In the process of grout penetration in sandy soil, the soil skeleton particles filtrate cement particle leading to an anisotropic diffusion, referred to as filtration effect. In this paper, Kozeny-Carman model is adopted to analyze the influence of filtration on soil porosity, permeability coefficient and slurry flow velocity. A constant pressure device is employed to investigate the grouting quantity variation of sand samples with different grading associated with different grouting parameters. Then, a quantitative relationship is obtained using multiple regression techniques. Test results show that filtration effect remarkably influences plugging behavior of cement grout and cement particle mass fraction is the main controlling factor of the grouting distance. According to Darcy’s law and regression relationship, theoretical values of grouting distance are determined with considering the filtration effect and ignoring the filtration effect respectively. Comparisons to the tested value indicate that omission of filtration effect lead to a considerable deviation of the theoretical value. Results can be used to instruct grouting design and offer references to similar projects.

To investigate the variation of matric suction of unsaturated soils under complex stress conditions, a series of isotropic consolidation and shear tests along the stress paths of different intermediate principal stress ratios b with constant water content is performed on intact loess with various initial suctions by using unsaturated soil true triaxial apparatus. The characteristics of suction of unsaturated intact loess under the condition of true triaxial stress are studied. The results show that suction decreases with the time in the process of isotropic consolidation. Suction decreases faster and the variation of suction is larger with the increase of initial suction and net mean stress. The suction after the completion of isotropic consolidation is smaller than the initial suction. The suction in the process of shear declines little or remains unchanged then rise rapidly with the time. The suction rising segment grows rapidly and the suction increment is larger with the increase of the net confining stress or ratio b. The suction after the completion of shear is greater than the suction after the completion of isotropic consolidation. The changes of the matric suction are closely related to the different combinations of the net confining pressure and the ratio b under the action of true tri-axial loading.

To investigate the performance of gypsum rock softened by water in the tunneling, gypsum rocks from Lower Triassic Jialing River Group were immersed by fresh water within different times for softening tests. The variations of uniaxial compressive strength, shear strength, elastic modulus and Poisson’s ratio of gypsum rocks with increasing immersion time were analyzed. The study indicated that, with the increase of the immersion time, the uniaxial compressive strength, shear strength and elastic modulus of gypsum rock exhibited a decline trend as a negative exponent, while Poisson’s ratio presented a linear increase trend under certain conditions. On the basis of damage mechanics, a softening damage evolution equation was deduced by considering time effect of gypsum rock. The relationship between the brittleness coefficient and softening-damage variable was revealed, and the softening-depth model at different immersion time was established. This study can be used to identify the softening area of surrounding rock caused by groundwater in the tunneling through gypsum layer.

The stress path triaxial tests were conducted to examine the coefficient of K0 of different stress paths and the undrained shear stress-strain under K0-conlidation. Formula of K0 coefficient was proposed for K0 overconsolidated soil. Based on anisotropic yield criteria, the undrained shear strength equations of K0 overconsolidated soil were developed. The validity of proposed procedure was verified by comparing calculated results and experimental results. The results indicated that the formula proposed by Mayne and Kulhawy overestimated the coefficient when the OCR was large. The undrained shear strength equations of K0 overconsolidated soil avoided errors which were caused by assuming a constant slope of rebound denoted by the effective overburden pressure, by considering the variation of K0 coefficient with OCR. Experimental results indicated that the proposed formula has the advantage to predict the undrained shear strength of K0 overconsolidated soil.

Energy pile is a new technology which supports the upper building load and extracts geothermal energy from its surrounding simultaneously. However, there is only little research available on the thermal-mechanical behavior of the piles subject to heating-cooling cycles and particularly repeated cycles. Model tests are carried out to examine the heat transfer performance and bearing characteristics of energy pile with embedded steel tube under working loading conditions over repeated temperature cycling, and especially the law of pile head displacement. Moreover, the displacement is also observed and analyzed under no load over one cycle of heating and cooling for comparative analysis. The results show that the heating-cooling cycles produce thermal strain in the pile shaft, heating generates compressive stress and cooling induces the tensile stress. Meanwhile, the positive and negative frictions are yielded in different parts of the pile side due to temperature. The upward displacement magnitude of pile top under no working load is 41% greater than that under working load, but the final displacement of pile top is approximately 10% of that under working load after one heating-cooling cycle. The repeated temperature cycling can lead to continuous accumulation of settlement.

