Based on the improved split Hopkinson pressure bar (SHPB) testing system with the rock coupled static and dynamic loads, dynamic experiments on serpentinite are conducted under the condition of different static axial compressions and frequent dynamic disturbance to research its dynamic deformation properties, dynamic peak stress and strain, energy variation and failure modes. The results show that in the process of frequent disturbance under the high static load, there is a positive correlation between the dynamic stress and strain of rock prior to the peak dynamic stress, but there are two different dynamic deformation phenomena, i.e. dynamic deformation rebound and non-rebound after the peak dynamic stress. With the increase of number of dynamic disturbance, the peak dynamic strain increases while the dynamic peak stress and the dynamic deformation modulus decrease, and the energy variation of rock changes from the release of energy from rock into the absorption of energy from dynamic disturbance. As the static axial compression is larger, the damage of rock during one impact disturbance is heavier, the impact disturbance number for rock failure is fewer, the failure mode of rock transfers from the tensile failure mode to the shear failure mode, and the fragmentation distribution of broken rock becomes more inhomogeneous, and fragments become larger. All these results will be helpful for us to reveal the failure mechanisms of deep rock mass under the high static stress and dynamic disturbance produced by frequent excavation and blasting, and prove the feasibility of adjusting static stress state and excavation blasting to maintain the long-term stability of surrounding rock in field site.

In order to evaluate the influence of bulkheads on the bearing capacity of bucket foundation, a series of comparative model experiments is conducted on two identical steel buckets, one with bulkheads and the other without. The load-displacement curves under the combined action of horizontal load and bending moment or under vertical load, the distribution of the soil pressure on the inner, outer and upper walls of the bucket are obtained by the experiment individually. Finite element simulation is performed at the same time. The experimental and numerical results show that under vertical load, the bearing capacity of bucket foundation with bulkheads is a slightly higher than that of the foundation without bulkheads. The bulkheads share the vertical loads from the cover and sidewall, cutting the load applied on the cover. Under combined action of horizontal load and bending moment, the bulkheads provide large pulling resistance for the foundation, improving the ultimate bearing capacity by 20.2%, and reducing the ultimate horizontal displacement and angle. In a word, bulkheads can not only improve the stability during towing, but also increase the ultimate bearing capacity of offshore wind turbine foundation during the operation.

The aim of this paper is to develop a method for simulating the macroscopic deformation of rock and the deformation of each component. Rocks can be regarded as the composition of voids part material and skeleton part material by considering the effect of heterogeneity of microdefects on macroscopic deformation of rock. Firstly, a deformation analysis method for voids part material is proposed by using the method of true strain description and the characteristic of nonlinear deformation. Meanwhile, on the basis of statistical damage theory and solid mechanics theory, a deformation analysis method for skeleton part material is developed. Then, a macroscopic deformation analysis method is established by analyzing the deformation process of rock and its components, which is used to obtain a statistical damage constitutive model for simulating the whole process of deformation and failure of rock. Moreover, the methods for determining the parameters of the constitute model are also given. Finally, experimental data and theoretical results of the proposed model and existing models are compared and analyzed. The deformation processes of rock and its components materials are further discussed. The current study indicates that the proposed model can not only reflect the main deformation characteristics of rock, but also overcome the difficulty of reflecting the characteristics of rock nonlinear deformation during initial voids compaction stage. In addition, the proposed model can be used to describe the deformation characteristics of components materials, and to reveal the relationship between macroscopic deformation of rock and the deformation of each component material. It has been shown that the deformation of voids part material is the fundamental reason for rock nonlinear deformation.

Rock progressive failure is usually accompanied by development of cracks. The propagation behaviour of rock cracks is closely related to the evolution of rock permeability. In order to investigate the permeability properties of rock during progressive failure, a series of hydro-mechanical triaxial coupling tests is carried out on tuff specimens under different confining pressures and seepage pressures using triaxial servo-controlled seepage equipment. Based on the experimental results, the following conclusions are drawn. At the stable crack growth stage, the permeability keeps low and stable and the permeability corresponding to crack initiation strength can be used to determine the minimum value of permeability. At the crack unsteady growth stage, the permeability increases significantly and the increase of permeability can be divided into two stages, which can be reflected by crack circumferential strain. The inflection point of crack circumferential strain curve is used to identify the initial permeability. At the softening stage after failure, the permeability decreases and enters into a residual stable stage, which can be reflected by crack circumferential strain rate. The circumferential strain rate can be used to determine the position of the peak of permeability.

