As cold region engineering expands to higher altitudes and latitudes，the application of fractured rock masses(FRM) in cold region engineering has increased significantly. However，existing  theories in frozen soil mechanics are insufficient to address the engineering challenges in these areas. This study investigates the mechanisms of freeze heave(FH) in FRM under low temperatures and freeze-thaw cycles. Through theoretical analysis，laboratory and field experiments，and numerical simulations，the research explores the formation mechanism and failure patterns of FH pressure in FRM. The study finds that FRM exhibit various FH mechanisms，including volumetric expansion，segregated ice，and mixed FH. The semi-elliptical open fissure mixed FH model provides a better description of these mechanisms. FH in FRM is a complex coupling of thermo-hydro-mechanical processes that involve moisture migration，multiphase heat conduction，and crack propagation. Factors such as fissure structure，saturation，sealing properties，freeze mode，ice-rock interactions and phase transitions significantly influence FH pressure and damage. The primary mechanism of damage in FRM is driven by FH pressure promoting the crack propagation，which is significantly influenced by the characteristics of the fissure and rock mass. Moreover，laboratory and field tests show differences in FH behavior，especially concerning freeze-thaw cycles，crack initiation temperature，and water absorption conditions. Future research should focus on micro- and meso-mechanisms，supported by laboratory and field experiments，to investigate moisture migration and ice-rock interactions. The objective is to solve FH pressure，explore the evolution of fracture networks using numerical methods，and develop an adaptive monitoring and decision-support system for predicting FH failure，integrating artificial intelligence and big data.

Accurate localization of acoustic emission(AE) sources and velocity tomographic imaging are essential for predicting rock deformation and failure in laboratory experiments. This paper proposes a novel multi-stencils fast marching method(MSFM) algorithm that accounts for anisotropy to compute the arrival times of seismic(or acoustic) waves in the source space. By matching the observed time difference matrix with the theoretical time difference matrix，this approach identifies the optimal source location. By integrating AE localization data，the algorithm achieves three-dimensional tomographic imaging of the anisotropic P-wave structure in rock samples. Furthermore，a coupled predictive method based on RA-AF value，Ib value，and AE multi-parameter critical slowing down is developed to forecast rock instability and failure. Additionally，an innovative wavelet coherence method is used to characterize damage evolution by analyzing the crack propagation modes of RA-AF parameters over time. Findings indicate that the new localization algorithm accurately identifies source locations within heterogeneous rock materials. The velocity tomography results reveal a sharp 5%–30% decrease in wave speed and a 5%–20% increase in anisotropy within the stress range of 95%–100% of peak stress. The stress peak range for AE multi-parameter prediction of rock instability is identified between 86% and 93%. These results and methods demonstrate substantial potential for enhancing failure localization and velocity imaging along nonlinear stress curves in complex geotechnical materials，providing insights into local changes associated with microcracks and failure.

Spalling is a predominant failure mode observed in underground tunnels subjected to explosion waves. The spalling laws of underground tunnels in homogeneous hard rock and fractured rock masses were analyzed using a one-dimensional wave model and a block instability model，respectively. The calculation formulas for spalling block size，spalling velocity and critical scaled distance for spalling under different conditions were obtained. Existing explosion tests conducted on underground tunnels were evaluated using the proposed theoretical methods. To obtain empirical spalling data and validate the theoretical methods，a prototype explosion experiment was conducted on a U-shaped tunnel in granite. The experiment demonstrated the spalling damage phenomenon of the tunnel at a scaled distance of 1.03 m/kg1/3，and measured the peak particle velocities at the tunnel top，the ejection velocities of spalling blocks，and the irreversible displacements of the tunnel wall. The measured results show good agreement with the theoretical calculation results，which verified the scientific reliability of the proposed theoretical method.

