Calculation of the deflection and tension of the membrane elements in geosynthetic-reinforced and piled (GRP) embankments is a controversial issue yet to be resolved. However, little attention was paid to the evaluation of the strain distribution in geosynthetic reinforcements and the three-dimensional displacements of the geosynthetic materials. Physical model tests of membrane effect of geosynthetic reinforcement in GRP embankments with different pile spaces were carried out using a self-designed equipment. In the model tests, the geosynthetic reinforcement was loaded by applying air pressure, the strains in geosynthetic reinforcement were monitored in different locations. It is shown that: the strain and tension of geosynthetic reinforcements are unevenly distributed, and the three-dimensional deformation of the geosynthetic above square-distributed piles can be represented by a parabolic face and a parabolic cylinder; the applied loads are carried mainly by 4 tensile strips between two adjacent piles, and the calculated tension in geosynthetic reinforcements is much larger than the experimental results. Based on the modeling test results, a new calculation method for the tension in geosynthetic reinforcements is tentatively presented in this paper.

To overcome the disadvantages of the traditional upstream dam-building method, an improved tailings dam-building method, i.e., upstream dam-building method using concentrated and classified tailings, is proposed. Based on the tailings dam model tests, and guided by the concentrated tailings technique, model tests of both the traditional and improved upstream dam-building methods were performed. Through these tests, the deposition condition and particle distribution of tailing materials were examined, and the variation of saturation line on dry slope was determined. The limit equilibrium method is used to calculate the minimum safety factors for different dam building methods. It is shown that: the building dam with highly concentrated tailings can lead to even particle distribution and obscure the particle classification, while reducing the inter-bed and fine-mud sandwiches in the sedimentary area; the improving upstream dam-building method favors the formation of a sedimentation type, which is good for drainage and flood control; the stability of tailing dam can be improved by 19% to 33% if the improved upstream dam-building method is adopted.

考虑不同应力条件和岩体节理属性，对静应力作用下含有长节理和局部短节理花岗岩的爆破破岩机制进行研究，探讨静应力、节理位置和充填介质物性等因素对爆破效果的影响。在节理间充填弹性弱介质和空气两种情况下研究半无限长节理面反射拉伸破坏效应和端部衍生翼裂纹扩展特性；在不同静应力作用下，改变爆源与节理间的距离，研究长节理面反射效应对爆破裂纹扩展的影响规律；在保持节理近端与爆源距离和节理长度不变的情况下，研究短节理端部衍生翼裂纹产生、扩展的规律。通过比较表明，当节理充填介质时，节理面的反射拉伸破坏作用弱于不充填情况，而前者的端部翼裂纹发育明显比后者充分；长节理阻断了爆破主裂纹的扩展，节理面反射拉伸应力波产生的拉裂纹与爆破主裂纹耦合贯通在爆源与节理之间岩体内产生密集裂纹网，改善了该区域岩体的破碎效果，但随爆源与节理之间距离的增大以及静应力增高，这种破岩效应逐渐消失；入射角对节理端部衍生翼裂纹的产生与扩展影响显著，而静应力对翼裂纹起抑制作用。

To research the long-term stability of engineering rock mass under triaxial loading, triaxial compression tests and periodic axial load fatigue tests are conducted on yellow sandstone under different confining pressures. Evolution laws of axial deformation, circumferential deformation and volumetric deformation are obtained. The tests results show that: the axial, circumferential and volumetric deformations of fatigue failure of yellow sandstone under conventional triaxial cyclic loading are controlled by the static complete stress-strain curves. The axial and circumferential peak deformations of fatigue failure and irreversible deformation present an inverted S-shaped law of three stages. The confining pressure can promote the development of the axial deformation, but restrain the development of the circumferential deformation. In fatigue failure deformation process, the development of peak strain ?max is similar to that of the reversible strain ?p, elastic strain ?e remains basically unchanged in three stages, and a change law of irreversible strain increment ??p can reflect the evolvement law of the strain rate in three stages. Deformation is more suitable to be used as a fatigue failure criterion of rock rather than strength, and the fatigue deformation properties of yellow sandstone obtained in this paper make the foundation for the establishment of the fatigue failure model.

Based on the theory of thermodynamic with internal variables, a soil water characteristic curve model is developed, which can comprehensively consider the effect of capillary hysteresis and volume deformation of unsaturated soil. Based on the second law of thermodynamics, it is shown that both capillary hysteresis and plastic deformation represent some kind of dissipative behavior in nature. The suction is not only related to the changes of fluid phases, but also controlled by the deformation of solid phase. There is mutual coupling effect due to the interaction between the soil-water characteristics and the solid phase deformation. The mechanism of hysteresis phenomenon and the effect of deformation on the suction force are expounded from a microscopic standpoint and a general model of soil-water characteristic curve is established. Then a boundary surface plasticity model is developed, from which a simplified model is developed to describe the influence of the capillary hysteresis and plastic void ratio change. The model is validated using the experimental data. The results show that the new model can address the effect of soil deformation, and accurately describe the capillary hysteresis phenomenon.

