Study of the coupled thermo-mechanical constitutive model of clays is an important issue in several specific fields, especially in the analysis of nuclear waste isolation. Based on the results of undrained triaxial tests on Boom clay at different temperatures, a coupled thermo-mechanical elastoplastic damage model is established. The yield surface is based on that of the Drucker-Prager cap model. A hardening law and a thermal/mechanical damage evolution equation are presented. The developed model is implemented into ABAQUS finite element code through subroutine USDFLD. Model parameters of Boom clay are then calibrated from experimental results by back analysis. Three dimensional coupled THM modeling of ATLAS III in-situ heating tests in HADES at the HADES underground research facility is performed by the proposed constitutive model. The numerical results are in agreement with in situ measurements, demonstrating that the model can reasonably depict the main features of thermo-mechanical behaviors of Boom clay.

A rigid retaining wall with a vertical back face backfilled with a cohesionless soil is analyzed. By assuming that the soil arch behind the retaining wall has a form of circle, the active earth pressure formulation of horizontally layered soil is modified with considering the shear stress between the soil layers; and a set of theoretical formulae are proposed for calculating the active earth pressure, the total earth pressure and its action point for the retaining wall under translation mode. In order to check their accuracy, the predictions by the proposed formulae are compared with the results of full-scale test and existing theories. The calculated results show that the mean shear stress between level soil layers is related to the wall-soil friction angle and the internal friction angle, and it is distributed nonlinearly along the wall height. The coefficient of lateral active earth pressure and the resultant of active earth pressure are both independent of the shear stress between level soil layers; while the position of action point of the resultant earth pressure by the proposed formulae considering the shear stress is higher than that of Coulomb’s solution, and lower than that of existing theories in which the shear stress is neglected.

Rainfall infiltration analysis is key to the prediction of rainfall-induced slope failure. Due to its simplicity and ease to use, the Green-Ampt model is potentially very useful for rainfall infiltration analysis in prediction of shallow landslides. However, the original Green-Ampt model assumes that the ground is level and that the initial distribution of water content does not vary with depth. In this study, the governing equation for infiltration analysis of sloping surfaces with arbitrary distribution of initial water content is derived. The Runge-Kutta method is used to solve the equation numerically. The modified model can be reduced to the previous Green-Ampt model when the initial water content is uniformly distributed. When the initial water content is not uniform, Richards’ equation shows that there is a transition zone between the saturated and unsaturated zone which is not affected yet by the rainfall infiltration. The wetting front calculated from the modified method is in this transition zone and is close to the bottom of the saturated zone. Overall, the distribution of pore water pressure predicted from the modified Green-Ampt model is close to that predicted based on Richards’ function.

In order to investigate the characteristics of ultrasonic wave velocity and acoustic emission activity of salt rock under triaxial loading, the triaxial loading tests are conducted. Employing a device composed of acoustic wave testing system, acoustic emission testing systems, we obtain the following results: Four stages are identified in the process, including stageⅠof the elastic compression, stageⅡof initiation and stabilized growth of fractures, stage Ⅲ of accelerated growth of fractures, stage Ⅳ of sustained strength. During phaseⅠ, the velocity of ultrasonic wave of salt rock has a slight increase，and a few acoustic emission(AE) events happen in this stage. In the stageⅡ, P-wave and S-wave remain stable, AE events become active. In phase Ⅲ and Ⅳ, AE activity of salt rock performs intensely, and the velocity of P-wave remains stable while velocity of S-wave reduces significantly. The level of confining pressure significantly influence ultrasonic wave velocity and acoustic emission activity. The rate of change of velocity of ultrasonic wave is large at a low level of confining pressure while small at a high level of confining pressure. When dynamic axial strain is 10%, the total amount of AE events is 16 271 under confining pressure of 5 MPa, the number decreases to 8 764 under confining pressure of 10 MPa due to the effect of compaction, 3 041 under confining pressure of 15 MPa, and 906 under confining pressure of 20 MPa, which indicates that the level of confining pressure greatly influenc the AE activity, showing “the effect of confining pressure densification”.

Permeability is the key control factor representing anti-seepage ability and anti-fouling property of clay layer, and the proper choice of permeability coefficient of clay liner is crucial to ensure anti-fouling effect of the liner. The purpose of this investigation is to characterize the variation of the permeability of saturated clay with mixed heavy metal ions solution as permeant fluid. The permeability tests are performed with flexible-wall permeameters. It is shown that the permeability of saturated clay is enhanced with the increase in proportion of ions concentration for the permeant fluid with soluble Cu2+ and Cr3+, Cu2+ and Mn2+. In addition, under the same experiment conditions, the permeability coefficient of clay liner permeated with Cu2+ and Cr3+ mixable solution is larger than that with Cu2+ and Mn2+ mixable solution; the existence of mixed heavy metal ions weakens the ability of clay liner, and deteriorates the hydraulic conductivity of clay layer. The scanning electron microscope reveals that the permeant fluid plays an important role in changing the soil microstructure. The effective transport pore fluid passage increases with increasing mass ratio of mixable ions in permeant fluid, and aggregates are found in the soil samples, which is in good agreement with variation of macroscopic permeability. The results can provide a reference for evaluating anti-fouling ability of clay as valid barrier.