Soil-rock aggregate mixture is one type of non-homogeneous dispersion material mixed by rock and soil. Its shear strength is mainly affected by the moisture content, rock content, particle gradation and the degree of compaction, etc. On the basis of previous studies and the lab large-scale direct shear tests, this study considers the effect of gradation parameters, attains the required degree of compaction level by controlling the quality and volume of samples, and analyzes the influence of the main parameters on the shear strength. The results show that, in well-graded case, as nonuniformity coefficient increases, the shear strength increases first and then decreases, and achieved peak when the coefficient is around 22.99, while little impact by the coefficient of curvature. The shear strength of soil-rock aggregate mixture increases with the increasing rock content and compaction degree respectively, but the increases less when the degree of compaction is more than 92.5%. The shear strength decreases after the first increase with the increasing moisture content, and peaks at near optimum moisture content. Based on the Matlab’ multivariate linear regression analysis, the order of primary and secondary influencing factors on the shear strength is rock content > moisture content> degree of compaction > nonuniformity coefficient > coefficient of curvature. Through the regression analysis of test results, fitting formulas of it’s shear strength are obtained with correlation index R2 bigger than 0.9.

The phase change evolution of pore water was analyzed in the freezing process. Based on the mechanism of sensible heat and latent heat, the frozen soil was divided into freezing soil and stable frozen soil. According to the physical meaning of latent heat, the specific heat of soil should be stepwisely calculated, namely melting stage, freezing state and frozen state as the temperature gradually drops from a general temperature to a larger negative temperature. In the melting stage, the pore water is always in the liquid phase. In the freezing stage, existing liquid water can be frozen into ice. In the frozen state, there exists no liquid water can be frozen into ice. Thus, the cooling of freezing soil would be accompanied by releasing of sensible heat and latent heat. In frozen soil, the further cooling was accompanied by releasing of sensible heat. Then, based on the weighted superposition of specific heat capacity of mixture, the specific heat of the frozen soil is investigated. A method is developed for calculating the specific heat capacity of frozen soil with consideration of latent heat.

This paper investigates the combined effects of dynamic stress, initial static shear stress, consolidation ratio and vibration frequency on growth of accumulative plastic strain of soft clay located in Lingang industrial district, Tianjin Binhai New Area. A series of cyclic triaxial tests is conducted with continental facies soft clay samples by GCTS triaxial torsional shear apparatus. It is observed that when shear stress increases, the ability of resisting cyclic loading enhances. The decaying rates of softening index become slow at the same time. Consolidation ratio affects similarly on growth of accumulative plastic strain. Empirical formulation between softening index and accumulative plastic strain is established considering various influencing factors above-mentioned. Based on previous results, referring to the relationship between softening index and vibration times of cyclic loading N, a new accumulative plastic strain growth model is proposed for soft clay in Tianjin Binhai New Area, including four types of influence factors. And applicability of this model is verified.

Penetration of the suction caisson closely depends on the suction pressure. If the magnitude of suction is set to be smaller, it will lead to a lower penetration velocity and increasing construction costs. The soil plug rises higher otherwise. Thus, the suction caisson cannot reach the desired depth, resulting in either insufficient bearing capacity for the requirements of foundation design, or the failure of sand foundation due to piping. The aim of the paper is to determine a reasonable suction range to meet requirements of engineering practice. The minimum suction is determined by using the static equilibrium method. The critical value of suction is obtained by the formulation of the effective stresses on the inside and outside of the bucket wall in terms of the Hencky’s stress equation. The maximum suction is obtained by using the mechanism of sand piping inside bucket. In addition, the maximum penetration depth of the caisson is determined in terms of the maximum and minimum suctions. Finally, the proposed calculation of suction are verified by the published data, and they have a very good agreement.

Compared with the solid pile, the vertical vibration of open-ended pipe pile is much more complex due to the complicated dynamic interaction between the soil plug and the inner wall of pipe pile. Therefore, the dynamic interaction of pipe pile and soil should be further investigated by considering the effect of soil plug. Firstly, by considering the lateral inertial effect and viscous property of pipe pile, governing equations of vibration for the pile surrounding soil-pipe pile-soil plug system under vertical dynamic loading are established based on the Rayleigh-Love rod model and additional mass model. Then, the analytical solution of velocity response in frequency domain at the pipe pile top is obtained by means of the integral transform technique and impedance function transfer method. The velocity response in time domain is also derived with half-cycle sine pulse applied on the top of pipe pile. Finally, the rationality of the theoretical solution derived in this paper is verified by comparing with the existing solutions and experimental results. Meanwhile, discussion about the influence of the design parameters of pipe pile on the vertical dynamic response at the pile top indicates that the dispersion effect of stress wave propagating in the pipe pile signifies with the increase of pipe pile section and the decrease of pipe pile length when the other parameters of soil-pipe pile system is constant. For the piles with same outer diameter, the reflective signals at the upper surface of soil plug and the pile tip are easier to be detected as the width of pile shaft increases, leading to the more reasonable assessment of the pipe pile integrity.

