterion, derive the analytical solution of collapse surface in the whole process considering pore water pressure, and draw the three-dimensional surface diagram of roof progressive collapse. Furthermore, we analyze the morphological characteristics of the collapsed body when the relevant parameters are varied singularly, and the influence of each parameter on the collapsed body gravity and tunnel support force under different pore water pressures. The results show that the dimensionless parameters characterizing the rock mass, unit weight, pore water pressure and tensile stress have significant effects on the morphology, gravity and support force of progressive collapse. In the process of progressive collapse of deep tunnel, the physical and mechanical parameters of surrounding rock gradually weaken, which is mainly reflected in the gradual attenuation of the rock strength with the development of deformation up to a residual value. The attenuation of the strength of surrounding rock and the residual strength have a certain influence on the collapse gravity and tunnel support force. The theoretically calculated collapse shape of tunnel roof is basically consistent with the collapse range of tunnel roof in the F₃ fault fracture zone of actual tunnel engineering, which verifies the applicability of the theoretical results for predicting the collapse range of tunnel roof progressive collapse. The research results can provide a theoretical basis for the construction design and safety protection of deep tunnel with weak surrounding rock.

To study the dynamic response characteristics of soil under impact load, flat dynamic load tests (FDLT) with different load levels and loading rates were carried out on two typical soils (sand and clay) in Guangzhou University Town by using the self-developed additional excitation force-type FDLT system, obtaining three characteristic curves of load-time curve, displacement-time curve, and load-displacement curve, establishing the empirical formula of dynamic deformation modulus and loading rate, and comparing the effects of loading rate of the two soils under impact load. It shows that: (1) There is a threshold of charging voltage in the FDLT test of sand. During the impact test on this threshold, the loading rate has a great effect on the dynamic strength and deformation of sand, while it has no obvious effect on clay under the same conditions. (2) The same point of the loading rate on the displacement response of clay and sand is that the maximum load and maximum displacement of the two soils both increase with the increase of loading rate. The difference is that the time for clay to reach the peak displacement increases with the increase of loading rate while the time for sand to reach the peak displacement shows a pattern of increasing and then decreasing with the increase of loading rate. (3) The dynamic deformation modulus varies logarithmically with the loading rate. The results can provide a reference for soil dynamic characteristics study, dynamic and static parameter conversion and engineering design.

The abundance of fissures is one of the important characteristics for expansive soils. The inherent fissures and fissure propagation during the loading process have a significant impact on the mechanical behavior of the soil. In order to investigate the influence of fissures on the deformation and failure mode of expansive soil, with the help of an improved true triaxial instrument, the undisturbed expansive soils with different inclination angles of inherent fissures (type I - the long axis of the sample is perpendicular to the dominant fissure direction, type II - the long axis of the sample is 45º oblique to the dominant fissure direction, and type III - the long axis of the sample is parallel to the dominant fissure direction) were used to conduct the consolidated drained plane-strain shear test. Meanwhile, the deformation was analyzed using digital image correlation (DIC) technology, focusing on the local deformation characteristics controlled by fissures. The results show that under the same confining pressure, the peak stress of the type-II fissured soil sample is the smallest, the stress−strain curve is of strain softening type, and its failure type is sliding failure. The peak stress of the type-I fissured soil sample is the largest, and the stress−strain curve is strain hardening type. The failure type of the type-I fissured sample is compression-shear failure, and the type-III fissured sample presents different failure forms due to different confining pressures. Under the same confining pressure, the inclination angle of the shear zone produced in the type-I fissured samples is smaller than that of other fissure types and basically does not change with the confining pressure. The shear zone of the type-II fissured sample develops along the original fracture surface, and its dip angle has no obvious regularity with the confining pressure. The confining pressure affects the number of shear zones developed in the type-II fissured samples. For type-III fissured samples, the confining pressure affects the development type of the shear zone, and as the confining pressure increases, the dip angle of the main shear zone decreases. Based on Roscoe’s theory, the inclination angle of the shear zone in the samples with different fissure directions is more in line with the inclination angle of the shear zone when the sample is actually damaged. This test lays a foundation for studying the anisotropic mechanical properties of fissured expansive soil.

