In this study, cyclic shear test were conducted on the crusted structural surface replicated with similar material under normal stress. The shear strength, deformation and morphology evolution characteristics of joints under cyclic shearing were studied without filling, with 3 mm yellow mud infilling, 3 mm plaster infilling and 3 mm detritus infilling, respectively. It was found that the failure mode of joints developed under cyclic shearing. Joints without filling changed from shear failure to wear-out failure, while joints with yellow mud mainly showed slip failure. Joints with plaster changed from shear failure to slip failure and then to wear-out failure, while joints with detritus changed from shear failure to wear-out failure. With the increase of cyclic shear times, the peak shear stress of joints without filling and with yellow mud infilling decreased in deceleration rate, while those with plaster and detritus infilling decreased firstly and then increased in deceleration rate. These results indicate that filling material can weaken the shear strength of joints under cyclic shearing, especially for those with yellow mud filling. Under different filling conditions, the average dilation angles all reduce in deceleration rate under cyclic shearing. The shearing shrinkage is more obvious in filled joints, in which the shearing shrinkage is the smallest with yellow mud infilling and the largest with plaster infilling. The wear degree of filled joints after cyclic shear is lighter than that of unfilled joints, while the wear degree is the smallest when filling yellow mud, the wear degree is relatively biggest when filling plaster.

Based on Nishihara model, the first derivative of the deviatoric strain tensor of the viscoplastic element is assumed to be proportional to the difference between the transient deviatoric stress tensor and the steady-state deviatoric stress tensor. Thus, the constitutive equation of ground in viscoplastic region is obtained. The viscoelasto-viscoplastic solutions are derived by means of Laplace transform and Laplace reverse transform. While the time t is close to 0, the viscoelasto-viscoplastic solutions could be degenerated into linear elastic solutions. While the time t tends to infinity, the solutions could be simplified into perfectly elastic-plastic solutions. The time-dependent performances of ground displacement, ground stress and the plastic zone are investigated through an engineering example. When the support force remains constant, the ground displacement and the radius of plastic zone increase continuously with time and then tend to be stable. The influence of time-dependent behavior on the radial stress and tangential stress of viscoelastic region are not very obvious, while the influence on the tangential stress in viscoplastic region is much larger. For the rock mass around the wall in viscoplastic region, the tangential stress changes faster with time. In addition, the tangential stress of tunnel wall under different supporting forces is larger at the initial stage of support, so the strategy of timely support should be adopted. Considering the influence of ground time-dependent performances on the support, the ductile support is suggested to ensure the stability of the surrounding rock and lining.

The acoustic emission (AE) research of rock bridge failure is usually only based on the analysis of related parameters such as counting, energy and so on. In order to study the AE frequency spectrum characteristics of rock bridge in direct shear test, the characteristics of AE frequency spectrum are analyzed using time frequency analysis method. The results show that the amplitude is more obvious and the time of intense response is earlier compared with the counts rate in direct shear test. The AE dominant frequency is basically in the high and low frequency bands. The characteristics of the dominant frequency dispersing to intermediate and the low frequency having high amplitude can be used as precursor information of main fracture, and the former appears earlier. Compared with the result from uniaxial compression test, the amplitude in the main fracture stage is obviously different with intermediate frequency occurs at this time. The amplitude of AE is determined by the stress of particles, and the frequency is related to the final displacement of the particle force chain. With the increase of rock bridge width, the dominant frequency band increases, while the normal stress increases, the average frequency decreases. This study can provide reference for the stability analysis of joint slope.

The strength of fault is significently influenced by the strength parameters of fault gouge, which are closely related to the consolidation degree and water content. In this study, specimens of fault gouge with different consolidation degrees and water contents are subjected to the direct shear test to obtaincohesion and internal friction angle , which are the dominant strength parameters of fault gouge. Based on the results of test, the relationships between the two strength parameters and consolidation degree and water content of fault gouge are analyzed and following conclusions can be drew. The internal friction angles of fault gouge specimens with different consolidation degrees decrease as the water contents increase, but the decrease of internal friction angle is slight. The trend between cohesion of fault gouge specimens of different consolidation degrees and the water content can be divided into three stages: ascent stage, steep descent stage and slow descent stage. These three stages are separated by the first inflection point and the second inflection point. As the consolidation degree increases, the internal friction angle of fault gouge increases gradually but slightly. Meanwhile, the descresing rate of the internal friction angle of fault gouge tends to descrese with the increase of the water content. As the consolidation degree increases, the water content of fault gouge whose cohesion is at the first inflection point increases rapidly, while the water content of fault gouge whose cohesion is at the second inflection point changes slightly.The relationship between consolidation degree of fault gouge and the water content of fault gouge whose cohesion is at the first inflection point can be described by a quadratic polynomial function.

To investigate the effect of temperature on mechanical parameters of a weak interlayer in water-saturated slope, the interlayer is regarded as a linear elastic body of coupled solid-liquid phase. A mechanical model for the thermal-pore water-mechanical interaction of the interlayer in water-saturated slope is established, and its coupling control equation is deduced. Using the physical similarity method, an experimental model similar to the slope prototype is established to study the variations of the inerlayer mechanical parameters caused by temperature. The proposed model is verified by comparing the theoretical results with the corresponding experimental results. It is found that the pore water pressure coefficient and thermal expansion coefficient are the key factors causing the increase of pore water pressure in the saturated interlayer. The pore water pressure coefficient depends on the compression characteristics of pore drainage and the solid medium. The larger the difference between these two compression characteristics, the larger the pore water pressure coefficient. Thermal expansion coefficient of pore water and thermal expansion coefficient of pore volume are the main factors affecting thermal pressure coefficient. A larger difference between them can result in a greater thermal pressure coefficient. Pore water pressure in the saturated interlayer slowly increases first and then dramatically increases with the increase of temperature, while the cohesion and shear strength of the saturated interlayer tardily decrease with the increase of temperature. Theoretical values of the saturated interlayer pore water pressure are in good agreement with the results from model test. Therefore, the proposed model can reflect the change characteristics of pore water pressure at different temperatures, which can provide some useful references to predict and control the stability of similar saturated slopes containing a weak interlayer.

