The shallow landslide of highway slopes in areas with expansive soil often occurs. Three representative types of expansive soil from the south and north of China are tested under wet-dry cycle in the laboratory. The collapse of slope surface in low pressure conditions, and the conventional process of saturation and shear test are considered in the experimental study. The purpose is to obtain the actual shear strength of expansive soil and its reduction. The results show that the shear strength of the expansive soil affected by different vertical load is not identical. The larger vertical load results in the smaller reduction ratio of shear strength. The sectional double line method is adopted to approximately reflect the nonlinear characteristics of strength of expansive soil. It can also be used to calculate the index of shear strength c and ?. The c reduces exponentially as the number of wet-dry cycles increases, and it is considered as the main reason for the shallow landslide. ? decreases as the number of wet-dry cycles increases. The amplitude of reduction decreases. Under the same wet-dry conditions, the ultimate shear strengths of different expansive soils drop to a certain range. Research conclusion can provide reference for analysis of shallow landslide of expansive soil slope effectively.

To improve the disaster prevention and mitigation of metro shield tunnel, model test of similar material and numerical calculation is studied in this paper to simulate the construction of building adjacent to foundation pit in Beijing. Under unloading-loading conditions of the foundation pit above, the mechanical characteristics of shield tunnel are studied, and the influence of spacing between the bottom of the foundation pit and the shield tunnel top and the loading strength of foundation pit are also analyzed. As shown by calculation, in unloading stage, the vertical displacement of shield tunnel gradually increases. When the foundation pit is excavated to the bottom, the vertical displacement reaches the maximum value. Under loading condition, the vertical displacement of shield tunnel can be restored. The greatest impact on shield tunnel always occurs at the end of the foundation excavation. At this point, the work of foundation slab construction should be completed as soon as possible. In the unloading-loading process of foundation pit , earth pressure of shield tunnel presents calabash type symmetrical distribution. At the top and bottom of the shield tunnel the earth pressure is larger than the soil pressure in the middle of the shield tunnel. In unloading stage, the earth pressure of shield tunnel vertical axis was significantly reduced. In loading stage, the earth pressure of shield tunnel can be restored. With the spacing between the bottom of foundation pit and shield tunnel top increases, the displacement of shield tunnel, the displacement difference between arch bottom and arch and the horizontal convergence are incrementally reduced. When the spacing greater than 3 times of the depth of deep excavation, the additional deformation of shield tunnel caused by foundation pit unloading-loading effect is slightly. When the loading strength of foundation pit 2 times than the unloading strength, the vertical displacement of shield tunnel can be restored to its original state.

When rock joints are disturbed by stress waves, the continuous stresses in joint ends change with various degrees of joints. To analyze the relationship between the dynamic change and the incident angle of stress wave, we introduce a combined model of the nonlinear normal normalised constitutive relation and the linear tangential constitutive relation of rock and the corresponding stress wave propagation equation of oblique incident joints in P wave oblique incidence joint. According to the theory of mode-I and mode-II crack-tip stress in fracture mechanics and the displacement field, a formula is obtained for calculating the stress field and displacement field with the change of the velocity of the joint in the combined form. This study investigates the effect of the simulated pulse signal with different incident angles on the simulation of stress and displacement at the end. Stress and displacement fields at the upper and lower ends of the joint present non-symmetric distribution under the oblique incidence of stress wave. With the increase and decrease of vibration velocity of the incident wave particle point, the position of the stress concentration changes. The data of 0.005 m position at the upper and lower sides of the joint end in the model are calculated, the change of joint normal stiffness bringing by the particle velocity of incident wave directly influences the transmission and reflection of stress wave, which leads to the lag of stress and displacement at the end of the joint. The lateral displacement and the incident angle do not change monotonically, but the vertical displacement decreases with the increase of incidence angle.

The grain-size formation in water/soil science is solely based on the soil characteristics (moisture and minerals) which ultimately determine the sizes of soil grains considering their interactions and surrounding environment. In this study, a methodology was utilized using different approaches (detailed observation, microscopic and macroscopic measurements) to determine loess evolution through time that affected its grain formation and structure. To explore the effect of freeze-thaw damage scale structure of loess environment, the loess Huangling county of Yanan city was taken as the research object with the initial moisture content and the number of freeze-thaw cycles as variables. Through the apparent structure test of scanning electron microscope and CT scanning, with variations of the loess body rate and freeze-thaw cycles, fine structure view and macro features in different initial water contents were explored. Based on Leica Qwin, Canny operator edge detection, fractal dimension, rainbow code pseudo color enhancement technology and saliency theory, the experimental results were analyzed. Testing results showed that with the increase of the number of freeze-thaw cycles, micro structure significantly changed. Specific changes include: particle size becomes uniform, skeleton connection mode changes, surface to surface contact mode changes to point to surface and point to point contact modes, and connection between grains weakens. Microscopic view showed a decrease of high density region and an increase of low density region, and an increase of the frost heaving force. Migration force continues to appear, enabling reduction of the loess body integrity and development of migration channel, which causes increase of ice volume from macro analysis . With the increase of the number of freeze-thaw cycles, surface porosity and fractal dimension increase in the beginning and become stable after 10 cycles. Under freeze-thaw environment, internal micro pores continue to transform into large pores. Enhanced dye technology significantly improves the recognition of CT image. The factors causing alteration of loess grain structure are not only the initial water content and the number of freeze-thaw cycles, but also the interaction (coupling) between water content and freeze-thaw cycles, which has a significant effect on grain structural alteration.

The bank slope of the reservoir is inevitably affected by the excavation unloading during the project construction. In this study, we selected sandstone from the typical bank slope in Three Gorges Reservoir Region. Water-rock cycle experiments were carried out to simulate the variation of water levels and immersion-air drying cycles. Especially, the water-rock interaction effect on unloading mechanical properties and microstructure of stratification sandstone was mainly analyzed. The results showed that in the course of ten times of water-rock cycles, the triaxial unloading and shear strength of sandstone presented initially a steep deterioration and later gradual deterioration trend. The first four times accounted for about 70% in the overall deteriorating degree, which indicated the degree in the last three cycles slowed down. As the water-rock cycle times increased, the deformation modulus of rock specimen decreased gradually before uploading at the same confining pressure. During the unloading process of confining pressure, the linear graded phase of deformation modulus gradually became shorter, whereas the nonlinear mutation phase became longer and slower. The brittle feature of rock specimen slowly weakened, whereas its plasticity enhanced. The repeated fluctuations in water pressure and the immersion-air drying cycle resulted in an irreversible slow and accumulated damage to the rock specimen. Microscopically, micro-cracks, cracks and pores of rock steadily developed and accumulated. Macroscopically, the porosity rate increased, and the triaxial unloading and shear strength decreased. This study can provide references for the analysis of excavation, unloading and deformation of slope rock mass in bank slopes.