To investigate the dynamic response of the track and saturated soil under moving train loading, a coupling model was developed to analyze the rail structure and underground railway tunnel in the saturated soil by using the analytical method. The train load was simulated by a series of moving constant or harmonic load according to the geometry of a real train. The tunnel was simulated as a thin cylindrical shell of infinite length, and the soil was treated as a saturated poroelastic medium using Biot’s theory. The tracks consisted of two beam units, i.e., an upper Euler-Bernoulli beam to account for the rails and a lower Euler-Bernoulli beam to account for the slab. The tracks and soil medium were coupled by the force and displacement compatibility conditions at the tunnel invert. The different dynamic characteristics of elastic soil medium and saturated soil medium were analyzed. The effects of soil permeability, moving velocity of load, and natural frequency on track responses, rail displacements, and pore pressures of saturated soil were investigated. It is found that the dynamic response of saturated soil medium with a low permeability is different from that of elastic soil medium. But soil model has limited effects on the track responses. The pore pressure increases with decreasing soil permeability. The track responses and pore pressure of the saturated soil are greatly influenced by the velocity and excitation frequency of the moving load.

The drained triaxial compression creep tests were conducted on the dispersive soil of Ningmute Hydraulic Project by using a stress triaxial creep meter to investigate the creep deformation law under different initial conditions and to develop a creep model and its parameters. Experimental results show that the dispersive soil is a typical creep material and its creep deformation characteristics are quite obvious. It experiences three creep stages which include instantaneous creep deformation, attenuation creep deformation and constant speed creep deformation. The data also indicate that the creep rate and creep magnitude of dispersive are significantly governed by the water content of soil, confining pressure and deviatoric stress level, and stress-strain curves exhibit obvious nonlinear. Mesri model and Singh-Mitchell model are employed to describe creep properties of dispersive soil. By comparing with experimental results, it is found that the creep properties of dispersive soil can be described more accurately and easily by Mesri model than Singh-Mitchell model. Moreover, Mesri model has the characteristic of less parameters and strong applicability.

This study is to investigate the mechanisms of fracture extending in coal rock by pulse hydraulic fracturing. Experiments are performed on coal rock samples using a pulse hydraulic fracturing experimental system under triaxial loading by changing the pulse frequency and viscosity of fracturing fluid. Experimental results show that the fracturing response of pulse pressure and acoustic emission (AE) indicates the process of fracture extending induced by pulse hydraulic fracturing can be divided into four periods, i.e., scattered initiation period, uniform growth period, sudden coalescence period and failure termination period. The average rates of the spatial distribution of AE locations at the uniform growth period and sudden coalescence period are 4.6 and 9.6 times higher than the scattered initiation period respectively. The general tendency of b-value curves demonstrates that microcracks mainly propagate at the uniform growth period and large scale cracks primarily propagate at the sudden coalescence period. The partial b-value curves present the characteristics by combining the shape of ridge line with ladder shape. Before the fracturing pulse frequency reaches the optimum state, the duration of uniform growth period and sudden coalescence period affected by high frequency pulse pressure is shorter than that by low frequency one. Under the same frequency, the duration of uniform growth period exhibits a shortening trend, as the viscosity of fracturing fluid increases, but the duration of sudden coalescence period shows the opposite way.

The phenomenon of zonal disintegration may occur in rock mass of deep tunnels. The zonal disintegration can be regarded as a continuous phase transition process of internal structure of rock mass near deep tunnels, and thus the continuous phase transition model is used to describe the main features of zonal disintegration. In this study, a nonlinear continuous phase transition model is developed for zonal disintegration based on the continuous phase transition theory presented in references [15-16], and classical elasticity-plasticity theories. By using the phase plane analysis method, the characteristics of solutions of the nonlinear continuous phase transition model are studied and three different types of solution are obtained. Numerical methods are applied to solve the nonlinear governing equation of the model, and results are compared with those of the linear model. Numerical results indicate that the spatial distribution of deformation of rock masses is represented well by the nonlinear model. For instance, nonlinear solutions can be used to simulate the variation of the distance between two adjacent fractured zones in the radial direction and the occurrence of inner rock burst. At last, the effects of coefficients on the solutions are given and its physical mechanism is clarified by using the elastic foundation string theory. The developed nonlinear continuous phase transition model can be served as a guidance to investigate zonal disintegration near deep tunnels.