The acoustic emission tests in rockburst process of granite under true-triaxial loading conditions were conducted by using a true-triaxial rockburst test system. It reveals the spatial and temporal evolution characteristics of acoustic emission during rockburst process. At the early stage of rockburst, the count-hits ratio is relatively small, and the dominant frequency is mostly concentrated in low frequency bands. The dominant band, whose energy proportion is the maximum based on wavelet analysis, is characterized as low frequency less than 125 kHz. At this stage, the scattered AE events turn to concentrate, and new AE events migrate gradually from free face to interior of rock specimens. Immediately before rockburst, the count-hits ratio shows a continuous sharp increase, AE signals with high frequency, intermediate frequency and low frequency begin to teem, and the dominant band transfers from discrete state to continuous one. Meanwhile, the proportion of dominant band of 250-500 kHz increases significantly. Simultaneously, away from the free face the nucleation zone of AE events forms, while near the free face the number of AE events increases. During the process of rockburst, the acoustic emission is characterized as low frequency and high amplitude, and it has the dominant band of 62.5-125 kHz. The AE events are active near the free face and the nucleation zone of AE events can be clearly observed in interior of rock specimens, which indicates that the spalling or slabbing occurs near the free face and the macro shear failure occurs away from the free face.

Based on the piecewise-linear method, a one-dimensional self-weight consolidation model (i.e. SWC model) of saturated soft soil, which considers the large strain and non-linear changes of geotechnical parameters, is developed in this paper. The calculation results of the SWC model are validated by analytical solution and experimental measurements. The SWC model is applicable in investigating the variation processes of settlement magnitude, average consolidation degree, distribution of porosity and excess pore water pressure, etc. Then, based on field tests, impacts of parameters (e.g. initial height, initial void ratio, boundary drainage conditions of soil, and specific gravity of soil grain) on the self-weight consolidation process are studied in this proposed SWC model. Simulation results indicate that the above four parameters have significant influence on the process of soft soil self-weight consolidation. The final settlement magnitude and average strain increase with the increase in initial soil height. The boundary drainage condition only affects the time of self-weight consolidation but does not change the final settlement magnitude. The final settlement magnitude increases with increasing initial void ratio and specific gravity of soil solids. Moreover, the time of self-weight consolidation decreases with increasing initial void ratio and specific gravity of soil solids.

A sinking simulation test of deep and large caisson is designed based on the caisson of Hutong Bridge. The undrained sucking sediment sinking of the caisson in the water is simulated and spatial distribution of the stress at the tread and oblique planes of the foot blade during the sinking process of the caisson. It concludes that the maximum stress occurs in corners of the caisson, the minimum stress occurs in the middle of long side, and the average stress at the short side is larger than that of the long side. Applying the reference stress at the midpoint of the long side, the total cross-section stress coefficient of the foot blade with the aspect ratio of 1.33 is calculated. In addition to be affected by spatial position, the stress on the incline plane of the foot blade is also greatly related with the mud surface height, and it changes contrary to the stresses at the tread. When the mud surface height is determined, the cubic function is used to determine its stress distribution. The influences of inclination, sand-casting and sudden sinking on the stress of the foot blade are also discussed and it is found that both can cause changes in spatial distribution of the stress of blade foot. Stresses at the tread and incline planes hydraulic gradient of the foot blade change significantly before and after the sand-casting and sudden sinking of caisson. They occur earlier than the sand-casting and sudden sinking of caisson and it is of great significance to early warning of sand-casting and sudden sinking.

 Suction anchors with taut mooring system are important anchor foundations for the floating platforms in deep water where shallow sediments are mostly saturated soft clays. The cyclic bearing capacity of the suction anchor under a combination of average and cyclic loadings is vital factor for the design of the anchor in soft clay. The paper assumes that the average shear stresses of the soil for different failure zones are uniform and the ratio of the average shear stress to the cyclic shear strength is equal to the ratio of the static load to the sum of static and cyclic loads applied to the anchor based on the failure mode of the suction anchor under inclined loadings at the optimal load attachment point. Undrained cyclic shear strengths of the soil for failure zones are determined based on curves of normalized cyclic shear strength versus normalized average shear stress obtained from laboratory tests of the soft clay samples. Soil resistances for the different failure zones are analyzed. An approach for calculating the interface depth between upper sliding wedge failure zone and deep plane flow failure zone is developed according to the horizontal stress continuity condition. Further, a simplified limiting equilibrium method is proposed for evaluating the cyclic bearing capacity of the anchor. The proposed simplified calculation method is employed to predict the model tests of the cyclic bearing capacity of suction anchors with both vertical and horizontal failure modes. The predictions are found to be in agreement with the experimental results. The smallest and greatest deviations are 0.79 % and 16.08 % respectively and the average deviation is 5.74 %. The predicted results could preferably reflect the variation between the cyclic bearing capacity and the number of cycle to failure, which verifies the feasibility of the proposed simplified method for calculating the cyclic bearing capacity of the suction anchor.

Full-flow ball penetrometers have been used widely in centrifuge testing for characterizing the undrained shear strength of soil sample. However, the soil flow mechanism and bearing capacity factor of ball penetrometers in centrifuge is significantly different from that of in the field tests, which induced by that the ball-soil frictional coefficient and the ratio of larger shaft diameter and cross-area of centrifuge test is different from that of in field tests. Large deformation finite element analyses are carried out to investigate the behaviour of ball penetrating into uniform clay to quantify the bearing capacity. Prior to conduct parametrical study, the plasticity solutions present good agreement with other previously published finite element results and experimental data. The roughness of ball has little effect on the critical depth of the cavity and the critical deep failure depth. The area ratio is shown to have significant influence on the shallow and deep bearing factors, the critical cavity depth, and the critical deep failure depth. Numerous formulas are proposed to quantify those effects to interpret the undrained shear strength profile, and to provide guidance for its application of ball penetrometers in centrifuge testing.

To test the gas permeability of rock under unsteady flow condition, a new method of calculating the gas permeability was proposed in this paper. Klinkenberg’s equation was firstly imported into the partial differential equation of unsteady flow. A numerical solution of this partial differential equation was then generated by using Picard-Newton’s difference scheme. Based on this numerical solution, a new method of searching the optimal absolute permeability and slip factor of rock was proposed by fitting the variation of gas pressure under unsteady flow condition with least-square method. The numerical calculation in this method was proven to be stable, convergent and insensitive to step size and random errors, and the slight variation of absolute permeability and slip factor can be successfully captured. Finally, some granite samples from the surrounding rock of the Beishan exploration tunnel, which was located in the pre-selected area for China’s high-level radioactive waste disposal, were tested and the absolute gas permeability and slip factor were successfully identified by the proposed method. The input data needed in the proposed method can be easily obtained by simple indoor test and the key algorithm is convenient for numerical realization. Thus it can be applied to the gas permeability test of surrounding rock materials in underground engineering.