Rubber-sand mixture (RSM) has recently been demonstrated to be an environmental beneficial geotechnical material with wide application. The purpose of this paper is to investigate the stress-strain characteristics, volumetric properties and the modulus reduction properties of RSM. United tests combining triaxial shear and simple shear apparatus, in which 2 confining pressures and 7 rubber contents are considered, are conducted. Experimental results indicate that: with the increase of rubber content, the modulus of RSM decreases, and the stress-strain curves extend to strain hardening as well as the linearity of stress-strain relationship enhances. Volumetric expansion of RSM decreases and contraction increases with adding rubber to RSM. The behavior of stress-strain of RSM could be simulated well by extended Duncan-Chang hyperbola model, and the model parameters are evaluated with rubber content in RSM. With the increase of rubber content, the Gs -? curves (where Gs is static shear modulus and ? is shear strain) of RSM decrease while the generalized Gs /G0 -? curves (where G0 is initial shear modulus) shifts toward to large strain domain. The stress-strain relationship and volumetric strain-shear strain relationship obtained from triaxial shear tests and simple shear tests are consistent and verified each other. The stress condition and stress path influence the experimental parameters. The stress-strain curve and modulus curve obtained from triaxial shear tests are higher than that from simple shear tests, which is possibly induced by higher mean stress in triaxial shear tests. An empirical relationship between the initial modulus of RSM and the reducing rubber content is proposed based on experiments.

The identification and analysis of geotechnical mesocharacteristic play a crucial role in the study of rock and soil mechanics. Based on the principles of Fourier transform, a two-dimensional outline of rock and soil particles can be transformed into Fourier descriptors and phases, and then a mesocharacterization method is established based on plural Fourier analysis. The corresponding relationships between different orders Fourier descriptors and mesocharacteristic parameter are investigated through counting Fourier mesocharacteristics of two groups of different particles. It is shown that the Fourier descriptors corresponding to -12-0 and 16-20 orders determine the shape of particles, and the Fourier descriptors corresponding to -23--11 and 1-15 orders are closely connected with the roughness of particles. There exists a significant linear correlation between the peak value of Fourier descriptor curve and the particle size of rock and soil particles. High-order Fourier coefficient and its order are approximately in a logarithmic linear relation, however, low-order coefficient is determined by different particle shapes. By using the established Fourier descriptor corresponding relationship, particle similarity remodeling is studied. It is shown that the outlines and Fourier statistical parameters of remodeled particles are very close to the actual particles, and the remodeled particles can reflect the mesocharacteristics of particles such as roughness and texture. The method can be used for the study of correlation between a large number of rock and soil particles and their mechanical properties.

Ground foundation adjacent to slope is already a universal foundation form. In view of the asymmetric failure characteristic of strip footing adjacent to slope, a bilateral failure model is developed. In this model, the double asymmetrical features are that the rigid sliding block sizes under both sides of strip footing or the geometry shape of the same slider is not identical. Based on the aforementioned mode, a new method for ultimate bearing capacity of strip footing adjacent to slope are put forward by adopting the upper-bound limit analysis method and optimization theory. The proposed precedure can reflect the influences of both the failure mode and the distance between strip footing and slope top on the bearing capacity of strip footing. At the same time, it could be degenerated well to determine the ultimate bearing capacity of the ground foundation. The feasibility, rationality and universal applicability of the approach are verified by the comparative analysis of results of engineering cases with existing methods in literature and the one in this paper.

A series of swelling tests is performed on a typical Nanyang expansive soil with medium swelling capacity compacted at various initial densities and water contents. The swelling tests are separately conducted using the conventional oedometer to confine the lateral swelling of the soil specimens, and using the GDS triaxial apparatus to allow the free volumetric swelling. The multiple nonlinear mathematical method is adopted to obtain the lateral swelling model (i.e. K0 model ), which fully considers the coupled effect of initial degree of compaction, moisture content and overburden pressure on the swelling strain. Also, an empirical model for the relationship between spherical stress and volumetric strain is proposed by triaxial swelling test. Based on the K0 swelling model, a formula is proposed to quantitatively evaluate the swell potential, and also a theoretical calculation method is derived to determine the processing layer thickness of expansive soil slope. Based on the assumption that volumetric swelling strain only changes with spherical stress and is not affected by the deviatoric stress, the correlations between the K0 model and triaxial model are analyzed, and a method to calculate the volumetric swelling strain by only employing the K0 model is given. Experimental results show that the proposed K0 model with multifactor coupling is reasonable to predict the swelling potential of compacted expansive soil. It is found that the key factor to link the K0 model and triaxial swelling model is assuming an average static lateral pressure coefficient. The average static lateral pressure coefficient tends to decreases with increasing overburden pressure by inversion method. This tendency of average static lateral pressure coefficient is believed to rely on the fact that lateral swelling pressure decreases with the increase of overburden pressure.

The failure of surrounding rock during excavation can be generally grouped into two categories: tensile failure and compressive-shear failure. The cracks induced by the two failure types exhibit different mesoscopic configurations, which lead to different permeability evolutions of rock with tensile and compressive-shear cracks. The tensile and compressive-shear cracks of granites are firstly generated by triaxial compression tests under different confining pressures and the permeability tests are conducted on the samples with tensile and compressive-shear cracks. It is shown that the permeability of tensile cracks is more sensitive to hydrostatic pressure. In order to analyze the mesomechanical structures of cracks, the scanning electron microscope tests are also performed on the surfaces of tensile and compressive-shear cracks. The tensile cracks dominated by axial compression tests and the crack surfaces are relatively smooth. With the increase of confining pressure, the mesomechanical structures of cracks surfaces change significantly, more and more compressive-shear cracks are generated on the crack surfaces, and the crack surfaces are relatively rough. Under hydrostatic stress, the tensile cracks can be easily closed, while the concave-convex configuration of compressive-shear cracks can prevent the closure to some extent. The evolutions of permeability of granites with tensile and compressive-shear cracks are essentially dependent on the mesomechanical structures of cracks surface. The above results are helpful in determining the permeability of the surrounding rock in underground engineering.