The failure and instability of locked-segment-type slopes are characterized by their rapid onset，unpredictability，and significant consequences. It is of great practical significance and an urgent social requirement to reveal the unlocked mechanism of such slopes. This study utilizes the Chana landslide as a geological prototype of a typical locked-segment-type slope. A physical model of locked-segment-type slopes is constructed using analogous materials，and the failure and instability processes of the physical model are conducted under gradual step-loading conditions. Furthermore，a numerical model of locked-segment-type slopes is established using the FLAC3D software platform，and its failure and instability processes are simulated through the strength reduction method. Based on the results from both physical and numerical simulations，the failure and instability modes of the locked-segment-type slope are analyzed，and its unlocked mechanism is finally revealed. According to the concept of the locked segment，an identification method for the locked zone is proposed，based on the strain concentration area observed in the physical simulation tests. Additionally，the damage evolution of the locked segment is analyzed. Utilizing the Fish programming language integrated within FLAC3D，an acoustic emission(AE) acquisition program is developed to monitor the locked segment area，and the AE evolution characteristics during slope instability are analyzed. Based on the simulation results，an effective identification method for characteristic stress during locked segment damage is proposed，and the evolution stages of slope instability are then divided. Subsequenly，the critical displacement criterion for the instability of the locked segment is employed to retroactively predict the instability process of the Chana landslide，achieving favorable predictive outcomes. The findings enhance our understanding of the instability mechanisms and improve the instability prediction capabilities for locked-segment-type slopes.

Numerous pre-existing fractures within deep engineered rock masses and the complex hydraulic environment，significantly affect the dynamic strength characteristics of these engineered rock masses. To investigate the effects of fractures and high water pressure on the dynamic strength characteristics of rocks，five types of red sandstone specimens with different fracture inclination angles were prepared. Impact tests were carried out under six water pressure gradients using a self-developed rock dynamics testing device capable of simulating high water pressure and high geo-stress conditions. The results indicate that with the increase of water pressure and fracture inclination angle，the dynamic stress-strain curve of the fractured rock is gradually transformed from plastic to elastic after-effects after the peak. The curve characteristics can be categorized into three stages. When the fracture inclination angle is fixed，the dynamic peak stress of the fractured rock initially increases with rising water pressure，followed by a slight decreases. The relationship between dynamic peak stress and water pressure follows a Gaussian distribution. As water pressure increases，the average strain rate first decreases and then increases. Under the same water pressure，the dynamic peak stress of fractured rock generally exhibits a trend of slow initial increase followed by a rapid rise as the fracture inclination angle increases，and this trend is influenced by the radial connectivity of the fractured rock and the oblique incidence of stress waves. Moreover，the average strain rate gradually decreases with the increase of fracture inclination angle. A clear negative linear relationship is observed between the dynamic peak stress of the fractured rock and the average strain rate.

The natural occurrence of numerous discontinuities in rock masses makes them prone to significant displacements and friction degradation under external dynamic forces，thereby increasing the risk of instability. To investigate the mechanical properties of jointed rock masses under cyclic dynamic conditions，this study focuses on natural red sandstone and explores the shear mechanical characteristics of rock discontinuities under cyclic shearing，with a specific emphasis on the dynamic properties of rock masses by setting a high shear rate. The evolution of shear stress and normal displacement during cyclic shearing is analyzed，and a peak shear strength prediction model for structural planes is established，incorporating the number of cyclic shear cycles based on plastic work theory. The results show that：(1) The applied cyclic shear rate (2 mm/s) reaches the dynamic range，and in comparison to quasi-static direct shear tests，this rate more accurately simulates the shear mechanical behavior of discontinuities under dynamic conditions. (2) During cyclic shearing，the number of cycles，type of discontinuities，and normal stress significantly influence the evolution of shear stress on the discontinuities. Normal displacement exhibits distinct patterns at shear displacements of 0 mm and ±5 mm. As the number of cycles increases，normal displacement gradually decreases，indicating a reduction in dilatancy and an increase in compaction. (3) A peak shear strength prediction model that accounts for the number of cyclic shear cycles is established，with an overall error of approximately 10% compared to actual conditions，demonstrating good applicability. The research findings provide theoretical support for the safety and stability assessment of rock engineering under dynamic disturbances.

