The growth of global population，the resource shortage and the environmental pollution have brought unprecedented challenges to the development human sociaty. The water is the source of life，the basics of production and the foundation of ecosystem. The issues such as the shortage of shallow surface water，the ineffectively-controlled water pollution and the groundwater level decline confront significantly our life. Under the deep surface of earth，however，there is abundant water resource. The underground hydraulic engineering (UHE) is aiming to solve the issue of surface water shortage，to develop and utilize the groundwater resources sustainably and to provide the reliable water resources for underground spaces. This paper extends the traditional water resources engineering from surface to deep underground，and proposes the strategies，connotation and development principles of UHE. Three key scientific questions are summarized as the existence and movement of deep groundwater，the regulation of groundwater resource and the safety protection of underground hydraulic engineering. Several critical technologies are proposed including the underground water corridor，the underground water reservoir，the underground water network and storage，the groundwater level regulation，the underground energy storage and power generation，and the groundwater eco-environment monitoring，protection and restoration. These strategic ideas and key technologies are expected to provide the scientific supports for the green utilization of underground space and the national development of ecological civilization.

Constant-resistance-large-deformation(CRLD) cables have been widely applied in a large number of slopes for monitoring and reinforcement. In order to analyze the constant resistance of the CRLD cable，it is considered as a thick-walled cylinder problem according to the shape and the dimension of the pipe. Differing from the classical Lame equation and thin-walled cylinder issues，this study took into account the effects of axial and shear stresses，and established a 3-dimensional elasto-plastic analytical model. The 3D analytical model became the Lame equation in 2D under the plane strain assumption. In order to verify this analytical model，static tension tests were performed to 3 samples with different geometric parameters. The experimental results were compared with 2D and 3D analytical models. The calculated results with 3D model was close to the experimental data. The 3D analytical model can be used to estimate the constant resistance of different CRLD cables and for optimal design.

 In order to study the dynamic behavior of crack initiation and propagation of borehole-crossed bedding blasting under the high stress condition，the experimental system of digital laser dynamic caustics is used to carry out laboratory model experiments together with the dynamic caustics method. The results show that in the blasting with the borehole cross the bedding，the blasting gas energy is difficult to be efficiently utilized，resulting in the unsatisfactory crushing effect of the rock around borehole. The blasting stress wave plays a dominant role in the initiation and the early period of propagation of crack at the end of bedding，while the static stress has a major effect on the behavior of crack during later propagation. The angle between the bedding direction and the static stress direction has a significant effect on the fracture mode of specimen, the dynamic stress intensity factor and the propagating velocity of the main crack. In addition，with the increasing of the angle, the propagating time and displacement along the bedding direction decrease，but the maximum deflecting angle increases，the propagating path becomes more flexural.

The stability coefficient of slopes is derived generally according to the static equilibrium equations. The sliding surface is simplified to be a line chain and the slope is sliced vertically in the method of residual thrust.  This paper presents a method to calculate the stability coefficient by means of shear strength reduction. A set of equilibrium equations for sliced slopes was established on the assumption of invariable horizontal acceleration within the sliding blocks. The sliding force of the slope was controlled by the dynamic distribution coefficient. The dynamic distribution process of sliding mass from stable to instable was described usually by the process of shear strength reduction. The unified expression between the dynamic and static equilibrium of the slope was derived firstly in this article. The dynamic distribution principle has not only inherited Panjiazheng principle but also developed Sarma method. The computed stability coefficient was identical to that of residual thrust principle，and the computed acceleration was similar to but more accurate than that from Panjiazheng principle. An important conclusion of this article is that the reduction coefficient was linear to the sliding acceleration of slope，which played a critical role in explaining the dynamic mechanism of landslide displacement.