With considering the difference between the horizontal and vertical mechanical properties of ground base, the torsional dynamic response of a pipe pile in transversely isotropic soil is investigated. Based on the stress-strain relationships of transversely isotropic material and soil-pile coupled vibration, differential equations and boundary conditions of the pile and soil are established. The solutions for the displacements of the outer and inner soil are obtained by Laplace transform and the variable separation method. The torsional response of the pipe pile is obtained based on the continuity conditions between the pile and both the outer and inner soil. The expressions of the complex impedance and velocity admittance at the pile head are obtained. The solution proposed in this paper is then compared with existing solutions to verify the validity. A parametric study is conducted to investigate the influence of the anisotropy of the outer and inner soil on the torsional vibration characteristics of the pipe pile.

The rock bolt foundation is a special type of wind turbine foundation. It shows a good pull-out performance, and can well demonstrate mechanical properties of rock mass in its original state. Recent studies indicate that the modified Mohr-Coulomb (M-C) failure criterion with a small but not zero tension cut-off is more applicable for describing failure characteristics of rock masses. According to the modified M-C failure criterion, an expression is derived with the energy dissipation rate per unit area. Furthermore, the planar translational failure mechanism of rock wedge is put forward, which is based on the upper bound theorem for the limit analysis of soil plasticity mechanics. A virtual power equation is deduced, when the work rate of external loads to self-weight of rock wedge equals to the rate of energy dissipation along the sliding surface of rock wedge. Then a formula is developed for calculating ultimate pull-out capacity of rock bolt. The prototype test results show that the calculated values by this formula are in good agreement with the measured values, which indicates the formula with certain reliability can meet the requirements of engineering application.

To investigate the influence of intermediate principal stress factor b and minor principal stress on stress-strain characteristics of cement-soil, true triaxial tests are conducted under the conditions of constant minor principal stress ?3 and intermediate principal stress factor b. The results show that both the initial tangent modulus and the peak strength of cement-soil have positive correlation with b; however, the correlation between major principal strain at failure state and b is opposite. What’s more, there is a linear relationship between the peak strength and b, and the straight slope increases with the minor principal stress. For the same value of b, the initial tangent modulus, the peak strength and the major principal strain of cement-soil increase with increasing minor principal stress. The peak strength linearly rises with confining pressure as b is increased. The stress-strain curve changes gradually from softening type to hardening type, along with the increasing of confining pressure and intermediate principal stress. Failure stress ratio becomes lower due to the increased ?3 and intermediate principal stress factor b. Finally, the parameters of the generalized Mises yield criterion and the generalized Tresca yield criterion are obtained. Comparative analysis shows that the former criterion is superior to the latter criterion to describe the mechanical behavior of the cement-soil.

The stability problem of the bored pile hole wall in saturated soft clay is taken as the unloading shrinkage problem of a cylinder in the semi-infinite space. According to the SMP criterion-based Cam-clay model, the elastoplastic solutions of the cylindrical unload shrinkage problem are derived using the stress transformed method, and the analytic expressions for the critical plastic pressure and the hole-wall shrinkage are obtained. On this basis, the methods are applied to calculate the minimum slurry density and the borehole safety factor, and the factors influencing borehole stability are analyzed in details. The results show that, after the bored pile hole-wall yield, the soil element around the wall is in the elastoplastic state, with the hole-wall shrinkage increasing and the borehole safety factor decreasing. Under the mud supporting condition, the stability of the bored pile hole-wall is dependent on the soil density, over consolidation ratio, internal friction angle and the slurry density, etc., and independent of the hole diameter and the hole depth. The larger the over consolidation ratio and the friction angle, the smaller the critical support pressure and the minimum slurry density for maintaining bored pile hole-wall stable. In practice, the slurry density must be properly chosen so that the minimum slurry density is satisfied.

Joint is the main factor influencing elastic modulus of rock mass. According to the statistics of joints parameters, a constitutive relation of jointed rock mass is deduced by combining with strain energy, which can reflect the weakening and anisotropy of elastic modulus. Then, the characteristics and influencing factors of elastic modulus of jointed rock mass are analyzed by using tunnel surrounding rock of Ji-Tu-Hui railway as an example. The results show that the weakening and anisotropy degree of elastic modulus vary with the angle between stress and structural plane normal, which can obtain the maximum value at the angle of 54.5°. Meanwhile, the weakening degree of elastic modulus is in negative correlation with the friction angle and cohesion of joints, but in positive correlation with the mean radius and density of joints. And when the mean radius is over 3 m, the elastic modulus tends to be stable. Confining pressure also has influences on the elastic modulus of rock mass. When confining pressure is increased, the value of elastic modulus increases, and weakening and anisotropy degree is decreased. When the confining pressure reaches 2.32 MPa, the weakening and anisotropy will disappear.

