In order to explore the damage evolution law of sandstone with different saturations under freeze-thaw environment，through freeze-thaw cycle，CT scanning and uniaxial compression test，and combined with three-dimensional visualization software and fractal theory，the evolution processes of porosity，permeability，pore parameters，throat parameters，fractal dimension and macroscopic mechanical indexes of natural water-bearing，incompletely saturated and completely saturated sandstone samples during the freeze-thaw process are quantitatively analyzed. The results show that the degree of freeze-thaw damage of rock samples is determined by the water saturation during freeze-thaw. Specifically，the damage deterioration of the completely saturated sandstone is the most severe while the damage degree of the natural water-bearing sandstone is the lowest. The meso-structure of the completely saturated sandstone is highly heterogeneous，and its porosity，permeability，pore parameters，throat parameters and fractal dimension all increase exponentially with increasing the freeze-thaw number. However，for the natural water-bearing sandstone and the incompletely saturated sandstone，these parameters increase linearly with increasing the freeze-thaw number. The porosity and the permeability of the samples are positively correlated with the fractal dimension，and freeze-thaw damage mainly presents in producing small pores and increasing the length of the original throat. The strength and the elastic modulus of three kinds of saturated sandstone samples decrease linearly with increasing the freeze-thaw number. After 150 freeze-thaw times，the strength of the completely saturated sandstone decreases by 98.53%，which is respectively 1.68 and 1.26 times higher than that of the natural water-bearing sandstone and the incompletely saturated sandstone. The macroscopic mechanical strength is negatively correlated with the porosity and the fractal dimension，and the meso-structure of rock is the key to affect the macroscopic mechanical properties. The initial freeze-thaw load damage variable is the freeze-thaw damage variable under the corresponding freeze-thaw number. The damage evolution curve has obvious stage characteristics，and the damage variable increases from concave to convex and then to gentle in the process of loading. The research results will provide a theoretical basis for the scientific evaluation of the long-term stability of rock mass engineering under the freeze-thaw environment in cold regions.

A model test system for simulating the evolution of frost penetration tunnels was developed to reveal the penetration mechanism and the formation of the frozen surrounding rock in cold regions with negative annual average temperature，and to investigate the thawing mechanism of the frozen surrounding rock due to global warming. The model test system mainly consists of a tunnel model，a ground temperature control system，a climate control system，and a data measurement and acquisition system. The main innovations of the test system are as follows：(1) The model test system can greatly shorten the test period for simulating the tunnel life-cycle temperature field，reducing the original test period from 40 days to 4 days. (2) The new climate simulation system can stably simulate the influence of the annual average temperature，the temperature amplitude，the wind speed and other combined conditions on the tunnel temperature field for a long period of time. (3) The developed ground temperature measuring and controlling system effectively solves the difficulties in data measurement and ground temperature control. The accuracy and reliability of the model test system were verified by comparing the tested temperature field with the field test result of a typical cold region tunnel. The preliminary test results show that，when the annual average temperature is －4 ℃，the frozen envelope of the surrounding rock will gradually advance along the longitudinal direction of the tunnel and penetrate into the deep of the surrounding rock on the cross section，causing the frost penetration of the surrounding rock and hence，forming so-called frost penetration tunnel. The experimental system，filling in the blank of the experimental study on the evolution law of the temperature field under the life-cycle operation condition of tunnels in cold regions，plays a positive role in promoting the frontier research of tunnel engineering in cold regions in China and provides experimental basis for the“one belt，one road”project and the Sichuan—Tibet railway project.

In order to study changes in permeability of coal in the process of damage and failure under stress loading and unloading，based on coal“cube”structure model，a triaxial stress loading-unloading damage coal permeability model was established by comprehensively considering the deformation of coal cleats and matrix due to the actions of damage，effective stress and adsorption/desorption under triaxial stress loading-unloading，and the permeability model was verified based on the experimental results. The research results show that the established permeability model can not only better reflect the change law of the damaged coal permeability under the action of triaxial stress loading and unloading，but also reflect the phenomenon that the effective stress has far greater influence on coal permeability than adsorption/desorption. When the load on the coal is lower than the peak strength，the coal is in a compressed state under the constant confining pressure and continuous axial compressive loading，and the gas permeability decreases；while reducing the confining pressure and continuously increasing the axial pressure，damage begins to occur inside the coal，and the gas permeability increases slowly. As a result，the overall gas permeability of coal exhibits a“V”-shaped development trend before failure. The increase in the permeability of coal after failure increases with increasing the initial confining pressure，and the initial permeability of coal increases with increasing the gas pressure. The research results can provide theoretical support for the high-efficiency gas drainage and gas disaster prevention and control in my country¢s deep coal mining face，and have guiding significance.