A discontinuum and continuum coupling numerical model was built based on centrifuge modeling of a reinforced soil wall with rigid/flexible facings on soft soil. In the numerical model, the particle flow code (PFC) and the fast Lagrangian analysis code (FLAC) were used to simulate wall backfill and soft foundation, respectively. The deformation of the wall, the distribution of reinforcement loads along the wall height, and the internal failure evolution in the wall were simulated and analyzed. The internal failure evolution of the wall on rigid foundation was also simulated to compare with that of the wall on soft foundation. It was found that the numerical results are in good agreement with those measured in centrifuge modeling. The potential internal slip surface of the wall on the soft foundation passes through the end of the embedded anchors, and the tension in each reinforcement layer is the maximum at the end of the embedded anchors. The reinforcements in the walls are on either soft or rigid foundation break at the end of the anchors from the bottom to the top in their limit states. The walls on soft foundation fail in a plane slip passing through the end of the anchors intersecting the circular plane slip in soft foundation. However, the walls on rigid foundation slide horizontally along the bottom anchors with a plane slip at the end of the anchors.

Fractured rocks in underground engineering are generally in the state of triaxial stress, which are characterised by large porosity and high permeability. Under the action of the ground stress and high-water head, it is most likely to result in the failure of seepage instability and induce the disaster of water inrush. In this study, a self-developed triaxial permeability testing system was employed to study the permeability characteristics of the fractured sandstone under different triaxial stress conditions. The permeability tests were carried out on fractured sandstone with five different sizes of fragments by the steady-state permeability method. Finally, the permeability variations of fractured sandstone under triaxial stress states were obtained, and the relationship between effective stress and seepage velocity was deduced. The result showed that the relationship between the effective stress and seepage velocity of fractured sandstone under triaxial stress was approximately linear. When the axial displacement increased, the decreasing amplitude of seepage velocity became smaller with the increase of the effective stress. The pore pressure gradients and the seepage velocities of fractured sandstone with five different diameters of fragments under triaxial stress conditions were coincidenet with the Forchheimer equation. Moreover, the correlation coefficient between these two was above 0.95. When the axial displacement was constant, the permeability k of fractured sandstone decreased and non-Darcy flow β factor increased with increasing confining pressure. Under various axial displacements, the relationship between the permeability and confining pressure of fractured sandstone presented an exponential function. With the decrease of porosity, the permeability of fractured sandstone with five different diameters of fragments exhibited a downward trend, whereas non-Darcy flow β factor increased as a whole. In addition, the magnitude of the permeability was from 10-14 to10-11 m2 and the magnitude of non-Darcy flow β factor was from 106 to 1012 m-1.

The key stratum theory widely applied to coal mining area has provided a significant theoretical foundation for controlling the strata and reasonably explained the strata behaviours after excavating coal seam. The theory contends that hinge structure formed after roof failure and instability controls the movement behaviour of overburden strata, but it ignores the influence of shear strength between strata on the movement. Based on the hypothesis of interlaminar shear stress, this paper establishes the mechanical model of composite beam in overlying strata, and then derives the theoretical solution of shear stress of nonhomogeneous beam. Meanwhile, a key stratum decision step is proposed by considering the inter-layer shear stress, and the equation for calculating the breaking interval of the key stratum is corrected. The results show that overlying strata controlled by key stratum probably does not fracture with key stratum at the same time. According to the double-layer rock mining model, we propose a concept of flexible control layer (FCS) which can hinder the deformation of one or more overlying strata above it after the key stratum broke. Additionally, the discrimination method for FCS is further deduced, and the effects of FCS on the composite fracture and strata behaviours are discussed. The existence of FCS weakenes periodic roof weighting to some extent and generates a secondary weighting. Finally, the rationality of this hypothesis is proved with the examples in practical engineering.

At the instant of the occurrence of an earthquake, the conventional prestressed anchor cables tend to break due to the insufficient deformation ability suddenly. Once the failure of the anchor cable occurs, the safety of the entire anchorage structure is endangered. This study aims to investigate the seismic response of the adaptive anchor cables and the corresponding dynamic characteristics of the anchored slope. The experiments for the rock slope with adaptive anchor cables are conducted based on a new type of anti-seismic anchor cable as a prototype and accordingly loaded by the shaking table test system. In the experiments, four kinds of seismic waves are applied, such as sinusoidal wave, Tianjin wave, EI wave and Taft wave. The real-time strain of the cables and corresponding seismic response of the slope are monitored. The results show that the axial force of anchor cables is closely related to the amplitude, type, seismic excitation frequency of the seismic wave. The acceleration and displacement on slope surface are enlarged with different elevations, and the peak acceleration and displacement decrease compared to the rock slope without anchor cables. With different values of preset slipping constant resistance of adaptive anchor cables, the anchor cables produce three slip modes such as non-slip, instantaneous slip and gradual slip. The corresponding time history curves of cable strain and dynamic safety coefficient of the anchored slope are identically different. Under the condition of the cable slipping, the safety coefficient of the slope decreases partially, whereas the safety capacity of the anchored structure is large enough to suit for the large deformation and transient impact loads on the rock slope. Therefore, the experimental results provide references for the study on the roadway slope supporting and anti-seismic design of the adaptive anchor cables in the strong earthquake area.

Using self-developed high-pressure oedometer system, we performed one-dimensional consolidation tests (OCT) on reconstituted clay samples at vertical pressures from 0.25 MPa to 12 MPa. In addition, we carried out isotropic consolidation tests (ICT) on the same samples with confining pressures from 0.3 MPa to 2.5 MPa using triaxial apparatus. The test results indicate that all the consolidation curves obtained from OCT have notable secondary consolidation portion, and the secondary consolidation coefficient of remolded clay changes nonlinearly with increasing consolidation pressure. Specifically, when the consolidation pressure is less than 2 MPa, secondary consolidation coefficient decreases significantly with increasing consolidation pressures. but it is almost unchanged for the case of consolidation pressure above 2 MPa. The secondary consolidation coefficients obtained from ICT are much less than those from OCT, and the change of secondary consolidation coefficient with confining pressure obtained from ICT is different from that from OCT. Moreover, the microscopic analyses of the test results show that the physical mechanism underlying secondary consolidation deformation of clay at high pressure is different from that at low pressure. At low pressure, the secondary consolidation deformation mainly results from the relative slip between clay particles while the creep of adsorbed pore water contributes much to the secondary consolidation deformation of clay at high pressure. Moreover, it seems that the deviatoric stress has a dominating influence on the secondary consolidation deformation of clay at low pressure while it has little effect at high pressure.