The bentonite/sand mixture is an optional buffer/backfill material in the geological disposal of high level nuclear waste. Understanding the volume change characteristics of this mixture under coupled hydro-mechanical condition is highly significant for evaluating the long-term safety of the repository. In this investigation, six groups of bentonite/sand mixtures were prepared with different contents of quartz sand, i.e. 0%, 10%, 20%, 30%, 40% and 50%. Then a series of experiments including uniaxial confined compression (to a final dry density of 1.7 Mg/m3), saturation (under 0.2 MPa vertical pressure) and reloading tests was carried out in sequence. The volume change behavior of bentonite/sand mixtures was understood, in addition, the effect of sand content was also analyzed emphatically. It is found that: (1) The compression behaviour of mixture significantly depends on sand content, saturation state and density. As the mixture is unsaturated and at relative low dry density (?d <1.7 g/cm3), the compression index decreases linearly with increasing sand content. However, as the mixture is saturated and at relative high dry density (?d <1.7 g/cm3), the compression index of mixture is independent of sand content, and the corresponding value is much less than that at unsaturated and loose state. (2) The final swelling strain of the mixture decreases exponentially with increasing sand content, while increases exponentially with effective bentonite dry density. As the sand content in the mixture is higher than 40%, volume collapse can occur during saturation, and collapse extent is enhanced by increasing sand content. (3) The observed volume change characteristics of the mixture under different hydro-mechanical conditions or at different sand contents are mainly related to the distribution of bentonite and sand in the mixture, and their dominant effect on soil skeleton. Generally, the higher sand content and higher density, the volume change behavior the mixture is more conditioned by sand. (4) The swelling pressure of mixture decreases with increasing sand content. A parameter terms as “effective bentonite dry density” is introduced to quantitatively describe the swelling pressure of the mixture, and an exponential relationship between these two parameters is developed which can be used to predict mixture swelling pressure. Moreover, the swelling mechanism of bentonite/sand mixture is further analyzed based on volume change ratio and mass distribution ratio of montmorillonite in the mixture. In conclusion, this work is a reference for the selection and performance optimization of buffer/backfill materials.

A large scale shaking table test was conducted on the bedding rock slope with soft interlayers in this paper. Based on the transfer function theory, we analyzes the relative transfer function and absolute transfer function of this slope. The effect of these two kinds of transfer function on function characteristics and the calculation of seismic parameters is studied. The feasibility and accuracy of estimating the frequency domain response by the transfer function is also analyzed. The results show that the transfer functions calculated by the acceleration near slope face and in the middle of slope body have similar form, and there is small amplitude difference. The real part of relative transfer function is less 1 than that of absolute transfer function. The imaginary part of two kinds of transfer function is the same, and both sides of imaginary part have a high degree of symmetry. Thus seismic parameters calculated by the imaginary part have high accuracy. When the frequency is greater than 10 Hz, the module of relative transfer function is close to 1, while the module of absolute transfer function is approximately 0. The natural frequency obtained by the imaginary part is similar to that calculated by the module of relative transfer function. The damping ratios calculated by the modules of relative and absolute transfer functions are accurate. This study shows that the damping ratio of bedding rock slope with soft interlayers is lager than that of homogenous rock slope. The acceleration mode calculated by the imaginary part is the same as that by module of two kinds of transfer function. The estimation of frequency domain seismic response of bedding rock slope with soft interlayers by absolute transfer function is more accurate than relative transfer function. The correction of spectral characteristics for input ground motion can guarantee the reasonable mechanism of input ground motion on dynamic response of bedding rock slope with soft interlayers.

This paper presents an elastoplastic constitutive model unifying thermo-hydro-mechanical modeling for unsaturated soils. It is based on existing hydro-mechanical models of unsaturated soils and experimental evidence of temperature effects on soils. The average soil skeleton stress, modified suction and temperature are selected as stress variables, while the soil skeleton strain, saturation and entropy are selected as strain variables. LY and TY yield surfaces and the hardening laws are adopted for simulating temperature effects on volume change. The predictions are performed on the test results in literature including the isotropic compression tests and triaxial shear tests under different net stresses, suctions and temperatures. The comparisons between measured and predicted results indicate that the proposed model can quantitatively predict the volume change behaviour of unsaturated soils under thermo-hydro-mechanical coupling conditions.

During the blasting excavation of deep rock mass under high in-situ stress, energy sources of vibration are composed of energy produced by the detonation of explosive and strain energy released by excavated rock mass. The precision of the prediction of vibration peak induced by blasting excavation under high in-situ stress is reduced by using Sodev’s empirical formula and its improved formulas which are based on the charge per delay. On the basis of energy conservation, a model is proposed for predicting the vibration peak by using the method of dimension analysis. By incorporating with blasting field test in the diversion tunnel of Jinping Ⅱ hydropower station, the monitored vibration data of upper part tunnel is used as the learning sample to calibrate the model, and monitored vibration data of lower part tunnel is used as the contrast sample to test the model. Predicted results indicate that the proposed model has a higher fitting correlation coefficient and a lower root mean square error than traditional ones, and thus it can be well used to predict the vibration peak induced by blasting excavation under high in-situ stress.

In many practices of underground engineering, rock mass is subjected to neither pure creep nor pure relaxation. In fact, both stress and strain change with the lapse of time, and thus rock mass has time-dependent property, which eventually lead to fracture of rock mass. It is hard to exhaustively explain this phenomenon by creep and relaxation. This study comprehensively describes the characteristics of generalized stress relaxation and the significance of engineering practices. Non-linear Maxwell model is applied to solve the exact solutions of variable compliance type constitutive equation under the generalized relaxation condition. In particular, we obtain the accurate value of parameter-n1 which governs the time-dependent behavior in the pre-failure region by alternately loading rate and loading-unloading tests. Generalized relaxation tests are carried out under 70% and 90% stress levels, and the time-dependent behavior under different stress levels and different direction coefficients are also discussed. Finally, experimental results of generalized relaxation under 70% and 90% stress levels are numerically simulated by the variable-compliance-type constitutive equation based on the non-linear Maxwell model. Calculated results are good agreement with experimental data, and the properties of generalized relaxation of Kawazu tuff are well explained.