The classical earth pressure theory is based on the assumption of semi-infinite space. However, the dimension of soil behind the retaining wall is limited, so that this theory is no longer suitable for calculating the earth pressure. The aim of this study is to develop the earth pressures for the limited cohesionless soil under complex conditions. The new solution for the active earth pressure is developed using the thin-layer element method. It is shown that the Ranking and Coulomb’s earth pressure theories are the particular cases of the proposed model. It is also shown that the limit rupture angle is not constant and it varies with the calculation parameters. The new solution approaches to Coulomb’s active earth pressure when the soil is infinite. When there are limited soils, the new solution tends to be consistent with the simulation solution, proving the reasonability of the new solution. In this case, it is significantly different from Coulomb’s solution.

To assess the variation of stress and deformation of surrounding rock during the process of tunnel water inrush, this paper aims at developing new types of similar material of fault and surrounding rock. The similar materials are further used in the solid-fluid coupling model test according to the similarity principle of geomechanical model tests. The developed fault similar material consists of sand, talc powder, gypsum, bentonite and paraffin liquid. The similar material of surrounding rock is composed of sand, barite powder, talc powder, white cement and latex. In these two materials, sand and talc powder are used as the coarse aggregate and the fine aggregate, respectively. Bentonite is used as cementing agents and paraffin liquid is used as a regulator in the fault similar material. The latex is used as bonder in the surrounding rock similar materials. A large amount of experimental tests are conducted to examine the deformation and failure behaviors of similar materials and prototype rocks. By adjusting the ratios of various ingredients of similar materials, the effects of sand, talc powder, gypsum, bentonite, paraffin liquid, barite powder, white cement and latex, on the uniaxial compressive strength, permeability coefficient, density, Poisson’s ratio and elasticity modulus are examined. Experimental results show that the proposed similar materials can be used for the solid-fluid coupling model test to model rock mass with different permeability coefficients with low and moderate intensity, due to their wide adjustment of parameters, stability of mechanical property, convenience of preparation and low cost. The developed similar materials are applied to simulate the water inrush occurred in the Yonglian tunnel of Jiangxi Province, China. The results show that the mechanical and seepage performances of similar materials satisfy the requirements of model tests, which can not only capture water inrush process of model tunnel, but also reflect the variations of physical fields.

A series of true triaxial tests is conducted by using TSW-40 type true triaxial apparatus to research the strength characteristics of silty clay under complex stress paths. The stress-strain curves change law of silty clay in true triaxial tests is researched. The influence of moisture content, confining pressure and intermediate principal stress ratio on the strength of silty clay is discussed. The difference among Mohr-Coulomb strength criterion, twin shear strength criterion and experiment results on the plane is compared and analyzed. The results show that the generalized shear stress-shear strain curves of silty clay mainly perform as strain hardening type. The strength of silty clay increases with the increase of the intermediate principal stress ratio and the consolidation confining pressure, but decreases with the increase of moisture content. The strength envelope on the plane gradually expands outward with increasing hydrostatic pressure, gradually shrinks inward with increasing moisture content. The comparative analysis results show that the strength of the soil is smaller than test results when Mohr Coulomb strength criterion, is larger when twin-shear strength criterion is adopted.

Shanghai 4th layer clay is a typical marine clay with distinct structure. It is helpful for numerical analysis to develop a constitutive model which uses one set of parameters to simulate the mechanical behavior of the clay subject to various stress paths. In this study, the evolution rules of structure and overconsolidation of UNIFIED constitutive model are modified. A new method is proposed to obtain the values of initial state and material parameters of undisturbed soils. In order to verify the modified model, the blocking sampling method is used to obtain undisturbed samples of Shanghai 4th layer clay. The conventional oedometer tests and drained/undrained triaxial compression tests are conducted on the 4th layer clay. A comparison between the test results and constitutive modelling results shows that the modified model and proposed method can simulate the stress-strain behaviors in the oedometer and triaxial tests of 4th layer clay properly, using only one set of parameters. It also shown that the structure of Shanghai 4th layer clay remains stable and the structure remains unchanged even after triaxial shearing at an axial strain of 35%.

Freezing tests are conducted on saturated silty soil samples with different boundary temperatures and heights to study water migration, water redistribution, frost heave and ice lens. The experimental results shows that, when temperature becomes steady, water keeps on migrating to freezing front and water content increases sharply near the freezing front and ice lens forms. The geometry of ice lenses in saturated freezing silty soil is regular without dendritic structure and frozen soil has massive structure without network of cracks. There is a starting time for frost heave, before which the water in soil is drained by the sudden freezing of the supper-cooled soil water. When the frost heave occurs water from outside supplies to the soil sample and the frost heave amount due to water inlet approximates to the total frost heave. The height of soil sample influences water migration and frost heaving amount. A higher soil sample has smaller frost heaving amount and less water content while water increment distributes more diffuse and the starting time for frost heave is longer.