To improve the understanding of the mechanical behavior of the gas-hydrate bearing sediments, a series of triaxial compression texts was performed on gas saturated CO2-hydrate-bearing sand. Samples were prepared using the excess gas method, in a modified high-pressure low-temperature triaxial apparatus. The hydrate saturation could be determined readily according to the gas assumption. The repeatability of sample making was analyzed in terms of the hydrate saturation. It was found that samples with high hydrate saturation could be formed from samples with high initial water saturation, however, the repeatability of making samples, which was affected by pore blocking effect, was low. It was shown that the repeatability of making samples with hydrate saturation lower than 40% using the excess gas method was relatively good. The triaxial test results show that the mechanical behavior of hydrate-bearing sand samples depends largely upon the hydrate saturation and the applied net confining pressure. The strength, stiffness, ductility and compressibility of hydrate bearing specimens increase with the effective confining pressure; as hydrate saturation increases, the strength and stiffness of specimens increase, while the ductility and compressibility decrease. In addition, the larger the hydrate saturation, the larger the apparent cohesion, whereas the friction angle remains practically unchanged.

Studying the seismic dynamic responses of loess-mudstone slope is important for the slope stability evaluation and engineering aseismatic design. The effects of seismic wave amplitude on the seismic dynamic responses of loess-mudstone slope are investigated by the centrifuge shaking table tests and numerical simulation. Results show that: the amplification of horizontal and vertical acceleration of loess mudstone slope increases nonlinearly from deep to shallow surface of the slope. The horizontal amplification is greater than the vertical, and it reaches the maximum value at the crest. The dynamic response of the loess layer is larger than that of the mudstone inside the slope. With the increasing of the amplitude of the input seismic wave, the dynamic response of the slope increases first and then decreases. When the input amplitude reaches 0.3g, the dynamic response of the slope reaches the maximum value. The deformation and failure process of the loess-mudstone slope could be summarized as: with the input of seismic wave amplitude increasing, tension crack formed at the crest, gradually expands, displacements develop towards the free face, and then overlying loess raises lightly, some soil accumulats at the slope toe.

Experimental research on cyclic behaviors of clay is performed based on undisturbed soil samples from the boreholes of oil field sites in northern region of South China Sea. Combined resonant column test and strain-controlled cyclic direct simple shear test, the dynamic shear modulus and damping ratio of the soils over a range of cyclic shear strain are studied. Additionally, the maximum dynamic shear moduli are compared with the calculation results on the basis of cone penetration test. Stress-controlled cyclic direct simple shear test is developed to capture the threshold value of cyclic strength by symmetric loading method. And then, based on the threshold value, the envelope curves of cyclic strength can be obtained by one-way and two-way loading methods. The test data are compared with Drammen clay and other clays obtained from different locations in the world. The cyclic performance and dynamic parameters of soils in the northern region of South China Sea are studied, which provides the basic data for structure design and evaluation of geological risks. Besides, experimental methods of cyclic tests are mastered, which can provide technical reference for dynamic experimental study and data analysis of geotechnical engineering.

Underground structures are severely damaged when subjected to strike-slip faulting during an earthquake. Mechanism behind inconsistent dynamic responses in soil stratum is not yet fully understood. The influence of pre-existing fracture in soil stratum on the dynamic response of the soil is still not clear. Based on a newly designed laminar box with simulation of strike-slip fault, two dynamic centrifuge tests were conducted on clay stratum to investigate the influences of the pre-existing fracture. Centrifuge test results show that inconsistent dynamic responses for the soil crossing bedrock fault is not obvious when there is no pre-existing fracture. With a pre-existing fracture, however, the magnitude of seismic wave at the moving block decreases with time during shaking. Inconsistent dynamic responses between the stationary and moving blocks are then observed. In addition, the excess pore water pressures developed at the stationary block are larger than what are observed at the moving block. The effects of distribution of the pre-existing fracture on the inconsistent dynamic responses in the soil stratum are also demonstrated. It is of significant importance to identify the tip depth of the pre-existing fracture in soil stratum.

Dynamic response of deposit slopes is complex with multiple influencing factors. The large-scale dynamic triaxial tests reveal that moisture content affects greatly on dynamic shear modulus and damping ratio of deposit. As such, two groups of large-scale shaking table model with scale of 1:12 of deposit slopes are tested with different moisture contents. Comparisons of acceleration and crest displacements distribution of 2 deposit slopes under multiple sequential ground motions show that, in either 2 models with different moisture conditions, the first order natural frequencies of model slopes decrease with the increasing loading sequence, while the corresponding damping ratios increase with its increasing. The slope model with 6.6% moisture content has larger numbers of first order natural frequency, damping ratio and peak ground acceleration (PGA) amplification coefficients than the numbers for model with 0.7% moisture content. The Fourier spectrum of slope crest accelerations of these 2 models are similar. The amplification effect of crest acceleration enhances evidently when the predominant frequencies of input seismic waves approach to the first order natural frequencies of the deposit slope. the input frequencies of seismic waves approaching the first order natural frequency from the lower frequencies side of slope amplifies PGA significantly than from the higher frequencies side. The permanent displacements of deposit slope crest with 6.6% moisture content are larger than that of 0.7% moisture content slope. In addition, the crest permanent displacements with higher moisture content are more sensitive to the spectra characteristics of input seismic waves.

The influences of the magnitude and distribution of horizontal earth pressure on reinforced body using unbonded prestressed reinforcement technique were studied through scaled model tests in the paper. The steel shot was adopted for its appropriate similarity ratio to simulate medium sand in the model test. Two cases were tested to compare the common reinforced body and the single unbonded prestressed reinforced body. The comparison shows that the settlement on the top of body with unbonded prestressed reinforcement before applying pre-tension is less than that on the top of the common reinforced body. The settlement on the top of reinforced body increases with the application of pre-tension. However, the vertical deformation of the reinforced body with single unbonded prestressed reinforcement is smaller under the heaping load. With the increase of pre-tension, the strain of fiberglass geogrid adjacent to prestressed reinforcement decreases and that of fiberglass geogrid at the bottom increases, particularly at the measuring-points near the panel. The pre-tension of the unbonded prestressed reinforcement remains unchanged during applying the heaping load on the top, and increases slightly after completing the heaping load.