The latest edition of ‘Technical code for building slope engineering (GB 50330-2013)’ employs a novel seismic active earth pressure formula of retaining wall. However, there exists a certain degree of deviation when seismic earth pressure is calculated by using the formula in the code, and the height of action point of seismic earth pressure has not been given. Based on research results in the literature, we develop a complete derivation of the formula and the action point height of seismic earth pressure, and point out the mistake existed in the code (GB 50330-2013). The transformation method with a rotating computational model of retaining wall is adopted in the formula derivation, in which the seismic active earth pressure computing problem is converted into non-seismic active one. Therefore, the active earth pressure formula for non-seismic condition is developed by applying mechanically derivation process of the active earth pressure. The earth pressure calculating parameters are utilized by rotating earthquake angle. This achievement simplifies the complex solution procedure for the active earth pressure.

The entity coal side of 1306 tailgate, Dongtan mine, showed evident large deformation and serious squeezing failure during excavation. The UDEC (Universal Distinct Element Code) model of roadway driving along next goaf in the fully mechanized top coal caving face of deep mines is established to analyze the failure location of main roof and the evolution law of stress and displacement of surrounding rock. The mechanisms of deformation and failure of the entity coal side are revealed that the rotation of key block in main roof results in the inside shift of its fracture line. The high abutment pressure caused by the movement of the roof structure is distributed on entity coal side. The high squeezing stress finally leads to strong extrusion and rheological behavior of entity coal side. Based on the analysis of failure mechanisms of the entity coal side, supporting measures are proposed by combining bolt-mesh and long cables with high strength and yielding capability. Numerical results and field monitoring data show that the large deformation of entity coal side is decreased effectively, and its displacement is well controlled within 300 mm in the support of ?22 mm×6 800 mm long cables with the strength of 600 kN and yielding capability of 260~300 kN. The stability of 1306 roadway is guaranteed by these supporting measures.

The effective simulation of soil-structure interaction and the nonlinear characteristics of soil is a key technical problem for conducting a seismic response analysis of soil foundations on nuclear island. A one-dimensional finite element model of nuclear structure is established on the software platform of SuperFLUSH by simulating the radial damping through setting viscous artificial boundary in the limited area of foundation to analyze the response of free field to the ground motion input, then to describe the nonlinear dynamic characteristics in the near field ground using equivalent linear method. Furthermore a case study of a simplified lumped mass model for CPR1000 nuclear reactor in the soil foundation is investigated about the effect of SSI and nonlinear layered soil foundation on the seismic response characteristics of nuclear island building structure. Numerical results verify the reliability and applicability of the proposed model in engineering application. With the consideration of SSI effect, the model of viscous artificial boundary can effectively absorb the energy of scattered waves and reduce structural response. With the consideration of the nonlinear characteristics of soil, the acceleration amplitudes decrease to a certain extent in different directions, and the peak frequency shifts to low frequency. Based on above analysis, the combination of SSI and the nonlinear characteristics of soil is necessary for seismic analysis of nuclear island.

The rock discontinuity morphology directly influences the macroscopic and microscopic characteristics of shear failure and shear strength. It is important to clearly recognize the stability of statistical parameters of morphology and anisotropy of nature rock discontinuities based on different interval point-cloud data. Thus, quantitative analysis is required to investigate the shear failure mechanism and to establish reasonable shear strength criterion. Three groups of natural rock discontinuity with different roughness characteristics were measured by using three-dimensional (3D) white light scanning system. The point cloud data of 15 different sampling intervals were further obtained according to the principle of sparse point cloud data density. Moreover, we studied the variation law of 2D statistical parameters (Z2、Rp、 ) and the corresponding 3D parameters (Z2s、Rs、 ) with the change of sampling intervals. The results show that these statistical parameters are decreased with the increase of the sampling interval, the attenuation amplitude increases gradually, and the quadratic polynomial function can be well characterized its relationship. Meanwhile, it is also found that the anisotropy of discontinuities at different sampling intervals also has interval effect. The constructed dimensionless parameter DAC (Discontinuity Anisotropic Coefficient) and the anisotropic characteristic parameters are used to quantitatively describe this phenomenon. The results indicate that the anisotropy of discontinuities has positive relationship with interval effect. Finally, comprehensive analysis is conducted on evaluating the stability of surface morphology parameters and anisotropic parameters of natural rock discontinues. These parameters are stable in the range of sampling intervals between 0.1mm and 0.5mm and their deviation is less than 3%. Therefore, the sampling interval is recommended to set to 0.5 mm when using 3D scanning to measure rock discontinuity.