In this study, a 1/30 model for a large section tunnel is firstly designed based on the extension project of Jinjishan Tunnel along Fuzhou 2nd ring road. Then the earthquake-simulating tests are conducted under 21 different loading conditions using a bidirectional earthquake shaking table apparatus at Fuzhou University. Through Fourier analysis of testing results, it indicates that the existence of the large section tunnel changes the seismic property of original ground significantly. It is also found that the 1st predominant frequency reflects the seismic properties of original ground, while the 2nd predominant frequency demonstrates seismic properties of the lining structure. The frequency-response analysis of testing results shows that the components augment significantly around the 2nd predominant frequency, when the seismic wave propagates from the bottom to top until the lining structure. The most-augmented frequency almost equals to the 2nd predominant frequency. The relationship between the seismic property of the large section tunnel and the amplitude of input ground motion is studied in detail. The apparent cracks emerge on the lining structure after the excitation by the large ground motion, which significantly reduces the overall stiffness of the large section tunnel. Moreover, both the 2nd predominant frequency and the most-augmented frequency show remarkable reductions. In conclusion, this study provides some useful suggestions for the earthquake fortification of the large section tunnel.

To explore the effect of structural damage on the time-dependent behavior of structural soils, a one-dimensional elastic visco-plastic (1DEVP) constitutive model is developed with introducing a new concept of constant visco-plastic strain rate lines, based on Bjerrum’s time lines. To describe the progressive destruction, a new structural parameter named “structural strain” is defined and a new mathematical formula is proposed based on characteristics of soil structure degradation under the one-dimensional loading condition. The method for calibrating the proposed model is discussed, and the proposed model is used to simulate conventional compression tests on Wenzhou natural clays, step loading fast-consolidation tests on natural Ariake clays, constant rates of strain tests (CRS) oedometer tests and long-term creep oedometer tests on structured Berthierville clays. The comparisons between simulations and experimental results show that the newly proposed 1DEVP model can reasonably describe destructuration effects on the time-dependent behavior of structured soft clays.

A series of triaxial undrained shear tests is conducted on the granular assemblies using discrete element method. The effects of initial fabric anisotropy caused by different sample preparation methods on the mechanical behavior and critical state of granular soil are investigated. The numerical samples are generated using the radius expansion method and the gravity deposition method, which can produce the samples with initial isotropic and anisotropic fabrics respectively. The fabric tensor is used to quantify the fabric anisotropy of granular soil and the influences of different initial fabric anisotropies on the fabric evolution are presented. The results show that initial fabric anisotropy has a great impact on the dilatancy of granular soil. The dilatancy of granular soil is more significant when the samples are more anisotropic or the direction of fabric anisotropy and loading direction are coincided. It is also observed that initial fabric anisotropy has no influence on the critical state of granular soil and the critical fabric tensor has a unique measure associated with the loading mode.

Scouring effect has a significant impact on the lateral vibration of caisson-piles composite foundation because of the change in stress history and the reduction of soils around the foundation. Based on the dynamic Winkler model proposed by author, a computational method of the dynamic impedance of caisson-piles composite foundation influenced by scouring effect with considering stress history is developed. Then sponge boundary dynamic FEM are used to validate the theoretical results. Finally, the proposed procedure is used to analyze a practical engineering. The results show that scouring effect has a significant impact on the dynamic stiffness and damping coefficient, and the more scour depth increased, the more remarkable influence on the results. This implies that it is not enough to only consider the loss of soil around the composite foundation for scouring effect, the change of stress history can further weaken bearing capacity of composite foundation and enlarge the dynamic response of structure, thus ignoring the change of stress history will result in an overestimate of value of dynamic impedance, and the change of stress history accompany with an apparent increase in amplitude of resonant response of composite foundation, while it has no effect on the value of resonant frequency.

Based on the logarithmic spiral collapse mechanism, and with considering the heterogeneity of clayey soil, the optimal upper bound solution of the face supporting pressure is obtained in heterogeneous soil in which the cohesive force varies linearly. The validity of the proposed solution is demonstrated by comparing the calculated results with the monitoring data and the results of model tests. In addition, the optimal supporting pressures and the logarithmic spiral sliding surfaces are analyzed in details for various soils. It is shown that in heterogeneous soils the soil cohesion and internal friction angle are the controlling factors. A design chart is developed for the limit supporting pressures in heterogeneous soils.