Natural rock mass usually contains multiple sets of cross-persistent joints, which greatly weakens its mechanical properties. This study aims to investigate the strength and failure characteristics of rock mass with two sets of cross-persistent joints. Under different confining pressures, compression tests are conducted at various inclination angles and spacing values of joints, respectively. Then obtained results are further stimulated by the elastoplastic numerical manifold method. The results show that the variation of strength with the joint inclination exhibits the multi-crest and multi-trough characteristics. The variation of strength with joint spacing confirms to a negative exponential function. According to the joint states, the failure modes of rock mass can be divided into three kinds, i.e., failure of rock block, slipping along a set of joints and slipping along two sets of joints. It is found that both sets of joints affected the strength of rock mass to certain degrees. Moreover, there is an interaction between these two sets of joints. Through the regression analysis of the simulation results, a new strength prediction model is established for rock mass with two sets of cross-persistent joints. The model is based on the strength prediction model of rock mass with one set of joints and the quantification of the interaction between two sets. The model is simple in form and easy to use, which can provide a guideline for evaluating the strength and failure characteristics of rock mass in practical engineering.

To explore the influence of slurry thickness and roughness on the mechanical properties of pile-soil interface, the distribution of protrusion size and roughness on the side of pile is obtained by statistical analysis based on the relationship between diameter and depth of borehole. The surface roughness of the concrete pile is simulated by the surface smooth and trapezoidal groove concrete respectively. A series of large-scale direct shear tests of concrete-sand interface is conducted with different slurry thickness and roughness. The results show that for the rough interface without slurry, the shear stress-tangential displacement curve is softened. For the rough interface with slurry at thicknesses of 5 mm and 10 mm, the shear stress-tangential displacement curve is hardened. The shear modulus G0.02 at the horizontal shear strain value of 0.02 decreases as the slurry thickness increases. For smooth interface, the peak shear strength is decreased exponentially as slurry thickness increases. For rough interface, the peak shear strength and the peak friction angle are decreased in an approximately linear way as slurry thickness increases. The interface strength becomes damped as result of the increasing initial thickness of the slurry and the thickness of the mixed layer after the test. Furthermore, a critical roughness on the peak intensity of contact surface under no slurry condition is found. When the roughness of concrete slab is less than the critical roughness Icr, the peak shear strength and the peak fiction angle of the interface will increase with the increase of roughness. When the interface roughness I is larger than Icr, both of them will decline with the increase of roughness. However, the existence of mud changes the relationship between roughness and peak shear strength.

The negative skin friction of composite foundation with rigid pile has a significant effect on the pile-soil interaction under embankment load. Based on the results of shear tests that ultimate shear displacement is less affected by stress levels but greatly influenced by specimen sizes, an ideal elastic-plastic load transfer model considering the variation of stress-dependent shear stiffness of pile-soil interface is developed. According to the three-section distribution pattern of pile-soil interaction, i.e., the upper plastic zone of negative friction, the central elastic zone of coordinate deformation and the lower plastic zone of friction bearing, a distribution model for skin friction on pile-soil interface describing the nonlinear elastic-zone and inhomogeneous plastic-zone is established. To investigate the equivalent reinforcement of single pile with equal area in rigid pile composite ground under uniformly distributed embankment load, the formulas of the depth of the neutral plane and stress concentration ratio are derived according to the force equilibriums differential equations of a simplified analytical unit cell and the compression deformation compatibility equations of pile-soil-cushion,. the effect of the embankment load and the flexibility coefficient of cushion on the depth of the neutral plane and the stress concentration ratio are analyzed. The analysis indicates that the depth of the neutral plane descends with the increasing of the flexibility coefficient of cushion, and then ascends with the increase of embankment load. The stress concentration ratio decreases with the increasing of the flexibility coefficient of cushion, and decreases afterwards with the increase of embankment load. Hereby, the technical ideas including the pile arrangement and the cushion design using the maximum stress concentration ratio are provided.

To study the strength and deformation characteristics of frozen clay under complex stress path, a series of directional shear tests on saturated frozen clay is conducted under different negative temperature conditions by using dynamic hollow cylinder testing apparatus of frozen soils (FHCA-300). Stress-strain curves of the axial and torsional shear components for frozen clay are analyzed under different shear directions. The shear deformation, anisotropy and shear band characteristics are discussed. And the effects of temperature, angle of major principal stress direction, mean principal stress and coefficient of intermediate principal stress in the strength and deformation behaviors of frozen clay are explored. The test results indicate that the value of the mean principal stress p has a significant effect on the stress-strain curves of frozen clay. When p=4.5 MPa, the frozen clay has a major shear strength and this stress value (the mean principal stress) may be the critical value of pressure melting. The anisotropy of frozen clay may be induced by the change of major principal stress direction. Shear strength decreases versus the increase of angle of major principal stress direction. The axial strength decreases with the increase of the coefficient of intermediate principal stress, but the effect of the coefficient on the shear strength is neglect. With the decrease of temperature, the strength of frozen clay increases gradually. Meanwhile, the samples produce brittle failure, and a shear failure surface appears and develops. The shear strength of frozen clay depends mainly on the cementation force between soils and ice particles.

The energy pile is a new type of ground source heat pump technology that places the ground heat exchanger in the building pile foundation to achieve underground heat exchange. However, the expansion and shrinkage of plie materials threat the long-term running and the building safety because that the energy piles often operate in cooling and heating thermal loads at different seasons. Grouting materials with proper thermal-mechanical characteristics are crucial for the safe use and widespread of the energy piles. This paper aims to investigate thermal-mechanical behaviours of concrete piles mixed with different contents of polypropylene fibre and steel fibre. The results show that thermal conductivity increased when adding steel fibre into concrete but decreased when adding polypropylene fibre. The maximum thermal conductivity was measured to be 2.44 W/(m·K) for the concrete mixed with 1.3% steel fibre. The thermodynamic step heating test showed the strain of the concrete reinforced both by steel fibre and polypropylene fibre was effectively reduced. The reductions of concrete reinforced by steel fibre and polypropylene fibre were about 62.43% and 61.11%, respectively. Furthermore, heating expansion and cooling shrinkage strain of concrete reinforced by steel fibre were both reduced during the heating and cooling load testing cycles, and the strain was the minimum in the whole process. Compared with three different types of concrete, steel fibre reinforced concrete is suggested as a suitable material for the grouting of energy piles.