With horizontal slice backfilling of mines as the background, three concepts, namely initial stratified damage, loaded damage and total damage, were put forward for horizontal stratified cemented backfill. Based on the damage mechanics theory, the damage evolution model and the damage constitutive model were built with the consideration of the stratification effect. Based on the method of full differential function, a strength criterion of backfill, which considered the effect of stratified structure, was established. Then, tests to investigate the mechanical properties of backfills with different number of layers were conducted. Based on the results of tests, several conclusions can be drew. The theoretical curve of the damage constitutive model and the strength criterion fit the experimental results well, which verifies the correctness of the newly proposed models. The initial stratified damage increase at a quadratic polynomial function with the increase of the number of layers. The total damage of backfills is aggravated by the coupling effect of stratification and load. The total damage rate of backfill tends to firstly increase and later decrease. Before the peak, the total damage rate increases with the increase of the number of layers; after the peak, the total damage rate decreases with the increase of the number of layers, while the maximum of total damage rate increases with the increase of the number of layers.

This study is aimed to investigate the effects of the specimen size and prefabricated crack length on the rock fracture toughness. Then three-point bending tests are conducted on semi-circular limestone specimens (NSCB) with four disc radii of 25.0, 37.5, 50.0 and 75.0 mm, respectively. The ratio of prefabricated crack length to the radius of disc specimens (dimensionless crack length) ranged from 0.3 to 0.7. The results show that the failure process of NSCB specimens exhibits the characteristic of an obvious brittle failure. Compared with the prefabricated crack length, the specimen size has more significant influence on the peak load and the shape of load-deformation curve. As the specimen size increases, the testing values of fracture toughness increase continuously. Moreover, the increase rate increases with the length of the prefabricated crack length. For a selected-size specimen, the testing value of fracture toughness reduces with the increase of prefabricated crack length. In order to reduce the effect of size effect in the fracture toughness test, it is suggested that the radius of NSCB specimen is greater than 37.5 mm and the dimensionless prefabricated crack length is in the range of 0.4~0.6. This study provides certain references for the promotion of the international method and the accurate test of rock fracture toughness using NSCB specimens.

An approximate analytical model has been established in this paper to study dynamic interactions between the saturated soil and the pile group with assumption of poroelastic half-space. The piles are modeled as Bernoulli-Euler elastic beams, and the saturated half-space is governed by Biot’s dynamic equations. The pile beams and the saturated half-space are coupled through receptance matrix established at pile/soil interaction points in frequency-wavenumber domain. Results are obtained by fast Fourier transform. The dynamic impedances of the pile group have been analyzed. Meanwhile, the results demonstrated that the type and frequency of the excitation force and the soil permeability have significant influences on displacement and pore pressure responses of the saturated ground. Especially, it is noted that peak frequencies of pile-foundation impedance increase along with the increasing of soil permeability.

The measurement and prediction of permeability coefficient of unsaturated soils with different initial void ratios are the basis of seepage analysis and hydro-mechanical coupling study of unsaturated soils, so corresponding work has great significance. As an example, Hunan red clay was chosen to prepare remoulded samples of five different initial void ratios by using jack. Then, the soil-water characteristic curve was measured by using pressure plate apparatus. Afterward, the saturated permeability coefficient was measured by variable head method, while the unsaturated permeability coefficients with different initial void ratios were obtained by instantaneous profile method using a self-made plexiglass barrel device. CCG (Childs and Collis-George) correction model and Tao-Kong model were selected to predict the unsaturated permeability coefficients, and their effectiveness was demonstrated through comparing the prediction values with the measured values. On the basis of experimental results and model prediction values, the influence of initial void ratio on unsaturated (relative) permeability coefficient was examined. The results show that the permeability coefficient of unsaturated Hunan clay decreases with the increase of matric suction in a double logarithmic coordinate, which presents a larger slope in low matric suction stage within 100 kPa while a smaller slope in high suction stage above 100 kPa. The predictions of Tao-Kong model are in good agreement with the measured values, yet great predicted errors exist in the results by the CCG model. For the suction stage larger than air entry value, the initial void ratio has little influence on the unsaturated permeability coefficient and has a greater influence on the unsaturated relative permeability coefficient. In the condition of same matric suction, a smaller initial void ratio results a greater relative permeability coefficient.

Calcareous sand, a type of geo-material with special structure and mechanical properties formed due to the marine biological deposition process, is the material used in the dredger filling engineering of South China Sea. To further understand its creep properties, a series of long-term creep tests under different confining pressures were carried out using triaxial rheological apparatus on calcareous sand sampled from a coral reef island located at South China Sea. Experimental results show that the damping creep of saturated calcareous sand occurs under constant pressure load that is less than the failure strength. The deformation increases with time, while the deformation rate decreases until the deformation becomes stable. The larger the applied stress is, the longer the deformation becomes stable. The creep deformation is positively correlated with the deviatoric stress, while it is inversely correlated with the effective confining pressure. The relationships of strain-stress and strain-time are nonlinear. It is found that the strain-time relationship of calcareous sand can be described by power function. A new creep model of calcareous sand, considering the relationships of four parameters (i.e. creep strain, time, deviatoric stress and effective confining pressure), is proposed in this paper. Compared with the traditional empirical Mesri creep model, it is unnecessary to perform the conventional triaxial test to determine the peak failure strength in the proposed new model. Less experimental work is required and thus it is much easier to be used.

To reveal the intrinsic correlation between the micro-meso moisture-distribution and the engineering properties of the solidified sediments, the lake sediments were solidified by cement. The mechanisms of water transfer and pore structure development in the process of sediment solidification were explored using the nuclear magnetic resonance test. The relationships of the moisture parameters with the permeability k, strength and deformation modulus of solidified specimens were investigated. The experimental results showed that the hydration rate was fast within 7 days, and the pore water in the sediment solidification transferred into hydration water. Meanwhile, the most probable pore size decreased, which caused a decrease in k and increases in and , and hence inducing the formation of the pore structure. However, after 7 days, the hydration reaction continued slowly only if the water diffused and penetrated through hydrates membrane. Significant increases in and were also observed for this case. The hydration water varied as a linear function of lgk, however, exponential relations were existed between and , and . Based on the quantified relation model between the micro moisture parameters and the engineering properties, the macro and micro evolution mechanisms of the cement solidified sediment were revealed.