Water phase percolation properties of coal with pore and fissure medium scale characteristics is of great significance for exploring cross-scale seepage of coal pore fissure structure. Using the concept of the representation unit to define the mesoscale lower bound of water seepage in coal pore fissure structure，and introducing the contribution rate correction coefficients of fissure and pore structures into the pore structure and fissure structure water seepage models，a cross-scale water seepage model of coal pore fissure structure was built. Through numerical analysis and water seepage experiment，the model was verified and its applicability was discussed. The results show that，with increasing the pore fissure rate，the correction coefficient of the contribution rate of the fissure structure increases while the correction coefficient of the contribution rate of the pore structure decreases，and that both show the rule of a power function. Both the model seepage velocity and the experimental seepage velocity increase with increasing the pore fracture rate. The consistency between the model calculation results and the experimental results is poor when the pore fissure rate is high while is high，when the pore fissure rate is low. The application of the stress in the experimental process has a great influence on the water seepage in the coal with a high pore fissure rate，while the influence for the coal with a low pore fissure rate is little. When the porosity is less than 3%，the variation coefficients of the data are all less than 15% with an average value of 9.70%，which indicates that the predicted values of the model are in good agreement with the actual experimental data and verifies the correctness of the model. The applicable conditions of the established prediction model are mesoscale and low pore fissure rate.

The bursting liability of coal is an important factor affecting the occurrence of rock burst and the degree of disaster. Bursting liability refers to the property that coal accumulates strain energy and produces impact failure，and hence，elastic energy accumulation at the peak strength is the key to evaluate impact propensity. Due to that there are enormous structural defects in coal，the failure process and energy evolution of coal are very complex. Uniaxial cyclic loading and unloading tests were carried out on standard coal samples to obtain the accumulated elastic strain energy under different stress states. It is found that the variation trend of the elastic strain energy density curve is almost consistent with that of the stress-strain curve，and the elastic strain energy reaches the maximum value at the peak strength moment. It is also shown that the energy input，elastic energy accumulation and energy dissipation have a nonlinear evolution law with the loading deformation of coal samples，but that at an arbitrary time，the stress square and the elastic energy accumulation show a good linear relationship. Based on the improved calculation method of the elastic energy accumulation at the peak strength moment，the elastic strain energy accumulation at the moment of the peak strength can be obtained more accurately. The effective elastic energy release rate index KET，which can comprehensively consider the coal strength，the energy evolution and the failure time，was proposed to evaluate the coal bursting liability，and combined with the existing indexes，the classification critical values of the bursting liability was given. Finally，the rationality of the evaluation results was verified by the fragmentation degree of coal samples(far-field debris mass ratio ω and average particle size da). The research results show that KET can effectively solve the limitations of large dispersion and conflict among existing evaluation results of various indicators. Compared with the failure state of coal samples，KET has positive correlation with ω and fragmentation degree and has negative correlation with da，which is more in line with the reality.

Three-point bending tests and grinding tests were carried out with notched semi-circular bend(NSCB) and 1–2 mm/0.425–1 mm particle size samples，and theoretical analysis was performed based on fracture mechanics and rock fragmentation. The theoretical analysis shows that the second power of the fracture toughness is directly proportional to the reciprocal of the newly added surface area after crushing. The fitting results on the test data also indicate that there is a good proportion between the second power of the fracture toughness and the reciprocal of the newly added surface area of the 1–2 mm/0.425–1 mm particle size samples after grinding，and the correlation coefficient R2 are 0.974 and 0.878 respectively，which shows that the fracture toughness of coal can be estimated by the newly added surface area calculated easily and then the surface energy can be obtained. The research not only provides a new test method for the determination of the fracture toughness of coal but also provides an experimental basis for further understanding of the energy dissipation in the process of outburst.