This paper applies Biot’s consolidation theory and an elasto-plastic constitutive model on the post-liquefaction deformation of sand to investigate the wave-induced dynamic interaction between pipeline and seabed. The numerical simulations can reasonably reproduce the accumulation of excess pore pressure and the process of soil liquefaction around the pipeline induced by linear waves in the liquefiable finite seabed. The numerical results show qualitatively consistent with Sumer et al.’s results, and demonstrate conclusions as follows: The distribution of liquefaction in the seabed is significantly influenced by the pipeline, liquefaction first occurs at the seabed surface and underneath the pipeline, then transfers toward the soil surrounding the pipeline. Eventually, the entire sandy seabed around the pipeline liquefies, leading pipeline flotation. The degree of liquefaction in the soil around pipeline is intensified by pipeline. The soil liquefaction is induced by waves which is found mainly the combination of accumulation of excess pore pressure and seepage forces. The elasto-plastic method of dynamic analysis presents more rational result compared to the existing elastic method in simulating the changes of seepage field and stress field under wave actions for realistic ocean environment.

The process of water inrush disaster accompanied by the information change of rock and soil within the preventing structure changed accordingly. This study conducted an in-depth fusion analysis of multi-physical field information for potential water inrush channels. The logical symbiosis relationship of multi-information was revealed, and a theoretical method for identifying the evolution state of water inrush was established. Two typical similarity model tests of water inrush were conducted to simulate the progressive failure of rock mass and filling structure instability. Then, we proposed the correlation between the multiple monitoring information and the quantitative characterisation function. Based on the theory of principal component analysis, this study revealed the logical symbiosis relationship between multiple parameters and determined the influence weight of the multi-physical field information in the surrounding rock, which provided a theoretical basis for the monitoring design and early warning of water inrush disaster. A comprehensive method for the evolution state of water inrush disaster was established, combined with the extreme point and the stagnation point of the fitting curve representing the changing trend of the function, Finally, the evolution stages of these two types of water inrush are divided into four periods: calming period, development period, abrupt period and post-disaster period. The energy discrimination method of the occurrence probability of water inrush was put forward at any moment. The theoretical method provides a reference for monitoring and early warning of water inrush disaster in tunnels and underground engineering.

The freezing method is adopted in Xinzhuang coal mine, and during the process of shaft excavation lateral, unloading leads to the unloading deformation of surrounding rock. The cretaceous media-grain sandstone is collected from Xinzhuang coal mine, and the samples are processed into saturated states. Then, the constant-axial compression and unloading confining pressure triaxial tests are carried out under freezing (-10 ℃) conditions using GCTS servo-controlled high-low temperature and high-pressure triaxial rock testing system. The stress change path of surrounding rock is simulated in the process of shaft excavation, and the deformation characteristics of frozen sandstone are explored. The results show that under the condition of lateral unloading, the frozen sandstone shows obvious elastic-brittle characteristics. There has compression deformation in the axial direction, but expansion deformation in the radial direction. Moreover, the radial deformation is approximately two times larger than the axial deformation. With the same unloading rate, the unloading deformation of rock specimens increases with the increase of initial unloading confining pressure, and especially the radial deformation is the most significant. This may be related to the unloading rebound deformation and the amount of energy accumulates within the sample. When the confining pressure is unloaded to the same stress level, the unloading deformation of rock specimens is small, and the deformation rate is greater under high unloading rate. The deformation modulus of rock samples decrease in the unloading process, but the ratio of lateral strain to longitudinal strain change increases. The smaller the unloading rate is, the larger the initial unloading confining pressure is, and the greater the variation of deformation modulus is.

A large model testing system was developed to simulate the rockfall impacting the slope of tunnel exit of Chengdu-Lanzhou railway which is under construction in the southwest area of China. The system is composed of the model testing bench, three-dimensional rockfall release device and high-speed photographic system. This system with high structural strength can meet the visual requirements in the testing process. Based on the measured physical and mechanical parameters of rock specimens, similar materials to phyllite and sandstone were developed to simulate slopes and rockfall by mixing with different proportions of vaseline, silicone oil, cement, fine sand, barite powder, gypsum and talcum powder. The effects of the initial slip angle, shape and mass of cuboid rockfall on the coefficients of restitution were studied through the physical model tests. Besides, a high-speed camera and coordinate callipers were adopted to record the complete trajectories of rock blocks during the rockfall process. At last, a video processing software was employed to analyze the variation of velocity before and after the collision between the rockfall and the slope, and the normal and tangential restitution coefficients of the rockfall were calculated by formulas. The testing results showed that with the increase of the initial slip angle, the normal restitution coefficient of the plate specimen increased, whereas the tangential restitution coefficient decreased and its variation was more obvious. The normal restitution coefficient had a remarkable decreasing trend with an increase of rock mass, but the tangential restitution coefficient seldom changed. For cuboid rock blocks, the normal and tangential restitution coefficients of cuboid specimens are larger than those of thick plate specimens. The normal and tangential restitution coefficients of strip specimens exhibited a polarisation trend due to different contact forms. Finally, the impact mechanism of rockfall was explored according to testing results and previous research results.

The behavior of a laterally loaded pile is sophisticated and diverse when it is loaded on the sloping ground. For the gravel soil as a typical strata of western Sichuan province of China, very limited research works have been carry out to study laterally loaded piles in gravel soil ground. In order to study the horizontal load-bearing characteristics and the interaction between pile and soil of the single pile embedded in the gravel soil slopes, a field test is conducted to investigate the single pile response varying gravel soil slope gradient. Slopes with gradient 0°, 15°, 30° and 45° are chosen, and the study includes pile deflection, bending moment, and earth pressure. The finite element analysis software FLAC3D is used to obtain the pile-soil stress field nephogram and displacement field nephogram for different gradient slopes. Furthermore, comparing field test results with numerical simulation results of critical load and ultimate load of pile in different slope ranges provides a reduction coefficient for the existing methods proposed in design. In summary, the study in this paper provides a reference for future practical engineering applications.

Karst collapse and shallow tunnel construction frequently cause partially ground settlement of shallow foundations, resulting in partially developed soil arching effect. Development of approaches to quantitatively analyze the influence of partially developed arching on vertical stresses is of much importance. Through shallow trapdoor tests, it is summarized that the shape of slip plane of shallow trapdoor may be tower-shape under plane strain conditions. Analytical solutions to coefficients of lateral and vertical stresses for any depth of the shallow foundation under the action of partially developed soil arching effect are deduced and proved, which are based on the principal stress rotation and the differential settlement of soil layers. The influence of main parameters about partially developed soil arching effect is investigated. The results reveal that the larger depth-width ratio and the displacement-width ratio of trapdoor are needed to fully develop the arch. In addition, the vertical stress of fill increases with the increase in the effective angle of internal friction and angle of slip plane. The corresponding results can provide theoretical guidance for the calculation of vertical stress of shallow subsoil subject to local settlements.