In order to study the mechanical properties and failure modes of the rod-grout interface，the direct shear tests of the rod-grout interface under different normal stresses and rib spacings were carried out. The characteristics of shear stress-shear displacement curves and dilatation curves were discussed，and the shear strength and the failure modes of the interface were analyzed and explained in detail. Four stages were observed in the shear stress-shear displacement curves and three stages were observed in the dilation curves. The experimental results indicated that the rod-grout interface had a strong characteristic of brittle failure，and the degree of brittleness decreased with the increasing of rib spacing. The failure modes of interface can be divided into the dilation slip failure，the shear through failure and the compound failure. With the increasing of the normal stress，the normal displacement decreased significantly and the failure mode transformed gradually from the dilation slip failure to the shear through failure. The normal displacement and the possibility of dilation slip failure increased significantly with the increasing of rib spacing. In addition，the shear stress in interface at dilation initiation was approximately 0.5 of the peak shear strength.

A limit analysis based on the horizontal slice method was proposed for the seismic stability of soil slopes reinforced with prestressed cables considering the inhomogeneity and anisotropy of multiple parameters. In this method，the sliding soil mass was divided into several horizontal slices，and the soil inside each slice was treated as a homogeneous part. The energy power of each horizontal slice was calculated based on the partitioning method and was substituted into the energy equation. According to the geometrical relationship among the polar angles of the slices，two polar angles were chosen to be the optimization variables in the computational program based on genetic algorithm theory. This method was applied in a case study，where the inhomogeneity was simplified as the linear distribution along depth. The reliability of this method was verified，and the influence of the parameters concerning the inhomogeneity and anisotropy on the seismic stability were analyzed. The results of analysis indicated that the yield acceleration factor of the slope is not sensitive to the inhomogeneity of cohesive force，it is therefore feasible to use the average value of cohesive force. The inhomogeneity of internal friction angle and unit weight have remarkable effects on the yield acceleration factor. The anisotropy coefficient influences greatly the yield acceleration factor.

Stability charts provide a simple and effective way to evaluate the preliminary slope stability. Using the Hoek-Brown strength reduction FEM，this paper presents a series of stability charts to determine the safety factor of rock mass slopes satisfying the Hoek-Brown criterion. Firstly，the explicit formula of instantaneous Mohr- Coulomb friction angle and cohesive strength are derived based on the normal and shear stresses relationship of the Hoek-Brown criterion. Then，implementation of the non-linear Hoek-Brown strength reduction method is achieved in slope stability calculations. An example analysis shows that the Hoek-Brown strength reduction method can obtain a reliable safety factor and slip surface，but avoid determining instantaneous friction angle ?i and cohesive strength ci through iteration analysis. On this basis，stability charts for calculating the safety factors with a specified slope angle ? = 45°and disturbance factor D = 0 are proposed. Then，the disturbance weighting factor fD and the slope angle weighting factor f? are established to illustrate the influence of the disturbance factor and slope angle on the stability of rock slopes，respectively，and the charts for determining fD and f? are then provided. Thus，Stability charts for rock mass slopes satisfying the Hoek-Brown criterion are produced. Finally，this method is applied to two real engineering cases，and the results show that the proposed stability charts can be used to determine safety factors of rock mass slopes reliably.

Excavation is routine operation in underground rock engineering. The unloading behavior of high stress rock differs from the rock under the loading conditions and results in engineering disasters such as large deformation or rockburst. In order to reveal the unloading effect，triaxial tests are conducted and the strengths of cubic red sandstone samples under loading/unloading conditions are measured. According to the tests，the samples firstly split in the adjacent area of unloading surface several seconds after the unloading，and then lose the bearing capacity as a whole. The test results indicate that under the rapid unloading condition，the unloading strength is 8.4% to 27.2% lower than the loading strength. The higher the confining stress is，the bigger the strength declines. When the confining stress is the same，the strength under double-direction unloading condition is lower than that of under one-direction unloading condition. A dynamic damping vibration model is established and the influence factors of unloading failure are discussed. The negative unloading effect of the rock strength revealed by the tests is of great importance for evaluating the rock parameters under unloading conditions reasonably.