The dynamic equations for the horizontal vibration of outer soil and inner soil of pipe pile are developed with considering the layered characteristics of soil. The stiffness and damping coefficients of pipe pile-soil dynamic interaction are obtained using potential function, boundary conditions and parity of the radial and circumferential displacements of outer soil and inner soil. The soil is simulated as a continuous-distribution spring-damper, and a horizontal vibration equation of the pipe pile is developed under the forces of outer soil and inner soil. The horizontal vibration is solved by the initial parameter method and the transfer-matrix method, and the dynamic horizontal impedance of pipe pile at pile head is obtained. The influences of shear modulus of soil layer, pipe pile wall thickness, outer and inner soil shear modulus ratio and thickness of soil layer on the dynamic impedance of pipe pile are obtained through numerical analysis. It is found that within the range of low frequency, the horizontal vibration of pipe pile in layered soils is mainly affected by the shear modulus of soil layer, pipe pile wall thickness and outer and inner soil shear modulus ratio, while the lateral vibration of pipe pile is influenced by the thickness of soil layer in a wide frequency range. As the thickness of the pipe pile wall and the shear modulus of soil around the pile increase, the absolute value of the lateral dynamic impedance of pipe pile increases.

Fluid flow tests are conducted on an artificial model constituted by two crossed fractures with an intersecting angle of 90°. The nonlinear fluid flow patterns within the fracture intersection are observed by using the visualization technique with a charge-coupled device (CCD) camera. Two discrete fracture network (DFN) models are established with and without considering the fracture surface roughness, respectively. The nonlinear flow behaviors of fluid in these two DFN models are characterized by directly solving the Navier-Stokes equations under two kinds of boundary conditions. Experimental results show that the obvious nonlinear flow behaviors in the segment connected to outlet_3 are observed, and measurements also exhibit a nonlinear correlation between the flow rate Q and pressure P. Numerical results demonstrate that, when the hydraulic gradient J is to some extent high (i.e., J > 0.1), the flow rate Q through the DFN is nonlinearly related to the pressure P between two opposite boundaries. However, when J is relatively low (i.e., J < 10-4), Q is linearly proportional to P. Based on these two DFN models in the current study, the critical condition of applying the local cubic law to calculate fluid flow in every fracture in the DFNs is J ≤10-4. The fracture surface roughness can significantly influence the permeability of the DFNs, however, it has negligible influences on the relative flow rate errors.

Centrifugal model test is designed to simulate basal heave failure of narrow-deep foundation pit in soft clay. The bending moment and horizontal displacement of retaining wall, the earth pressure distribution as well as basal failure mechanism of excavation at different groundwater levels are analyzed. Based on fundamental parameters and excavation procedures of the centrifugal model test, a finite element model is established to analyze the basal heave stability factors and failure modes of foundation pit, comparing with the centrifugal test results. The results of centrifugal model test show that, with the excavation depth increases and groundwater level rises, the bottom uplift and deformation increase. Horizontal displacement towards pit occurs at the bottom of retaining wall, because of the lateral restraint reduction and greater uplift, at the same time, the axis force of internal struts increases. Eventually, retaining wall rotates around a braced point of its own, resulting in a skirting failure. The results of numerical simulation show that, with the increase of groundwater level outside the pit, the safety factor of basal heave stability gradually reduces. When foundation pit collapses due to basal heave failure, the uplift displacement at bottom is greater. The basal failure mechanisms revealed by centrifugal model test and numerical simulation are similar.

A new method for analyzing the debonding problem of a single prestressed anchor is proposed based on the catastrophe theory. A simplified nonlinear shear slip model of the anchorage interface is proposed. The total potential energy of the anchorage system is obtained using the elastic theory based on the assumption that the anchor is an elastic body. The catastrophe theory is introduced to simplify the energy function of anchor to the normal form of cusp catastrophe model. A failure criterion of anchor is established and the debonding analysis is conducted. The results show that the softening behavior of the anchorage interface can be reasonably described by the established nonlinear shear slip model. The distribution of the shear stress between the rod body and the surrounding grout gradually evolves into a single peak curve with the increasing drawing load, and the whole anchorage interface softens until damages occur at the ultimate state. The proposed theoretical formula for critical loose displacement is simple and practical and may provide theoretical tools for analyzing the debonding of a single prestressed anchor. Finally, the rationality and feasibility of the method proposed is also examined through an engineering project.