Multi-line overlapping shield tunnel has a complex form in underground space layout. The interaction mechanism between tunnels and soil is very complex. The new tunnel caused ground settlement and made adverse effects when traversing the existing tunnels. The drainage method model test was applied according to the technical requirements of shield tunneling and control requirements of approaching construction. The ground surface subsidence and the existing tunnels longitudinal deformation caused by strata damage and excavation unloading are analyzed for three typical construction methods, above-crossing, under-crossing and above-under crossing. Finite element model was built to simulate the dynamic construction process of model test. The results show that ground settlement turns to be larger, and the existing tunnels have an upward moving during the construction of up-crossing. While under-crossed, a smaller settlement with a tendency of subsidence is found in the existing tunnel. The under-crossing first and then above-crossing construction leads to a uniform variation in ground settlement during various construction stages, and the final settlement is relatively small. The above-crossing first and then under-crossing construction results in a great curvature in deformation curve of existing tunnel, and the deformation of existing tunnel fluctuates constantly. The research results can provide some theoretical basis and prophase guidance for similar projects in the future.

The Yangzhou Slender West Lake tunnel passes through the hard plastic clay stratum by slurry shield machine in full section. During the construction process, surface subsidence accidents occurred many times after halting the slurry shield machine. The disintegration and softening of hard plastic clay in soaking condition are investigated by experimental study. The analysis of reason of surface collapsing by combining with field data, and the discussion about the procedure of surface collapsing propose corresponding preventive measures on collapsing in slurry shield construction in the hard plastic clay stratum. This study concludes: disintegration and softening of hard plastic clay occur when the slurry shield shutdown in slurry soaking condition, which generates the local cavity above the excavation face. With the time prolonging, the local cavity gradually develops to the surface, and ultimately forms the ground collapse. It is suggested to avoid halting machine when slurry shield go through full section hard plastic clay stratum. When halting machine is inevitable, reducing the halting time, avoiding the building region, meanwhile strengthening the monitoring and grouting reinforcement in the stratum above the excavation face are suggested to prevent collapsing.

It is the key to ensure smooth connection and safe mining of fully-mechanized top coal caving face by setting a reasonable segment pillar width. To guarantee the stability and effective water resisting capacity of bilateral segment pillar, an analytical model is established to calculate pillar stress. Pillar stability is assessed by means of numerical simulation and a borehole camera. The reasonable scheme of cutting top blasting successfully reduces the effect of mining-induced stress on the pillars in the vicinity of the 3107 working face, accomplishes pillar stress optimization, and ensures safe mining. The study has a certain significance for determining the width of coal pillar and stress release of segment pillar with hard roof under similar mining conditions.

By combining with the fast multipole expansion method (FMM), a fast and high precision multi-domain indirect boundary element method (IBEM) is developed for solving plane SH wave two-dimensional (2D) scattering by large-scale inclusions in the elastic half-space. Numerical results show that the FMM-IBEM can solve the problems of large-scale multi-domain scattering efficiently and accurately, and the storage capacity is also dramatically reduced. Then the rapid solution of millions degree-of-freedoms (DOFs) of plane SH wave scattering problem in the multi-domain is achieved on a personal computer. Finally, in the case of SH wave multiple scattering by inclusions with uniform distribution and random distribution in the half-space, the effects of stiffness and inclusion shape on plane SH wave scattering are investigated. Due to the multiple coherence scattering of elastic wave, the plane SH wave scattering by inclusion group is significantly different from that by single inclusion, and the characteristics of spatial distribution and displacement spectrum of the total wave field become complicated. The scattering properties of plane SH waves mainly depend on material hardness, geometric characteristics, the angle and frequency of incident wave. In addition, this study to some extent provides new techniques and theoretical basis for forward and inversion analysis of elastic wave scattering by complex inclusions in the half-space.

Stress changes induced by tunnel excavation and static pipe bursting inevitably lead to soil movements, and cause adverse effects on existing pipelines. Although the pipe-soil interaction has attracted increasing research attention worldwide, a simplified design code has not been developed to directly estimate bending strains of existing pipelines due to tunnel excavation and static pipe bursting. In this study, we use the commercial software of ABAQUS to systemically simulate the tunnel-soil-pipeline and pipe bursting-soil-pipeline interactions, and develop a single dimensionless plot for calculating bending strains of existing pipelines due to static pipe bursting and tunneling by using equivalent soil-pipe stiffness. Results of field monitoring and centrifuge tests are adopted to verify the proposed dimensionless plot. The ratio of the maximum pipeline curvature to the maximum ground curvature correlates well with the proposed equivalent subgrade soil stiffness. By given the ground displacement profile, pipe dimension, pipe material properties, and soil properties, engineers can apply the proposed dimensionless plot to directly estimate maximum pipeline curvature and/or strains due to tunneling and pipe bursting.