Through analyzing the relationship of function in wavelet transforms and the wavelet data, and meeting the conditions of the admissibility of Morlet wavelet, Marr wavelet, DOG (difference of Gaussian) wavelet, Haar wavelet and Daubechies wavelet, it is challenging to simultaneously achieve the smoothness and compactness. Moreover, it is difficult to find a suitable wavelet basis function for acoustic emission (AE) signals of rocks by wavelet analysis. According to the characteristics of AE signals of rocks, the impulse function of the 2nd-order underdamp vibration unit pulse is used as the basic function of the wavelet analysis. In order to fit the AE signals, three tectonic conditions are proposed, and the optimization is proved by each other. Finally, the wavelet based functions with rock characteristic parameters are constructed. It is proved that the new wavelet basic function has substantial advantages in comparison with normal wavelet based functions by dealing with AE signals of rocks. This method provides a theoretical foundation for the application of wavelet analysis in the field of rock acoustic emission.

Fracturing grouting could effectively improve elastic deformation properties of soils, and thus the soil deformation induced by excavation will be greatly reduced, so it is important to study the equivalent elastic properties of the composite soils. By investigating the fracturing-grouting mechanism and some case histories comprehensively, a three-dimensional geometric model is firstly proposed for composite soils with fracturing grouting based on homogenization theory. A volume ratio and stress distribution diagram of grout-soil is developed following equivalent principle. Then the analytical solutions of the equivalent elastic modulus and equivalent Poisson’s ratio of the proposed model are derived based on transversely isotropic constitutive law. The same elastic parameters of this model for specific conditions are calculated by finite element method (FEM). At the same time, the FEM results are compared with the analytical solutions. The Flac3D is employed to predict the ground settlements of a heat-supplying tunnel using the proposed model and its analytical solutions. The predicted results are then compared with monitoring data. The study shows that: the analytical solutions of the proposed model highly agree with FEM results; the proposed model and its analytical solutions are reliable and valid for engineering practice.

By the compacting test on saturated crushed mudstone with different gradations, the curves of axial displacement and grain size distribution with variable axial stress are obtained; and the influence of Talbol exponent on the compaction deformation and grain size distribution is analyzed. In the end, the relationship between compression modulus and axial stress is established. The experimental results indicate that the compacting process of saturated crushed mudstone can be divided into two stages: the rapid deformation stage (0-4 MPa) and the slow deformation stage (after 4 MPa). And about 80% of the deformation occurs in the first stage. Axial displacement, compression modulus and fractal dimension can be fitted by the exponential functions of axial stress respectively. Under same axial stress, the axial displacement increases and the fractal dimension decreases with the Talbol power exponent increasing. After the axial stress exceeds 8 MPa, the fractal dimensions of samples with different grain size distributions are very close. In addition, with the further increase of the axial stress, the fractal dimension difference between samples with different gradations will continue to decrease; and finally, the fractal dimension will tend to be consistent. In the initial stage of loading, the compaction deformation of samples mainly results from movement of rock particles, and in the late loading, the compaction deformation mainly results from rock particles crushing.

Bending failures are common phenomena in thin-bedded rock mass during underground construction. In order to investigate deformation feature and failure process of rock under bending conditions, three-point bending tests were conducted on single-layer limestone samples and bedded specimens with weak planes. Our results illustrate that a close relationship exists between flexural capacity of the specimen and bed thickness, and the variations of the peak moment per width and the deflection at the peak force with specimen thickness can be reasonably predicted by the beam theory. It is shown that various fracture modes appear during testing, including mode I cracks perpendicular to bedding, zigzag cracks with changing direction at interfaces(weak planes), and oblique cracks across bedding. Theoretically, crack paths in thin-bedded rock mass are mainly controlled by the relative magnitudes of strengths (or fracture toughness) between rocks and interfaces. Tensile modulus is obtained by iteratively using the double modulus theory. Iteration process converges relatively slowly when the difference between compressive and tensile moduli is significant. These findings prove to be helpful in understanding the deformation characteristics and failure mechanisms of thin-bedded rock masses under bending conditions.

A series of cyclic simple shear tests was conducted on the rubber-sand mixtures (RSMs) with 7 different rubber contents, under 4 different vertical consolidation pressures (p). The shear moduli and damping ratios of RSMs are evaluated for different shear strain amplitudes and rubber contents. It is shown that: with the increase of the rubber content of RSM, the stress-strain curves drop significantly at the same consolidation pressure, meanwhile the shear modulus of RSM decreases, and the damping ratio, especially small strain damping, increase significantly; the stress-strain curve and the shear modulus of RSM increase with the increase of consolidation pressure while the damping ratio decreases slightly for RSM with a certain rubber content. Increase of rubber content causes earlier attenuation of the generalized curve, proving to be more conducive for the applications of RSM as isolation material. The experimental results agree well with the results available in the literature, implying that the type of testing methods influences insignificantly the shear modulus and damping ratio of RSM.