A mechanical model of macro-rigid sliding planes with the finite element stress load without the previous assumptions was put forward to resolve the problems in the calculation of anti-slide stability of deep and complex sliding planes of gravity dams. The principle of extreme projection direction of vector forces was proposed. The objective and quantitative method of stability calculation that combines the micro finite element stress with the macro limit equilibrium was established. The limit equilibrium equation and the energy extreme condition equation of macro-rigid sliding planes of the gravity dam against deep sliding under the elasto-plastic finite element stress load were given in the extreme projection direction. The objective safety factor accurately obtained not only satisfies the equilibrium condition but also has the extreme value of energy. The calculated results of examples are reasonable and reliable.

For the rock slope without a definite sliding surface，finding the position of the most dangerous sliding surface with planar failure and critical inclination is beneficial to accurately assess the stability of the slope and to the design of slope engineering. The mechanical model of rock slope with a planar failure considering the water pressure is established，and the unified analytical solution of the critical inclination is obtained by minimizing the value of a certain function. The reasonable form of water pressure distribution on the sliding surface is adopted and the detailed analytical solution of critical inclination is obtained considering the facts whether the front edge of the slope is blocked，whether a tension crack exists at the back edge of the slope，whether the sliding surface is filled with water，etc. The analysis results indicate that within a certain range，the critical inclination increases with the the height of the slope，the slope angle，the internal friction angle of the sliding surface，the unit weight and the height of the water table. The higher the cohesion of the sliding surface，the lower the critical inclination. The critical inclination is firstly increased and then decreased with the depth of the tension crack. There are differences in the critical inclination obtained with the different models of hydraulic pressure distribution. The drainage conditions of the slope have a significant influence on the critical inclination. A program calculating the critical inclination was written according to the established analytical solution.

The strain rate is an important factor affecting the mechanical and acoustic emission(AE) characteristics of rock. The AE monitoring experiment of granite specimen with a circular hole under biaxial loading was carried out to study the influence of strain rate on AE characteristics during the elastic deformation process of rock. The relationship between the strain rate and acoustic emission rate(AER) during the rock deformation was studied. The mechanism of influence of the strain rate on AE characteristics was discussed. The trends of the strain rate curve and the AER curve were found to be remarkably similar，both with an increase-decrease-steady tendency in the process of elastic deformation. The relationship between these two curves in the process of plastic deformation was very weak，with the strain rate curve being steady and the AER curve increasing gradually. The variation of the AER was mostly caused by the interior frictional strength of rock，which was mainly determined by the fictional area(related to the interior crack) and the fictional rate(depending on the strain rate). In the elastic stage，the variation of AER was mainly determined by the strain rate，for the fictional area was regarded as constant with no interior cracks formed. In the plastic stage with the strain rate to be constant，the AER trend was determined by the increase of fictional area，which was caused by the appearance of local fissures. The AER variation caused by the interior friction was belong to continuous signal，while the AER variation related to the fracture development could be regarded as the mutational signal. Based on this difference，the concept of mutation rate of acoustic emission(AEMR) was put forward as a new AE index. The AE information caused by fracturing can be extracted effectively and the fracture development can be reflected comprehensively by the analysis of AEMR，combining with the tendency of AER.

To reveal the failure characteristics and crack propagation law of brittle rock under triaxial compression，mechanical parameters and deformation-failure characteristics of oil shale were obtained through laboratory compression tests under various confining pressures first. Based on the Voronoi block division logic，Fish language was used to compile a program，and a numerical model composed of trigon blocks was established in UDEC with the program. The micro-parameters was determined through calibrating properties，and an optimal trigon edge length average was chosen to construct standard compression numerical specimens. A series of compression simulations under various confining pressures were performed. According to the contact failure criterion，some FISH programs were compiled to calculate the number and length of contacts which broke in the compression process through iterations. The failure of contacts represents crack initiation and propagation，the compressive failure modes and crack development law were gained at last. It turns out that the shear angle increases slightly with the increasing of confining pressure，and the failure mode is brittle shear containing a set of conjugate shear planes，which is in agreement with numerical simulation results. The shear failure plays the main role in triaxial compression. For the same specimen，the number of shear cracks is 6 times of the tensile cracks，while the length of shear cracks are 40–70 times of the tensile cracks. With the increasing of confining pressure，the maximum length of shear cracks increase gradually from 77.3% to 94%，but the maximum length of the tensile cracks decrease from 16.3% to 12.1%.