From the point of view of insufficient statistical data, a non-probabilistic reliability method for implicit performance function of geotechnical engineering is developed, which is a useful complement to the probabilistic reliability method. Firstly, we use the series of interval extension functions to reduce the interval correlation, and establish the iterative program for implicit function interval solution based on the subinterval method. Then, the experiment design is utilized to establish the statistical optimization model from the point view of comprehensive design to search the optimal solution and further to achieve a target range. Finally, by setting the confidence interval for the interval solution to stress function in the performance function, the non-probabilistic reliability model of the implicit performance function with limited computational cost is obtained. Moreover, through the solution for external function of support pressure, the detailed operation procedures of the non-probabilistic method are demonstrated. The results of the engineering examples show that the relative error between the maximum support resistance obtained by the proposed method and the maximum value of Monte-Carlo simulation is 0.83%, and the proposed method has a preference for absolute security.

Water and mud inrush is a common geological disaster in the completely weathered granite tunnel during construction. The key factor leading to the expansion of water inrush channel is the loss of soil particles, where the starting velocity is the critical condition of its loss. By considering the water erosion and particles adhesion, the particle 3D force in the water inrush channel cross-section was analyzed, and the function of particles starting velocity was established. The function includes the following factors, namely, the particle diameter, the particle position of channel cross-section and relative exposure degree of the particle. Meanwhile, numerical simulation was conducted to analyze the effect of the above three factors on the starting velocity. Numerical results showed that the starting velocity decreased initially and then increased with the increase of particle diameter. Furthermore, the starting velocities of channel cross-section at different positions were symmetry, and they were increased with the increase of relative exposure degree generally. By combining with orthogonal tests to analyze the sensitivity to starting velocity for the above three factors, the results showed that the diameter was the most sensitive factor. Accordingly, by considering the minimum starting velocity standard of channel cross-section and the randomness of relative exposure degree, the factors of particle position of channel cross-section and relative degree of exposure were simplified, and a simplified theoretical formula of the starting velocity was developed. Finally, the rationality of the starting velocity formula was verified by the laboratory test and field investigation.

From the study of the particle size boundary, rock distribution and particle frequency, a three-dimensional (3D) stochastic model of soil-rock mixture is generated by using the direct method. A set of relatively complete and easily operational method is developed for building the 3D stochastic model. Moreover, a finite-difference model is established with the same distribution as natural soil-rock mixture by combining the newly developed building method and FLAC3D random modeling method. Then, direct shear tests and its numerical simulations are performed to obtain shear stress-displacement curves under different vertical pressures, which is further applied to explore shear failure characteristics of soil-rock mixture and the interaction mechanism between soil and rock. Due to different levels of shear deformation of soil and the horizontal or rotational motion of stone under the friction action, the shear band of soil-rock mixture shows obviously irregular and discontinuous features in the shear process, and the shear stress-displacement curve exhibits significant strain hardening behavior. This study shows that the 3D stochastic model can be used to well represent macroscopic mechanical properties and microscopic failure mechanisms of soil-rock mixture, which can be served as a model to explore the mechanical property.

For a deep circular tunnel excavated by the full-face millisecond delay blasting, a two-dimensional mechanical model is firstly developed to calculate the vibration caused by dynamic loads in the process of rock blasting fragmentation. Dynamic loads include transient release of in-situ stress occurring on excavation faces and blast loading. By using this theoretical model and analyzing the vibration signals measured in the field, characteristics of the vibration frequency and their influence factors are subsequently studied. The results show that the vibration frequency in deep tunnel during blasting excavation is mainly affected by the duration of stress release, rising time of blast loading and the dimension of excavation faces. Rock vibration in the vicinity of blasting source is primarily caused by blast loading. Since the rising time of blast loading is much shorter than the duration of stress release, the vibration caused by blast loading thus has a higher frequency than that induced by the transient release of in-situ stress, and the peak velocity of vibration attenuates faster with the distance. At the far distance, the vibration generated by the transient release of in-situ stress may become a primary component. It is also found that the vibration caused by the combined action of dynamic loads has two dominant frequency bands. The transient release of in-situ stress is mainly responsible for the low-frequency vibration, while the blast loading accentuates the high-frequency vibration.

To investigate seepage characteristics of a single rough fracture, based on lattice Boltzmann method, a numerical model is established to analyze the water flow in laminar under pressure. The D2Q9 model is applied to simulate the discrete velocity direction. On the macroscopic-scale, the upper and lower boundaries (ux =uy =0) are assumed to be impermeable, and the left and right boundaries are controlled by pressure (i.e., the pressure on the left side is larger than that on the right side). On the microscopic-scale, the non-equilibrium extrapolation scheme is set on the pressure boundary and smooth fracture surface boundary, and the standard bounce-back scheme is set on the rough fracture surface boundary. The corresponding program is compiled to verify the classical cubic law of smooth plate fracture flow. The rough fracture surface is generated by setting piecewise random length and random width, and seepage characteristics of different rough fracture surfaces are discussed in detail. The results show that fracture surface roughness greatly affects seepage characteristics. With the increase of relative roughness, the deviation from cubic law becomes more obvious. Therefore, by considering the effect of relative fracture surface roughness, the modified cubic law is proposed according to numerical results of different schemes of rough fracture surface. This study lays the foundation for further research on the complex hydraulic characteristics of rough fracture.