Model tests on a single pile in sands with different densities are carried out by using a self-developed cyclic loading apparatus for a long term. A prediction model for cyclic cumulative displacement is established considering the influence of sand density, and its application scope is discussed. The regression analysis of model tests results reveals that the first cyclic cumulative displacement increases exponentially with cyclic load ratio, and the model parameters are related to the relative density of sand. The predicted values of cyclic cumulative displacement obtained by the power function model are consistent with the measured values within a certain level of cyclic cumulative displacement rate, but the cyclic cumulative displacement rate is associated with the sand density, therefore, the influence of the sand density on partition rule of cyclic cumulative displacement under cyclic axial loading should be considered.

To investigate the performance of reinforced embankment under cyclic and/or static loading, a series of model tests is performed on the geocell-reinforced embankments with various reinforced layers, geocell heights and welding distances, using the loading device of USTX-2000 developed by the GCTS company. The ultimate bearing capacities of the reinforced embankment subjected to both long-term cyclic loading and fixed-number cyclic loading are determined under various operating conditions. The experimental results show that geocell can significantly increase the ultimate bearing capacity of foundation soils and observably reduce the normal accumulated deformation at the critical failure of the top and the middle of slope; for a certain reinforcement spacing, with the increase of the number of reinforcement layers and the height of the geocell, the ultimate bearing capacity can be improved to different degrees and the normal accumulated deformation of critical failure of the top of slope can be reduced; the decrease of the weld space of geocell can also improve the ultimate bearing capacity, and the weld space of geocell has trivial effect on the normal deformation of slope. The number of reinforced layers in fixed spacing has little effect on the vertical cumulative settlement of embankment and can influence the normal accumulated deformation of the top of slope under long-term cyclic loading; the increase of the height of geocell and the decrease of weld space of geocell can reduce the vertical cumulative settlement of the embankment and the normal accumulated deformation of the slope to different degrees. The less the distance to the point of loading, the more significant the influence of reinforcement on the soil pressure of embankment, and the reinforcement increases the stiffness and compactness of soil, so that the soil pressure of the reinforced embankment increases more obviously than that of unreinforced embankment. For the unreinforced embankments, the change of dynamic load amplitude and vibration number can result in the reduction of the ultimate bearing capacity after vibration. For the reinforced embankment, when the dynamic load amplitude is greater than or equal to 30 kPa or the number of dynamic load vibration is greater than or equal to 1 000, the ultimate bearing capacity after vibration is improved.

A model experiment is conducted to study the bearing mechanism of geosynthetic encased stone columns. Bearing capacity, end resistance, deformation, destruction of the columns with different stiffnesses and lengths of encasements are analyzed. The radial deformation of the columns is measured by self-made radial deformation measurement instruments in the experiment. It can be found from the experiment results that the fully encased stone columns effectively improve the bearing capacity and stiffness of the stone column installing on hard soil layer. And the one with a larger modulus of elasticity of encasement material has the higher bearing capacity and stiffness. However, the partially encased stone column with an encasement length of 60% of the column length does not have significant advantages over the stone column. This is because the load capacity characteristic, deformation property, and failure mode of the fully encased stone column are different from those of the partially encased stone column. The load transmitted to the fully encased stone columns tip is higher than that of the partially encased stone column and the stone column. The deformation along the fully encased stone columns distributes more uniform compared to that of the partially encased stone column. And the maximum radial deformation along the fully encased stone column is lower at a certain pressure. What's more, the fully encased stone columns fail from penetrating into the upper cushion; while the uncased part of the partially encased stone column undergoes excess bulging.

In grouting reinforcement of water-rich and broken rocks, it is difficult to use the traditional cement-based material to simultaneously achieve water plugging and reinforcement. To solve the above problems, a kind of cement-based composite grouting material (CGM) is developed by using sulphoaluminate cement clinker and steel slag powder. The performances of traditional material and CGM are compared, based on the experimental results of the scanning electron microscope and X-ray testing methods. The effect of preparation methods, constituent contents and slurry preparation conditions on the performance (strength, volume stability and slurry viscosity) of CGM is systematically analyzed, and the feasibility and management effect of CGM are also examined with combination of field applications. The results show that the CGM should be prepared by the process of mixed grinding (less than 45 min), so that its hydration activity can be significantly improved. With the increase of the amounts of steel slag powder and the water-cement ratio, the required setting time increases and the strength decreases. It is found that there exists a late dry shrinkage phenomenon when water-cement ratio exceeds 1.2. Compared with traditional cement materials, the setting time and viscosity of the new CGM material are less sensitive to the environment, showing the good performance of the CGM material.