In this study, the variation characteristics of normal stress and normal closure of granite joints were studied. The effect of chemical solution with different pH values (1, 3, 7, 12), the corrosion time t (10, 30, 100 d) and various initial openings b on the evolution of the specific normal stiffness of the specimens were revealed through normal load compression tests taking into account constraint circumferential stiffness boundary conditions. The experimental results indicate that the normal stress-normal strain curves of specimens with mismatched joints are different from the intact specimens and specimens with matched joints. Due to the crushing of the asperity on the surface of the joints, the local stress drop phenomenon occurs during the overall growth process of the stress−strain curves. The normal closing curve of mismatched joints exhibit nonlinear characteristics, in which the increasing rate of normal displacement first increases and then decreases with the increment of normal stress. Under the same normal stress, both the normal displacement and the closing rate increase with the increment of the initial opening. By introducing several classical joint closure models, it is found that the normal stress and stiffness of joints are affected by the initial opening, pH value and corrosion days, among which the initial opening is the most import. When pH = 1 and t = 10 d, with an increase of normal displacement from 1−3 mm, the normal stress of the specimens with b = 1.65 mm increase by 46.26 %−149.46 %, compared with the specimens with b = 3.4 mm. The specific normal stiffness of the mismatched joints decreases with the reduction of pH value. The normal stiffness of joints in different chemical solutions decreases first and increases then as the corrosion time increases.

Capillary action has an important impact on the deformation and stability of various geotechnical engineering structures. It is challenging to accurately monitor the dynamic changes in the capillary water rising process by traditional methods. In this study, the relationship between RGB (red, green, blue) information and water content of aeolian sand images was firstly investigated. The methodology and technology for measuring water content and obtaining centimeter-level resolution moisture fields via image RGB information were developed. According to this method, the spatial-temporal variation of water content during capillary water rising in aeolian sand was further explored, and the wetting front profile was precisely distinguished. Furthermore, the wetting front transport law was analyzed. The results show that a negative linear relationship exists between the normalized color feature ξRnor and the water content θ of aeolian sand, which brings about a good prediction for the water content of aeolian sand. The proposed technology for measuring the water content field based on image RGB information, with centimeter-level resolution and high accuracy, can visualize the spatiotemporal changes of the water content during the rising of capillary water in the aeolian sand column. An improved k-means clustering segmentation method was adopted to analyze the sand images of the capillary water rising process, allowing precise identification and quantification of the profile information of the wetting front. This method is more accurate and reliable than the visual inspection method. Under the action of surface tension and inertial force, the water content at each height of the aeolian sand in the initial capillary water rising period has obvious overshoot phenomena. After several small fluctuations, it falls back, and the stable water content is about 1.5% lower than the peak value. The rising height of the capillary water is proportional to relative density D_r. The rising process of the aeolian sand wetting front can be well fitted both by the power and double logarithmic quadratic polynomial functions. The research provides an accurate and rapid novel approach for the capillary water rise test.

Expansive soils, known for their considerable swelling pressure upon wetting, have been identified as potential instigators of instabilities in retaining walls. The incorporation of expanded polystyrene geofoam (EPS) inclusions between the retaining wall and the backfilled expansive soil has been found to considerably mitigate the lateral pressure on the wall, which results from the water absorption and expansion of the expansive soil. This substantial reduction is due to the impressive compressibility of the EPS inclusion. To explore the implications of the EPS inclusion on the lateral pressure distribution on retaining walls, and to analyze the factors influencing this pressure, a comprehensive model test and a corresponding lateral pressure theoretical analysis were performed. The results show that (1) the total lateral pressure acting on the retaining wall was reduced by about 50% by the EPS inclusion with a density of 12 kg/m³ when the expansive soil is saturated in the model test; (2) in the absence of the EPS inclusion, the lateral pressure distribution acting on the retaining wall escalated along its depth, whereas with the EPS inclusion, it remained largely uniform throughout the wall’s depth; and (3) the lateral pressure reduction due to the EPS inclusion was enhanced with increasing thickness and decreasing density of the EPS inclusion.

The strength parameters of fault gouge is significant for the fault sealing of underground gas storage. In this paper, direct shear tests are performed for fault gouge composed of different contents of kaolinite(K), montmorillonite(M) and quartz(Q), thus obtaining the shear strength parameters (cohesion c and internal friction angle φ ) of various types of fault gouge. The results reveal that: (1) the cohesion of fault gouge is positively correlated with the content of montmorillonite and negatively correlated with the content of quartz. The cohesion of fault gouge specimens with high content of montmorillonite (≥40%) is positively correlated with the content of kaolinite; (2) The internal friction angle of fault gouge is negatively correlated with the content of montmorillonite and positively correlated with the content of kaolinite; (3) The cohesion of fault gouge is positively correlated with the plasticity index I_P, while the internal friction angle is negatively correlated with the plasticity index I_P; (4) The predicted shear strength parameters of fault gouge obtained by statistical methods match well with the experimental results, indicating that statistical method is feasible in predicting the shear strength of fault gouge.