The interfaces between the anchor cable，anchoring agent and surrounding rock represent critical weak points in anchoring systems. This study focuses on the debonding failure characteristics under cyclic disturbance and established an equivalent shear mechanics model for the anchoring system，conducting cyclic shear tests. The influences of rock lithology，normal load(Fs)，cyclic shear displacement amplitude(ud) and cycle number(N) on the shear mechanical properties of the anchoring system were investigated. The experimental results reveal that the anchoring system experiences progressive accumulation of damage during the cyclic shear stage，with the maximum shear force(Fhpeak) generally presenting a decreasing trend as N increases. As ud increases，the location of the maximum shear force in the anchoring system gradually shifts from the direct shear stage to the cyclic shear stage. The peak shear force(Fhmax) and residual shear force(Fhres) decrease by 28.46%–54.13% and 21.22%–49.25%，respectively. An increase in the rock strength and normal load enhances the shear resistance characteristics of the anchoring system. During the cyclic shear stage，the sustained wear and degradation at the interface result in a more pronounced normal shear contraction within the anchoring system. Conversely，in the direct shear process，the climbing effect of the through fracture plane leads to a gradual increase in normal shear dilation. The overall shear contraction of the anchoring system is gradually enhanced with increases in ud and Fs，while it diminishes with increasing rock strength. Three typical shear failure modes of the anchoring system are identified：internal shear fracture through the coal matrix，debonding failure and fracture at the anchoring agent-sandstone interface，and debonding and slip at the anchor cable-anchoring agent interface in the limestone anchoring system. The entire process of shear fracture surface formation and the evolution of debonding failure in the anchoring system is discussed based on digital photography.

To reveal the deformation characteristics of soil in the shield tunnels section during proximity pile foundation excavation and to establish the convergence model for the tunnel excavation section. A model test of shield tunnel excavation adjacent to piles was carried out in this paper. Based on the data collected using a developed monitoring device for soil displacement in the tunnel section，the soil deformation law of excavation section under various conditions was analyzed. Additionally，a loss contour for the tunnel section exhibiting a non-uniform convergence mode was proposed. A mapping function expression for the Laurent series of the tunnel excavation loss section was established based on the theory of complex functions. The results indicated that the deformation of the tunnel arch bottom lagged behind that of the tunnel vault，and the deformation of the soil around the tunnel section was restrained by the adjacent piles. Comparisons between the calculated and measured values were made to evaluate the validity of the mapping function expression for the tunnel excavation loss section.

When a proposed tunnel is located in close proximity to a concealed karst cave，the superimposed effects of construction disturbance and cave pressure may lead to the collapse of the rock mass situated between the tunnel and the karst cave. This study constructs a three-dimensional failure mechanism for the rock mass of the tunnel roof，taking into account the failure characteristic associated with the concealed karst cave existing above. The spatial discretization technique and upper bound theorem are employed to develop this failure mechanism. Utilizing the constructed failure mechanism，the stability number of the rock mass is proposed，with the objective function derived from calculating the internal and external energy consumption of the failure mechanism. The upper bound solution for the stability number and the corresponding three-dimensional collapse surface of the rock mass are obtained through optimization calculations. By comparing the stability numbers and collapse surfaces derived from theoretical analysis with results obtained from numerical simulations and model tests，the validity of the proposed theoretical approach is validated. Parameter analysis indicates that the stability number of the rock mass decreases with the increase of the geological strength index(GSI)，material constant of the rock mass( )，the distance between karst cave and tunnel(H)，and unit weight of the rock mass( )，while it increases with the increase of cavity pressure(T).