In order to obtain the law of moisture migration in unsaturated soil due to thermal gradient effect, experiments of vaporous water migration and moisture migration are carried out under different thermal gradients. Experimental results show that the temperature fields in all soil samples have been achieved steady within 24 h and changed linearly along the length direction of the soil samples. Migration quantities of vaporous water and liquid water both increase with the increase of temperature gradient, however the former is significantly greater than the latter. Temperature effect in vaporous water migration experiments increases significantly with the increase of initial water content of soil sample. By contrast, temperature effect in liquid water migration experiments has little relation with initial water content of soil sample. Finally, a function of water content gradient, which contains temperature gradient and initial water content, is developed.

To study the catastrophic process of the inrush of water and mud in a fault fracture zone tunnel, a 3-D geologic model test system is established. Based on the water and mud inrush catastrophe of the F2 fault zone in the Yonglian tunnel in Jiangxi Province, a new similar material of fault and surrounding rock is developed by examining a large number of different material ratios and the physical and mechanical parameters of the materials. The temporal effects of displacement, seepage pressure, stress-strain and mass of gushed particles around the tunnel are analyzed. The seepage pressure increases with time, and the smaller the distance between excavation face and monitoring interface, the greater the fluctuation range. Before the inrush of water and mud, the mass of gushed particles decreases temporarily and then increases rapidly. The displacement of surrounding rock vault is mainly in vertical direction, and the displacement of hance is mainly in horizontal direction. At the same stress state, the strain of hance is larger than that of the vault. Comparative analysis is made between the catastrophe characteristics of test and the process of field catastrophe evolution, showing consistent results.

A series of impact experiments is conducted on natural, soaked and saturated red sandstone samples. The effect of hydro-dynamic coupling on dynamic strength, deformation characteristics and energy mechanism of samples is analyzed by comparing with that of dry samples. Experimental results indicate that dynamic compressive strength and peak strain of samples are mainly affected by the water content and impact load, while the peak modulus is obviously influenced by the water content. The water softening effect leads to loosen expansion of granular structure, which further results in cementation weakening. In addition, the meso-mechanical effect of pore water might lead to low strength and high strain of samples at high strain rate. The strength of samples is strengthened and the plastic deformation is decreased by the effect of strain rate. A large amount of total absorbed energy U prior to the peak stress is transformed to the releasable strain energy Ue, but a small portion of U is transformed to the dissipation energy Ud. The strain energy of each part increases with the increase of the pulse intensity, however the variation with the water content is largely different. A modified brittleness index BIM (ratio of Ud to Ue) shows that both the water content and pulse intensity have a threshold resulting in the response of plastic deformation quite different. The effective impact energy index (Keff) shows that the strain rate effect significantly influence the impact tendency of dry and saturated samples. Water softening effect reduces the effect of strain rate, but the strain rate effect is enhanced when the saturation is reached.

Through on-site large-scale shaking table tests, the transfer function theory is used to derive the relative transfer function and the absolute transfer function, and these two transfer functions are compared with regard to their characteristics and the adopted dynamic parameters, and then the feasibility and accuracy for estimating the frequency domain response are analyzed. It is shown that the imaginary parts of the two transfer functions are equal to each other, and the dynamic parameters estimated by using the imaginary part of the two transfer functions are the same. The imaginary part of the two transfer functions and the module of the relative transfer function can be used to calculate the natural frequency of site, while the imaginary part of the two transfer functions and the module of the absolute transfer function are suitable for the calculation of the damping ratio. The acceleration vibration modes calculated by the real part, imaginary part and module of the two transfer functions are the same. It is feasible to estimate the seismic response of site in the frequency domain by the transfer function and the relative transfer function yields better results than the absolute transfer function.

A modified strain wedge (SW) method for analyzing the behavior of laterally loaded single piles in sand is proposed. The modified model assumes that the lateral displacements of a pile behind the three-dimensional passive soil wedge are nonlinear, which makes the horizontal soil strain variable with depths instead of a constant value in the original strain wedge model, and also employs two different hyperbolic models, one for describing horizontal stress increment-strain behavior of soil in the wedge, and the other for describing the shear stress-displacement property at the interface between soil and pile shafts. An example is analyzed to demonstrate the effectiveness of the modified method, and a good agreement is obtained. Finally, the effects of modifications on the lateral bearing capacity of pile shafts are discussed. The results show that the problem of overestimating the lateral bearing capacity of piles with strain wedge method can be ameliorated by introducing the assumption of nonlinear lateral displacements of piles. It makes the SW method more convenient and effective in analyzing the behavior of laterally loaded piles by introducing the new relationships of horizontal stress increment-strain and shear stress-displacement.

Sliding-zone soil is one of the most critical inherent factors that can affect the stability of landslide. A series of X-ray diffraction and mercury intrusion tests is conducted on the undisturbed sliding soil samples to obtain the composition and content as well as structure in the sliding-zone soil. The results show that the sliding-zone soil is mainly composed of quartz, labradorite, illite, etc, and the pores include mainly micropores and transition pores. Based on a practical landslide in which the pore water chemistry of the sliding-zone soil is variable, different chemical solutions are introduced as the saturating fluid. Through a series of tests including shear, permeability and scanning electron microscopy, variations of strength parameters, permeability and microstructure of sliding-zone soil are obtained. Variations of mechanical properties and permeability of the sliding-zone soil are analyzed. The results indicate that at the same pH value, the peak and residual strengths of soil soaked in the salt solution are larger compared to distilled water. However, under different pH conditions, introducing acid and alkali solutions as the saturating fluid can significantly reduce the strength parameters of sliding-zone soils. The results provide important data for both the analysis of the variation law of mechanical properties and the multifield coupling theoretical models of the sliding-zone soil under water chemical conditions.