It is a challenge to deal with the scrap tyre all over the world. Reusing the scrap tyres in the geotechnical engineering provides an effective way of controlling pollution. The aim of the tests is explore the application of the scrap tyres to improving soils. A series of tests at different high stress levels is conducted on completely decomposed granite (CDG) sand and tyre rubber (granular rubber and rubber fibre) mixtures. The compression behaviour of the mixtures with different rubber percentages and different shapes of tyre rubber particle are analysed. The effect of the rubber particle on the breakage of sand particle in the mixture is investigated during compression. It is concluded that the normal compression line (NCL) can be found in each the mixture. The compression and rebound significantly increase as the rubber percentage is over 20%, which are not affected with the shape of rubber particle. There is less breakage in the mixture with higher rubber contents. Meanwhile, the larger size of the sand particles, the more breakage of the sand in the mixtures.

An analytical solution is presented for Biot’s consolidation induced by surface loading within a finite two-dimensional (2D), isotropic, homogeneous and fluid-saturated poroelastic media. Here, the consolidation is assumed to be governed by the incompressible porous media model. The Dirichlet boundary conditions are adopted for the pore pressure field, and the displacements field on the upper and lower surfaces conforms to physical boundaries. However, the boundary conditions of the displacements field on the lateral sides are specifically given. Finite sine and cosine transforms, Laplace transforms and Laplace numerical inversions are implemented; and the analytical solutions in the closed-form of double summations of series are obtained. Finally, an example of plane strain consolidation is studied. The proposed exact solution is validated against the numerical solution by the finite element analysis software ABAQUS. And based on the presented analytical solution, the spatiotemporal evolution of the pore pressure field and the displacements field is briefly examined. The presented analytical solution can be of great potential to further understand the behavior of flow and deformation coupling in a finite 2D fluid-saturated porous medium.

The static drill rooted nodular pile is a new type of composite pile foundation which consists of precast nodular pile and the cemented soil along the pile shaft. The bearing capacity of this composite pile is promoted via the cemented soil around the pile shaft which improves the skin friction and the enlarged cemented soil base which promotes the tip bearing capacity. A field test by self-balanced method is conducted to investigate the tip bearing capacity of this new type of pile; the theoretical tip displacement formula is also adopted to calculate the tip displacement of the composite pile. In addition, a model test is conducted to study the tip bearing capacity of the new type pile with different relative positions of the precast nodular pile base and enlarged cemented soil base. The results of the field test and model test show that: the enlarged cemented soil base can truly promote the tip bearing capacity of the new type pile; the tip bearing capacity of this new type of pile seems to be fully mobilized when the nodular pile tip is at the middle of the enlarged cemented soil base; the theoretical tip displacement formula can be applied to this new composite pile foundation; and the theoretical tip displacement-tip load curve fits well with the measured curve when the nodular pile tip is at the middle of the enlarged cemented soil base.

Currently, most definitions of the damage variable for jointed rock mass only use geometrical factors of joints such as the length and dip angle, but cannot consider joint shear strength. Due to this limitation, a new damage variable formulae is firstly deduced to calculate rock mass with a non-persistent closed joint, based on the connection between the increment of additional strain energy caused by the existence of one joint in fracture mechanics and the emission of damaged strain energy in damage mechanics. Secondly, the calculation method of the stress intensity factor (SIF) of a single joint tip under biaxial compression is studied according to fracture mechanics theory, and then the calculation SIF formulae of KⅠ and KⅡ are obtained. Thirdly, the calculation formulae of the tip SIF of a set of one or more-rowed joints are given by considering the interaction among the joints. Finally, a biaxial compression damage constitutive model for the jointed rock mass is developed, which is further employed to analyze an example. It is found that for rock mass with a single non-persistent closed joint, the strength of rock mass is the same as that of the intact rock, and the damage is 0 when the joint dip angle is less than its internal friction angle. Furthermore, with the increase in joint dip angle, the change laws of rock mass strength and the damage with the joint dip angle are parabola with the hatch up and down, respectively. The strength of rock mass is the lowest and its damage is the highest when the joint dip angle is about 60°. With the increase of joint length, the damage of rock mass increases; while with the increase of the internal friction angle of joint, the damage of rock mass decreases. For the rock mass with a set of one-rowed non-persistent closed joint, when the total length of the joint is the same, the damage of rock mass gradually decreases with the decrease of a single joint length and the increase of the joint number, however the relationship between the decrease amplitude and the joint number exhibits nonlinear response.

An isolated similar design method is developed for a scaled model test, and then is applied to simulate a shaking table test for the slope-anchor cable-lattice beam system. The functional equation of dimensional analysis is defined as the primary characteristics equation, which is used to describe the relationship of variables in the shaking table test. According to the difference of characteristics of the slope, anchor cable, lattice beam and seismic wave, all the system variables can be divided into four sections, which is further used to derive the scaling relations for the slope, anchor cable, lattice beam and seismic wave. The functional equations for characterizing the scaling relations for these four sections are the secondary characteristics equations. The importance degree of variables in the secondary characteristics equation is different for the scaled model. The functional equations used for characterizing the scaling relations of more important factors are the tertiary characteristics equations. Based on the tertiary, the secondary and the primary characteristics equations of these four sections, the scaling relations for the primary variables, relevant variables and irrelevant variables can be derived, respectively. Then the primary variables are selected to design the scaled model. The problems that all the scaling relations in the previous studies cannot be satisfied in the scaling model have been resolved.