Establishing an accurate and reasonable contact surface constitutive model is the premise to investigate the bearing capacity, settlement and deformation of a pile foundation. By considering the depth effect of initial shear stiffness of the pile-soil interface, a zero-thickness contact surface constitutive model is developed based on the contact surface statistical damage constitutive model. Parameters of this model are fitted by shear experimental data. Simulation curves under different normal stresses are obtained and compared with experimental results. A corresponding calculating program is developed by implementing the FISH language into FLAC3D software to realize a secondary development of the pile-soil improved contact surface constitutive model. The developed model demonstrates that it is reasonable for calculating the nonlinear behaviour of the pile-soil interface stiffness, which is further used to realize numerical simulation of the pile-soil shear test. Numerical results indicate that the developed model can be successfully applied to simulate mechanical behavior of the pile-soil interface, and to describe the nonlinearity of the pile-soil contact surface stiffness well. Furthermore, it is suitable for programming calculation, and it broadens the application of the statistical damage constitutive model and establishes a foundation for the further study on characteristics of pile foundation bearing and deformation.

The existing optimal designs of tunnel support are all based on the Mohr-Coulomb (M-C) criterion, which cannot truly reflect the effect of the intermediate principal stress on rock strength and the three-dimensional unequal stress state of surrounding rock. Theoretical solutions for both the optimal support pressure of soft rock tunnel and the allowable maximum displacement of surrounding rock are derived based on the steady creep criterion expressed by the third invariant of deviatoric stress. The generalized spatially mobilized plane (SMP) criterion is employed to evaluate the strengthening effect of intermediate principal stress on rock strength. The obtained solutions in this study can be well generalized to the M-C criterion and the circumscribed Drucker-Prager (D-P) criterion, meanwhile they can be validated by an engineering instance. It is shown through parametric study that the differences among these three strength criteria actually reflect different considerations on the effect of the intermediate principal stress; the results of the M-C criterion and the circumscribed D-P criterion represent the two extremes, and then the results of the generalized SMP criterion is recommended; long-term rock strength is a key parameter for the optimal design of tunnel support, and the effects of surrounding rock cohesion and its internal friction angle are also significant, so that the rock strength parameters must be properly determined and their variabilities should be fully taken into account.

To evaluate the settlement problem of loess-filled venues, a series of high pressure consolidation tests was conducted on the remolded loess of different initial water contents and dry densities. It is found that the ratio of load to dry density p/?d is linearly correlated with the applied loading, based on which a settlement prediction model of filled venues of different thicknesses (loads) is established. The model has two parameters, namely, k and b. Parameter k decreases with the increase of initial moisture content w0 or initial dry density ?d0, satisfying a relationship among the three variables (i.e., ?d0 k + w0=1); whereas parameter b depends mainly upon the initial dry density and decreases with the increase of initial dry density. Compared with the previous compressive deformation calculation models based on the relation between void ratio and loading, the proposed method which has a higher precision fitting with experimental data shows a stronger correlation, showing that it is more suitable for the calculation of compression deformation of remolded loess.

Grouting lifting technique is considered as an effective method to dispose uneven settlement of shield tunnel; but the overlying earth pressure on the shield tunnel during lifting is complex and has not been investigated deeply, which makes it hard to estimate the stress and deformation condition of the shield tunnel lining. A small scale model test system is firstly developed to simulate shield tunnel lifting in the laboratory，in which the transparent soil and PIV technology are applied. The limit overlying earth pressure as well as velocity field of the overlying soil layer are obtained through the tests. Then the shear-slipping features are studied, which indicate that the overlying soil layer could be divided into four different motion regions. A calculation model of limit overlying earth pressure is established based on the associated flow rules together with velocity field of overlying soil layer; then the calculation formulas are derived on the basis of virtual work equation. It turns out that the calculated limit overlying earth pressures are 4%-15% higher than the results obtained from model tests.

A fractal model for the rockfill particle-size distribution is employed and the range of fractal dimension is estimated through fitting the gradation curves of 32 rockfill materials from different rockfill dams. The gradation of rock materials possesses good fractal behavior, so that the fractal dimension D can be used as a new index to describe the gradation characteristics of rockfill materials. Based on the analysis of statistics, it is shown that fractal dimension D varies from 2.348 to 2.699, and it is about 2.6 for most dam materials. Six sets of gradation curves for cushion layer of Gushui rockfill dam are designed based on the fractal model. By employing the stochastic granular discontinuous deformation method (SGDD), the effects of fractal dimension on compaction capacity and the macroscopic and mesoscopic mechanical properties of rockfill materials are analyzed. As fractal dimension varies from 2 to 2.8, the void ratio decreases first and then increases, the material achieves its maximum compaction at D = 2.7. The degree of heterogeneity of force chains increases with the increasing of fractal dimension, and attains the maximum at D = 2.8. By considering both the compaction capacity and the force chain heterogeneity, it is found that the gradation at D = 2.7 is the optimum gradation.