To study the mechanical properties，seepage characteristics and essential acoustic emission characteristics of collapse column skeleton sandstone under different confining pressures and osmotic pressures，triaxial compression seepage tests were conducted by triaxial seepage experiment system combined with AE21C acoustic emission monitoring system，and the total stress-strain curve and the permeability curve of sandstone during deformation，as well as the evolution laws of deformation，permeability and acoustic emission signals of sandstone，were obtained. The results show that collapse column skeleton sandstone has obvious brittleness，that under the same osmotic pressure，the peak stress strength，the elastic modulus and the peak strain of sandstone increase with increasing the confining pressure，and that the confining pressure has an apparent effect on the macroscopic failure characteristics of sandstone with the failure mode gradually changing from multi-crack shear failure to single inclined-plane shear failure. The sandstone has low permeability characteristics，and the evolution law on the permeability of sandstone，generally characterized by a three-stage evolution of gradual decrease followed by steady development and rapid increase，is closely related to the stress-strain relationship under triaxial loading. It is also revealed that the variation characteristics of acoustic emissions are essentially consistent with the characteristics of stress-strain and permeability curves. In the initial stage，the ringing counting rate decreases with increasing the confining pressure；in the fracture development and propagation stage，the ringing counting rate of acoustic emissions presents a densely active state and gradually increases；during the unstable failure stage，the ringing counting rate increases rapidly and then quickly decreases. The test results provide an essential reference for studying the variation laws of a karst collapse column¢s stability and permeability.

Most of rock bursts in China are closely related to the hard roof overlying coal seams. In order to improve the effect of presplitting and pressure relief of the hard roof，a directional presplitting and pressure relief method of water jet prefabricated slot of the hard roof was proposed. Aiming at a sandstone roof，the rock was cut by high pressure water jet in the laboratory，and the characteristic index values of the fracture groove，such as the depth，the width and the erosion volume of the rock under different hydraulic parameters，were measured. The field test was also carried out. The results show that the depth，the width and the erosion volume of rock slot characteristic indexes are positively correlated with the jet pressure，the nozzle diameter and the repeated cutting times，while negatively correlated with the transverse velocity of the nozzle. The characteristic index of the rock slot first increases and then decreases with increasing the target distance，and the optimal target distance is 5d. In the four level orthogonal experiment，five key factors affecting the optimization of pressure relief of the seam slot are the nozzle moving speed，the repeated cutting times，the jet pressure，the nozzle diameter and the target distance. Through selecting parameters based on the laboratory research rules，the underground experimental research on high-pressure water jet cutting of a roof was carried out. It is revealed that，for the roof rock with an uniaxial compressive strength of 68 MPa，the cutting radius is 10 times higher than that of the traditional cutting tool and the fracture initiation pressure is reduced. Under the guidance of the energy accumulation of the roof rock water jet prefabricated joint，the maximum single fracturing radius can reach 15 m，and the effect of anti-impact and pressure relief is good.

When a tunnel passes through the deep soft-rock formation with high in-situ stresses，excavation and unloading disturbances will cause severe squeezing deformation and loose failure of the surrounding rock，resulting in large deformation of the supporting structure. This paper takes the Wanhe tunnel of Yumo railway，which goes through a fracture zone of granite with high in-situ stresses，as the research object. First，based on the classic Kastner and Caquot formulas，the characteristic curve of the surrounding rock considering loose effect was determined. Then，the stabilities of the surrounding rock and the supporting structure were analyzed using the convergence-confinement method to clarify the necessity of considering the loosening effect and controlling the degree of support deformation. Finally，through the comparison of working conditions，the support time and the support parameters of the second arch were determined. This research indicates that the surrounding rock pressure is dominated by the deformation pressure in the early stage. With the release of the stress，the proportion of the loose pressure exceeds 30% when the deformation reaches 0.8 m. At this time，the surrounding rock pressures acting on the support structure considering and without considering the loose effect are respectively 1.094 MPa and 0.765 MPa，indicating that the influence of the loose effect cannot be ignored in the stability analysis and the design of the supporting structure. After the initial support deforms to a certain degree，a second layer of steel arch should be added in time to resist the deformation of the surrounding rock. The best support time for the second arch is the time when the deformation reaches 0.45 m，and the best support parameter is 0.6 m spacing of the I22b steel arch. On-site monitoring data show that the large-deformation support scheme has achieved a good control effect. The research results can accurately reflect the large deformation and catastrophic process of the tunnel in the fracture zone with high in-situ stresses，and have clear guidance for the optimization of the support design in similar projects.