The unloading-reloading conditions during an asymmetric foundation excavation usually are harmful to the underlying subway tunnel. To verify the disturbance of foundation excavation, the modified centrifugal system was used to simulate different excavation steps, and the load transfer of foundation-soil-tunnel, considering the relative position of tunnel and foundation pit. The results show ground loss due to the excavation, tunnel float and offset induced by asymmetric unload. The disturbance of first excavation step was small, and became stronger as the excavation depth increases; the left tunnel close to foundation center was disturbed more significantly. Longer distance to the foundation and tunnel generates weaker the disturbance by unload-reload. The force of tunnel lining was redistributed because of the foundation load influence after loading. The disturbance of the section lying between piles was obvious. The section lying one-sided pile was small, the disturbance of pile group effect was more obvious.

Mutual construction of underground caverns on the right bank of Baihetan hydropower station caused complex microstructures and rupture mechanism of surrounding rock of intersecting caverns. In this study, the IMS system was introduced to investigate microseismic characteristics of surrounding rock of caverns and rupture mechanisms under high stress, multi-faceted unloading stress conditions. According to the lithology, fault zone and fault distribution in the underground powerhouse area, the sensors with perfect monitoring effect were chosen and rationally arranged. The accuracy of the source location was analyzed by using the velocity inversion through the fixed point percussion test. Based on the case of failure of the #10 omnibus cave, this study analzsed the spatial-temporal evolution of microseismic events and energy release of the intersected chambers. Meanwhile, according to the energy ratio (ES/EP) criteria, the evolution mechanism of rock mass rupture was discussed, which included a large number of tension events (blasting shock caused), pull-shear and pressure-shear events, tension and shear events alternately (the effect of concentrated stress and joint) and failure (the effect of gravity and joint) in the order of occurrences. This study is helpful to optimise the excavation scheme of the underground section on the right bank of Baihetan hydropower station, and it is of significance for the excavation and support of similar projects.

High-speed railway embankment as the supporting body of the train and track is directly exposed to the complex climate conditions such as the rainfall, evaporation and other environmental events. These conditions change the water content, suction and shear strength of the filling soil, and further affect the stability of the embankment. A moisture migration model of the high-speed railway embankment with considering environmental effects is proposed. The model is verified by comparing with the measured data of a highway embankment in Rouen, France. Based on the meteorological data of Beijing and Shanghai in 2013, the volumetric water content (VWC) response of the high-speed railway embankment is calculated to investigate effects of climate conditions and soil types. The result shows that, the soil type and climate condition significantly affect the moisture variation of high-speed railway embankment. The influential depth of atmospheric environment on the VWC of sandy embankment is small, but the shallow embankment is affected by the atmospheric environment significantly with a large amplitude of VWC fluctuation. The influential depths of atmospheric environment on the VWC of silty and clayey embankment are relatively deep, but the amplitude of VWC fluctuation is small at every depth. The VWC of embankment under weather condition of Beijing has a huge seasonal variation and the VWC of embankment under weather condition of Shanghai fluctuates significantly with each rainfall.

The natural repose angle of sand has important guidance for mound construction, foundation design and slope stability analysis. The natural repose angle is mainly affected by friction, particle shape, grain size and water content of sand. In this paper, the natural repose angle of calcareous sand under various conditions is tested. In the three common particle shapes of calcareous sand, the flake particles present the largest repose angle; the rod stick particle shapes present less, and the block shape ones show the smallest value of angle. The natural repose angle of calcareous sand increases with the increase of grain size. When the mean particle size is the same, the natural repose angle increases as the inhomogeneous coefficient increases, and decreases as the curvature coefficient increases. The natural angle of repose of quartz sand is smaller than that of calcareous sand. It is mainly determined by the mineral composition and particle shape. The underwater repose angle of calcareous sand is smaller than that of natural repose angle in general. The maximum angle of foundation ditch slope is larger than the tested natural angle of repose. The results are instructive for coral reef mound constructions and design.

In western China, many irrigation canals have been built to encourage the development of agriculture. At the same time, the failures of loess landslides are provoked by irrigation, and caused great loss and heavy casualties. Taking the failure of loess landslide caused by irrigation in Gaolou village channel as an example, the irrigation of the canal is proposed. The water infiltration produces a certain depth of water pressure in the sliding body, the loess cohesion decreases and the friction angle becomes smaller, which causes shear failure of the landslide body at a certain depth, and then the shearing damage occurs at the trailing edge of the sliding body, causing the trailing edge loess to be in the state of the post-destruction zone and generating unbalanced shear stress. The unbalanced shear stress drives the sliding body to move forward until only one point of the sliding surface is in a critical state, and then the entire landslide is destroyed, and the body is disintegrated to generate mudflow. This failure process can be summarized as follows: the landslide first produces shear failure, and then the trailing edge produces tensile shear failure. When the driven shear stress generated in the post-destruction zone is greater than the frictional resistance of the sliding body, it will push the landslide forward until the landslide occurs damage completely. The correctness of this failure mechanism is demonstrated by theory and experiment, and the rationality of a new shear stress constitutive model is verified.

The water injection into coal seam is an effective means of increasing and discharging pressure during coal mining. The sealing length is a key factor affecting the effect of water injection fracturing. To determine the reasonable sealing length, six different elastic constitutive equations were selected according to the basic theory of Lippmann coal seam instability. The expression of nonlinear dynamic gas pressure distribution was introduced, and the mechanical theory model of gas-containing coal seam was constructed. Accordingly, this study obtained the formulas for the vertical, horizontal stress and plastic zone width of coal seam. Combined with the Gaojiapu coal mine, this study compared and analysed the stress distribution curves and the plastic zone widths of the coal seam calculated by three elastic constitutive equations. Moreover, the field sealing experiment was carried out. The results show that the hole leakage is small and not abnormal when the sealing hole length is 9.5 m, which indicates the sealing effect is good and a reasonable sealing length is 9.5 m. Compared with the other three elastic constitutive equations, the calculated results by i, v and vi elastic constitutive equations well confirm with the actual length of the sealing.