Loose sand is highly susceptible to liquefaction, and small changes in stress state can affect its liquefaction characteristics. Based on multi-directional cyclic simple shear tests, this study conducted cyclic simple shear tests on loose sand under different magnitudes and directions of static shear stress, and complex shear paths. The cyclic simple shear characteristics of loose sand under complex initial stress states are studied. The main conclusions are drawn as follows: (1) As the static shear stress ratio increases, the peak shear stress of the specimen increases, the increment of pore water pressure in the first cycle increases, and the specimen is more prone to liquefaction. The effect of the magnitude of initial static shear stress on excess pore water pressure is more significant at the early stage of shearing. (2) With the increase of the angle between the initial static shear stress and the main direction of dynamic shear stress, the peak shear stress of the specimen in the X direction decreases, and the pore water pressure of the specimen accelerates to increase. In addition, the increment in pore water pressure in the first cycle and the last cycle increases, and the difference between the cycles increases. The specimen is more prone to sudden liquefaction. (3) The specimen with 8-shaped shear path has the largest area of stress−strain hysteresis loops, which consumes the most energy per cycle, followed by the specimen with the circular shear path, and the specimen with the straight shear path has the smallest area. Complex shear paths can induce a sudden increase in pore water pressure at the beginning of shearing, increasing the increment in pore water pressure in each cycle and making it more prone to liquefaction. (4) The sequence of factors affecting the liquefaction of loose sand is the angle between the initial static shear stress and the dynamic shear stress, the shear path, and the magnitude of the initial static shear stress.

In this study, the “Xiaguiwa” landslide in Jinsha River Basin of Qinghai-Tibet Plateau is taken as a prototype, and the shaking table model test on bedding rock slope with weak interlayer is carried out. The dynamic response of bedding rock slope with weak interlayer under earthquake is studied from the aspects of peak ground acceleration (PGA) amplification factor and Hibert-Huang transform (HHT) time-frequency characteristics. The results show that the slope exhibits obvious “elevation effect” and “surface effect” under the action of input seismic waves. The PGA is larger at the 1/4 height of the slope surface from the bottom of the slope, the top of the slope, and the weak interlayer. With the increase of the intensity of the input seismic waves, the slope stiffness and natural vibration frequency decrease gradually. When the input wave amplitude reaches 0.7g, the slope cracking and structural deformation occur. When the input amplitudes are the same, the PGA amplification coefficient is positively correlated with the elevation, and decreases gradually with the increase of the input amplitudes at the same measuring point. The influences of different input wave types and time scale factors on the slope dynamic response are significantly different. The Hilbert spectrum shows that the elevation and weak interlayer amplify the energy of seismic waves, especially the high-frequency part. The Hilbert marginal spectrum shows that the weak interlayer could amplify the energy of the high-frequency part. The Hilbert marginal spectrum indicates that the cumulative energy of the high-frequency part is significantly amplified under the influence of soft interlayer, and the energy of the measuring point at the 1/4 height of the slope surface from the bottom of the slope suddenly increases, which is similar to the conclusion of the acceleration amplification effect. The results of Hilbert marginal spectrum shows that with the increase of the amplitude of the input seismic wave, the cumulative energy of the high-frequency part and the part representing the natural vibration frequency of the slope gradually decrease, and the energy of the main frequency part of the input seismic wave gradually dominates, indicating that the modal characteristics of the slope gradually disappear.

In situ mining organic-rich shale by heat injection is a complex solid fluid thermochemical coupling process. The stope and wellbore are subjected to shear stress during the orebody pyrolysis. However, the mechanical response of shale after temperature action is completely different from that under real-time high temperature action. In order to study the shear mechanical properties and deformation evolution law of shale under real-time high temperature, a real-time high-temperature rock variable angle shear test system is designed. By combining acoustic emission and digital image correlation technology, the shear strength and deformation field distribution characteristics of shale under different temperatures and shear angles are thoroughly investigated. The results show that: (1) With the increase of temperature, shale shows a transformation from brittle failure to ductile failure, and the shear strength decreases as the shear angle increases. (2) The shear strength of shale changes in a “V” shape with the increase of temperature. The shear strength of shale decreases to the lowest (2.93 MPa) at 400 ℃, which can be regarded as the threshold temperature of shale shear properties. When the temperature exceeds 400 ℃, the shear strength of shale continues to increase due to the transformation of mineral lattice inside. (3) The evolution of the strain field of shale in the shear process at room temperature −600 ℃ can be divided into two stages. At room temperature−400 ℃, the shale will form an obvious strain localization zone along the bedding structure during shear failure and undergo direct shear failure along the localization zone. In the range of 400−600 ℃, the shale shows obvious strain softening characteristics, and mutual “stagger shear” occurs along the bedding, instead of shear failure directly along the bedding structure, it shows progressive shear failure characteristics.