In the underground mining engineering activities of coal resources，coal is in the complex stress environment with the coupling of the axial pressure，the confining pressure and the gas pressure. In order to explore the damage deformation and energy evolution characteristics of coal during the compression process，the triaxial compression test of coal under different confining pressures and different gas pressures is carried out by using a triaxial servo-controlled seepage device. Based on the theory of continuous damage mechanics，the damage stress-strain function composed of heterogeneous Weibull function，energy dissipation function and plastic strain function is deduced theoretically from the angle of energy，and the plastic damage constitutive model of coal based on energy dissipation is constructed. The results show that the change trend of stress-strain and plastic deformation behavior of coal under different confining pressures and different gas pressures have stage characteristics. Corresponding to different deformation and failure stages，the energy evolution trend of coal and rock shows periodic changes. It is also shown that，at the stress peak point，the total energy absorbed by coal as well as the elastic energy and the dissipated energy increases with increasing the confining pressure. However，the total energy absorbed by coal and the elastic energy decrease with increasing the gas pressure，while the dissipated energy increases. A plastic damage constitutive model of coal based on energy dissipation under the effect of the confining pressure and the gas pressure is constructed，and the rationality of the model is verified by experiments. Under different confining pressures and gas pressures，the energy dissipation and damage evolution of coal experience a S type evolution trend.

The equivalent method of shear mechanical parameters(cohesion c and internal friction angle) is used in the arching theoretical model. To investigate the influences of different shear mechanical parameters on the arching theoretical model，the triaxial tests under constant confining pressures were firstly carried out to study the shear behaviors of the surrounding rock-CPB(cemented paste backfill) interface and the triaxial compression behaviors of the CPB. Based on the framework of Mohr-Coulomb strength theory，the shear mechanical parameters of the CPB and the surrounding rock-CPB interface were compared and analyzed under different curing times. Then，The scanning electron microscopy(SEM) and the thermal differential analysis(TG/DTG) methods were used to explain the difference in the shear behaviors of the CPB and the surrounding rock-CPB interface. Finally，based on the experimental results，the engineering application of Terzaghi's theoretical model was also discussed. The results show that with increasing the confining pressure，the shear displacement-shear stress curves of the surrounding rock-CPB interface present a multi-peak phenomenon and the axial strain-deviatoric stress curves of the CPB show an obvious elastic strengthening stage. The shear mechanical parameters(cb and ) of the CPB are significantly larger than the corresponding parameters(ci and ) of the corresponding surrounding rock-CPB interface. The ratio of the cohesion(cb/ci) between the two types(CPB and surrounding rock-CPB interface) is 2.26 to 2.67 times，and 1.65 to 2.76 times for the corresponding ratio of the internal friction angle(). Moreover，using the shear parameters(ci，) of the surrounding rock-CPB interface to predict the vertical stress() inside the CPB structure can effectively overcome the limitations of the Terzaghi¢s theoretical model. The conclusions can provide a reference for the prediction of the arching effect of the CPB structure and the selection of mechanical parameters in the numerical analysis.

The boulder-cobble mixed(BCM) stratum is mainly composed of boulder and cobble and mixed with sand gravel and clay，in which the tunnel excavation is prone to cause a large-scale collapse at the arch due to the poor cementation of the BCM stratum. To systematically study the mechanical response mechanisms and deformation laws of surrounding rock during the process of excavation and instability failure，the model materials with similar particle size distribution were prepared based on a tunnel crossing through the BCM stratum in Tibet，and laboratory similar model test was implemented to analyze the stress and displacement fields comprehensively in the case of no support system. The results show that the BCM surrounding rock is of relatively good self-stability after excavation，but the capacity decreases sharply when encountering water. Three progressive arch collapses expand gradually in turn and develop upward in the form of a sharp arch. At the excavation stage，the radial stress relaxation of the surrounding rock is obvious while the circumferential stress concentration occurs within a certain depth range. The radial and circumferential stresses of the surrounding rock increase sharply before the first collapse and decrease continuously at the arch in the later collapse. For sidewall rock，the radial stress almost changes in a sinusoidal form and the circumferential stress increases all the time or increases first and then decreases. At the excavation stage，the deformation of the surrounding rock is generally low，and the settlement within the range from 15 cm in front of and 25 cm in back of the monitoring section caused by excavation accounts for more than 90% of the total settlement. The settlement of the vault increases sharply once encountering water，and the settle zone of the vault is exactly consistent with the collapse area in each stage. However，the deformation of the side wall is insignificant at the stages of excavation and water-induced failure. The results of model test and field monitoring are basically consistent with each other. The stress and displacement transfer of the surrounding rock is synchronous and interactive，and the staged failure process is mainly attributed to that the surrounding rock forms a protective circle similar to an arch structure around the tunnel through self-regulation. The formation of each arch structure has a certain hindrance to the development of surrounding rock deformation，and the first arch structure¢s obstruction effect is greater than the subsequent ones. The research results are valuable to guide the determination of support design parameters and the prevention and treatment of the BCM stratum tunnel collapse.