The uncertainty in the development patterns and combinations of structural surfaces makes it difficult for the effective prevention and control of disasters such as landslides and block falls during the excavation of jointed rock tunnels. It is crucial to consider the impact of weak-scale joints on the mechanical properties of the surrounding rock，as well as the controlling effect of large-scale structural planes on local stability，in order to provide reasonable support. To this end，based on the morphological distribution law of two-dimensional trace lines in sequential tunnel face images，the rapid extension and prediction of dominant structural planes in three-dimensional space have been achieved. Using the embedding coefficient method in Sadovsky?s theory，the theoretical values for different structural plane scale divisions were determined. By employing the secondary development function of 3DEC，a multi-scale DFN-DEM equivalent modeling method for jointed rock masses was proposed，constructing an equivalent rock mass model that retains only the dominant structural planes and large-scale joint sets. This method was validated and analyzed through a series of tests conducted in the Erlangshan Tunnel project of Linzi-Linyi Expressway in Shandong Province. The results indicate that the multi-scale equivalent modeling method for jointed rock masses，which incorporates the dominant structural plane extension method，more effectively reveals the collapse evolution laws of the local surrounding rock in tunnels，especially the morphological distribution characteristics of key block groups. The rose diagram of hazardous rock groups provide a theoretical basis for understanding the spatial distribution of overbreak caused by dangerous rockfall in tunnels from a probability and statistics perspective. By integrating these methods，targeted support schemes can be provided for structure-controlled surrounding rock，offering both a theoretical basis and field guidance for the prevention and control of local collapse instability in jointed rock mass tunnel engineering.

The instability of rock slopes in cold regions is closely linked to the development of dominant fractures，with the freeze-thaw environment and load conditions serving as direct triggers for the expansion of these fractures. Therefore，understanding the frost heave characteristics of water-bearing fractured rock masses under the combined effects of freeze-thaw and load is a crucial prerequisite for preventing and controlling slope disasters. Using a self-developed loading device，the freeze-thaw tests on fractured rock mass under compressive and shear forces were conducted to identify patterns in the temperature field，stress field，frost heave force，and end strain during the freeze-thaw process. The numerical model was established to simulate the freeze-thaw process in fractured rock masses under compressive-shear loads，systematically analyzing the frost heave expansion law of rock mass fractures under the coupled action of freeze-thaw and various combinations of compressive and shear loads. Finally，the three-dimensional schematic diagram illustrating the zoning of fracture rock mass failure types was summarized，and the key indicators such as loaded freeze-thaw failure surface and compression shear promotion zone were proposed. The results show that，the variation of temperature and frost heave force during a single freeze-thaw cycle exhibits similar characteristics，which can be divided into seven stages. In repeated freeze-thaw cycles，the peak frost heave force decreases with an increasing number of cycles，but it decreases rapidly in the early stages of the cycles，remains stable in the middle stage，and dissipates due to damage in the later stage. With the increase in the number of freeze-thaw cycles，the strain value at the crack tip of the rock mass under compression and compression-shear action gradually increases. Compression and shear loads exert inhibitory and promotive effects on the frost heave deformation of fractured rock，respectively. The application of shear load accelerates crack development，with a higher load resulting in faster crack propagation. After the application of compressive load，crack development accelerates in the early stages of freeze-thaw cycles but gradually slows down in the middle and later stages. However，the number of freeze-thaw cycles required for ultimate failure increases. The shear load in compression-shear load combination is the dominant load. The properties of the rock mass，freeze-thaw environmental conditions，and load magnitude not only affect the freeze-thaw failure law of loaded fractured rock masses but also alter their failure mechanisms.

To reduce the correlation among feature parameters in rock burst sample data，eliminate outliers and balance the number of various rock burst levels in the dataset，a method combining principal component analysis (PCA)，cluster-based local outlier factor(CBLOF) and support vector machine synthetic minority over-sampling technique(SVMSMOTE) is proposed to enhance the quality of the rock burst database. Initially，a total of 343 rock burst cases are collected from both domestic and international sources to build the original dataset. PCA is employed for dimensionality reduction，CBLOF is used to identify and eliminate outliers within each rock burst level，and SVMSMOTE synthesizes new minority samples near the boundaries of each rock burst level. The processed and original rock burst databases are subsequently used to train six different machine learning models to validate the effectiveness of the PCA，CBLOF，and SVMSMOTE combination. The results show that the accuracy of models trained on the processed database significantly improved：AdaBoost by 29%，CatBoost by 28.5%，LightGBM by 34%，Gradient Boosting by 28%，ExtraTrees by 26.5% and Random Forest by 24%，compared to models trained on the original database. Therefore，processing the rock burst database using the combined PCA，CBLOF and SVMSMOTE algorithms effectively enhances the quality of the database and im-proves the performance of machine learning prediction models.