Mengdigou Hydropower Station is the fifth cascade hydropower station in Yalong River’s middle reaches. Due to the maximum dam height of 200 m and complex geological conditions, it is of significant importance for engineering safety to ensure the dam’s good workability and global stability after impounding. Based on deformation reinforcement theory (DRT), a 3D finite element model of Mengdigou is built to analyze the behavior of the dam’s displacements, plastic zones extension and global stability under both normal and overloading conditions. The results show that the displacements and overloading capacity meet the stability requirement. The introduction of fillet significantly reduces the plastic complementary energy of both the dam and base, which is beneficial to the global stability. Two ways of replacement treatment for f4 fault are investigated. It turns out to be more favorable using shallow replacement together with grid replacement for the dam’s global stability.

It is important to ensure the soft soil ground stable during the rapid filling of an embankment. The controlling indexes adopted in current embankment specifications are the displacement or its rate. The fact that ground failure repeatedly occurs implies that the controlling indexes need to be reviewed and improved. In this paper, the concept of critical state line is proposed. This critical line is obtained from the figure coordinated by the two indexes of the embankment central settlement rate and horizontal displacement rate of slope toe based on a series of field data and centrifuge model test data. The critical line separates the embankment into stability and instability regions. Hence this critical line with its mathematical expression can be taken as a critical index, which can be used to identify the stability or instability states. The three indexes, i.e. the embankment central settlement rate, slope toe horizontal displacement rate and the critical line, are combined together to form an improved stability controlling standard, which is more reasonable than the current stability controlling standard, which adopts independently the above two indexes of displacement velocities.

Since stochastic discrete fracture network (DFN) model can only consider statistical properties of discontinuities, it is difficult to incorporate other significant factors, such as geological genesis, structural pattern and feature. To fully characterize the true DFN of rock mass, non-negligible biases are introduced into the subsequent mechanical and seepage calculations based on a pure stochastic model. The Beishan granite, a well-studied rock type, from a Chinese high-level radioactive waste repository, is selected for the analysis. A new stochastic-deterministic DFN model is built up for Beishan granite in terms of the statistical properties of measured discontinuities, the inherent hierarchy of discontinuities with different scales, and their hydraulic interconnections. The deterministic part of this model is implemented after combining the unbiased stochastic model with manually identified structural patterns. Model validations are conducted using the graphical comparison and seepage simulation techniques. The results show that the amount of discontinuities in the new model is consistent with the measured data, with an increase in model accuracy by 48.8%. The flow path and flux calculated by the new model are more realistic and more consistent with testing data, with approximately a one-third increase in consistency compared with the traditional model results. In addition, in-situ seepage-pressure tests are conducted on different boreholes. Testing results show that the deterministic discontinuities have great influence on the seepage simulation compared with the stochastic ones. The regional seepage is primarily controlled by these deterministic discontinuities. The developed new model is beneficial for the future DFN simulations.

The deep-seated reservoir ancient landslide often shows intermittent reactivation which corresponds to seasonal rainfall and periodic reservoir water level fluctuation. In this kind of reactivation, sliding and sliding dormancy occur alternately. This means that the sliding zone soils may experience shearing at different shear rates and consolidation with different durations. The kinetic equations are established based on three aspects, i.e., shear rate effect of residual shear strength in sliding, peak shear strength recovery in sliding dormancy and the initiation of sliding from sliding dormancy caused by pore-water pressure. Ring shear tests are performed on the soil samples from slip surface of Tangjiaocun-1 landslide in this paper. The results show that the residual shear strength is negatively dependent on the small range of shear rates and there is no decrease of residual shear strength after a larger shear rate even with a trend of small increase; the peak shear strength can be recovered within a short consolidation duration after the shearing ceased whereas the strength recovery is lost after a small shear displacement; and shear failure triggered by pore-water pressure lags behind the increase of pore-water pressure applied on sample, the lag implies pore-water pressure diffusion. The kinematic mechanism is then discussed based on the results of ring shear tests and the kinetic equations established above. These results may provide help for the prediction of deep-seated reservoir ancient landslides due to rainfall and/or reservoir water level fluctuation.