With the orthogonal optimisation algorithm, a fluid-solid coupling numerical model for subaqueous tunnels was established to analyse the arching laws of the pressure arch in the vertical direction above the vaults of the soil-rock compound strata and the full-section rock strata. The reasonable overlying thickness of subaqueous tunnels can be further derived based on the critical arching thickness criterion. Subsequently, a regression model for the overlying thickness of the subaqueous tunnel was proposed and further evaluated by BP neural network model. Finally, the regression model was applied to Foguan Shiziyang intercity tunnel to calculate its overlying thickness. The results showed that the height of the pressure arch was mainly affected by the overlying thickness when the lithology of the overlying strata was good, while it was affected by the thickness of soft soil when the lithology of the overlying strata was poor. The height of the pressure arch increased linearly with the increase of the overlying thickness, and then gradually decreased to achieve a stable state when reaching the critical arching thickness. When the lithology of the overlying strata became poorer, the critical arching thickness became greater. The variation of the pressure arch height in the full-section rock strata was less than that in the soil-rock compound strata. The example analysis shows that the reasonable overlying thickness model can be predicted according to the critical arching thickness criterion. Hence, the obtained results can provide references for the design and construction of the subaqueous shield tunnel in the soil-rock compound strata.

Dyke intrusion is a relatively common geological phenomenon. The intrusive dykes and their weathered layer have great effect on the seepage field and stability of the slope. However, few studies have been reported on this aspect. The intrusive volcanic dykes and their weathered crust are widely distributed in the volcanic rocks in areas of about 500,000 km2 along the southeastern coast of China, which greatly affect the stability of the residual soil slopes in this area. In this paper the Zhonglin landslide was chosen as the object to study the impact of dyke and its residual soil on seepage and stability of residual soil slope. Zhonglin landslide is mainly composed of tuff and its residual soil and intrusive granite dyke and its residual soil which was found at the crest of a slope. Firstly, by soil column tests we obtained the difference between two types of residual soils with regard to their infiltration property and rules of two in the rainfall. Then the Seep/W in Geo-studio software was applied to retrieve soil column rainfall experiment to obtain the unsaturated seepage parameters of both types of soil to apply it to the numerical analysis of the whole slope seepage field. At last, Slope/W in Geo-studio software was used to analyze the slope stability, which revealed the impact of weathered crust of granite dyke on slope stability. The results showed that the permeability of the granite residual soil in weathered layer of intrusive dykes was better than tuff residual soil. Under the effect of weather conditions like typhoon and rainstorm, the rain water infiltrated rapidly through weathered layer of the intrusive granite dyke and flew along the strong weathered-residual soil interface into the tuff residual soil layer, which resulted in the rapid rising of the groundwater level at the junction of the intrusive granite dyke and the tuff residual soil layer and the pore-water pressure grown from negative to positive rapidly so that the slope was more likely to lose stability at this location. The results in this paper may provide theoretical background for the stability evaluation and treatment for the residual soil slope invaded by granite dykes and weathered crust in the China southeast coastal region.

Nowadays the number of deep tunnels is greatly increased. Since there are lacking successful numerical methods and experiences from relevant engineering construction and design of deep tunnels, massive on-site monitoring and testing should be conducted to support the design and safety control. The present monitoring and testing design methods are proposed according to the characteristics of shallow underground engineering. However, when they are introduced into the deep tunnels, great challenges are encountered. In this study, the failure of surrounding rock mass in the field was presented and then the deformation and failure responses of surrounding rock mass were obtained during excavation by various advanced observation techniques. Based on the acquired information, this study analyzed the mechanical response characteristics of surrounding rock mass during excavation and lining at the operating period. Then we discussed the problems arising from the application of current modern monitoring methods. Six suggestions were put forward relating to the monitoring and testing design of the representative section in the deep tunnel. This study can provide helpful guidance to improve the design methods in similar projects.

This paper used the analytical layer element method to analyze two-dimensional dynamic response of multilayered saturated subsoils subjected to a vertical time-harmonic load based on Biot consolidation theory. Starting with the governing equations of plane strain problem in Cartesian coordinates, ordinary differential equations can be obtained from partial differential equations through the Fourier-Laplace transform, and the analytical layer element for a single saturated soil layer is established. The global stiffness matrix for multilayered saturated subsoils is assembled according to continuity conditions of adjacent layers and boundary conditions. Actual solutions are obtained by taking Fourier-Laplace inverse transform. The results of numerical calculation are in good agreement with the existing solution in literature. More numerical examples are designed to analyze the effects on vertical displacement due to the circular frequency, force depth and soil stratification. Numerical results show that the vertical displacement increases firstly and then decreases with the increases of circular frequency; the vertical displacement at the load point presents a crest and is influenced by the characteristic of topsoil greatly.

A 3-D discrete element method is presented on the basis of new definition of distance potential function, which is suitable for the arbitrary polyhedral elements. In this approach, the potential function is characterized as a distance function between the contacts using a normalized computational algorithm. Meanwhile, the calculation formulations for contact interaction are deduced with the new function. Thus this new distance potential function implies sound physical meaning, and is reasonable and concise to avoid different contact situations. This method overcomes the problems of the original potential method in which the physical meaning of potential function is not clear and the calculation of contact force is affected by the element form. Furthermore, it gets rid of the restrain of tetrahedral shape and can be applied to arbitrary polyhedrons. Several examples are analyzed to demonstrate the accuracy and efficiency of this new method. This method shows good capability of handling the problem of contact transformation for arbitrary polyhedrons and the motion process of the complex discontinuous media.

Energy piles are piles equipped with heat exchange pipes through which a heat-carrying fluid circulates and exchanges heat with the ground. This technology couples the structural role of classical pile foundations with heat exchangers for energy supply. In order to analyze the bearing characteristics of energy piles in clay, it is necessary to determine the temperature response of the pile and the ground, so as to investigate the effects of temperature on mechanical behavior of the soil. With the use of the finite element software ABAQUS, a numerical approach has been developed to evaluate the thermal response. In this method, the heat transfer process is simplified as the thermal convection between the fluid and the pipe wall, the thermal conduction in the pile, and thermal conduction in the ground. Validation of the proposed method is carried out by comparing the computed results with the theoretical solutions and data from literature. A thermo-bounding surface model is implemented into ABAQUS. The ability of the model to describe the temperature effects on mechanical behavior of clays is verified by several examples. Based on the numerical heat transfer method and the implemented constitutive model, the long-term performance of a free-head energy pile in normally consolidated clay is analyzed. In the analysis, three different load levels are considered. The effects of the seasonally cyclic thermal loading on pile settlement, shaft friction, and axial force are studied in detail.