In order to study the dynamic progressive failure characteristics and crack propagation mechanisms of pre-fabricated flaw sandstone，the impact compression tests were conducted on sandstone specimens with pre-fabricated double flaws using the split Hopkinson pressure bar(SHPB) device. The real-time monitoring of crack propagation and dynamic failure processes was carried out using digital image correlation(DIC) technology. An analysis of the strength characteristics and failure modes under different flaw angles was performed to reveal the corresponding crack strain evolution mechanisms. The results show that：(1) the crack initiation load，determined by capturing the surface crack whitening，ranges from 70% to 83% of the peak load，with both peak and initiation loads exhibiting a“U”-shaped trend，initially decreasing and then increasing with the flaw angle. (2) The energy dissipation rate，inferred from the number of fragments，shows a significant positive correlation with the degree of fragmentation，with the lowest and highest energy dissipation rates observed in the 45° and 90° specimens，respectively. (3) During crack propagation，tensile cracks generally form prior to shear cracks. The former are distributed along the direction of the applied load，while the latter appear within an angular range of approximately 50° to 130° relative to the load direction. Specimens with different flaw angles exhibit three primary failure modes：“一”-shaped tensile splitting failure，“X”-shaped tensile-shear mixed failure，and“spindle”and“X”-shaped composite failure，involving six distinct crack aggregation forms. (4) The progressive failure process under impact loading is divided into three stages：elastic deformation，yield deformation，and macroscopic failure. The failure process is primarily determined by the strain concentration factor( ) at the crack tip and the crack initiation time. The slope mean square index(K) of the curve provides a theoretical basis for evaluating the dynamic damage rate of fractured rock masses.

A fracture process zone(FPZ) dominates the fracture initiation and propagation in rock and quasi-brittle materials. The complete development curve involves two FPZ characteristics：the fully-developed FPZ length and its corresponding loading stage. However，due to limitations in observation techniques and processing approaches，it is rare to have the full curve of the FPZ development in the literatures. In this study，a series of mode I sandstone beams under three-point bending are employed to induce the mode I fractures for laboratory testing，where acoustic emission(AE) is used to observe the fracture process. By employing an improved approach to analyze the AE cluster and determine the FPZ boundary，a full curve of the FPZ development in the mode I fracture is obtained. The main conclusions are as follows：(1) the derived curve of the FPZ development aligns well with the theoretical model，which not only provides accurate information of the FPZ(i.e.，the fully developed FPZ length and its corresponding loading stage) but also indicates that the FPZ is related to the material properties. (2) Based on the full curve，the FPZ can be divided into two parts：FPZ initiation and FPZ propagation，with the transition position occurring when the FPZ reaches its fully-developed stage. During the FPZ initiation，the FPZ length increases when the external load increases. Once the peak load is reached，the FPZ continues to grow until the fully-developed FPZ occurs. For the FPZ propagation，the length FPZ remains relatively constant，though fluctuations are observed in the curve，probably due to instability in FPZ propagation. However，these fluctuations are centered around the length of the fully developed FPZ，allowing it to be treated as a constant. Experimental results demonstrate that the FPZ in mode I fractures attains its fully developed stage when the external load exceeds the peak(at post-90%，indicating that the load is at 90% of the peak in the post-peak regime)，with its length nearly twice as the length observed at the peak(17 mm vs. 9 mm). This study confirms that the fully developed stage of the I-type fracture process zone is directly correlated with the cohesive force model through experimental measurements of local acoustic emission.