This paper aims to investigate the laws of large deformations of surrounding rock around a deep mining roadway. A true triaxial model experimental device is applied to simulate the strain characteristics of surrounding rock under different loading gradients based on geomechanical estimation and strata behavior test. The results show that, under shallow hydrostatic pressure conditions, surrounding rocks express a characteristic of tensile strain at shallow depth and zero strain at deep level. The tensile and compressive zonal regions alternately appear under deep hydrostatic pressure and initial excavation stress conditions. The nonlinear large strain phenomenon of surrounding rocks is observed when mining stress concentration factor is greater than 2. A comparative simulating of the displacement and distribution of plastic zonal regions in surrounding rock is conducted using the Mohr-Coulomb constitutive model and the Mohr-strain-softening constitutive model in FLAC3Dsoftware. It is found that under strain softening condition, the zonal disintegration of surrounding rock occurs due to tensile and compressive effects, and displacement after softening agrees with the observation. Based on aforementioned results, the zoning mechanism of surrounding rock induced by deeply mining are revealed. Some preliminary measures including grouting and shotcreting are suggested to control large deformation of surrounding rocks induced by immoderate strain softening.

The numerical simulation method is used to analyze the reinforcement effect of dynamic compaction. The volumetric strain of reinforced foundation and the required dry density are linked. The required volumetric strain is proposed as an evaluation standard to determine the reinforcement range. It can be found from both the numerical simulation and test data that the larger impulse leads to the better reinforcement, while a fixed tamping energy can lead to different reinforcement effects. Hence, it can be concluded that the tamping impulse is more suitable to be used as a control standard of construction parameters than tamping energy. In order to study the hammer size, a modified method is proposed. Besides, it is shown that if the parameters of soil are given, settlement will be only concerned with impulse per unit hammer area; while reinforcement depth will be concerned with both the impulse per unit hammer area and the radius of hammer. The influence of tamping number is also studied. On the basis of setting the required volumetric strain as the evaluation standard and analyzing the influences of major parameters, the formulas of reinforcement range are developed with dimensional analysis method. The proposed procedure is used to predict the reinforcement range of a foundation by dynamic compaction, it is found that the calculated dry density distribution of reinforced foundation is in agreement with measurements, which demonstrates the proposed procedure is applicable.

The material point method (MPM) is applicable to simulate large deformation of continuous medium, such as slope failure process. The strength reduction method (RSM) is applied to material point method for evaluating slope stability. In a slope stability evaluation example, the safety factor and slip surface position are got respectively by limit equilibrium method (LEM), finite element strength reduction method (FEM-RSM) and material point strength reduction method (MPM-RSM). Compared with LEM, the caculated results of safety factor value and slip surface position by MPM-RSM are basically the same with LEM and FEM-RSM. Compared with FEM-RSM, the physical meaning of MPM-RSM’s failure evaluation criterion is more clear. Taking advantage of its superiority in calculating large deformation, MPM-RSM can evaluate the consequences after slope failure. A case study shows its performance in evaluating debris shape and sliding distance for different safety factors, especially the advantage in estimating the impact degree of landslide on adjacent buildings. MPM can be used for the evaluation of slope stability and evaluation of slope failure consequence.

Mechanical properties of rocks at different loading rates have positive effects on their inner failure mechanisms under dynamic loading conditions. Numerical simulations of uniaxial compression and Brazilian tests are conducted on granite specimens using the bonded particle model (BPM) in particle flow code (PFC2D). To quantify rate-dependent behavior of compression strength, tensile strength, failure modes, strain energy rate and acoustic emission, four different loading rates are chosen from 0.001 to 0.500 m/s. The uniaxial compression strength (UCS) and its corresponding strain, and the tensile strength and its corresponding strain increase nonlinearly with loading rate. The failure pattern and damage degree are also influenced by the loading rate. It is found that specimens under compressive loading have a single inclined section at a lower loading rate, while have several oblige section forms at a higher loading rate. In addition, crack numbers and high horizontal strain rate distributions show that the failure zone area and damage degree increase with increasing loading rate. In Brazilian tests, the failure modes change from one primary crack to several major cracks as loading rate increases. Cracks extend to the edge of circle specimens, and loading rate exacerbates damage degree. Additionally, acoustic emission counts and strain energy rates of specimens in both tests increase in nonlinear forms with increasing loading rate.

Although lined tunnels usually pass through the alluvial valley; the study of their seismic responses is limited. The SH-plane wave scattering induced by an elastic-half-space lined tunnel in alluvial valley is studied with MATLAB program and the indirect-integral boundary element method. The influences of frequency, tunnel depth and incident angle are investigated on the surface displacement and the dynamic stress concentration factor of the internal and surface displacements of alluvial valley. The difference between tunnel and semi-space tunnel in alluvial valley is compared. The results show that the dynamic stress concentration of lined tunnel in alluvial soft clay is significant, and due to the high frequency incident waves, the effect of tunnel depth in alluvial area on the displacement distribution and lined stress spectrum as well as surface displacement is also significant.