On the basis of the spatially mobilized plane (SMP) theory, a theoretical solution for the spherical cavity expansion pressure of belled piles is derived. Energy dissipation principles are applied to analyze the whole process of spherical cavity expansion. A yield criterion of spherical cavity expansion is obtained by using stress invariant. Then an elastic-plastic zone stress expression is developed by simplifying differential equation to derive the displacement and strain expressions. The solutions of the expansion pressure are derived by the volume conservation and energy conservation separately. The elastic deformation of the plastic zone is considered and the relation among the pressure p, plastic zone radius R and the expansion radius is established. The analyses demonstrate that the calculated value of expansion pressure agrees well with field testing results. The results also show that the plastic zone radius and expansion pressure increase with the increase of the expansion radius, but the growth rate gradually slows down and tends to be stable. The dilatancy angle of soil has significant influence on the plastic radius and expansion pressure. The plastic zone radius and expansion pressure increase greatly with the increase of dilitancy angle.

This paper develops an efficient random finite element method (RFEM) using subset simulation (SS) for slope risk assessment. Equations are derived for integrating SS with RFEM to evaluate slope failure probability and risk, and corresponding implementation procedures are illustrated. The proposed method is validated by using a soil slope example. The results indicate that the efficient RFEM based on SS can be viewed as a development of the original RFEM based on Monte Carlo simulation, and it significantly improves the computational efficiency of evaluating failure probability and risk as well as the ability to generate failure samples, particularly at small probability levels, which enhances the applications of RFEM to the slope reliability analysis and risk assessment. The proposed efficient RFEM expresses the overall slope failure risk as a weighted aggregation of slope failure risk at different probability levels and quantifies the relative contributions of those to the overall risk. In this method, slope reliability analysis and risk assessment are decoupled from the deterministic finite element analysis of slope stability, which highly simplifies the calculation procedures. In addition, it is found that the vertical spatial variability of the undrained shear strength affects the slope failure risk significantly.

The spatial and temporal distribution of the horizontal displacement and settlement of a metro tunnel adjacent to the foundation pit is analyzed. Due to the creep effect of the soft soil, the space-time effect caused by the excavation of the foundation pit should be considered, in addition to the limit of the horizontal support structure of the foundation pit supporting system. The influencing scope of excavation on adjacent metro tunnel is 2.5 times the depth of excavation, and for the right line far away from the foundation pit, the influence scope is smaller, even lower to 1.5 times the depth of excavation. Based on the monitoring data of foundation pit and adjacent tunnels, the ratio-horizontal distance relation curve is obtained. Then the horizontal and vertical displacements of tunnel are estimated based on the monitoring data of foundation pit and fitting formula. The ratio between horizontal displacement of tunnel and that of the adjacent soil at the same depth, which corresponds to the maximum accumulative displacement, generally ranges from 0.60 to 0.65, and becomes stabilized at 0.60 after the foundation slab casting is finished. The ratio between settlement of tunnel and that value of the adjacent earth’s surface, which corresponds to the maximum cumulative settlement, is largely located within the range from 0.50 to 0.60, and becomes stable at the level of 0.52±0.05.

Foundation pit excavation can produce additional deformation and internal forces on the adjacent piles of buildings, and it will reduce their bearing capacity. This will threaten the security of the upper structure if the deformation of piles is too large. To avoid the drawbacks of three-dimensional numerical analysis, such as complexity modeling and time consuming, and to take full use of foundation pit supporting displacements provided by the in-situ monitoring technique, a two-stage method based on virtual image technique is presented for determining the behavior of adjacent single-pile induced by foundation pit excavation. An improved calculation model, Kerr foundation model, is proposed, which makes up for the defects that Winkler foundation model cannot consider the continuity of soil. The virtual image technique is used to estimate the free-field soil displacements induced by excavation in the first stage. The foundation pit supporting displacements are employed to calculate the free-field soil displacements. In the second stage, the free-field soil displacements are applied to the piles, the displacement governing equations of piles are established based on Winkler and Kerr foundation models, respectively. The simplified formula for pile response induced by excavation is obtained, including horizontal displacements, bending moments, and shearing force. The results are compared with that from existing calculated results, measured data, and 3D FEM. Good agreements are obtained. The simplified solution based on the Kerr foundation model is more accurate than the solution of Winkler foundation model. It indicates that the simplified calculation method can provide a theoretical base to analyze the impacts of excavation on the adjacent piles effectively.

Anchor-net-cable coupling support is widely used in roadways and particularly in large broken zones. Since the anchor is not long enough to support the whole range of surrounding rock in a large broken zone, it should consider the influence of anchor support on the design of a combined supporting when using the anchor-net-cable coupling support. Through systematically analyzing the support mechanism of surrounding rock bolt anchorage in large broken zones, the bolt supporting of anchorage bearing coefficient is proposed. Then, a calculation method to quantitatively determine parameters for anchor-net-cable coupling support is derived. The results show that the strong fractured surrounding rocks are divided into a shallow anchored zone and a deep fracture zone. The former can bear part of the surrounding rock as the cable-pallet, and its stability condition when the bearing capacity is greater than the axial force and the fracture zone is effectively controlled. The bearing coefficient of an anchor is in the range of 0.45-0.75 for the large broken zone, which is depending on the anchorage zone, fracture zone, anchor pre-tightening force and anchoring force. In the field test, the convergences of roof-to-floor and two walls are 273 mm and 393 mm, respectively, and the surface displacement tends to be stable. The proposed method for determination of parameters of anchor-net-cable coupling support effectively ensures its regular applications to the roadways.