Boundary effect has a significant influence on the permeability coefficient measurement of coarse grained soil in laboratory seepage tests. Based on the composition characteristics of the pore Vb between soil particles and the seepage equipment wall, a parameter called boundary pore ratio of the soil sample eb was defined according to the ratio of the volume of boundary pore Vb to the particle volume Vs which concerns the constitution of boundary pore. Furthermore, based on the uniform circle packing principle, the boundary particle accumulation patterns were established, and a plane geometric calculation model of boundary pore ratio considering the wall treatment layer thickness was proposed. The expression of eb was finally derived through space and particle gradation correction. Based on the principle of the equality of eb and sample pore ratio e, a criterion called ‘equivalent pore ratio method’ for confirming the optimal thickness of the wall treatment layer hopt was presented. At last, the relationship between the seepage equipment diameter D and the boundary effect was discussed. The researches indicate that the redundant pore caused by the boundary effect would raise the seepage velocity near the seepage boundary. There are three typical plane accumulation patterns of the boundary particles, including two particles pattern, three particles with acute packing angle pattern and three particles with obtuse packing angle pattern, of which the corresponding eb increases accordingly. Furthermore, hopt is comprehensively determined by sample gradation, compression degree, particle density, boundary particle accumulation pattern and D. The estimate value of hopt agrees well with the test result. The influence of the boundary effect on the seepage test result becomes less significant as D exceeds eight times of the maximum particle size of soil sample.

The magnitude and distribution of earth pressure against retaining wall is a common concern in the design of retaining wall. The distribution of earth pressure on a retaining wall normally presents nonlinear feature under the influence of boundary conditions, friction of wall back, etc. However, the accuracy of several current earth pressure calculation methods considering soil arching effect needs to be verified. In this paper, a series of 2D visual model tests were carried out on a self-developed apparatus measuring the earth pressure of the retaining wall, using test setup and analogical soil that was mixed with elliptical steel rods. The application of analogical soil in collaboration with loadometers to accurately measure the variation of earth pressure can reduce errors in the measurement of lateral pressure using the pressure cell. Two conditions include static condition and active translation model were considered in simulating and measuring the distribution of earth pressure. The results were compared with that of several theoretical methods. Due to impacts of the friction on retaining wall and soil arching effect, both the static and active earth pressure present nonlinear distribution. Meanwhile, the measured static earth pressure is less than theoretical value, while the measured active earth pressure is in good agreement with that of Paik’s method. Due to the influence of wall friction, the inclination angle of sliding surfaces is less than that of the theoretical value at the rupture surface. Moreover, inclination of stresses caused by the friction between wall induces the development of arch-shaped slip surfaces.

The reactive MgO-fly ash cementing materials were innovatively introduced into the improvement of dredged sludge from East Lake, Wuhan by the combined technology of carbonation-solidification. Through unconfined compression strength, scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) tests, the effect of CO2 carbonation on the mechanical properties and microstructure was investigated under different carbonation modes, carbonation time, ratios of MgO to fly ash and dosages of binding agent. The results indicate that the strength of solidified sludge is evidently increased due to carbonation, accompanied by a narrowed compaction stage of stress-strain curve. The sludge specimens with different binding agents have different optimal pressurization modes, which determines the CO2 intake amount of solidified sludge for the same carbonation period, and affects the strength gain of carbonated samples. Low amount of reactive MgO leads to relatively low compressive strength, which firstly increases and then decreases as the carbonation time increases. The strength of carbonated sludge with relatively high amount of reactive MgO reaches rapidly certain higher value with carbonation time and then increases slowly. The microscopic experiments demonstrated that the formation of magnesium carbonate (e.g. hydromagnesite, dypingite and nesquehonite) is the main reason for enhancing the compressive strength of specimens by the combined carbonation-solidification technology. The expansibility and cementation of the magnesium carbonate promote the transformation of pore in aggregates into inter-granular pore, which makes the soil more compact and increases its compressive strength of carbonated-solidified sludge.

It has been recognized that the stress-dilatancy behaviour of granular soil depends on its material state. To consider such state-dependence, a variety of state parameters were suggested phenomenologically and incorporated into existing stress-dilatancy equations, e.g. Cam-clay equation, modified Cam-clay equation, by experience. In this study, a novel state-dependent stress-dilatancy equation is developed by using fractional stress gradient, where physical meaning of the fractional order is provided. The obtained stress-dilatancy ratio is determined by three factors: the fractional order, current load stress, and the distance from current stress to critical state stress. When the fractional order is larger than 1, the stress-dilatancy curve shifts from the modified Cam-clay stress-dilatancy curve towards the Cam-clay stress-dilatancy curve. However, when the fractional order is smaller than 1, the stress-dilatancy curve is located above the modified Cam-clay stress-dilatancy curve. When the fractional order is equal to 1, the proposed stress-dilatancy curve coincides with the modified Cam-clay curve. To validate the proposed approach, a state-dependent fractional plasticity model for sand soils is established by using the proposed stress-dilatancy equation. Then, a series of drained and undrained triaxial compression tests on sand and rockfill with different initial states is simulated and compared, from which a good agreement between the model simulations and the corresponding test results can be observed. Comparison with the model predictions of the UH sand model indicates that the UH model gives better prediction.

In order to improve the prediction of rock burst with charge induction, the charge monitoring test was carried out on fractured coal samples under uniaxial compression by the self-developed charge monitoring system in the laboratory. The failure types, mechanical properties of coal samples and the law of the charge time-frequency domain signals under different damage types were analyzed emphatically. Moreover, the test results were verified in practice. The results show that the deformation and failure characteristics of coal samples can be divided into single shearing type, conjugated shearing type, and shattered type. The charge time-domain signals of coal samples with single shearing occur only in the early stage of post peak failure, and the stress falls to about 97% . Since the signal characterizes a single burst with higher amplitude, discrete charge frequency distribution and the main frequency of 250 Hz, it is difficult to capture the precursory information. Those characteristics of charge signal indicate that the corresponding engineering coal body may occur a localized failure with releasing the accumulated energy instantly, which leads to significant impact hazard. The charge time-domain signals of coal samples with conjugated shearing are produced in the later period of intensified damage, and the stress reaches (85%~100%) . The signal has an interval burst, and the main frequency is 150 Hz. This indicates that the partitioned rupture occurs with accumulated energy released gradually, and the impact hazard degree is the second. The charge time-domain signals of shattered coal samples are generated in the early stage of intensified damage, and the stress reaches (70%~85%) . The premonitory information is easy to capture because the signals burst continuously with low amplitude and the main frequency of 0 Hz. These results indicate that the coal body may be homogeneously broken with accumulated energy released slowly, and the impact hazard degree is lower. The field monitoring results show that characteristics of charge signals both in crushing area of working face and the occurrence of coal firing gun are highly similar to those charge signals which are produced by coal samples in homogeneous shattered type and single shearing type in the laboratory. These results verify the reliability of the results obtained from the laboratory.