This paper aims to investigate the promotion-efficiency and law of the dissipation process of excess pore water pressure by using permeable pipe pile based on indoor model test. The contrast tests between the permeable pipe pile and normal pipe pile indicate that the permeable pipe pile is beneficial to evidently promote the dissipation of excess pore water pressure. The promoting effect increases along the depth direction and decreases along the horizontal direction. Permeable pipe pile has the most obvious effect in the early period of the dissipation of excess pore water pressure, greatly speeding up the construction progress. Through changing the initial drainage time of the permeable pipe pile and observing the variation of the excess pore pressure dissipation in the soil around the pile, it is found that the peak value of excess pore pressure is evidently reduced in the soil around the pile when the initial drainage time is set at the finished moment of pile driving. This study shows that the permeable pipe pile can improve the construction schedule and reduce the influence on the surrounding environment in the process of pile driving,

Surrounding rock stability of underground caverns in the deep resources development are facing deformation and failure problems of post-peak fractured rock. The deformation and failure mechanisms of the deep fractured surrounding rock in strong unloading excavation are unclear, which normally leads to large volume collapse, large deformation and other major engineering accidents. In the current study, a large scale three-dimensional experimental model testing is conducted to investigate the characteristics of deformation and failure of deep rock mass with multi-cleftiness of different angles under the excavation condition at high in-situ stress. When the crack angle is small, the surrounding rock on upper and lower sides mainly presents large deformation phenomenon, and the rock on the left and right sides shows layered fracture phenomenon. With the increase of crack angle, the rupture zone from the left and right sides of tunnel gradually extends to the whole tunnel, and the top rocks become more prone to collapse. The displacement and stress of surrounding rock show the fluctuation from inside to outside. The plastic zone around the tunnel increases with the increase of crack angle. The larger the crack angle, the more the plastic zone around the tunnel is easily connected with the prefabricated crack at the top and bottom of the tunnel. The research provides some experimental data for ensuring the safety of construction and operation of deep underground engineering.

The depth where the maximum lateral wall deflection occurs during excavation is one of the key indicators for evaluating excavation deformations. To date, there are few studies involving the relationships between the depth of the maximum lateral wall deflection and the adverse influences on adjacent environment. Based on analysis of field monitoring data and finite element numerical simulations, the influences of depth of the maximum lateral wall deflection on both ground displacement and uneven settlement of buildings adjacent to excavation are investigated. It is found that the deeper the location of the maximum lateral wall deflection, the wider the ground displacement field and the greater the uneven settlements of buildings. The soil strata undergo various vertical displacements at different depths, and the displacement patterns are governed by depth of the maximum lateral wall deflection. As the maximum lateral wall deflection goes downwards, the ground displacement field of soil adjacent to pit extends to deeper level and the corresponding influence zone expands as well. The investigation results indicate that the excavation-induced adverse effects on adjacent environment could be mitigated effectively by controlling the location where the maximum lateral wall deflection occurs.

The permeability coefficient of the aquifer in the first terrace of the Yangtze River in Wuhan is inevitably anisotropic under the effect of the sediment dynamics sorting and consolidation pressure of the overlying soil. Therefore, the dewatering design based on the isotropic permeability coefficient of aquifer and the hydrogeological parameters obtained from a single well and steady flow pumping test is inevitably different from the actual engineering. Based on the stratified pumping test, the group drilling pumping test and combined with three-dimensional numerical inversion simulations, the basic laws and characteristics of the aquifer hydrogeological parameters are concluded in the first terrace of Yangtze River in Wuhan. The permeability coefficients gradually increase from top to bottom and show anisotropy in different aquifers. The horizontal permeability coefficient is greater than the vertical permeability coefficient, and the ratios of them are between 1.6 and 2.6. The case study shows that the dewatering design of deep foundation pit made in accordance with the result of the anisotropy of hydrogeological parameters and the three-dimensional seepage analyses is much more approximate to the actual situation. In addition, this method reduces the project cost and minimizes the impact of foundation pit excavation on surrounding environments. Moreover, it is also conducive to save precious groundwater resources.

Due to the distinctive formation mechanism of columnar joints, columnar jointed basalt exhibits strong discontinuity and anisotropy. The columnar jointed basalt rock mass in Baihetan project is selected as a case study. The structural control effect on the equivalent modulus of deformation is discussed by using jointed finite element method (JFEM). The results show that, on the transverse plane, the equivalent modulus of deformation increases with the column irregularity, which increases about 10% from completely irregularity to completely regularity. This trend also agrees with the increase of the column size. The equivalent modulus of deformation corresponding to a side length of 0.1 m is 5.36 GPa, and this value increases up to 23.4 GPa for a side length of 0.5 m. Columnar joints rock mass may be considered as isotropic on the transverse plane, and basically, the equivalent modulus of deformation varies correspondingly with joint stiffness. But when it comes to the longitudinal plane, the wider joint spacing of the joint set 2 is, the higher equivalent modulus of deformation is. The equivalent modulus of deformation increases firstly then decreases, and reaches its maximum value when the cross-bedding ratio is 50%. The equivalent modulus of deformation on the longitudinal plane shows a similar correspondence relationship with the joint stiffness on the transverse plane. The simulation results obtained by JFEM agree well with the existing field observations and 3D numerical studies.