In order to study the characteristics and failure mechanism of crack propagation in open rock bridge， the uniaxial compression tests with AE monitoring system were conducted on specimens with granite bridge. The crack propagation process was analyzed according to the stress-strain curve and the parameters and spectrum characteristics of AE results. The results showed that the crack was initiated from the tip of the prefabricated crack and its direction was controlled by the stress direction. The phenomena of pre-peak fluctuation rise and post-peak ladder descent occurred in the stress-strain curves. The sudden increase point and relatively calm period of AE count rate and energy rate appeared intermittently and the cumulative count rate and cumulative energy curves rose in steps. Each point of sudden increase corresponds to the formation of a macro-crack，and the energy for the propagation of macro-cracks was accumulated in the relatively calm period. The AE spectrum was more sensitive to the crack propagation. The high frequency and low amplitude and the low frequency and low amplitude signals corresponds to the inter-granular cracks or trans-crystalline micro-cracks. The high frequency and high amplitude signals corresponds to the mesoscale cracks or small scale trans-crystalline cracks. The low frequency and high amplitude signals corresponds to the macro-cracks. The middle frequency and low-amplitude signals are related to the friction of existing section.

This paper is concerned with the effect of control hole on the crack initiation and propagation under the cumulative blasting in coal seam. The propagation characteristics of explosive stress wave，the expansion characteristics of the main coal seam cracks caused by explosion，and the stress state and displacement characteristics of coal seam in the area around the control hole were investigated theoretically and numerically. The results demonstrated that the superposition of the incident compressive stress wave and the reflected tensile stress wave changed the stress field around the control hole and the main crack tip significantly and caused the directional expansion of main cracks and the formation of annular cracks around the control hole. In addition，the displacement compensation space provided by the control hole and the surface properties of the hole wall enhanced the tangential tensile stress between the coal particles on and around the control hole wall，which promoted the development and expansion of the radial cracks around the control hole. Furthermore，during the process of cumulative blasting and breaking，the combination of main cracks created by the explosion，the radial cracks and circumferential cracks around the control hole constitutes a large-scale fracture network under the effect of control hole. Besides，the deep-hole cumulative blasting tests were carried out to study the influence of presence and absence of control hole in coal seam on the degree of permeability increasing. The test results showed that the presence of control hole in coal seam under cumulative blasting improved significantly the permeability of coal seams. The increasing of the mean concentration of extracted gas in the area influenced by the control hole after blasting in the tests carried out in two different time periods are 1.78 and 2.48 times respectively that in the area without influenced by control hole.

It is well known the seepage flow in fractured rock has significant negative effect on the underground mining and cave excavation. While the opening of fractures in surrounding rock is wide and the fractures connect to the water rich stratums under the medium-high pressure，water-related disasters such as the water inrush could occur. A number of works on the seepage flow in fractures have been conducted by previous researchers. However，they were largely based on a single fracture or the simple fracture network，with the simplified cubic law (only suitable for low speed laminar flow) assumed. Therefore，further works need to be conducted on the seepage flow in fractured rocks to consider the effect of the complex fracture networks，the roughness of fractures，the high-speed nonlinear seepage flow and the local turbulent vortex. In this study，the characteristics of the high-speed nonlinear seepage flow through large opening fracture networks in rock is systematically investigated based on the two-phase flow theory using the open source platform OpenFoam (a package of computational fluids dynamics). Computational results precisely captured the process of water entry，air flow and water exit in the fracture networks with large opening. The distribution and variation characteristics of flow velocity are also revealed. Meanwhile，the existence of boundary layer is confirmed. The magnitude of the viscous shear stress in the boundary layers，the Reynolds number of the seepage flow，and the equivalent permeability are all estimated. This study clearly demonstrates that it is feasible to characterize the nonlinear seepage flow in fractured rock at the meter scale adopting the two-phase flow theory.