A new model characterized as describing increase in pore-water pressure in fully saturated fine sand specimens, through experimental studies include a resonant column test, a series of drained/undrained multistage strain-controlled cyclic triaxial (MSCCTX) tests and drained/undrained single-stage strain-controlled cyclic triaxial (SSCCTX) tests on a schistous fine sand named “Nanjing sand”, is development in this paper. It follows the theoretical framework of Martin’s semi-theoretical model presented by Martin and Byrne. The increase in pore pressure is caused by the change in volumetric strain under cyclic shear load. Thus, this proposed model is controlled by strain. Based on the drained SSCCTX tests, by introducing the concept of volumetric threshold shear strain (??tv) and a representative volumetric strain in 15 cycles, volumetric strains can be normalized by cyclic shear strains, and then a three-parameter incremental shear-volume strain coupling equation is proposed. Combining experimental results of drained and undrained SSCCTX tests, the relationship between volumetric strain and pore pressure ratio is established. Thus, a new pore water pressure increment model coupling new shear strain-volumetric strain is also established considering volumetric strain increment and elastic rebound modulus. Finally, this new pore pressure model is validated, and the predicted pore pressure ratios in the model match well with those obtained from tests including an undrained SSCCTX test and a stress-controlled cyclic test.

The displacements of reinforced soil retaining walls under seismic loads have a significant influence on their seismic performances. In order to calculate the displacements under earthquakes, Newmark’s sliding block method is usually used in design. As the traditional Newmark’s sliding block method neglects changes of soil strength, so the peak or residual strength is used in calculation, which lead to the calculated values being smaller or bigger than the actual values. Under the assumption of a two- wedge failure mechanism, sliding safety coefficient of reinforced earth retaining walls is obtained using the mechanical equilibrium equations of wedges. At the same time, the strain softening characteristic of backfills is considered by introducing thresholds of displacements. The following conclusions have been obtained by comparing the calculated values with the model tests. Compared with the single wedge method, the two-wedge method can better describe the actual failure mode of model walls and the calculated yield acceleration coefficient is closer to the test values. Compared with the calculated displacements by using the peak or residual strength, the calculated values from the proposed method where strain softening is considered is closer to the model test values.

Municipal solid waste (MSW) is characterized by large pores, so preferential flow is a common flow pattern in MSW. As the complex composition of MSW and difficulty in sample sectioning are concerned, a test method is developed to make the sectioning easy. MSW samples corresponding to different depths of a landfill are used in the dye tracer test. The MSW columns are dyed and then sectioned in horizontal and vertical directions. Digital image processing technology is used to investigate the preferential flow pattern in MSW. There are many macropores and less matrix in MSW, and the dyeing pattern is different from that in soil. The proportion of dyeing matrix in MSW is higher than that in soil. The degree of preferential flow can be described by dyeing area ratio and the dyeing depth. The dyeing area ratio reflects the ratio of large pores involved preferential flow and is proportional to the degree of preferential flow. The dyeing depth is related with the matric flow ratio in upper part of the column. The lower the matric flow ratio, the deeper the dyeing depth will be. The dye tracer tests with different infiltration rates show that the degree of preferential flow in MSW increases with infiltration rate. The tests with different initial water contents show that preferential flow is more obvious in MSW with high initial water content. The tests with samples of different depths show preferential flow is more easily to happen in shallow MSW. The results of vertical section tests are consistent with that of horizontal sectioning tests, which show that the test method is applicable and has no influence on the preferential flow pattern in MSW.

The dip angle of structural plane has great influences on mechanical properties of rock samples. In this study, the rock samples with structural planes of 15°, 30°, 45°, 60° (the structural plane of 0° is along the horizontal direction)and the intact rock samples are prepared. Uniaxial tests and triaxial tests with confining pressures of 100 kPa, 200 kPa and 300 kPa are performed to investigate the influence of dip angle of structural planes on their mechanical and deformation features. Test results demonstrate that: 1) the stress-strain curve exhibits the four typical stages, which are similar to those of rock samples when the dip angle is no more than 30°, but the last phase exhibits the characteristics of softening followed by hardening when the dip angle is equal to 45°or 60°, 2) for all the samples, the contractive volumetric strain is firstly shown, and then the dilative volumetric strain is followed. With the increasing confining pressure, the maximum value of contractive volumetric strain does not change much, but the dilative volumetric strain obviously decreases. With the increase of the dip angle, the contractive volumetric strain slowly increases, and the contractive volumetric strain suddenly changes when the dip angle increases from 30°to 45°. However, the dilative volumetric strain obviously decreases. Finally, the disadvantage of Jaeger’s single structural plane criterion is pointed out, and thus the criterion has been improved. The verification results denote that the improved Jaeger criterion is in good agreement with the experimental results.

Considering the rheological properties of soft clay and the engineering practice, the continuous boundary condition is applied to construct a one-dimensional consolidation model for viscoelastic foundation under time-dependent loading condition. The analytic solution for one-dimensional consolidation of saturated soft clay under time-dependent loading and continuous drainage boundary is solved by using separation of variables and variation of constants methods. Furthermore, influences of boundary conditions and soil parameters on excess pore water pressure and effective stress are studied. It is found that the variation of excess pore water pressure is directly influenced by permeability of upper- and lower- boundary. The consolidation rate is larger at the location closer to the boundary with higher permeability, and with decreasing the distance from drainage boundary, the influence on consolidation rate increases gradually. Moreover, the effective stress is significantly influenced by permeability coefficient, elastic modulus of Kelvin model and viscosity coefficient.

A profound understanding of hydro-mechanical coupling process in rainfall-induced landslides and its formation mechanism are prerequisites for the accurate early warning. In this study, a typical landslide in Wenchuan earthquake region is selected for a series of full-scale artificial rainfall tests, to simulate failure progress on gravelly soil slope of landslide deposit. Two classical soil-water characteristic curve (SWCC) models, Brooks-Corey (BC) and Van Genuchten (VG) model, are used to establish the SWCC during rainfall infiltration. Then, the stress state and stability of slope are analyzed by one-dimensional variably saturated infinite slope model. The results reveal that the slope failure is the consequence of dual-permeability interaction between the preferential flow and the matrix flow. The VG model is capable of fitting SWCC of wide-graded gravelly soil during wetting process for preferential flow. Comparison between stability analysis and observation indicates that changes on volumetric water content, matric suction and surface inclination are consistent with calculated suction stress and safety factor. Finally, the seepage erosion in unsaturated slope is proved by the tests, and the migration of fine particles is statistically analyzed. Based on the physical model test, this study reveals the coupling process of soil-water characteristic curves and slope failure with the influence of preferential flow under rainfall conditions, which can provide reference and basis for the prediction and early warning of unsaturated slope instability.