In order to study the mechanical response behavior of overlying strata in steeply inclined coal seams，the stress distribution patterns and energy accumulation characteristics of the overburden were analyzed through theoretical analysis，numerical simulation and field measurements. The evolution law of destruction in the overlying strata was established，elucidating the interaction mechanisms of the overburden structures in different regions during the rotation process. Additionally，he fracture rotation behavior of the overlying rock strata in steeply inclined coal seams was discussed. The results show that the bending momentin the short side direction of the roof in steeply inclined coal seams is large，with the maximum principal bending moment in the middle area exhibiting a “trapezoidal-heart-shaped”distribution. Conversely，the maximum principal bending moments on both long sides display an“oblique trapezoidal”characteristic. Notably，the elastic energy is released substantially in the middle upper parts of the overlying strata，and the overall energy characteristics of the roof show an evolutionary trend of release in the middle upper parts，oblique extension at the upper end，and dip expansion at the lower part as the advancement distance increases. The advanced elastic energy in the middle upper parts of the roof is relatively concentrated，reaching a peak value. When the mining distance reaches 150 m，the peak value of the advanced elastic energy of the middle upper parts is 2.7 times than that of the upper area. With the increase of mining distance，the advanced elastic energy of the roof in steeply inclined coal seams experiences sequentially accumulation in the middle lower areas and rapid growth in the middle and upper areas. The fractured overlying rock structures squeeze each other and undergo rotation along the dip，with the compressive stress on the fracture surface of the overburden structure increasing with the increase of the dip angle. The slip parameter of the upper fracture surface is relatively small，and it is easy to experience slip instability during the rotation process. In contrast，the middle fracture surface remains stable，with the slip parameter directly proportional to the rotation angle. The lower fracture surface exhibits significant stress concentration but remains stable due to the filling effect. An increase in overburden load will aggravate the instability of the fractured rock strata，while the extension of gangue filling enhances the stability of the overburden structure. This study provides a theoretical basis for analyzing the migration characteristics of overburden in steeply inclined coal seams，and it is of great significance for the stability control of surrounding rock in steeply inclined coal seams.

The investigation of the long-term strength of surrounding rock is of significant engineering importance for the stability assessment of underground spaces. The user-developed visual triaxial compression servo control test system is used to conduct generalized stress relaxation tests on sandstone under seepage-stress coupling conditions. The results indicate that the large strain zone gradually converges to the macroscopic fracture region，exhibiting a tension-shear failure trend when the samples are damaged under the conditions of = 3 and directions. An exponential function effectively fits the strain rate of sandstone，demonstrating strong fitting performance. The permeability of the sample decreases rapidly at first and then increases gradually during the rheological progress. The porosity of the sample after the rheological test increases continuously with the rise in seepage pressure，with values of 1.72%，3.52% and 12.86%，respectively，and degree of the damage to the fracture surface is significant. It is observed that the hydrolysis and erosion of clay minerals in the sandstone are intensified，leading to the loosening of mineral particle structures and the formation of macroscopic fractures. Additionally，pore pressure induces hydrolysis and physical effects on the mineral particles，causing substantial amounts of rock debris to be detached from the crystals and accumulate on the fracture surface. From a fracture mechanics perspective，the increase in seepage pressure accelerates the crack propagation rate in sandstone，making the influence of stress levels on the generalized stress relaxation behavior of sandstone more pronounced. Consequently，the results in the samples becoming more susceptible to fracture under high seepage pressure. The pore pressure promotes the dissolution and destruction of the cementing material in the sandstone，causing it to detach from the crystal surface and form distinct cracks. Consequently，the failure mode of sandstone under seepage-stress coupling is predominantly characterized by intergranular fractures.