In order to study the cyclic shear characteristics of soil-rock mixture-geogrid interface under different gravel contents, monotonic direct shear tests, cyclic direct shear tests and post-cyclic monotonic direct shear tests were carried out under different vertical stresses, shear displacement amplitudes and shear frequencies by using cyclic direct shear apparatus. The effects of five gravel contents (0%, 25%, 50%, 75%, 100%), three shear displacement amplitudes (1, 3, 6 mm) and three shear frequencies (0.2, 1.0, 2.0 Hz) on the peak shear stress and volume change of soil-rock mixture-grid interface were investigated. The test results show that when the gravel content is between 0% and 75%, with the increase of gravel content, the cyclic peak shear stress and maximum shear stiffness of the soil-rock mixture-grid interface increase, and the absolute value of final vertical displacement and the maximum damping ratio decrease. When the gravel content continues to increase to 100%, the maximum peak shear stress and maximum shear stiffness of the soil-rock mixture-grid interface decrease, and the absolute value of final vertical displacement and the maximum damping ratio increase. Under the same gravel content, a great shear displacement amplitude and a high shear frequency lead to a high maximum peak shear stress and the absolute value of final vertical displacement of the interface. Compared with the monotonic direct shear test, the internal friction angle and quasi-cohesive force obtained by the monotonic direct shear test after the cycle increase. In the range of 0%−75% gravel content, the interface friction angle and quasi-cohesive force increase linearly with the increase of gravel content. When the gravel content continues to increase to 100%, the interface friction angle and quasi-cohesive force decrease.

The swelling deformation and swelling force of bentonite in contact with water are caused by the repulsive force between particles, and the magnitude of this repulsive force is related to the fixed negative charge density of the soil (represented by the cation exchange capacity, CEC) and the concentration of the pore solution. The previous research on the swelling deformation of bentonite is mainly focused on the pore solution rather than the cation exchange capacity of the soil. In this paper, sodium bentonite is taken as the research object, and Li⁺ fixation method is used to control the fixed negative charge density of bentonite so as to reduce the cation exchange capacity of bentonite. Finally, the free swelling rates of reduced charge bentonite are obtained under different concentrations of LiCl solution. The results show that the free swelling rate of soil samples decreases with the increase of solution concentration; when the concentration is less than 1 mol/L, the decrease is more significant, and then slowly decreases; the free swelling rate of soil samples also decreased as CEC decreases. The physicochemical interaction between the soil particles and pore solution depends on CEC of soil particles and salt solution concentration. Lower CEC and higher salt concentration mean lower repulsive force between soil particles and lower free swelling rate of soil sample. The intergranular stress with the consideration of the physicochemical interaction between particles and pore solution is introduced to describe the experimental results, showing that the swelling deformation and swelling stress under different CECs and different salt solutions are merged on the same curve, which further verifies the effectiveness and applicability of the intergranular stress in simulating expansive soil.

After the impoundment of a high dam reservoir, the water pressure environment of the rock masses in dam base and reservoir bank changes, which may easily induce engineering problems such as bank slope instability and dam collapse. In order to investigate the influences of different constant water pressures on the rock mass of dam base, triaxial compression tests were conducted on sandstone with initial damage under different high constant porewater pressures, and the multidirectional fracture mechanism was analyzed by combining computed tomography (CT) and scanning electron microscopy (SEM). The test results show that: (1) Under the confining pressure of 80 MPa, the greater the pore water pressure, the more brittle the sandstone, the lower the peak strength, and the smaller the volume expansion stress. The pore water pressure increases from 10 MPa to 50 MPa, and the peak strength decreases by 33%. (2) Under different pore water pressures, there are significant differences in sandstone internal deterioration range and deterioration effect as the fracture surfaces of sandstone specimens have various forms and directions. Due to CT scaning results, with the increase in pore water pressure, the deterioration effect spreads from the middle of the specimen to both ends. When the ratio of the water pressure to the confining pressure is less than 25.0%, the deterioration of pore water pressure is mainly concentrated in the middle of the specimen. When the ratio of the water pressure to the confining pressure is larger than 62.5%, the pore water pressure has obvious deterioration effect on the whole specimen. (3) The SEM test reveals that with the increase in pore water pressure, the microgranular structure of sandstone changes from shear slip failure to shear fracture failure, and the microcrystalline structure of sandstone changes from cauliflower to rice granules. The macroscopic failure mode changes from plastic failure to brittle failure, and the multidirectional fracture plane is formed, which is related to the migration of fine particles and the fracture of large particles in the meso-particle structure under pore water pressure. The formation of the multidirectional fracture plane is directly related to the uneven accumulation of the microscopic crystal structure.