The demarcation of homogeneous structural domains within rock mass is normally the first step of rock engineering design, since mechanical and structural properties vary largely from one homogeneous domain to another. Most of researches focus on occurrence or the other joint parameters to statistically identify homogeneous regions or structural domains in jointed rock mass. However, only few studies have been conducted by integrating both of them. In this study, by considering the factors of joint occurrence and trace length, a novel method is developed for dividing the statistical homogeneous area of jointed rock mass, which is known as the Wilcoxon rank sum test. It is used to evaluate the degree of the similarity between two samples and then to realize the division of this region. Based on the data of occurrence and trace length of rock mass from four adits at two levels of the left abutment of Maji hydropower station in Yunnan province, the proposed method is applied to statistically identify different homogeneous domains, and to analyze the influence of multi parameters of jointed rock mass on a statistical homogeneous division. The results indicate that, the rock masses within the adits of PD241 and PD253, PD201 and PD241 are incorporated into a statistical homogeneous area, respectively, but those of PD201 and PD251, PD251 and PD253 are situated at different structural domains. Compared to Miller method, the results obtained by the proposed method are more likely to be reasonable and reliable. In comparison with field investigation, it is found that the results from the new method are more agreement with the actual situation.

To solve the problems of the sharply appeared mine pressure and determination of the support working resistance in mechanized caving mining process of the ultra-thick coal seam, a structural model of cantilever beam-masonry beam is considered. When the support resistance is low at the time of face pressing, the subsidence of masonry beam results in the rotational deformation and the failure of cantilever beam. Furthermore, it leads to the slide breaking and pressing frame of the cantilever beam and the violent behavior of mine pressure. A mechanical model of breaking cantilever beam structure with a central crack is thus established. The effect of a serious of factors on the failure of the cantilever beam is analyzed by using fracture mechanics theory, and the expression of the support load is derived. The results show that the angle and length of the crack are the main factors influencing the failure of the cantilever beam. In addition, the support load is also relevant to other factors such as the falling step distance of the masonry beam. Based on the geological conditions of the fully mechanized working face 8102 in Tashan mine, the calculated support loads are used to verify the theoretical analysis. The results are of practical significance for analyzing the support load of ultra-thick coal seam mining and for providing reasonable selection of support to reduce the roof accident.

In view of shield tunnel projects in soft soil area of Shanghai, the advancing speed of the shield machine is adjusted, and the water pressure, soil pressure and strata stability are observed in the whole process of the shield crossing the monitoring section. According to the measured soil consolidation ratio, the pressure of artesian water and other strata soil properties, the influences of the varied advancing speeds are comprehensively evaluated by the mechanisms of soil disturbance. The analysis indicates that the shield tunneling in soft soil area is approximately 4 times of the diameter (D=6.2 m). By keeping the chamber earth pressure and grouting amount invariant in the shield advancing process, the change of the hoisting speed of the jack to great extent influences the soil pressure and deformation. Reducing the advancing speed results in the decrease of the disturbance range of water and earth pressures and then the disturbance is generated by the pressure increment, but leads to more stable values of the induced surface heave. The corresponding change value is determined by the ratio of depth to the advancing speed. Experiment results provide a certain valid reference to ensure the effect from the shield speed of the jack at the depth of 4D.

In large deformation tunnel, the supporting bolt tensile failure occurrs frequently due to the excessive deformation. In order to extend the application of the bolt in large deformation tunnel, a yielding bolt with constant resistance and extrusion slip is developed. It realizes the function of yielding by extruding and stretching the bolt to consume the energy of surrounding rock deformation caused by high ground stress. By adjusting the position and grip of the extrusion sleeve, the yielding pressure and the maximum yielding distance can be adjusted based on the specific characteristics of the surrounding rock in large deformation tunnel engineering. Through the indoor experiment, it has proved that the yielding bolt with constant resistance and extrusion slip is reliable. The mechanical characteristic of the yielding bolt is obtained. The yielding bolt is applied to the Maoyu tunnel engineering at Lanzhou-Chongqing railway, the experimental results show that the yielding bolt in large deformation tunnel engineering can exert an effect of yielding support reliably, but in soft rock tunnel with large deformation, the grouting and other means should be used to guarantee the yielding bolt effective support. Yielding bolt can effectively control the deformation of surrounding rock, and ensure the convergence effect of large deformation of surrounding rock in tunnel.