The geographical distribution of discontinuous graded soil-rock mixtures is extensive，and these materials are commonly used as backfill materials in subgrade and slope engineering. The discontinuous grading effect on their shear mechanical properties is a critical factor influencing compaction quality control. This study proposes a discontinuity index that effectively quantifies the discontinuous grading characteristics of soil-rock mixtures. Large-scale triaxial standard specimens of soil-rock mixtures with different discontinuity indices were prepared，and triaxial shear tests were conducted using the DJSZ–150 large-scale triaxial testing apparatus. The influence of and confining pressure on the strength and deformation characteristics of soil-rock mixtures was analyzed. Subsequently，intrinsic relationships between the parameters of the composite power-exponential model and the deviatoric stress strength， and were investigated，leading to the development of the CPE model，which accounts for the effects of varying interrupted grain fractions and stress states. Finally，the impact of changes in discontinuous particle size fractions on the mesoscopic structural characteristics of soil-rock mixtures was explored，elucidating the discontinuous gradation effect on their shear mechanical properties. The results indicate that under triaxial compression，discontinuous graded soil-rock mixtures typically exhibit strain hardening characteristics. The deviatoric stress strength increases with the increase in ，while the internal friction angle initially increases and then decreases with the increase in . In contrast，the cohesion exhibits an opposite trend. The parameters k，b，and n showing linear relationships with the . Both k and n show a linear relationship with the ，while b follows a quadratic function relationship with . The improved CPE model not only accurately predicts the strength and deformation characteristics of soil-rock mixtures，but also effectively reflects the influence of discontinuous grading features and stress state on their deformational mechanical behavior of soil-rock mixtures. Different discontinuous particle groups constitute distinct mesostructural characteristics of soil-rock mixtures. Under axial compression loading，an increase in the content of the largest particle group enhances the overall deformational mechanical properties of the soil-rock mixture. However，excessive amounts of large particles can increase the voids between coarse particles，leading to an initial increase in deviatoric stress strength followed by a decrease.

The conventional triaxial and cyclic loading-unloading tests with acoustic emission monitoring at different confining pressures were conducted to investigate the fracture process and failure characteristics of granite under confining pressure. This study compares the evolution of mechanical parameters，acoustic emission characteristics，energy dissipation patterns，and the effects of confining pressure under two stress paths. The results indicate that：(1) under both stress paths，peak strength，residual strength，axial peak strain，elastic modulus，and Poisson?s ratio increase with the confining pressure. The elastic modulus during cyclic loading-unloading is lower than that observed in conventional triaxial tests，while the Poisson?s ratio is higher. (2) The ringing ratio in conventional triaxial tests correlates well with the crack propagation stage of the rock samples，whereas the acoustic emission activity during cyclic loading-unloading exhibits a clear Kaiser effect，with a higher peak ringing ratio compared to conventional triaxial tests. (3) As axial strain increases，the input energy，dissipated energy，and elastic energy for cyclic loading-unloading initially increase and then decrease，while for conventional triaxial tests，both input and dissipated energies gradually increase，and the elastic energy follows a similar trend. (4) Under both stress paths，input energy，dissipated energy，and elastic energy increase with confining pressure，with both paths exhibiting similar input energy levels. However，the dissipated energy in conventional triaxial tests is greater，while the residual elastic energy in cyclic loading-unloading is higher. The failure patterns revealed in this study provide significant insights for the stability analysis of deep underground engineering，the degradation patterns of rock parameters，and disaster prevention and control strategies.

In order to study the impact of roof and floor rock mass strength on the overall mechanical properties of rock and backfill body，the uniaxial compression tests were conducted on the rock-backfill body-rock combination under conditions of both uniform and different strengths of the roof and floor. The study focused on the failure modes，stress-strain curves，uniaxial compression strengths，and acoustic emission characteristics of different combinations under uniaxial compression conditions. The results show that when the strength of the backfill body is lower than that of the roof and floor rocks，the failure of the combination is primarily caused by the instability of the backfill body. When the strengths of the roof and floor rock are low，splitting damage occurs in both the roof rock and the backfill body. As the strength of roof and floor rocks increases，the backfill body exhibits obvious shear damage while the ends of the rock remain relatively intact. When the strength of the roof and floor rock are identical，the elastic modulus and peak strength of the combination tend to increase with the increase of rock strength. Conversely，when the strengths of the roof and floor rocks are different，the elastic modulus and peak strength of the combination initially increase and then decrease as the strength of the floor rock increases under the condition that the strength of roof rock remains constant. In contrast，when the floor rock strength is held constant，both of them show an increasing trend with the increase of the roof rock strength. The rock- backfill body-rock combination exhibits an obvious acoustic emission response during the loading process，with the acoustic emission ringing count of the combination at the peak strength continuously increases with the increase of rock strength. The cumulative ringing count curve of the specimens presents a trend of rapid growth-stable growth-rapid growth. The energy migration instability model of the combination under different roof and floor conditions is developed，which reveals the energy accumulation and distribution law of the combination during the loading process. These findings provide theoretical insights for understanding the collaborative bearing mechanical characteristics of the roof and floor rocks and backfill body under the condition of constructional backfill mining.