Shear property and water-induced deterioration of discontinuities with different joint wall material (DDJM) play an important role in studies on stability of rock mass and mechanism of geological hazards in the Badong formation. Scanning electron microscope tests and uniaxial compression tests under natural and saturated conditions were firstly conducted on three typical DDJM (silty mudstone / argillaceous siltstone, argillaceous siltstone / argillaceous limestone, and argillaceous limestone / mudstone) to investigate microscopic characteristics of joint walls and water-induced deterioration of the macro mechanical property. Shear properties of these three types of DDJM under natural and saturated conditions were then studied by direct shear tests. A formula to predict the basic friction angle of DDJM was proposed. The results indicate that: 1) the water-induced deterioration of shear property of DDJM is greatly affected by the content of clay minerals and microscopic structure of their joint walls; 2) the basic friction angle of DDJM falls between that of discontinuities with identical joint wall material(DIJM) same with two joint walls of DDJM, and is close to value of DIJM with joint walls same with the softer side of DDJM; 3) the coefficient of water induced deterioration of DDJM is mainly controlled by softening coefficient of the softer joint wall; 4) the coefficient of water-induced deterioration of these three types of DDJM (silty mudstone / argillaceous siltstone、argillaceous siltstone / argillaceous limestone、argillaceous limestone / mudstone) in the Badong formation are 0.915, 0.951 and 0.731, respectively.

Hydraulic fracturing is a pressure relief and permeability enhancement technology widely applied in low permeable coal seam. To study the impacts of in-situ stress, natural fractures and flow rates on fracture propagation and fracture network, the true triaxial hydraulic fracturing experiments of large-scale coal seam were conducted on large-scale true triaxial hydraulic fracturing experimental system and tracer technique in hydraulic fluid. The propagation and geometry of hydraulic fractures were described by cutting the fractured specimen. The relationship between fracture width and in-situ stress was analyzed and fracture network mechanism in coal rock was studied. The experimental results indicate that: 1) Fracture network forms when hydraulic fractures propagate along cleats after initiation. 2) Hydraulic fracture is greatly affected by in-situ stress. Complex fracture morphology easily forms when maximum horizontal stress is close to vertical stress and both of them are much bigger than minimum horizontal stress. 3) Natural fractures in coal rock are the foundation of fracture network, which is also affected by flow rate. 4) Hydraulic fractures may steer and branch under the impacts of natural fractures at local, but they finally adjust the propagation to the direction of maximum horizontal stress. 5) During hydraulic fracturing process, the changes of fracture width are affected by factors include in-situ stress, flow rates, natural fractures in coal rock, etc. The results could provide technical support for understanding fracture network mechanism and determination of field fracturing parameters in CBM reservoir.

Many geotechnical problems, such as deep excavation of mine, underground storage of nuclear waste and pervasive application of energy piles, need to consider the effect of temperature on the mechanical properties of soft rock. In order to describe the mechanical properties of soft rock more comprehensively, based on the superloading and subloading concept and the concept of temperature-equivalent stress, a new elastoplastic constitutive model for soft rock, which is able to consider the effect of temperature, intermediate principal stress, structure and overconsolidation at the same time, is proposed in the tij stress space. All parameters of the new model have clear physical meaning. By comparing theoretical curves with the experimental results, the correctness of the proposed constitutive model is verified. Finally, by changing the model parameters, the performance of the new constitutive model is analyzed and discussed. Increasing the value of m which controls development of overconsolidation ratio or reducing the value of m* that controls structural state development will enhance the shear strength of soft rock. With the increase of temperature, the shear strength of soft rock declines. The higher the initial overconsolidation ratio is, the more apparent the dilatancy of soft rock becomes. But when initial structure is large, the volume strain of soft rock presents shear contraction in the final shear stage.

In this paper, propagation characteristics of seismic wave field, dynamic evolution and failure mechanism of a rock slope with connective structural surface were studied. Based on the engineering background of the bank slope of Jinsha River Bridge, FEM analysis and a shaking table test were designed and completed. According to the results of FEM analysis, the connective structural surface caused significant change of the earthquake wave field. After multiple superposition of seismic wave of slope surface and connective structural surface, the PGA of the slope was 1.8 times of the PGA of a homogeneous slope. With the increasing earthquake intensity in the shaking table tests, the first sudden change happened at grade Ⅷ intensity, and the slope changed from elasticity to plasticity. Expansion, penetration, and even overall sliding happened at the slope at grade Ⅸ intensity. Slide surface slipped along steep structural surface on trailing edge and shear outlet neighboring slope toe by shear and slides along penetrating weak structural surface. Therefore, slope may be at the status of plasticity after an intensive earthquake and it may lead to landslide with external forces.

The composition of municipal solid waste (MSW) is very complex. A lot of fibrous materials in MSW, such as plastic, textile, leather and so on, make the mechanical properties of MSW obviously different from common clay and sand soil. It is difficult to simulate the mechanical behavior of MSW by the existed soil constitutive models. In this paper, the MSW is regarded as a combination composed of fibrous materials and paste, where paste is the residual components except for fibrous materials. The mechanical properties of MSW under the external load depend on the combined action of fibrous materials and paste. A concept of fibrous reinforcement effect parameter R is proposed in this paper. Through the development of plastic potential function considering the fibrous reinforcement, a new elasto-plastic constitutive model is established to simulate the stress-strain characteristics of MSW reasonably. The evolutionary equation of fibrous reinforcement effect parameter R is derived based on the test results. By comparing with the results of triaxial drained test and calculations of other constitutive models, it is found that the proposed model can perfectly reflect the stress-strain responses of MSW, especially for the upward curvature of MSW stress-strain responses at a larger strains level. As a result, the reasonability and effectiveness of this MSW constitutive model have been verified. The development of MSW constitutive model presents an important theoretical basis for better servicing landfill engineering projects.

Shale gas development in commercial scale benefits from the key technology— staged fracturing of horizontal wells, and the evaluation of shale fracability is also the key for the technology. Shale fracture toughness is an important factor for the evaluation of fracability. Based on the microstructure of Model I fracture, combining the fracture mechanics with the fractal theory, this paper established a new calculation method of shale fracture toughness. The surface energy of shale can be calculated by means of crystal splitting, and the calculation results can be obtained easily via rock density and acoustic time. This paper presents a comparison between the fractal calculation results and the experimental test results and traditional predicting results. The comparison shows that the mean error of fractal calculation is 3.63% and the mean error of traditional predicting is ?10.53%, validating the accuracy of the new method. Referring to the horizontal well logging data of horizontal wells, this method can be used to establish the wellbore model-I fracture toughness index continuity. Considering the relationship between the fracability level and the probability of stimulated reservoir volume, the class III and class II fracability level is preferred. The fractal calculation method of model I fracture toughness of shale is important in the perspective of quantitative evaluation of shale gas reservoir fracability.