Water is one of the most active factors causing the weathering of and salt damage to the surrounding rock of the cave. Experiments of liquid water adsorption(contact) and gaseous water adsorption(non-contact) on the support conglomerate of Mogao Grottoes at Dunhuang were carried out. The water absorption characteristics of and its influential factors to the conglomerate under two types of conditions were compared and analyzed. The influence of each factor on the water absorbing capacity of conglomerate under liquid and vaporous water conditions was quantitatively analyzed through the mathematical models of the water absorption rate of conglomerate under the different influential factors. The experimental results showed that，for two different conditions，the characteristics of water absorption of the conglomerate samely varied with the time. The double logarithmic curves of water absorption have an upward convex shape. The average water absorption of liquid water is 14.9 times higher than that of gaseous water. The influence of the kaolinite on the water absorption capacity has a negative correlation under the liquid and gaseous water conditions. Especially，the kaolinite has the greatest influence on the rate of saturated water absorption under the gaseous water absorption. The illite has a negative correlation with the water absorption capacity under the liquid water condition，but has a positive correlation under the gaseous water condition. The porosity has a positive correlation with the water absorption capacity under the liquid and vaporous water conditions，and has the greatest influence on the water absorption capacity under the liquid water condition.

In order to explore the stabilities and differences of roadways in mining coal rock at different depths， the multiple linear regression and neural network model based on the Data Mining technology was used to perform the formula fitting and the weight analysis to the influencing factors of roof separation according to the monitoring data at different depths(700 m，850 m，1 050 m) within the area of Pingdingshan mine. The time series predictions to the roof separation and bolt stress have been carried out and the range of mining disturbance，roadway deformation and stress variation characteristics at different depths are revealed preliminarily. The distance from the mining face and the bolt stress was found to have the greatest influence on the roof separation，but the weights of them decrease by nearly 50% with the increasing of depth. The time series predictions of the roof separation and bolt stress show that the separation displacement of roadway roof and the bolt stress increase sharply at the depth of 1 050m with the advancement of mining face. At this depth，the predicted maximum value of bolt stress is 15 MPa，which is 2–3.5 times of the other two roadways，and the predicted maximum value of the roof separation displacement is 80 mm，which is 6–8 times of the other two roadways. With the increasing of depth，the variation of deformation and stress of roadways caused by the mining disturbance are more and more intensified， indicating that the close monitoring of stability and the proper measures of stability control should be applied for the roadway at depth over 1 000 m during the mining process.

In order to control and reduce rockfall，chamber instability，rockburst and other engineering dynamic，a new energy absorbing bolt is developed，taking advantage of the excellent properties of high damping rubber such as resisting impact effect，resisting shearing and absorbing energy. In this paper，to evaluate the energy absorbing characteristics of high damping rubber on impact protection of rock，a series of impact tests were conducted by SHPB system of ?50 mm. The energy absorbing characteristics of rubber samples，granite samples and rubber-granite samples under different loading rates were obtained. Results shows that the energy absorbed by high damping rubber increases linearly with incident energy，which indicates that the energy absorbing rate is nearly constant. For granite samples，the growth rate of fractal dimension is 5.5 times that of rubber-granite samples，which reflects good energy absorbing characteristics of high damping rubber under dynamic loading. In addition，the optimal relationship between rubber thickness and absorbed energy is confirmed. When the incident energy is between 200–250 J，the recommended rubber thickness of rubber is 15–20 mm. The research results can provide design guidance and theoretical foundation for the development of energy absorbing bolts.

In order to investigate the stress-strain characteristics of soil and the effective range of reinforcement under dynamic compaction，a two-dimensional dynamic model based on the capped yield hardening model was established and implemented through finite element method. A subroutine was programmed to simulate the parameters such as the soil modulus changing with the ramming times. With the proposed numerical model，the soil stress path in p-q plane( p is principal stress，q is deviatoric stress) and the stress-strain relationship in both vertical and horizontal directions were firstly obtained in the simulation of successive dynamic compaction. Results indicated that the cap surface expanded with the soil densification but that the reinforcement processes varied with the soil locations due to the different characteristics of the stress-strain relationship. Soils right under the pounder were mainly compressed in the vertical direction caused by the higher vertical stresses. However，the development of horizontal deformation at the shallower depths lagged behind the vertical deformation due to the effect of lateral inertia，while the deformation at the deeper locations developed simultaneously in both horizontal and vertical directions. Soils beside the pounder were mainly compressed in the horizontal direction due to the much larger lateral stresses. Then the reinforcement effect was quantitatively characterized by the relative distance between the peak value of the deviatoric stress and its front cap surface. When the deviatoric peak stress located above the cap surface，soils experienced the yield hardening. Otherwise，soils would not be compressed. Finally，the effective improvement ranges determined by the criterion of 90% compaction degree were found to be more conservative than those by the criterion of 5% relative density increment. Empirical formulae were put forward to predict the effective ranges of improvement for the sand foundation.