The energy accumulation induced by abutment pressure and its burst release is one of the root causes of strain-mode rockburst. Firstly, the coal seam ahead of a coal face is divided into three zones, i.e., hindering zone, driving zone and unobvious influenced zone. The strain-mode rockburst from gestating to occurring is divided into three stages: energy steady accumulation, energy balance and energy unstable release. The basic condition of rockburst occurrence is discussed in detail, and then an energy criterion of abutment pressure induced strain-mode rockburst is established. Based on criterion, it is stated that the ratio of elastic strain energy released by the driving zone to energy consumed by hindering zone is higher than one when the mechanical equilibrium of coal and rock mass is destroyed. Secondly, a concept of critical width of the hindering zone is defined from the point of view of preventing rockburst, and an occurrence criterion of strain-mode rockburst is developed with a factor of the width of hindering zone. Finally, theoretical results and in-situ investigations show that both the energy criterion and the factors have reasonable accuracy and reliability in engineering application. This study can provide basis for forecasting strain-mode rockburst induced by abutment pressure and for implementing relevant preventive measures.

Quantitative determination and explanation for geometric characteristics of joints in deep rock mass are fundamental to investigate mechanical properties and permeability, and in particular they are the main factors to determine the site of nuclear waste reserve and evaluation. Based on the modified data of joint number from three boreholes of BS17, BS18, BS19 in Beishan, Gansu province, a negative exponential-like distribution mathematical model is established. Furthermore the model is applied to evaluate the integrity of deep rock mass in rock quality designation (RQD) system. The reliability of the model is verified by comparing with rock integrity designation (RID) results. It should be pointed out that there is a piecewise function of a linear relationship between the joint number and spacing, and the fractal dimension increases with the spacing expanding. According to the power-law between the joint number and spacing, there is a multi-fractal relationship between them, as well as the joint number and depth. The multi-fractal spectrum is suggested to analyze the statistical data of joint geometric shape and distribution. The results provide a reference for 3D simulation of joints in deep rock mass based on geometrical characteristics of borecore joints.

The existing formulae of roof bearing layer thickness for punch damage and shear damage are modified by reasonably simplifying the model for bearing layer of karst cave roof. A formula of roof bearing layer thickness under tensile and bending damage conditions is proposed with considering the deadweight of roof. With clear physical significance, the proposed formula can well address the effects of the effective width and the deadweight in one-way slab and two-way slab models, and it is straightforward to be applied. Based on the observed results in practice, the range of parameter variation is determined for various failure modes such as punching, and it is concluded that the net thickness of roof bearing layer is more than 2.5d-3.5d for punching and shearing (where d is pile diameter), while the total thickness is more than 5.0d-5.5d for bending, so that the design thickness can meet the design requirements. The results can provide reference for design and construction of bridge piles in karst area.

The side friction will decrease due to unloading rebound deformation after excavation, though it occurs only within a certain depth range near the bottom of pits. Unloading action can be considered as a uniformly distributed upward pressure on the bottom of pit when the side friction of piles after excavation is calculated. The additional stress of soils can be expressed by Mindlin solution when the load is under the ground. Excavation calculation depth Hc is defined as the depth at which change ratio of side friction is 5% between after and before excavation, and all the influential factors on Hc such as excavation depth, excavation wide, internal friction angle and Poisson’s ratio are considered. In addition, the calculation formula of Hc is derived. Although the result of vertical additional stress from Boussinesq solution is bigger than that from Mindlin solution, but the difference becomes smaller with the depth. The side frictions after excavation for 216 cases are calculated. It is shown that the maximum ratio ? of results from Boussinesq solution to that from Mindlin solution is 1.124, and the minimum value is 1.001. Therefore the side friction of pile after excavation can be calculated by Boussinesq solution for safety.