Incorporating flake graphite into bentoni te enhances its thermal conductivity but compromises its impermeability. To address this trade-off，spheroidal graphite was selected as a substitute for flake graphite. Thermal conductivity and saturated permeability tests were conducted on spheroidal graphite-bentonite mixtures to investigate the effects of graphite content and shape on both thermal conductivity and permeability properties of bentonite. The results show that a spheroidal graphite content of approximately 20% significantly enhances the thermal conductivity of bentonite(1.7 W/(m•K))，while ensuring the permeability coefficient(9.44×10－13 m/s) remains below the buffer layer?s threshold of ≤1×10－12 m/s. Concurrently，the maximum swelling pressure attained is 5.76 MPa，which falls within the recommended range of the buffer layer(1–10 MPa). Furthermore，the concept of critical graphite content is introduced，with the critical contents for spheroidal and flake graphite determined to be 27.3% and 16.4%，respectively. At a spheroidal graphite content of 20%，the graphite has not yet formed a skeleton within the mixture，indicating that it remains suspended in the bentonite. Conversely，at a flake graphite content of 20%，a skeleton forms with bentonite filling the spaces within the graphite structure. By analyzing the microscopic morphology and pore distribution characteristics of graphite-bentonite，this study elucidates why flake graphite enhances thermal conductivity while reducing impermeability. Additionally，it explains the physical mechanism by which spheroidal graphite outperforms flake graphite，providing scientific guidance for improving the operational safety of the repository.

To facilitate the reutilization of solid waste and the solidification of soil，this study integrates ground granulated blast-furnace slag(GGBS) with microbial induced carbonate precipitation(MICP) technology，referred to as GGBS-MICP. It aims to investigate the engineering properties of silty-sandy specimens solidified with various particle sizes and dosages of GGBS combined with MICP technology through a series of tests，including unconfined compressive strength(UCS)，triaxial compression，erosion resistance，Ca2+ utilization quantity，pH values，and scanning electron microscopy(SEM). Compared to silty-sandy specimens treated solely with GGBS and MICP，those solidified by GGBS-MICP exhibit a significantly increased UCS. Finer GGBS particles yield a more pronounced enhancement，especially with＜10 GGBS-MICP. For ＜10 GGBS-MICP solidified silty-sandy samples with/without  ，the peak values of effective bonding strength c? and internal friction angle occur at 1% and 3% GGBS dosages，respectively. Adequate  effectively enhances the erosion resistance of GGBS-MICP solidified samples against the saline solutions，and the erosion resistance follows the sequence：water＞1%Na2SO4＞1%NaCl. SEM analysis reveals a denser C-S-H gel structure in GGBS-MICP samples，with GGBS being capable of regulating the crystal morphology of bio-CaCO3，where the crystal growth of vaterite predominates over that of calcite and aragonite. The difference in pH values(ΔpH) between＜10  GGBS-MICP and original GGBS-MICP solidified silty-sandy samples indicates that ΔpH1 without Ca2+ is greater than ΔpH2 with . The increase in utilization quantity with increasing GGBS dosages in GGBS-MICP(with ) samples highlights the importance of sufficient calcium sources for the effective solidification using GGBS-MICP technology. Therefore，the adoption of＜10 3%GGBS-MICP(with ) solidification approach greatly enhances the engineering properties of silty-sandy soil，achieving the efficient utilization of solid waste，and aligning with the strategies of green and sustainable development.