In order to investigate the strengths and deformation properties of frozen sandy soils，a damage constitutive model under different confining pressures was established to describe the strain softening under lower confining pressures, the strain hardening under higher confining pressures，and the transition process from the internal micro-fissures to macro-shear bands. The strength criterion for the micro-element of frozen sandy soil is described by the modified Mohr-Coulomb yield criterion. Due to the discreteness and randomness of the internal micro-defects and micro-fissures，an assumption is made that the distribution of the micro-element strength for frozen sandy soil obeys Weibull function. Meanwhile，a stochastic damage constitutive model is established to reflect the whole breakage process under the ambient loads，based upon the theoretical knowledge of the statistical theory and the continuum damage mechanics(CDM). The parameters are determined from the cryogenic triaxial compression tests. The distribution parameters of n and F0，along with their variation with confining pressures，are discussed in detail. In addition，the cohesion c，the internal friction angle ? and the damage variable D are analyzed under the different confining pressures. Finally，comparisons between the predicted results and experimental data demonstrate that the modified model can simulate well the whole stress-strain curves，reflect the transition process from strain softening to strain hardening, predict reasonably the volumetric strain-axial strain curves，especially both the dilatant behaviors under relatively low confining pressures and the compressive characteristics under high confining pressures.

Freezing technology is frequently used in the construction of connection tunnel between metro tunnels and in the operation of shield break-in and break-out. The freezing-thawing action has the negative effect on the stiffness of soil. When the subway comes into service，its safety could be undermined under metro loading. In this paper，the undrained dynamic triaxial test was conducted in laboratory to simulate the dynamic stress state of frozen-thawed soil under metro loading. The variation of soil stiffness was obtained and a stiffness softening model was established. A piecewise function was found to describe the stiffness softening of frozen-thawed soil under cyclic loading. When the metro loading is firstly loaded on the sample，there is a semi logarithmic correlation between the stiffness G and the loading times N. When the value of lnN reaches to 8.25 to 9.5，a turning point of the softening curve appears，and the stiffness declines little after this turning point. The freezing-thawing operation also accelerates the stiffness softening process under the metro loading. The stiffness can be increased by raising the consolidation degree. The higher the consolidation degree of the frozen-thawed soil is，the lower the stiffness loss ratio and softening speed are.

The p-y curve and calculation methods of piles in sloping ground need further study. The p-y curves for cohesive and sand slopes were established by improving the p-y curve proposed by Matlock for clay and the hyperbolic p-y curve for sand，and the deflection differential equation of piles in sloping ground were deduced accordingly. A finite difference method for analysing the internal force and displacement of pile shaft was proposed. The methods were verified through Matlock tests and indoor model tests. The calculated results of the horizontal displacement and bending moment of piles with the proposed method are in good agreement with the experimental results. The maximum calculation error is less than 20%，and the error increases with the horizontal loads and slope angles. The maximum bending moment of piles occurs within the depth of 3 to 6 times the pile diameter below the ground surface. The locations of the maximum and zero bending moment increase with the slope angles and horizontal loads. In order to improve the bending capacity of piles in sloping ground，reinforcement should be increased in the range of 6 times the pile diameter below the ground surface，and the full length reinforcement along pile shaft is recommended.