The existing smoothed particle hydrodynamics (SPH) method is normally difficult to establish the particle model in the simulation of landslide. This paper aims at developing a new particle arrangement and insertion method based on geographic information system (GIS). Firstly, the SPH particle model is established and its corresponding particle generation is programmed. Secondly, the Bingham fluid model combined with the Mohr-Coulomb failure criterion is used to achieve the three-dimensional movement process of the damaged slope. Finally, the SPH model is verified by simulating Tangjiashan landslide and is further applied to predict the influence of the Jinpingzi landslide. The results indicate that the established landslide SPH model based on spatial GIS data is feasible and applicable. Therefore, the simulation study of landslide disasters can not only enhance greatly the analysis of geological disasters, but also provide reference for the forecast and control of landslide hazard.

The generation of three-dimensional (3D) manifold element is a critical problem for analyzing with 3D numerical manifold method (NMM). The aim of this paper is to investigate the generation of 3D manifold element and then develop the corresponding program with the C++ programming language. With the method of generation of 3D manifold element, the vertices, oriented edges, loops, faces and shells are considered as the fundamental data structure of 3D block based on the oriented theorem of topology. The Boolean intersection operations of blocks and mathematical meshes are conducted to validate new blocks. The manifold elements are generated once the validity of new blocks is satisfied. The vertices of each mathematical mesh are considered as the mathematical covers of new manifold elements, and then the physical covers are generated by subdivision of mathematical covers. Case studies for blocks with concaves, hollow or finite structural surfaces are conducted by using Boolean intersection operations with corresponding mathematical meshes. Moreover, a selected rock slope with many finite discontinuities is used to verify the developed method, with which shows that complicated block with concaves, hollow or finite structural surfaces can be well dealt. This study provides an effective and reliable way to analyze complicated structures.

First of all, weakened material is generated by introducing strength reduction factor(SRF) into bonded-particle model(BPM), hereafter, uniaxial compressive strength tests on weakened specimens are performed. The results indicate that: the weakening effect both reduces uniaxial compressive strength of the specimen and leads to a gradual decline in elastic modulus of the specimen, which is consistent with findings of related laboratory tests. In order to further examine the effects of the proposed weakening method, a numerical model for direct shear tests on weakened rock joint is developed by means of a computer program, which is written by a built-in language named FISH in particle flow code (PFC); and then direct shear tests are conducted at various normal stresses. Tests results indicate that: the shear character of weakened joints corresponds well to that of real rock joints, in addition to the dependencies of shear strength and peak dilatation angle on normal stresses are also conform well to the classical JRC-JCS model. It is suggested that the rock strength reduction method adopted in this study could preferably simulate the effect of strength reduction of rock material. Finally, investigation of evolution feature of microcracks within specimens shows that: attenuation of joint wall will lead to a rapid growth and development as well as rapid expansion of distribution range of microcracks, and will also result in a rapid increase in the proportion of shear cracks in specimens. As a result, the reason for a weakened joint is more susceptible to arise shear failure is revealed from mesoscopic angle. Research results in this article may provide a reference for the study of shear strength of rock joints，and for the study of stability of large-scale rock slopes.

Based on the framework of continuum damage mechanics (CDM), mechanical properties of jointed rock mass are characterized by using damage tensor and the effective stress. A numerical code named CD-FEM is developed to study the equivalent mechanical behaviour of rock mass. To investigate the uncertainty mechanical behavior of jointed rock mass, random input fields are represented using the Karhunen-Loeve expansion, and random output fields are expressed by polynomial chaos expansions. Probabilistic collocation method is used to generate collocations, and then the numerical code CD-FEM is adopted to solve the deterministic equations. Finally the output domain statistical results are obtained, from which a method for uncertainty analysis of jointed rock mass is proposed by the coupling of random analysis and CDM-based finite element method. The uncertainty analysis of mechanical behaviour of a typical jointed rock mass under external loading is conducted by this method. The computational efficiency of the proposed method is improved to a greater extent than that of the Monte-Carlo method. The developed method can be used to analyze efficiently the uncertainty of mechanical behavior of jointed rock mass under different loading conditions.

Large deformation of soil will occur when the structure moves through the soil, which is more appropriate to be analyzed by large deformation finite element method. Coupled Eulerian-Lagrangian (CEL) method is one of the most suitable method of the large deformation finite element methods. In China, there is no research on the ultimate capacity of plate anchors using the CEL method. Two pull-out model tests including square anchors in uniform and linear clay are taken as prototypes to build numerical models with the CEL method. Comparative studies are conducted between the model tests and numerical models on the ultimate capacity of plate anchors and the failure mechanism of soil. With a user-defined subroutine, the linear shear strength of soil can be updated with the mobilization of the anchor. The results show that the ultimate capacity of the anchor increases with the Young’s modulus, when the pullout force of the anchor increases with the displacement to a peak value at the beginning and followed by a decrease. If the embedment ratio is lower than the critical value, general failure mechanism can be observed. Contrarily, localized failure mechanism will develop. The numerical results agree well with model tests, demonstrating that the CEL method can well simulate the complete response of the anchor of large mobilization.