Rock hardness is an important indicator reflecting the relative hardness, drillability, and blastability of rock formation. Accurate interpretation of the response of percussion drilling tools can provide new possibilities for in-situ, rapid, and quantitative evaluation of rock hardness. High-precision digital displacement, hydraulic pressure, and rotational speed sensors were used to monitor the key transmission parts of the percussion drill, and a digital monitoring system for the drilling process was then established. Rock hardness measurement and orthogonal tests of percussion drilling were parallelly carried out, and a standard database containing response data of the drill when drilling various types of rocks was established. Functional relationships between propelling pressure, percussion pressure, rotational speed, and impact energy dissipation were developed, and a novel method to quantitatively evaluate rock hardness based on impact energy dissipation index was proposed. The performed tests and analyses revealed that there is an inverse linear relationship between the propelling pressure and impact energy dissipation. The fitting curve of percussion pressure and drilling energy dissipation is an open upward parabola, and the curves for various rocks have the same symmetry axis but different curvature radii. The influence of rotational speed of drill rod on drilling energy dissipation is negligible. By dimensionless treatment of drilling energy dissipation per unit volume, the influence of mechanical parameters of percussion drilling is removed, and the impact energy dissipation index , which has low discreteness and a good correlation with rock hardness, was defined to characterize the rock hardness. The response data of the drilling process in conventional boreholes were obtained and interpreted by the digital sensing technology, and the obtained data were used to determine the rock hardness without additional survey or test, which puts forward a new way for direct evaluation of rock mass parameters on engineering sites.

The dam anti-seepage body with cohesive soil as the main source often has perforated cracks due to extreme temperature difference, uneven settlement and other reasons, which eventually induces seepage failure. Hydraulic filling and clogging is an economical and effective repair method. In order to understand its repair mechanism, this paper takes the remolded cohesive soil as the research object, prefabricates cracks in the soil to simulate the cracking of the impermeable body, and studies the effects of muddy water concentration, water head, seepage direction (i.e. horizontal, vertical, oblique), non-uniformity coefficient of the filter layer and other factors on crack repair. The motion state of the fluid in the crack during crack repair was determined, and the evolution law of the crack under the action of muddy water seepage was studied based on digital image processing technology. The results show that under the action of muddy water seepage, the crack repair process in cohesive soil can be divided into three stages: transition period (stage 1), repair period (stage 2) and stable period (stage 3). The crack repair process will affect the seepage characteristics of the fluid in the crack. The fluid motion law at stages 1 and 2 satisfies the Forchheimer flow, and the fluid motion law at stage 3 conforms to the Darcy flow. The starting point of stage 3 is the critical point of flow transition, and the stage of fracture repair can be judged according to the change of flow state in the crack. Muddy water concentration, water head, crack type and uneven coefficient of filter layer are the main influencing factors of crack repair. The time required for crack repair is the shortest under high muddy water concentration and low water head. The seepage quantity Q and flow velocity ν after the crack repair are significantly lower than those before the repair, which are reduced by 99.83% and 99.98%, respectively, while the hydraulic gradient J is significantly increased by 27.92 times, and the impermeability performance of the soil after the repair is significantly enhanced. The research has certain theoretical guidance value for the evolution mechanism and prevention measures of the cracks in the dam anti-seepage body.

The implanted drainage pile can eliminate the excess static pore pressure in the adjacent soil layer and improve the soil strength on the side of the pile. It is a new type of pile foundation technology suitable for deep soft soil ground in coastal area. In order to study the dissipation law of excess pore water pressure and the development of consolidation degree in the adjacent soil layer, the cement-soil is regarded as the smearing zone, and based on the assumption of elliptic column drainage body and the coordinate transformation method, an analytical solution of consolidation of implantable drainage pile under vacuum negative pressure was derived. The reliability of the theoretical solution was verified by model tests, and then an example analysis was carried out. It is found that the vacuum negative pressure transfer is mainly concentrated near the vertical drain, and the plane distribution of excess pore pressure in the soil layer shows a shape of “8”. The duration of vacuum negative pressure and permeability of cement-soil have significant influence on the reinforcement effect of drainage pile. The permeability of cement-soil decreases rapidly under the action of drainage consolidation and hardening, and thus the duration of vacuum drainage should be controlled within 5 days. The research results lay the foundation for the further development and application of the implanted drainage pile technology.

The influence of cyclic loading frequency f on the liquefaction behaviors of saturated quartz sands during undrained cyclic laboratory tests has been investigated extensively, but rare studies reported on the effect of f on the liquefaction behaviors of saturated coral sands. To better understand the effects of the f on the liquefaction behaviors of saturated coral sand, a series of undrained cyclic shear tests is performed on isotropically consolidated, medium dense saturated specimens subjected to the 90º jump of principal stress in the range of f = 0.01−1.0 Hz. For all loading patterns investigated, the growth rates of the excess pore water pressure ratio ru and the generalized shear strain γ_g decrease with the increase of f; the number of liquefaction NL required to cause the initial liquefaction increases with f; a unique tangent correlation function exists between the generalized shear strain amplitude γ_ga and the peak excess pore pressure ratio rumax (r_umax is independent variable); and an increase in f weakens the dilatancy of the specimen. By introducing a unit cyclic stress ratio USR as a proxy for cyclic stress level acting on the specimen during the cyclic testing, USR versus N_L curves moves to the upper right with f, a unified form of power-law function of the correlation exists between USR and N_L regardless of f. This indicates that liquefaction resistance of saturated coral sands increases with f.