The existing microseismic source location methods assume that the seismic wave propagates along straight lines，and the positioning results often fail to meet the accuracy requirements. In this paper，a new location method of microseismic source named as minimum travel time surface method of seismic wave was established. This method considers Snell¢s law of seismic wave propagation and assumes seismic wave minimum travel path ray propagation plane. By rotating the coordinate system，the three-dimensional propagation path of seismic waves can be transformed into a two-dimensional plane propagation path，and the relatively accurate calculation of initial arrival and travel time of micro-earthquakes in parallel layered rock mass can be realized. The improved genetic algorithm was used to optimize the microseismic source in the three-layer horizontal model to verify the validity of the hypothesis，and the factors affecting the positioning accuracy were analyzed. The results show that the travel time obtained by the minimum travel time surface method is significantly better than that obtained by the traditional shortest geometric path travel time calculation method，and that the difference between them is a function of the inclinationand the distance L of the connection line between the microseismic source and the sensor. The new location method has a great positioning accuracy in multi-layer media and the positioning accuracy is not affected by the wave velocity ratio of the medium.

The coefficient of earth pressure at rest is one of the important parameters in the design and analysis of bracing structures，and its value determines the magnitude of the initial horizontal earth pressure acting on the structures. Taking the Fujian standard sand as the object，the centrifugal model tests and the triaxial consolidated drained tests are performed to investigate the variation rules of both the coefficient of earth pressure at rest and the peak strength of the sand with the initial relative density. Based on the above test results，a NHRI equation，which can truly reflect the relationship between the earth pressure at rest and the peak friction angle of sand，is proposed. To validate the correctness of the NHRI equation，a conceptualized model for microstructure of particles to depict the mesoscopic stress characteristics is put forward. The relationships between the coefficient of earth pressure at rest of sand with the contact angle and the deviation angle of particles are obtained through derivation of the NHRI equation. The effect of the initial relative density on the internal stress transfer mechanism of sand is preliminarily discussed. It is pointed out that with increasing the relative density of sand，there exists the possibility for the microstructure of particles to transform from the high-order circular unit structure to the low-order steady unit structure，while that the dual-bracing action of the low-order steady unit structure will promote the vertical stress of particles to laterally transfer. The research results are of a certain reference value to improve the earth pressure theories of sand.

The composition contents of the quaternary soil-rock mixture(SRM) have a significant impact on the structural characteristics and play a decisive role in the strength of the typical two-phase geo-material. The increase of the rock block proportion(RBP) leads to the emergence of the soil-rock interfaces with a weak strength，and in the meanwhile，the added rock blocks change the transmission path of the force system and are more capable to bear force effect than the soil matrix. The studies on the empirical strength theory of SRM are still limited. This study picks the SRM in the density state as the research target，and pays attention to the influence of the soil-rock-interface system on the strength. Base on the similarities in mechanical properties and constituent structures between the medium and high RBPs and fractured rock masses，the calculation format and empirical parameter selection method of Hoek-Brown criterion were considered to build the non-linear strength criterion of SRM for describing the mechanical characteristics of the material. The RBP，the unconfined compression strength，the characteristic parameter A，the disturbed parameter  and the geological parameter G were chosen to reflect the impacts of the RBP，the strength of the soil-rock interface and the mechanical properties of the compositions on the SRM strength. The fitting formulae of the experience parameters ms and a were promoted based on the large triaxial test data collected from previous studies，and the accuracy of the established criterion was also proved using the test data in the database. The results show that the non-linear criterion proposed in this study can describe the non-linear strength characteristics of SRM with various rock block proportions. The predicted major principle stress values are divided into test fitting value and formula fitting value according to the approaches for obtaining the experience parameters of the proposed criterion. The test fitting values are much more precise than the formula fitting values compared with the test results. The predicted values of the two methods are consistent with the test results with coefficients of correlation greater than 0.9. The non-linear criterion developed in this study is much more suitable and accurate for the SRM with a higher rock block proportion，while the accuracy is relatively poor for the SRM with a low proportion in contrast with the Mohr-Coulomb criterion. The research results can provide calculation references for the strength estimation of natural SRM in engineering practice.