It?s hard to correctly determine the seepage coefficient of clay by experiment，and the existing empirical methods are difficult in parameter acquisition and poor in calculation accuracy. On the other hand, the empirical methods to derive the seepage coefficient of coarse-grained soil are mature. Therefore，the empirical methods of clay and coarse-grained soil were summarized and listed firstly. The ineffective and effective pores were defined for clay，putting it equivalent to the coarse-grained soil according to micro-bound water solidification. The parameter of equivalent void ratio was introduced to establish the equivalent method of calculation. The liquid-plastic limit method to determine the volume of ineffective pore was explained，and the feasibility of equivalent calculation method was also evaluated in three examples. The results show that the introduction of the equivalent void ratio unifies the calculation of the seepage coefficient for clay and coarse-grained soil. The maximum micro-bound water content is about 0.9 time the size of liquid limit. The seepage coefficients from Mesri Equation of clay and the method of Terzaghi or Curson-Karman after the application of equivalent method are in agreement with the experimental values.

A modified Shen Zhujiang?s elasto-plastic model with double yield surfaces(Shen?s model) in terms of the ratio of bioenzyme-based soil stabilizer was presented for the double yield surfaces elasto-plastic constitutive relation of the bioenzyme-treated expansive soil. The triaxial consolidated-drained shear tests of remolded expansive soil samples with the different ratios of bioenzyme-based soil stabilizer were conducted and the stress-strain relationship of bioenzyme-treated expansive soil was studied. The mechanism of the ratio of bioenzyme-based soil stabilizer affecting the relevant parameters in Shen?s model was analyzed using the parameter analysis method of Duncan-Chang model. The correlation expressions between the parameters of Shen?s model and the ratio of bioenzyme-based soil stabilizer were put forward and a modified method of tangent volume ratio ut based on the hyperbola relationship of - curve was presented. The physical meaning of the parameters of the modified Shen?s model is well defined，and the test methods for determining the parameters of the modified Shen?s model are the same as that of the Shen?s model. The comparison between the experimental curves and the modified Shen?s model curves of - and q- respectively shows that the modified Shen?s model describes reasonably the elasto-plastic stress-strain characteristics of expansive soil under the different ratios of bioenzyme- based soil stabilizer，and reflects the mechanism of the ratio of bioenzyme-based soil stabilizer in influencing the elasto-plastic stress-strain relationship.

A total of 36 groups of direct shear tests to samples with the sand-concrete interface were conducted under three different initial normal stresses and three roughness values of interface to analyze the mechanical properties of the interface and the effects of the unloading ratio and roughness on its softening and shear dilation. The soil for test is the gray silt in ③2 layer from Shanghai. The test results show that the initial normal stress influences indirectly the peak shear stress of interface by affecting the density and moisture content of the soil at interface. The peak and residual shear stress of interface decrease linearly with the increasing of the unloading ratio. The roughness of interface has no significant effects on the decreasing rate of peak and residual shear stress varied with the unloading ratio. The residual shear stress of interface increases gradually with the increasing of roughness and tends to reach the same value as the applied normal stress when specimens were sheared. The softening phenomenon of the interface becomes more obvious with the increasing of the unloading ratio. There exists a critical value for the roughness r of interface，i.e.，the softening of the interface is little affected by the roughness when r＜10 mm and is affected by the roughness when r＞10 mm. The shear contraction of interface appears under the loading conditions，and the shear dilation of interface transforms gradually to the shear contraction as the unloading degree decreases. The maximum value of shear dilation of interface increases with the increasing of unloading degree，whereas the maximum value of shear contraction decreases under the same roughness of interface.

In this paper，a model of tubular piles with variable section is proposed to simultaneously investigate the three-dimensional effect of variable section and the damping effect of tubular piles on the dynamic response. The analytical solution of the dynamic response is derived in Laplace domain，and its semi-analytical solution in time domain is obtained by inverse Fourier transform. A parametric study of the static drill rooted method and the pile structure is conducted. In order to emphasize the structure information of piles，the parameters of static drill rooted method should be restricted for the dynamic test，according to pile?s inside diameter and nodular segments? outside diameter，height and interval. On this basis，some conclusions are summarized to provide a practical guidance for engineering implementation.