Earthquake-induced liquefaction can cause significant damage to geotechnical structures, and the aftershocks accompanying the earthquakes may cause the sandy soils to liquefy again. To investigate the effect of free-field subsurface seismic history on the liquefaction resistance of saturated sandy soils at various depths, a series of shaking table tests was designed and conducted. Four shaking events with different accelerations, subdivided into seven smaller events, were input to the sandy soil in the tests. The excess pore pressure ratio, acceleration response, and soil settlement were calculated and compared for each vibration event to investigate the variation laws of soil liquefaction resistance at various depths under different seismic histories. The test results show that the magnitude of the input seismic wave acceleration is positively correlated with the acceleration response coefficient. Liquefaction of saturated sandy soils subjected to minor and moderate earthquakes is more likely to occur at shallow rather than superficial layers. In superficial soils, moderate aftershocks after strong earthquakes make the soils less sensitive to seismic intensity, pore water pressure dissipates rapidly after peak acceleration, and liquefaction time is shortened. Soils with strong seismic histories reduce the liquefaction resistance of shallow soils, while the liquefaction resistance of deep soils is enhanced after strong earthquakes, and the affected depth range depends on the earthquake intensity. Quantitative formula of the liquefaction-resistant lifting ratio of soil at different depths under different levels of cyclic earthquakes is obtained through data fitting. The test results can reflect the different effects of different seismic histories on the soils at various depths in the actual earthquake.

Artificial ground freezing technology has the advantages of reliable water sealing, low environmental impact and strong applicability. It is one of the effective methods for underground excavation in urban subway to achieve the waterless construction. The cooling power is continuously depleted by groundwater flow, resulting in the delay of closure of frozen wall and the thickness less than expected. A physical model test with groundwater seepage velocities of 0, 5, 10 and 15 m/d was conducted based on the freezing project of a subway cross passage in Beijing Subway Line. The influence of groundwater seepage velocities on the formation of frozen wall was investigated, and the variation of cooling power in the freezing process was discussed. The results show that under seepage condition, the bottom and top plates of the cross passage intersect first, followed by the downstream sidewall, and finally the upstream sidewall, which is the weak part. At the groundwater seepage velocities of 0, 5 and 10 m/d, the closure times of the frozen wall are 60, 114 and 444 minites, respectively, while it is impossible to form continuous frozen bodies at 15 m/d. The cooling power increases with the groundwater seepage velocities. The energy consumption at the groundwater seepage velocities of 5, 10 and 15 m/d increases by 67.0%, 88.5% and 120.1%, respectively compared with no-seepage condition. The results can provide a basis for the design and construction of cross passage freezing under seepage conditions.

To optimize the width of protective coal pillar of shallowly buried pipeline in the coal mining subsidence area, the gas-coal superposition area in Ordos Basin is taken as the engineering background. The theoretical algorithm of pipeline tensile strain is proposed using the probability integral method. According to the tensile strain limit, the prediction model of the width of the pipeline protective coal pillar in the subsidence area is established and verified by comparison through numerical analysis and engineering application. Then the distribution and evolution of tensile, volumetric and shear strains of adjacent buried pipelines during the advancing process of the panel are analyzed. The results show that the prediction method greatly shortens the width of the pipeline protective coal pillar and improves the recovery rate of coal resources under the premise of ensuring safety. With the decrease in the horizontal distance between the panel and pipeline, the tensile amount and the tensile strain of the whole pipeline grow exponentially. However, the tensile increment shows a trend of increasing first and then decreasing. As the panel becomes closer to the pipeline, the volumetricstrain distribution shows a ‘V’ shape in the axial direction and an ‘M’ shape in the circumferential direction. However, the shear strain distribution of the pipeline shows a ‘W’ shape in the axial direction and ‘’ shape in the circumferential direction. As the panel advances, the volumetric and shear strains are exponential functions of the horizontal spacing of the panel pipeline. The locations where the pipeline is mostly prone to damage are the center and edge of the subsidence area, and similar working conditions require focused attention. The application of our method is conducive to the accurate and coordinated mining of oil, gas and coal resources. Meanwhile, the investigation of multi-parameter mechanical characteristics is helpful to provide decision support for the correction of prediction results and preventive maintenance.