To evaluate the mining-induced surface movement and deformation in the coal field covered with thick alluvium, the parameters used to characterize the surface displacement are determined with the process of mining, and a formulation for calculating the subsidence velocity is developed based on the measured data from the surface observation stations. The results show that when the working face is pushed to the maximum settlement point of 170 m, the point of the sinking speed reaches the maximum value of 22.85 mm/d. The longer distance the panel advances, the higher value the surface maximum subsidence velocity reaches, so with its lagging distance. However, after the distance of the panel advancing is over about 600 m, the increase of these values is gradually much smaller, and then these values reach stable state of 22.00 mm/d and 165 m. In the following mining process, the surface subsidence velocity curve constantly moves forward with a fixed shape and lagging distance to panel. With domestic mining cases in the area covered with thick alluvium, relativity between dynamic subsidence parameters and the geological and mining technical parameters is analyzed with multiple regression method. At last, based on dynamic subsidence parameters variation, subsidence velocity prediction formula is obtained, which can predict subsidence velocity of any position at any time along the strike major cross direction. Comparing the expected value with measured value, it is proved that the expected results can meet the needs of the engineering practice.

The soil piece with finite width between the adjacent foundation pits in synchronous excavation is different from the one adjacent to existing building foundation pit. Their destruction direction is unknown, the destruction forms are various. Therefore it is difficult to apply the existing methods to calculate the active earth pressure on both sides of the soil piece. The silo balance principle is intruduced. By the mechanical equilibrium analysis and solving differential equation, a formula is developed for calculating the active earth pressure on the supporting structure installed on both the sides of soil piece (sand and clay) with a finite width. The formula has fewer parameters, less conditions and is convenient to use. By comparing with results from the Rankine active earth pressure calculation formula and the finite element simulation, it is found that the results by proposed procedure are consistent with those by Rankine’s method and finite element simulation. But the earth pressure by this paper is far less than Rankine earth pressure at the same depth.

A classification indicator is proposed and the hazard levels are classified for the carbon dioxide (CO2) dispersion when a well blowout is encountered at a CO2 capture and storage (CCS) site. Taken the topography and climate conditions of Yijinhuoluo Qi in Erdos city, Inner Mongolia, China, as the environmental background, the CALPUFF model is used to simulate a supposed well blowout with a reference import from the Shenhua CCS demonstration project. The dispersion distance of released gases is calculated, then the relationship between the dispersion distance and the blowout speed is chosen to classify the hazard levels of CO2 dispersion, and finally its influence is conducted. According to the maximum dispersion distance of the CO2 concentration in 50 000 ppm, the hazard levels of CO2 dispersion are classified into three levels by using an indicator of blowout speed. A well blowout with a release rate of CO2 between 0 and 2.5 m/s is identified as Level 1 corresponding to the dispersion distance of 0-130 m; a release rate of CO2 between 2.5 and 10 m/s is as Level 2, and the dispersion distance is 130-180 m; a release rate of CO2 between 10 and 20 m/s is as Level 3 with the dispersion distance of 180-420 m. According to the experimental research, it is shown that the persons around the leakage well within 420 m are affected when the hazard level reaches Level 3, which means the workers in the operational site, the residents around and the pedestrians passing on the nearby county road are all affected. Our research on the classification of hazard levels not only contributes to the supervision of relevant government agencies to the CCS projects but also is beneficial to the promotion of coming CCS demonstration projects in China.

The search for non-circular slip surface is a key to slope stability analysis, and its essence is to find a slide path with a minimum safety factor. Using radial movement optimization, a new global optimization algorithm, to search a path has advantages in calculation, storage and step, but the result is unstable. To overcome the instability of the algorithm in the searching process, its data structure is adjusted and an improved radial movement algorithm is developed, which make the route search stable. A safety factor corresponding to a searched path is calculated with unbalanced thrust method. Under the premise of feasibility analysis, the dichotomy is employed to solve the safety factor quickly and accurately. Based on a calculation example of a reservoir bank slope, the effectiveness of the improved radial movement algorithm is verified. Analyzing and comparing the search results obtained by improved radial movement optimization algorithm and particle swarm optimization, it is shown that the former is apparently superior to the latter in convergence and stability.

Based on the in-situ observations of a high-steep rock slope in Baiyun Ebo eastern open-pit iron mine, and the data of mechanical properties of rocks and joints, an equivalent rock mass (ERM) model is developed to describe the joint distribution characteristics at multi scales from laboratory tests, in-situ tests to engineering applications. Uniaxial compression tests are conducted on rock mass ERM models to investigate the scale effect and anisotropic properties of uniaxial compressive strength and elastic modulus. The results show that the existence of joints results in the size effect and anisotropic behaviour of rock mass, and these properties tend to gradually weaken with the increase of rock block size. The representative elementary volume, uniaxial compressive strength and elastic modulus of rock mass in the selected research area are 20 m×10 m×10 m, 1.46 MPa and 3.91 GPa, respectively. The relation between uniaxial compressive strength, elastic modulus and the axial size of rock mass is approximately a gradual exponential function which can directly represent mechanical parameters of rock mass.