The Wudongde hydropower plant under construction is located where there exists strong regional tectonic movement. Faults, folds and joints are well developed, and most of the river valleys are steep and narrow in this area. Therefore, the distribution characteristic of geostress is quite complex, which exhibits a significant nonlinearity. To accurately obtain the distribution rule of initial geostress field in the dam area, a new two-stage back analysis method is presented, which considers the optimized distribution characteristics of geostress. Firstly, the geostress is decomposed into three components including gravitational stress, tectonic stress and nonlinear stress. By considering the influence of topography, geological structure, and inhomogeneity of the rock mass, a model in numerical software FLAC3D is developed. The superposition method is applied to the first-stage back analysis of initial geostress field. Then, by considering the small scale geological structures and local excavation in the vicinity of the dam area, a refined three-dimensional discrete element model using 3DEC is established. The preliminary lateral pressure coefficient and gravitational stress modified coefficient of the refined model are obtained from the developed model. Moreover, the calculation for the second-stage back analysis of geostress field is conducted by optimizing the coefficients based on a uniform design. During the first-stage back analysis, the calculated results agree well with the measured data on the whole, but the difference between them is large in the vicinity of the local geological structures. The calculated results are in agreement with measured data in the second-stage back analysis by considering the small scale geological structures, local excavation and the results in the first-stage back analysis. The commercial FLAC3D and 3DEC software can be combined to obtain the initial geostress field under complex conditions. The distribution characteristics of geostress field shows continuous in a large scale and discontinuous within a small range due to local geological structures.

Based on a continuous-discontinuous element method (CDEM), this paper is to discuss fracture mechanism of rocks under compressive loading applied by a single tooth of roller bit. The CEDM is then applied to investigate the fracture evolution process of rock mass under compressive loading, and to analyze the effect of cohesion, internal friction angle on rock fragmental volume and the patterns of fragmental pit. Numerical results show that the shape of fragmental pit is basically a semi-ellipsoid and the ratio of width to depth of the fragmental pit is only affected by the internal friction angle. With the increase of internal friction angle, the ratio of width to depth decreases gradually. The width and depth of fragmental pit and the volume of fragment can be described by the single tooth pressure F, tooth radius r, cohesion c and internal friction angle ?. By considering the rock damage influenced by several bits which are in contact with rock simultaneously, the correction coefficient of single tooth fragmental volume is introduced. Then the relationship between operating parameters of roller bit (i.e., axial force of drill pipe, rotational speed and drilling speed) and rock mass strength (i.e., cohesion and internal friction angle) is established. Field experiments are conducted at the south mining area of Anqian mine, and the rock fragmental volumes of different rock properties are obtained. The corresponding cohesive strength and internal friction angle of the rock mass are obtained by laboratory tests. When the correction factor is 0.363, the field testing results coincide well with the theoretical results, which demonstrates the validity of numerical analysis and formula derivation. The results will be used for the field test of dynamic strength of rock mass.

The stress wave expressed by the double exponential function is often used as the dynamic input in the numerical analysis for blasting engineering. A solution of stress wave propagation in elastic rock based on the characteristics method is an important benchmark for engineering numerical analyses, and for the validation of a dynamic numerical algorithm. The problem of blasting wave propagation in the surrounding infinite medium is solved using the given unified approach for elastic waves based on the characteristics method. The solution procedures for the radial and circumference stresses and the velocity and displacement waves during the compression blasting wave propagation are given, and these procedures are coded in MATLAB. The results of the radial and circumference stresses and the velocity and displacement waves are discussed. The characteristics method provides a powerful tool for solving the hyperbolic partial differential equations, and the solution of compression wave, expressed by the double exponential function, plays an important role in the analysis of the wave propagation problem.

Numerical manifold method (NMM) is established based on the two cover systems (including mathematical cover and physical cover) and the contact loop, and can be used to solve the continuous and discontinuous deformation problems in the geotechnical engineering in a unified way. Similar to other numerical methods based on the partition of unity (PU) theory, the NMM can also improve the computational accuracy by increasing the orders of local displacement functions freely without mesh refinement, though this may cause the global stiffness matrix singular, leading to the linear dependence issue. In this study, a new local displacement function of high-order polynomials is proposed. The new function is applied to solve the general elastic problems. The results show that the linear dependence is eliminated. Compared with the traditional NMM based on the first order polynomials, higher precision is reached. Stresses at nodes are continuous. All the degrees of freedom defined on a physical patch are physically meaningful, with the third to fifth simply being the strain components at the interpolation point of the patch. As a result, the stresses at the interpolation points can be directly obtained. The proposed procedure can be easily extended to other PU-based methods.

To explore the failure process and mechanical properties of soil-rock mixture (SRM) which contains soft and hard rock blocks, a modeling method for SRM is proposed based on the template database of soft and hard rocks. In the particle flow code in two dimensions (PFC2D) the template database of soft and hard rocks is established, which conforms to in-situ structural characteristics. The particle flow models of SRM with different rock block proportions (RBP) are generated, and then a series of uniaxial compression tests is simulated by the particle flow code. The simulation results indicate that the failure of SRM initiates from the interface between rock block and soil, and the separation of soil-rock medium is mostly caused by shear failure. The microcrack grows rapidly at the earlier stage of loading, and the growth rate of microcracks slows down at the middle and later stages, while the macroscopic deformation gradually becomes significant. The uniaxial compressive strength of SRM decreases with the increase of RBP. The rotation and lateral movement of rock block can aggravate the integral destruction of SRM. At a certain RBP, the uniaxial compressive strength of SRM shows a decreasing tendency with the increase of soft rock block proportion. The rupture rate of specimen and the rupture degree of single soft rock granule are maintained at a high level after the test, and the rupture of soft rock granules is an important reason for the decline of SRM’s uniaxial compressive strength.