The influence of the vibration wave generated during the operation of the high-speed railway train on the building structure along the line can not be ignored, and the vibration evaluation method suitable for the building structure around the high-speed railway line is of great engineering significance. Firstly, a calculation formula of equivalent particle peak vibration velocity (PPV) is developed based on Sadowski formula and the principle of building structure response under harmonic excitation. Then the field soil monitoring test is carried out to analyze the attenuation characteristics of particle peak vibration velocity of vibration signal; and wavelet packet transform is used to analyze the frequency band energy characteristics of vibration signal so as to theoretically derive the equivalent PPV. Finally, the allowable vibration value of the building under the existing traffic vibration is taken as the evaluation standard to estimate the vibration safety of the building structure around the high-speed railway line. The results show that except for the position with significant local magnification, the attenuation formula of the peak particle velocity of the high-speed railway vibration wave in the pier foundation of the viaduct has a good fitting result to the actual peak particle velocity as a whole. Based on the vibration evaluation method of equivalent PPV, it is found that at 45 m distance from vibration source, the 0−10 Hz equivalent PPV value of G924 train is greater than 1 mm/s, which exceeds the vibration tolerance value of vibration sensitive and protective buildings. Compared with the existing vibration evaluation, the equivalent PPV vibration evaluation method has the advantages of high vibration sensitivity, vibration control for specific building structures, and vibration evaluation before the location of the buildings around the high-speed railway line.

Based on an ultra-deep circular shaft project in Shanghai, the field data of soil heave at the bottom of the pit during the construction were collected, and the vertical distribution pattern, evolution law and main influencing factors of soil heave at the bottom of the pit were summarized. An axisymmetric numerical model was then built and verified to investigate the effect of the excavation-induced unloading, dewatering, diaphragm wall and soil mechanical properties on the soil heave, then the mechanism of soil heave was revealed. Soil heave was the combined result of the soil mechanical response under the excavation-induced unloading, dewatering, and diaphragm wall restraint, in which the excavation-induced unloading and the deflection of the diaphragm wall caused the soil heave, and the dewatering and the negative frictional resistance inhibited the soil heave. Excavation-induced unloading had a prominent influence on the depth, and the unloading rebound mechanism dominated the soil heave within that depth, while the shear deformation controlled the soil heave beyond that depth range. Soil rheology and dissipation of negative pore water pressure jointly led to the time dependence of soil heave. The soil heave at the pit bottom of small-diameter ultra-deep shafts in soft soil areas decreased approximately linearly along the depth, and its maximum value was located at the center of the excavation face. The soil heave first increased slowly and then increased near linearly and rapidly with the increase of excavation depth. However, the soil heave tended to increase slowly with time in the non-excavation stage.

In order to understand the one-dimensional transient transport process of organic contaminants in a triple-layer composite liner composed of geomembrane (GMB), geosynthetic clay liner (GCL) and compacted clay liner (CCL) under the non-isothermal distribution condition, the factors such as advection, diffusion, mechanical dispersion, adsorption, degradation and thermal diffusion are considered, and the corresponding mathematical model is established, in which the variations of parameters such as diffusion coefficient and permeability coefficient with temperature are incorporated. The numerical solution for the proposed model is obtained by using the finite difference method. The correctness of the established model is verified by comparing the calculation results of the model with the experimental data, the calculation results of the existing analytical model and the numerical results of COMSOL software. Based on the defined breakthrough time t_b, toluene is adopted as a representative organic contaminant to analyze and discuss the influences of different factors on the transport behaviors. The results show that: (1) The non-isothermal distribution condition changes the parameters such as the permeability coefficient, the diffusion coefficient and the linear adsorption coefficient. The variations of the permeability coefficient and the diffusion coefficient accelerate the transport process, especially for the diffusion coefficient, while the variation of the linear adsorption coefficient slows down the transport rate. (2) The increase of the temperature difference ΔT between the upper and bottom of the composite liner reduces the breakthrough time t_b and increases the bottom transport flux Jb. When ΔT values are 0, 10, 20, 30 and 40 K, the corresponding tb values are 65.7, 58.5, 54.3, 52.0 and 51.2 years, respectively. (3) When the diffusion coefficient D_m,R of GMB changes from 3.0×10⁻¹⁴ m2/s to 3.0×10⁻¹³ m2/s, the effect of Dm,R on t_b is obvious. It is found that the existence of GCL has a little effect on the transport behaviors by comparing t_b in the two cases.