There is an inherent relationship between deformation and strength parameters for a typical rock mass. From the perspective of material properties, the essence of the strength reduction method (SRM) is to search another appropriate rock mass, which could lead to a slope at the critical state under the current slope conditions. It is obvious that when the cohesion and friction angle are reduced but other mechanical parameters remain unchanged, the new mechanical parameters may be contrary to the inherent relationship, and thus the newly found rock mass might not be exist objectively. Therefore, it is necessary to adjust the mechanical parameters of rock mass during the reduction of cohesion and internal friction angle. Based on the above analysis, we propose a strength reduction method which considers the deformation and strength parameters of the coordination fold reduction. Then, the deformation parameters (i.e., elastic modulus and Poisson's ratio), tensile strength of safety coefficient, plastic zone are discussed in detail. The results indicate that the safety factors obtained by the proposed method are very close to those calculated by the limit equilibrium method. In addition, the distribution of the plastic zones in the failure slope is in agreement with that of the real slope state. The plastic shear zones mainly appear near the free face of a slope, while the plastic tensile zones largely locate at the top of the slope.

Based on the cavity expansion theory, the influences of the geometric parameter (edge angle), the rock mechanical parameters (compressive strength and Poisson's ratio) on the radius of elastic-plastic interface in rocks under wedge cutting conditions are studied. The results show that the smaller the blade angle, the larger the damaged area and the larger the elastoplastic radius when the contact width between the blade and rock is constant. When the other factors remain unchanged, the increase of the Poisson's ratio or the increase of the compressive strength leads to a decrease in the radius of elastoplastic interface. The sensitivity of the three factors is studied by the sensitivity analyzing method. Under a certain contact width, the blade angle has the most significant influence on the damaged area. A numerical model of the penetration of indentation is developed by using the software ABAQUS. The numerical results show that the changes of the blade angle can significantly influence the spatial position of the point with the maximum tensile stress and the direction of minimum gradient.

To realistically simulate stress states of coal in tunneling laneway frontal and other stress environments, a true triaxial gas-solid coupling testing system is developed to examine the damage deformation and gas seepage of coal under three-dimensional (3D) stress condition. The device is composed of a true triaxial pressure chamber, servo-hydraulic system, gas seepage system, and monitoring and control system. The suitable size of a specimen for the equipment is 200 mm×100 mm×100 mm. The maximum pressures of the axial compression (?1), the principal lateral direction (?2), and the secondary lateral direction (?3), and gas are 70, 35, 10 and 6 MPa, respectively. The characteristics of the device are shown as follows: the axial pressure (?1) and the principal lateral pressure (?2) are loaded by rigid head loading methods, and the secondary lateral pressure (?3) is applied by flexible loading. The 3D stresses are applied independently. To ensure the heads at σ1 and the ?2 directions without disturbing by each other at the same loading time, the rigid-flexible head and pressure transmission slider are designed. The loading control functions of the servo-hydraulic system are more likely to be stable and reliable, and the stress and displacement loading can be controlled accurately. Honeycomb holes are used for ventilation of specimen surface in the gas seepage system to ensure the uniform distribution of gas pressure at the inlet of the specimen. A variety of high-precision sensors are used to monitor and record the stress, the deformation and the gas seepage discharge of coal in real-time. Seepage experiments in two different loading paths are performed to verify the accuracy and reliability of the experiment system which performs well. The device can be used to reveal the coupling mechanism of coal and gas under 3D stress conditions, and to provide a reliable experimental basis for prevention and control of gas disaster and investigations on gas extraction.

In the rainfall model test of the bedding slope, since it is challenging to install the monitoring components and to measure stress of retaining structures, a fiber grating technology is adopted to solve the above problems. Based on the deductions of different measuring principles of fiber grating sensors, the corresponding experimental components have been developed by combining with the fiber Bragg grating (FBG). Two kinds of the bedding rock slope with a dip angle of 40 ° of rock layers are constructed to perform model tests under rainfall conditions with and without retaining. Physical and mechanical characteristics including slope displacement, frame-anchor internal forces and moment of anti-slide pile are monitored. The results show that components of fiber grating technology maintains well in difficult environments. Interlayer dislocation phenomena appear in the bedding rock slope with high dip angle without retaining during rainfall. When the retaining structure is used in the model slope, the displacement model changes from a steep variant to a slow variant, and the interlayer dislocation phenomena are also restrained. The locations of retaining structure where the internal force grow fast transfer from the rear retraining structure to the front and deep portion of the slope. The internal force increases slowly in the anchor at the top of the slope, while increases rapidly at the lower part of the anchor and the pile. The internal force reaches a steady state until about 2 hours after rainfall. It can be used to determine the key monitoring location and time for the bedding slope under rainfall conditions. This study provides new insights and methods for the bedding slope monitoring and evaluation of the operating characteristics under rainfall conditions.