The mathematical mesh of numerical manifold method (NMM) does not have to accommodate to various boundaries of physical domains, and thus the mathematical coverage is always built by the regular structured mesh. However, for most problems, it is wasteful to use the mesh with the uniform density on the the whole physical region. Therefore, it is necessary to study the implementation of the local refinement on structured mesh, and a method of refining physical patches is proposed to solve the problem. For a practical problem, firstly we determine each region in which the mesh needs to be refined, and it is found that the physical patches entirely is contained by the refined mesh. Then, an interpolation on the refined mesh is constructed inside each physical patch, and the original local approximation of the physical patch is replaced by the new interpolation. Thus, the order of the local approximation is improved. Numerical results show the proposed method has good convergence. In addition, for two-dimensional analysis, the stiffness matrix obtained by the proposed method is positive definite if local approximations on all physical patches are constant.

The construction of fine 3D geological modeling is one of challenges in realizing true 3D visualization of geological body. Based on Triangulate Irregular Network (TIN) and Corner Point Grid (CPG) spatial data model, we propose a hybrid spatial data model for constructing fine 3D geological model. Firstly, a tectonic-stratigraphic framework model is set up based on TIN. Then, the framework model is converted into fine volumetric 3D geological model based on CPG voxel data model, and thus the fine expression of inner space of 3D geological body is achieved. During the converting process, it is necessary to define the order of strata. The mesh of the geological body is further realized by calculating the spatial coordinate of each grid cell and validating and saving data file in GRDECL format. Finally, the fine geological model is displayed by inputting the file into QuantyView3D system. A fine 3D hydrogeological model of the key area of a coastal city is constructed successfully by using the proposed method, and the feasibility of this method is verified.

To handle the interactions between discrete medium and fluid in macro scale problems, the paper develops a fluid-structure coupling algorithm using smoothed particle hydrodynamics (SPH) method and discrete element method (DEM) and based on the principle of Darcy permeability test. A numerical model for landslide surge applying this coupling algorithm is developed by Fortran. The proposed procedure is used to simulate surge triggered by underwater block landslide, and numerical results are compared with the experimental data to verify the validity of the coupling algorithm. In addition, the generation, propagation and dissipation processes of surge induced by discrete medium landslide are also studied. Numerical maximum surge height is compared with the results of empirical formulas. Research shows that: simulated underwater block landslide surge is in good agreement with the tested results. In dealing with discrete medium landslide surge problems, the coupling algorithm can clearly reflect the process of surge formation, infiltration and deformation of landslide medium. Moreover the empirical formulas deduce different maximum surge heights. The process of discrete medium freely sliding along the inclined slope into water is closer to the vertical motion mode of Panjiazheng method, so does the numerical result.

Analysis of active earth pressure under seismic condition is one of important contents in the anti-seismic design of retaining walls. In order to expand the applicability of classic Rankine earth pressure theory, diamond unit body is investigated according to the mechanical characteristics of inclined back?ll. On basis of Rankine's conjugate stress concept, the formulation of seismic active earth pressure considering cohesive soil and back?ll slope is proposed based on Rankine classic earth pressure theory and Mohr-Coulomb yield criterion. The results show that the formulas of classic Rankine active earth pressure and Rankine active earth pressure of cohesive soil considering back?ll slope are special cases of the proposed formulation. The proposed formulation yields comparable results to several existing calculation formulations for cohesive soil. The maximum error caused by different calculation methods does not exceed 10%. In addition, the vertical seismic coefficient and values of cohesion and soil internal friction angles have also been explored.

Predicting the internal forces and deformation of the adjacent pile is significantly important for the impacts of tunnel excavation. The traditional elastic foundation beam method ignores the plastic deformation of pile-soil interface and its interaction, resulting error between actual results and predicted values. Based on Mindlin solution, we assume the interface is perfectly elastic-plastic for the pile-soil interactions, and derive the maximum shear stress and lateral soil pressure according to Mohr Coulomb criterion. From the constitutive equations, we derive the plastic solutions considering internal force and deformation of the pile due to tunneling. The validity of this new method can be verified by comparing the plastic solutions and the centrifugal test results. By accounting the plastic deformation in pile-soil interface, additional deformation by proposed procedure is greater than that by traditional foundation beam methods. Finally, the effects on the internal forces and deformation of adjacent pile are analyzed. Results show that large deviation of internal force and deformation in pile occurs in case where plastic deformation of pile-soil interface being ignored, and this deviation increases with the increase of ground loss rate.