The extended finite element method (XFEM) is presented to simulate natural fracture and hydraulic fracture. The generalized nodal shape functions are used in a cluster of nodes around cracks. To improve the accuracy in the vicinity of crack, a ramp function is used to avoid the blending element problem. An uncoupled model is used for hydraulic fracturing, which assumes uniform water pressures at crack faces. The mortar method (segment-segment contact approach) in combination with the augmented Lagrange’s method is adopted to establish contact conditions between the closed natural fracture faces. Several numerical examples are analyzed in detail. In comparisons with analytical solutions, the proposed method obtains higher accuracy on computing the stress intensity factors of the pressurized crack and the hydraulic fracture with the uniform water pressure on the crack faces. The influence of hydraulic fracturing on natural fracture face is simulated, and the contact stress and contact condition on the natural fracture face are further analyzed.

In this paper, the blasting vibration of rock is analyzed under the condition of multi-free surfaces, which is based on the excavation of a collapse cavity section in the extension of the existing tunnel. The effect of simulating blasting vibration on the reinforcement area of the collapse cavity and surrounding rock is investigated by a fully restarted numerical method and Lagrangian algorithm method. Furthermore, vibration velocities and the attenuation law of characteristic points are obtained. It is found that the maximum vibration velocities calculated by the numerical method are in accordance with the requirement of the safety of blasting vibration. It also verifies the feasibility of blasting design, and can be used a guide for the blasting construction. Meanwhile, the blasting vibration monitoring of collapse section excavation is also conducted. The obtained vibration velocities from field experiments with the numerical simulation are in a good agreement. The results indicate that numerical methods can be used to describe the rock fragmentation and blasting vibration well in the millisecond delay blasting. Therefore, the current study shows a good example for the similar engineering conditions.

The three-dimensional crack propagation is analyzed by the three-dimensional numerical manifold method (3D NMM) and the corresponding program with C++ is established. By taking the advantages of 3D NMM in crack propagation, only crack shells and manifold elements are required to be updated without the use of Heaviside function. According to the stress obtained from 3D NMM, the nonlocal crack tracing method is used to analyze the failure state of every crack tip. If failure occurs, the crack propagates along the direction perpendicular to the maximum principal stress. Quadrilateral or triangle tracing method is selected according to the situation during propagation. In order to make the deformed surface remain plane, triangulation of new generated face is needed. The edge crack, horizontal penny-shaped crack and inclined penny-shaped crack are simulated. It is feasible that simulating crack propagation with 3D NMM. Furthermore, the proposed method is suitable for cases of nonclosed and closed crack tip line and is also effective for the nonplanar 3D crack growth where the crack tip line is located in an element.

Natural gas hydrate deposits, as a new type of green energy, have attracted global attention for broadly commercial exploitable prospect. Methane hydrate (MH) has different formations in the pore of methane hydrate-bearing sediment (MHBS), i.e. pore-filling, cementation, and so on. A pore-filling type MHBS with a specific MH saturation is firstly generated based on the hydrate morphology. Subsequently, a series of consolidated-drained true triaxial tests with a constant mean effective stress is carried out using the distinct element method (DEM) under the same π-plane with different intermediate stress ratios (i.e., b = 0, 0.25, 0.50, 0.75, 1.00). The influence of intermediate principal stress on the macro-micro mechanical properties of MHBS is analyzed by linking the evolution of micro-mechanical parameters (i.e. the contact fabric and the percentage of sliding contact) to the macro-scale behavior of granular material. The results show that there is a good correlation between the strong fabric changes (i.e. major, intermediate, and minor principal fabrics) and the variations in principal stresses (i.e. major, intermediate and minor principal stresses) and principal strains (i.e. major, intermediate, and minor principal strains). The three-dimensional strength of MHBS can be approximately described by Lade-Duncan failure criterion. The percentage of sliding contact increases with increasing intermediate principal stress ratio and reaches the maximum when MHBS is in the triaxial extension state.

Reduction of sliding resistance of unstable rock mass after an earthquake and heavy rainfall, is one of the main causes of failures. The instability and failure of unstable rock masses depend on the condition of the surface structure. As cracks open and the cohesive force of the bonding surface decreases, rock blocks become unstable, and then collapse occurs. If the conventional displacement-based monitoring method is adopted, it is generally difficult to achieve the goal of monitoring and early warning. In contrast, natural vibration frequency can effectively demonstrate the changes in physical and mechanical parameters of rock, and thus the safety assessments of unstable rock can be easily made. Through model experiments, the internal cohesive force and friction force of rock under constant sliding force are analyzed based on the natural vibration frequency. The friction force is compared to the maximum static friction force, in order to check whether or not the rock is prone to failure. It is shown that the calculated friction force can be used to effectively represent the safety of rock and the rock fails when the natural vibration frequency is below 6 Hz. The increase of the frictional force is up to 80% of stabilizing force, so that the friction increases rather than decreases in some cases. By monitoring the natural frequency, a quantitative assessment of the rock damage is made, and the static friction force index is identified, providing a new way for safety monitoring and damage evaluation of rocks based on natural vibration frequency.