The pile dynamic response under high-strain condition(HSC) is the basis work of pile capacity prediction and pile driving analysis，which involves the soil strain history，elastoplastic characteristics，nonlinearity and slippage of the pile-soil interface. In order to simulate pile base soil under HSC，a discrete-element fictitious soil pile model was proposed in this paper，and a numerical differentiation solution was given. Firstly，the pile base soil was regarded as a fictitious soil pile with a certain diffusion angle，and the real pile and the fictitious soil pile were then discretized into a series of elements. The interactions between the pile elements and the soil elements were simulated by the linear viscoelastic model and the nonlinear spring-viscous damper-slippery parallel model，respectively，and the pile shaft resistance was calculated with a rheological model. The semi-sinusoidal impulse was applied for simulating the dynamic load acting at the pile top，and the velocity and displacement responses under different impulse widths and amplitudes were calculated combined with 3-dimensional finite element model(FEM) and the proposed model. Finally，the sensitivity analysis of related parameters was carried out. The results show that the fictitious soil pile model can consider the base soil vibration and the vertical propagation of the stress wave and is applicable for both low-strain and high-strain conditions. The proposed model can accurately predict the displacement and velocity response of the pile head under various impact conditions. The displacement amplitude increases approximately linearly with the increment of the impulse width and amplitude，whereas the velocity amplitude is not sensitive to the impulse width. The calculation results by FEM and the virtual pile model are consistent. Increasing the length of the fictitious soil pile and the soil nonlinearity degree or reducing the diffusion angle，the elastic modulus and the modulus ratio of the pile base soil would increase the peak velocity，the peak displacement and the residue displacement of the pile head. The fictitious soil pile model makes it possible to study the effect of soil distribution under the pile toe and has special value for simulating pile base soil and improving the dynamic calculation theory of pile foundation under HSC.

To solve the issues of finite mesh size sensitivity and mesh locking in applying finite element method to simulate strain localization in geotechnical engineering，a three-dimensional Mohr-Coulomb elastoplastic model was developed by introducing Cosserat continuum theory as a regularization mechanism. The secondary development function of ABAQUS and Euler backward implicit stress update algorithm were adopted to realize numerical simulation，and the reliability and effectiveness of the model and program were verified by numerical simulation of uniaxial compression. The relationship between the bending characteristic length and the shear band width was discussed，and the value standard of the bending characteristic length was proposed. The three-dimensional anti-sliding stability analysis of a gravity dam was carried out repectively by the classical elastic-plastic model and the proposed model. It is shown that the classical elastic-plastic model has obvious mesh dependence，more specifically，an overestimation of the actual anti-sliding stability resulted from sparse finite element mesh，while that the anti-sliding stability safety factor and the sliding mode by the developed model almost keep unchanged for different density finite element mesh.

Dumping method is one of the key factors affecting the failure mechanism of waste dumps of open-pit mines. To reveal the failure mechanism of mine waste dumps with different dumping methods，a real waste dump was taken as a background and a series of base friction tests were conducted on four models. The failure process was comparatively analyzed through the graphical point-tracking technology. The results show that：(1) under the same base condition，the forwardly downslope dump has the highest stability level，followed by the forwardly covered dump，the reversely covered dump and the reversely downslope dump. (2) The block size of the waste material has a significant impact on the dump stability，and the stage mainly composed of small-size blocks is easier to collapse than that mainly composed of large-size blocks. (3) Failure mechanisms of waste dumps with different dumping methods are different. Exactly，the failure of the forward dumps mainly occurs at the upper and middle slope，among which the forwardly covered dump shows slip-cracking failure at the upper and middle stages while the forwardly downslope dump presents extrusion-slip failure in the upper and middle stages. In contrast，the failure of reverse dumps locates at the lower and middle slope，among which the reversely covered dump shows slip-cracking failure at the lower and middle stages but the reversely downslope dump shows slip-cracking-traction composite failure at the lower and middle stages with the largest damage area. The research results can provide reference and guidance for the stability evaluation of open-pit mine waste dumps and the selection of dumping methods.