Peridynamics is a non-local mesh-free numerical method which has great advantages in simulating dynamic damage and fracture. This work proposes to use the peridynamics to simulate rock fragmentation by TBM disc cutters and seven tests are simulated to verify this idea. The feasibility of peridynamics in simulating rock fragmentation by TBM disc cutters is demonstrated by comparisons with the punch penetration tests. A nonlinear short-range force contact model is proposed to better reflect the variation of the disc cutter normal force with respect to penetration depth, and can also calculate the changing contact area between the disc cutter and the rock during the penetration. In addition, the proposed non-linear short-range force contact model is used to simulate five 3D TBM disc cutter penetration tests with different bi-directional confining stress conditions, and the numerical results are compared with the experimental results. The simulation results show that the proposed contact model is able to calculate the variation of the cutter peak force with different confining stress and to obtain a reasonable form of rock surface damage. Peridynamics approach is easy to operate and requires only two calibration parameters to simulate the 3D rock fragmentation by disc cutters, which is promising for practical engineering applications.

In order to meet the requirements of refined analysis, a novel hydro-mechanical coupled multi-physical modelling approach is developed to qualitatively evaluate the grout loss in the soils and predict the ground movement induced by tunnelling and backfill grouting. The process of grout transport is described by a group of non-linear transient partial differential equations based on the mechanics of the continuous medium. The blockage of grout in the soil pore space and the resulting changes in the permeability of the soil are considered by introducing the mass exchange term into the mass equilibrium equation. The hydro-mechanical coupling process is further considered by incorporating the momentum balance of the mixture system. Then, through the secondary development of ABAQUS, the governing equations are solved by defining a new plane strain element of 7-degree of freedom, making it possible to analyze engineering-scale initial boundary value problems of grouting. After that, the excavation and grouting process of a typical shield tunnel has been modelled as an example. The results demonstrate that the proposed novel multiphysics modelling approach is able to describe the spatial and temporal changes in grout pressure, ground settlement and the grout penetration range during and after grouting. It is also found that when the permeability coefficient of the soil is greater than 1.0×10⁻⁶ m/s, the grout loss due to grout penetration needs to be considered; meanwhile, the grout blockage in the soil pore space will lead to a reduction in soil porosity and permeability, thus further reducing grout loss, but the ground deformation is barely affected. Therefore, in engineering practice, grout blockage may not be considered in order to analyze the maximum grout loss and ground deformation.

In this study, the effects of shear displacement and the number of cycles on the nonlinear hydraulic characteristics of three-dimensional rough fracture during cyclic shear are numerically studied. First, a three-dimensional fracture surface is generated based on the elevation data of the natural rough fracture surface, and the normal displacement during the process of cyclic shear is determined through previous studies. Sixteen fracture aperture fields are established so as to obtain the contact area and aperture distribution under different shear displacements and cycles times. It is found that when the number of cycles is constant, the upper and lower surfaces of the original mesh will stagger with the increase of shear displacement, the average mechanical aperture and anisotropy increase, and the contact area decreases; when the shear displacement is constant, with the increase of the number of cycles, the upper and lower surface undulations are smoothed under the action of shear stress, the contact area increases, and the average mechanical aperture decreases. Finally, the model is imported into the numerical simulation software to calculate the flow rate data. The results show that the variation of hydraulic gradient with flow rate conforms to Forchheimer’s law. The linear Fochheimer coefficient decreases by 88.3% and the nonlinear Fochheimer coefficient decreases by 95.2% as the shear displacement increases from 4 mm to 10 mm under 1 cycle; the linear Fochheimer coefficient decreases by 95.4%, the nonlinear Fochheimer coefficient decreases by 99.7% as the shear displacement increases from 4 mm to 10 mm under 2 cycles, and the reduction range of the two coefficients increases with the increase of the number of cycles. Further analysis shows that the increase of shear displacement boosts the fracture permeability, while the increase in number of cycles inhibits the permeability. In addition, the variation of normalized transmissivity coefficient with Reynolds number conforms to the research results of Zimmerman et al., and the numerical simulation results of this study are fitted with the normalized transmissivity coefficient prediction formula proposed by Zimmerman et al. The fitting coefficient approaches to 1, where Fochheimer coefficient β decreases with the increase of shear displacement and increases with the increase of cycle times.

Granular materials are associated with the irreversible behaviors during the loading-unloading-reloading path, the mechanism of which can be analyzed in terms of complex network evolutions. In this paper, the discrete element method is used to conduct triaxial tests for granular materials under the loading-unloading-reloading condition. The macroscopic responses of the granular materials and its corresponding complex network characteristics of different strong contact systems are analyzed. It is found that the macroscopic “hysteresis loop” is formed during the loading-unloading-reloading path of dense granular materials, and the microscopic coordination numbers and clustering coefficients also exhibit irreversible behaviors. Different thresholds are used to classify the strong and weak contact systems for granular materials under different loading and unloading stress states, and the difference in the mesostructure corresponding to each stress state decrease as the threshold increases. The maximum cluster of the strong contact system reaches the limit of percolation when the force threshold falls in a range, and the irreversible behaviors of granular materials can be reflected in the geometric and topological features of strong network during the loading-unloading- reloading path.