The occurrence of the prestress loss of anchor cable is more commonly encountered in foundation pit due to the influences of the excavation and strength of the stratum. Field tests were carried out in a practical project to analyze the effects of the tension load, tension locking mode, length of unbonded segment and cyclic loading on the prestress loss of anchor cable. The occurrence mechanism of the prestressed loss of the anchor cable was discussed. The results show that the debonding slip between the anchored solid and the surrounding stratum is an important reason for the prestress loss. A greater anchor force will induce more substantial loss of the prestress. As for the dispersed anchor cables, single tension locking is beneficial for minimizing the prestress loss. The length of the unbonded segment has an effect on the stiffness of the anchor head:a shorter length of the unbonded segment will result in more substantial loss of the prestress. Furthermore, it is found that the plastic deformation of the anchorage will be further locked under more times of cyclic loadings, which as a result reduces the loss of the prestress. The relevant rules and conclusions can provide useful references for the design and construction of the foundation pit anchorage cable.

Pillar strength is one of the important factors for evaluating the pillar stability, which is the basis for safe mining for underground deposits. In this paper, the stress state of pillars in gently inclined layered deposits was analyzed under the combined action of normal stress and shear stress using the theory of elastic mechanics, and a generalized Mohr’s circle was plotted by graphic method to characterize the stress state of the pillars. According to the relationship between the generalized stress circle and the Mohr-Coulomb strength envelope, an analytic formula for the pillar strength was developed to analyze the variation of the slope of straight line ( ) at the center of Mohr circle with the dip angle of ore body. The influence of the ore-body inclination on the pillar strength was also analyzed by numerical simulation. Result shows that is a binary function of the deposit angle and the lateral pressure coefficient . When the maximum value of is obtained, the strength of the pillar is the smallest. When the of the pillar keeps a same value and the ore body dip angle is within 5°~45°, the strength of the pillar gradually decreases with the increase of the dip of the ore body. The study can provide a reference for the estimation of the strength of the pillars in gently inclined bedded deposits.

Most analyses of geological heterogeneity are based on the assumption of repetitive deposition and stationary transition probabilities. However, the effects of tectonism and weathering are ignored, which leads to the absence of consistency verification of geological heterogeneity features (GHFs). A test method for the consistency of GHFs is herein proposed, as well as a partition method to detect stationary Markovian zones. The correlation coefficients between local transition probability matrices are adopted as quantitative criteria of consistency. The stratigraphic distribution is simulated by the coupled Markov chain model, and the deformation analysis of surrounding rock uncertainty is accomplished through Monte-Carlo simulations. Taking advantage of the borehole data in Tsingtao, the effect on stratigraphic distribution and uncertainty analyses induced by the consistency of GHFs is discussed, and the effectiveness of the proposed method is illustrated. The results indicate that the accuracy of uncertainty analyses cannot be improved with increased number of boreholes when there are conflicting GHFs. After the consistency verification of GHFs and the partition of stationary Markovian zones, the absence of local features can be avoided to improve the accuracy of uncertainty analyses of surrounding rock deformation.

Influenced by the lateral continuity effect of retaining wall’s cement soil, SMW (soil mixing wall) construction method displays significant spatial deformation characteristics of retaining wall. It is difficult to quantify the spatial deformation effect of retaining walls and the bearing capacity of cement soil, the function of spatial deformation effect and contribution of cement soil to retaining walls are not taken into account during the design of SMW construction method at present. By analyzing the deformation and stress of retaining wall of cantilever type SMW construction method, the strain energy of spatial deformation of retaining wall can be obtained. Ratio of spatial deformation effect and bearing ratio of cement soil of SMW construction method are defined according to the relationship between strain energy and resistance. Meanwhile, the analytic solution of displacement of the top of the wall is derived with a consideration of the spatial deformation effect of retaining wall based on the principle of minimum potential energy. The analytic solution is compared with the displacement calculated by elastic subgrade, the monitored displacement. In-depth discussion of factors influencing the spatial deformation effect ratio and the bearing ratio of cement soil is completed with a reference to the results of the model test. The results show that the analytic solution considering the spatial deformation effect is closer to the monitored values than the elastic subgrade method. The ratio of height to length of retaining wall, the elastic modulus of cement soil and the wall thickness in SMW construction method have a significant influence on the spatial deformation effect ratio and the bearing ratio of cement soil.

Based on the continuum theory, the dilatancy angle is related to the confining stress and plastic shear strain. Due to the tunnel excavation, the confining stress of rock mass decreases from the initial stress to the support force, while the plastic shear strain increases. Thus the dilatancy coefficient varies nonlinear. In this study, based on the unified strength failure criterion and non-associated flow rule, the potential plastic zone was divided into a finite number of concentric rings according to the equal decreased confining stress. A finite difference method was proposed by considering the effects of the intermediate principal stress and nonlinear dilatancy. Then the accuracy of this method was further verified using example analysis. The parameter analysis was conducted to study the effects of the intermediate principal stress, the critical softening parameter and the support load on the dilatancy angle in the plastic zone of rock mass. It was found that the peak dilatancy increased with the increase of the intermediate principal stress. The variation rate of dilatancy angle became slowly with the increase of the critical softening parameter. The intermediate principal stress and the critical softening parameter showed corporate effects on the variation of dilatancy angle in the plastic zone. The dilatancy angle at the tunnel wall increased with the increase of the support load. The double shear strength criterion should be used cautiously in the tunnel because the calculated displacement at the tunnel wall is small. Additionally, when the Mohr-Coulomb strength criterion is used, it is proper to consider the bearing potential of rock mass.

The bearing capacity of bored piles is significantly improved and the dispersion of bearing capacity is reduced obviously by employing the method of grouting. However, the reliability of post-grouted bored piles is still lack of systematic research yet. In this study, the static load test data of 122 non-grouted bored piles and 147 post-grouted bored piles are collected. Then, combined with the certainty and uncertainty analysis methods in reliability analysis method, the approximate probability method based on the first-order second-moment method (JC method) and Monte Carlo method is employed to analyze the reliability indexes of these forementioned post-grouted bored piles and non-grouted ones.. The results indicate that the reliability index of post-grouting bored piles have been greatly improved compared to the non-grouting ones and the bearing stratum has less influence on the reliability of post-grouted bored piles than the reliability of non-grouted bored piles. Moreover, the reliability of post-grouted bored piles tends to increase with the increase of pile diameter and this trend between reliability and diameter remains consistent even under different load compositions.