The foam concrete serving as a tunnel damping material plays an important role in resisting seismic compression-shear load. This study aims to investigate failure characteristics and shear stress evolution law of the contact surface between the lining and shock absorption layer under seismic loads. Thus, direct shear tests are carried out under different normal stress levels by using RMT-150C electro-hydraulic servo testing machine. We obtain the variations of shear stress-shear displacement curves, peak strength, residual strength, shear stiffness as well as the failure characteristics and physical state changes of the contact surface through tests. Meanwhile, according to the FEM simulation results, the distribution and evolution laws of shear stress in contact surface are studied. The results show that the failure characteristics of the contact surface between the lining and foam concrete are affected by the normal stress and the density of foam concrete. In addition, shear stress-shear displacement curves are divided into three types which could be converted to each other with the changes in foam concrete density and the normal stress. The influences of the foam concrete density and normal stress on the peak strength, residual strength and shear stiffness of the contact surface are mutual, and the influence of normal stress weights higher. Finally, according to the variation of shear stress-shear displacement curves, the composite index model is proposed to reflect the damage evolution and friction-slip process of the contact surface between the foam concrete and lining under the compressive-shear load. Therefore, this study can provide certain references for the effective aseismic design.

When dynamic time-history calculations are carried out by using the particle discrete element method (DEM), the absorption boundary condition must be applied to avoid the reflection of outward propagating waves back into the model at artificial boundaries. By considering the various radius of particle elements on the artificial boundaries and their uneven boundary surfaces, the equivalent equations for DEM is obtained based on the boundary conditions of the viscous and viscoelastic continuum and free field. Calibration factors are introduced into the equivalent equation of viscous boundary condition for DEM, and a ratio-iterative method is proposed to determine the values for optimum waves absorption quickly. Numerical models for the viscous, viscoelastic and free-field boundaries are established using the 2D particle flow code (PFC2D). We also analyze the effects of particle distribution patterns on the radius and velocity of particles on the viscous boundary and the process of the ratio-iterative method. The validity of the setting method for viscoelastic boundary condition is verified with examples of the external source problem and the Lamb problem. The free-field boundary for DEM is applied to a tunnel example for the validation.

In order to resolve the technical problem of tunnel face collapse in the tunnel which is overlaid by quicksand stratum, a typical tunnel below the dynamic water-rich and weakened sand stratum of Qingdao metro No.2 line (the section of Pi-Miao), was taken as a research model, and a failure mechanical model was established according to the actual destructive characteristics. Based on transformation equilibrium from work to energy, the upper bound analysis of tunnel face stability was conducted. By the upper bound analysis and the strength reduction/gravity loading method, the tunnel face safety factor was obtained, and the critical soil rupture range given different soil cohesions, frictional angles, densities, water-resisting layer thicknesses and tunnel excavation heights were obtained as well. The theoretical research showed that tunnel face safety factor increased and stability was enhanced with the increase of cohesion, frictional angle and water-resisting layer thickness, and safety factor and face stability decreased with the increase of density and tunnel excavation height. A series of numerical calculation models was established to verify the accuracy of theoretical research, which obtained typical failure modes and critical parameters of tunnel face overlaid by quicksand stratum. Results from this research provide theoretical guidance and scientific solutions to the transfixion of Qingdao No.2 metro line and other cases with similar geological conditions.

An improved limit analysis method was proposed for the bedding rock slope reinforced by pre-stressed anchor cables under seismic loads, according to their typical models. Especially, this improved method considered the force variation of anchor cables when the slope was in a limit equilibrium state. First, based on the assumption of the rigid rock mass and the limit equilibrium method, the formula of the force variation of anchor cables was derived according to the force changes of anchor cables during the sliding process. In this study, the earthquake effect only referred to the influence of the transmitted waves on the stability of the slope when the transmitted waves were generated in the propagation process of the seismic wave into the sliding surface. Second, from the perspective of power, a theoretical formula was derived for calculating the dynamic safety factor of bedding rock slope by combing with the upper bound limit analysis method and the strength reduction method. Finally, in the case study, we analyzed the difference of dynamic safety factors by using two methods with or without considering force variations of anchor cables in the sliding process. Moreover, this study discussed the calculation results by using these two methods under the variations of amplitudes and incident angles of P wave, and the variations of cohesion and friction angles of sliding surface, respectively. The results show that the variation laws of dynamic safety factors obtained by these two methods are consistent. However, the variation degree of dynamic safety factors by the improved method is significantly lower, which indicates that the anti-seismic effect of anchor cables has been fully reflected. This study can provide references for the design and the stability analysis of bedding rock slope reinforced by pre-stressed anchor cables under seismic loads.

In this paper, a general particle dynamics (GPD) was proposed to reveal the mechanism of rock fragmentation by TBM cutters. The GPD method is regarded as a novel meshless numerical method by introducing the particle failure criterion. The numerical models were established for intact rock fragmentation by a single TBM cutter and double TBM cutters using the GPD method, and the whole processes were successfully investigated. It was found that numerical results obtained by the GPD method were in good agreement with experimental results and previous numerical results, which verified the validity of GPD method for rock fragmentation by TBM cutters. Moreover, the GPD method was adopted to establish numerical models by considering the influences of joints and high confining pressure on rock fragmentation by TBM cutters. The results showed that the inclination angle of joints significantly affected the efficiency of TBM cutting, and the effects were quite different with increasing the inclination angle of joints. The high confining pressure greatly influenced the characteristics of crack initiation, propagation and even the depth of rock fragmentation by TBM cutters. Therefore, the GPD method demonstrates great potential to investigate rock fragmentation by TBM cutters.

In order to obtain the critical slip surface of the three-dimensional slope, the construction method of six parameter-controlled three-dimensional slip surface is proposed, and the objective function based on the minimum potential energy stability analysis method is established. The genetic algorithm is used to search the critical slip surface, and the corresponding search process is developed. To validate the rationality of this method, results from this method are compared with the critical slip surface and the minimum safety factor obtained by other methods, and the critical slip surface obtained from the indoor model test is compared against the theoretical search results. The results show that the search method in this paper can achieve steady convergence. Simultaneous change of multivariate search is performed for the examples in this paper, and the results are close to the results from limit equilibrium method, which indicates that the proposed search method is reasonable. The theoretical calculation results are close to the critical slip surface of the slopes obtained from the model loading tests, which again proves that the critical slip surface searching method of the three-dimensional slope is feasible.