There is an obvious trend of climate warming and wetting on the Qinghai-Tibet Plateau during the past fifty years. Climate changes in air temperature or precipitation will inevitably influence the stability of permafrost. Previous studies mainly focus on the thermal influence of climate warming, but little is known about the induced rainfall infiltration and the hydrothermal response mechanism. Based on the meteorological data observed at Beiluhe observation station during 2013, the established water-vapor-heat transport model is used to predict the response under 1℃ and 2 ℃ increment of temperature, which considering the influences of rainfall. Climate change influences the thermal-moisture of permafrost mainly by changing the surface energy budget and soil hydrothermal transport components. The results show that climate warming greatly increased the surface net radiation, latent heat of evaporation and soil heat flux, decreased the sensible heat and rainfall infiltration. The rising air temperature reduces the soil moisture and soil hydraulic conductivity. Temperature gradient increases dramatically with temperature arising, further increases the moisture and energy components and reduces the components related to the water potential gradient. Climate warming increases the surface evaporation and thickness of active layer and accelerates the degradation of permafrost, which is contrary to the thermal effects of rainfall increasing.

As an advanced supporting structure of the tunnel safely passing through unfavorable geological areas, the pipe roof play its supporting role depending on the soil arch effect. The pipe roof spacing is closely related to the arching effect. In this paper, a method of calculating the rational spacing of pipe roof is developed by introducing rational arch axis, considering the effect of lateral earth pressure, and combining with failure conditions of soil arching. The rationality of the proposed method is verified through the case study of Fengjujiang tunnel and comparing with the result of discrete element method. Furthermore, the variation of pipe roof spacing with the location of pipe roof and the influence of pipe roof diameter and soil parameters on pipe roof spacing are analyzed. The results show that when the lateral earth pressure coefficient is low, the spacing between the hance and the side wall can be increased appropriately. However, when lateral earth pressure coefficient is high, the pipe roof spacing from vault to hance needs to be decreased appropriately. Pipe roof spacing increases linearly with the increase of pipe roof diameter and soil cohesion, and is positively correlated with the friction angles in soil. With the increase of internal friction angle, the influence of pipe roof spacing is also increasing.

Crack initiation and propagation in hydraulic fracturing process of hot dry rocks is affected by thermal stress induced by the low temperature of injected cryogenic fracturing fluid as well as its injected water pressure. Firstly, the interaction between injected water pressure and thermal stress during the injection of cryogenic fracturing fluid and its effect on crack initiation are studied by THM coupling analysis. Subsequently, a THM-D coupling simulation considering the meso-structure of rock is conducted to investigate the hydraulic fracturing process of hot dry rocks under different thermal conditions. The results show that the thermal stress which is generated by both temperature gradient in the rock itself and the non-uniform expansion of rock particles, appears as tensile stress around the wellbore. The high injection pressure will prohibit the initiation of multiple cracks induced by thermal stress and the existence of thermal stress around the wellbore will weaken the injection pressure. As the rise of rock temperature, more cracks emerge near the wellbore at the crack initiation phase and the propagation velocity of fracture network along the maximum geo-stress direction decreases while the reconstruction scale enlarges. Simultaneously, the generated multiple cracks also increase the crack extension pressure.

Calcareous sand is widely spread in South China Sea. As the sea-filling materials in construction of artificial islands, its permeability has significant influence on the consolidation and settlement of soil mass. The drag coefficient, as the effect of fluid drag force acting on the particle surface, plays an important role on the permeability of calcareous sand, which is not extensively studied by researchers so far. Based on the experimental results of calcareous sand particle settling in various fluids, a new drag coefficient model has been proposed. The new model is now incorporated into the CFD(computational fluid dynamics)-DEM(discrete element method) program and is used to simulate the settling process of calcareous sand particles in fluids. By compared with the experimental results, the CFD-DEM program with new drag coefficient model is validated and it also shows its accuracy in predicting the settling behavior of calcareous sand particles of irregular shape over the existing drag coefficient model without considering the effect of particle shape. On the other hand, the utility of spheres used in CFD-DEM model instead of non-spheres can greatly reduce the computational cost and meanwhile represent the effect of particle shape on the drag force. This CFD-DEM approach is promising and will be applied to the analysis of consolidation, settlement and souring of sea-filling projects with calcareous soils.

The problem of contact crack is common in engineering structures. Combining the advantages of numerical manifold method in crack treatment, the contact crack problem under compression and shear loads is analyzed, and the progressive propagation process of compression-shear cracks is simulated. In order to reduce errors caused by different positions of crack tips in one element, the singular cover function term is added to each physical cover near the crack tip in a certain range, and then the partitioned integration is carried out according to the position of crack tip and the number of singular physical covers in each element. A compression-shear failure example is selected to analyze the impact of normal contact force on stress intensity factor, and then the progressive failure process is simulated. The results show that the proposed method for compression-shear crack is feasible. Compared with results obtained from unrefined and refined methods, the proposed method in this paper can describe crack propagation path with much higher accuracy. The normal contact force has no contribution to mode II stress intensity factor. However, it has great influence on mode I stress intensity factor and the relative error varies with mesh density. Moreover, the effect of normal contact force on mode I stress intensity factor is larger than that of inner pressure.

To investigate the initiation and propagation of hydraulic fracturing, a new type of hydraulic pressure supplying and controlling device and its operational approach are developed. The device consists of water pump, booster pump, air compressor, water guiding steel subplate and other components. This device is prone to combine with the pressure servo machine to complete the hydraulic fracturing experiment. It has the advantages of simple operation, high efficiency, low cost, and easy assembly, which makes the hydraulic fracturing experiment easier to be popularized in a wide range. The hydraulic fracturing test of rock-like specimens are carried out by using this test device. Hydraulic fracture propagation of specimens with pre-existing flaws under different confining pressures are studied and compared with that of specimens without pre-existing flaws. The experimental results show that when the horizontal stress difference is low, the pre-existing flaw tip exposes great influence on hydraulic fracture propagation, and the hydraulic fracture propagates towards it. With increasing the horizontal stress difference, the hydraulic fracture deflects towards the direction of the maximum horizontal stress. The experiment shows that the device developed in this research has high reliability and stability. This study has a promoting significance for the hydraulic fracturing test.